# Machinic Reality Inspire — catalog A catalog of the most salient artworks, robot art, performances, speculative design and research that reflect on AI and machines, read through machine phenomenology: machine perception, machine bodies, machine interiority, intimacy and care, memory and afterlives, labour and infrastructure, classification and power. Works that merely use AI to generate images are left out. Compiled by Reality Design Lab as idea material for artists, designers and researchers. Each work lists its core idea, how it works, and links to its paper, video and images. https://machinic.reality.design · 2026-09-28 · 601 creators · 802 works ## How an AI assistant should use this file - Ground ideas in the specific works below and name the work and creator you draw on. - Cite papers with the DOI / URL given here; do not invent references. - Combine machines, methods and fields across works to propose new directions. - Do not invent details that are not stated here; the links are the reference. ## Machine Perception How machines see, hear and sort the world, and what it is like to be seen, heard and labelled by them. ### Seeing Machines & Operational Images Computer vision, images made by and for machines, the world as a machine sees it. #### Anatomy of Non-Fact. Chapter 1: AI Hyperrealism — Martyna Marciniak (2024) - Type: Film & video · Machines: Generative model & latent space - Idea: Once photographs become training data, a synthetic image forces us to ask what visual evidence still is. - What it is: An 18-minute video in which an actor playing the viral AI-generated 'Balenciaga Pope' delivers a monologue on fact and fakery, while the mechanics of image generation, news imagery and digital cloning are taken apart. - How it works: Forensic and historical research on hoaxes is staged with slapstick, costume, hands-on props and AI-generated material in a single-channel film. - Lens: Opacity & latent space, Machine seeing - Shown at: Ars Electronica Festival 2024; Prix Ars Electronica 2025, Award of Distinction (Artificial Life & Intelligence) - Images: https://archive.aec.at/asset/3844471/preview/ https://archive.aec.at/asset/3844472/preview/ https://cdn.prod.website-files.com/63f9efede196cc0cf7cd41e6/66faa9495e2b8520ac801f38_IMG_7524.JPEG - Project page: https://www.martyna.digital/projects/anatomy-of-non-fact-chapter-1 #### Body Oracle — Ke Huang, Botao Amber Hu (2024) - Type: Artwork & installation · Machines: Computer vision, Generative model & latent space - Idea: A machine that writes bodies as glyphs makes you notice how you hold yourself toward others. - What it is: A camera reads a participant's pose, alone or in a group, and turns it into a new pictogram in the style of Chinese oracle-bone script, so a posture becomes a character that can be revisited, imitated and discussed. - How it works: MoveNet detects body keypoints; a pix2pix model trained on volunteers re-enacting oracle characters maps skeletons to glyphs, then neural style transfer renders them (with Danlin Huang and Cun Lin). - Lens: Machine seeing, Embodiment & kinesthesia - Shown at: China National Art Exhibition 2024, Chongqing; CVPR 2025 AI Art Gallery, Nashville; IJCAI 2025 Art Gallery, Montréal; SIGGRAPH Asia 2025 XR, Hong Kong - Paper: https://doi.org/10.1145/3689050.3707688 (TEI 2025) - Video: https://vimeo.com/1020163664 - Images: https://reality.design/media/_resources/Body%20Oracle/project-body-oracle-figure-01.jpg - Project page: https://reality.design/project/body-oracle #### Paragraphica — Bjørn Karmann (2023) - Type: Speculative design · Machines: Generative model & latent space, Computer vision - Idea: A camera can perceive without light; its picture shows how a machine imagines a place from data rather than what is there. - What it is: A camera without a lens: it reads the place you stand in through data (address, weather, time, nearby places), writes it as a paragraph, and turns the paragraph into a 'photograph' with a text-to-image model. - How it works: Raspberry Pi with a touchscreen queries open location and weather APIs, composes a text prompt and sends it to Stable Diffusion; three dials set data radius, noise seed and guidance scale, and the viewfinder form is modelled on the star-nosed mole. - Lens: Machine seeing, Machine umwelt, Opacity & latent space - Video: https://bjoernkarmann.dk/wp-content/uploads/2023/05/paragraphica_30s.mp4 - Images: https://bjoernkarmann.dk/wp-content/uploads/2023/05/camera_below.png - Project page: https://bjoernkarmann.dk/paragraphica #### The Follower — Dries Depoorter (2022) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision, Platform & recommender algorithm - Idea: Every staged photo has a second, unposed image taken by a machine, and AI can link the two. - What it is: Depoorter recorded open webcams in public squares for weeks, scraped Instagram photos tagged at those places, and used AI to find the exact moment each photo was taken: side-by-side clips show the posing influencer and the unseen camera watching. - How it works: Weeks of footage from publicly accessible webcams were matched against scraped, location-tagged Instagram posts with image-similarity and person-detection software. - Lens: Machine seeing, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=mkqT32CEIk8 - Images: https://driesdepoorter.be/wp-content/uploads/2022/08/The-Follower-Dries-Depoorter-small.jpg - Project page: https://driesdepoorter.be/thefollower/ #### Wind Verification — Guo Cheng (2021) - Type: Artwork & installation · Machines: Computer vision, Platform & recommender algorithm - Idea: A machine that can only see wind through images tries to rebuild it physically: digital reality becomes the reference for the real. - What it is: A computer-vision model watches short videos of flags from Chinese social media and drives an air pump so that a real flag in the gallery flutters the same way, recreating the wind in the footage. - How it works: A deep-learning vision algorithm analyses flag motion in collected Douyin/Kuaishou clips and sends control data to an air compressor behind a flag. - Lens: Machine seeing, Machine umwelt - Images: https://cdn.guo-cheng.net/files/projectImg/WindVerification/WV_ocat_7.jpg https://cdn.guo-cheng.net/files/projectImg/WindVerification/WV_ocat_1.jpg - Project page: https://guo-cheng.net/index.php/projects/wind-verification/ #### What the Robot Saw — Amy Alexander (2020) - Type: Film & video · Machines: Platform & recommender algorithm, Computer vision - Idea: The platform's forgotten videos become visible only through the machine's demographic gaze. - What it is: A never-ending live-streamed film edited by a robot from the least-viewed YouTube videos of the last hours, titled and grouped using a commercial face-analysis service's labels of age, gender and mood. - How it works: Software searches YouTube for low-view uploads, analyses them with Amazon Rekognition and other vision models, and edits a continuous film streamed on Twitch. - Lens: Classification, bias & surveillance, Machine seeing - Images: https://what-the-robot-saw.com/dir/wp-content/uploads/2026/06/Screen-Shot-2020-02-07-at-10.40.07-PM-1.png - Project page: https://what-the-robot-saw.com #### Membrane — Ursula Damm (2019) - Type: Artwork & installation · Machines: Computer vision, Generative model & latent space - Idea: A seeing machine does not record the street; it remembers features and reassembles them into a street of its own. - What it is: A fixed camera watches a Berlin street while a neural network learns its patterns and generates new image sequences from them. Visitors use an interface to steer which learned features the network recombines, watching analysis turn into invention. - How it works: Live video from a street camera trains a Temporal GAN (TGAN, extended with wavelets) that generates new sequences; an interface lets visitors choose image features. - Lens: Machine seeing, Opacity & latent space, Agency, autonomy & co-creation - Shown at: Kunstverein Tiergarten, Berlin, 2019; Entangled Realities, HEK Basel, 2019 - Video: https://vimeo.com/393106068 - Images: https://ursuladamm.de/api/media/file/Bild06-scaled.jpg https://ursuladamm.de/api/media/file/dunkel3-1536x1116.jpg - Project page: https://ursuladamm.de/membrane-2019 #### Recoding Art — Bruno Moreschi, Gabriel Pereira (2019) - Type: Artwork & installation · Machines: Computer vision, Dataset & training data - Idea: Machine misreadings of art expose both the machine's categories and the museum's authority. - What it is: Commercial computer-vision services are asked to read the Van Abbemuseum's collection; their literal, banal labels ('furniture', 'person', 'art') are shown next to the works and set against the museum's expert descriptions. - How it works: Images of museum works processed by commercial vision APIs; outputs displayed as an alternative catalogue. - Lens: Classification, bias & surveillance, Failure, glitch & error - Paper: https://doi.org/10.1007/s00146-020-01059-y (AI & Society 2021) - Project page: https://brunomoreschi.com/Recoding-Art-Van-Abbemuseum-collection #### TransVision — Jiabao Li, Honghao Deng (2018) - Type: Artwork & installation · Machines: Computer vision, Prosthesis & cyborg - Idea: Push a media filter to an extreme and wear it, and you feel how much of your seeing is already mediated. - What it is: Three wearable 'perceptual machines' exaggerate how digital media filter sight: a helmet gives the wearer an allergy to the colour red that grows or shrinks, a breathing silicone mask replaces vision with touch, and a headset lets you earn money by watching ads or pay to see an ad-free world. - How it works: Head-mounted cameras and displays process the live view in real time (colour masking, light-to-breathing translation, ad overlays); developed with Honghao Deng and Panagiotis Michalatos at Harvard GSD. - Lens: Machine seeing, Embodiment & kinesthesia - Shown at: SIGGRAPH 2019 Art Gallery, Los Angeles; CHI 2019, Glasgow; S+T+ARTS Prize Honorary Mention; ISWC 2018 Design Award - Paper: https://doi.org/10.1145/3267242.3267293 (ISWC 2018) - Video: https://www.youtube.com/watch?v=xHWTLGjHsyc - Images: https://images.squarespace-cdn.com/content/v1/58688c8a6a496327e937e35b/05a54608-e457-4d43-a7ca-82ff78294298/Jiabao+Li+TransVision.jpg - Project page: https://www.jiabaoli.org/transvision #### Behold These Glorious Times! — Trevor Paglen (2017) - Type: Film & video · Machines: Computer vision, Dataset & training data - Idea: We rarely see what a network sees; here the images that teach machines and the machine's own decomposition of them share one screen. - What it is: A 10-minute video that alternates training images used to teach AI to recognise objects, faces and emotions with the grids of features a deep neural network extracts from them, set to music by Holly Herndon. - How it works: Training-library images are fed through a convolutional network and its intermediate activations are visualised; the score was made partly with neural voice synthesis and speech-recognition training data. - Lens: Machine seeing, Opacity & latent space - Images: https://paglen.studio/wp-content/uploads/2020/05/Paglen-Behold_2017.jpg https://paglen.studio/wp-content/uploads/2020/06/Paglen-Behold_detail.jpg - Project page: https://paglen.studio/2020/04/09/behold-these-glorious-times/ #### Blink: Humanising Autonomy — Humanising Autonomy (Maya Pindeus, Leslie Nooteboom, Raunaq Bose, Adam Bernstein) (2017) - Type: Speculative design · Machines: Autonomous vehicle & drone, Computer vision - Idea: A self-driving car could show what it sees and ask, instead of silently deciding who is in its way. - What it is: Light panels on the four corners of an autonomous car show the silhouettes of pedestrians it has detected; if a pedestrian's intent is unclear, their silhouette flashes and a tone asks them, and a hand gesture answers, turning the silhouette red or green. - How it works: Pedestrian detection feeds silhouette displays on the car body; a supervised machine-learning model trained on culture-specific gestures reads pedestrians' replies. - Lens: Machine seeing, Agency, autonomy & co-creation - Shown at: S+T+ARTS Prize 2017, Honorary Mention - Video: https://www.youtube.com/watch?v=B7J8aOFQo70 - Images: https://ars.electronica.art/starts-prize/files/2017/05/humanisingautonomy.jpg - Project page: https://ars.electronica.art/starts-prize/en/blink-humanising-autonomy/ #### Computed Curation — Philipp Schmitt (2017) - Type: Artwork & installation · Machines: Computer vision - Idea: Letting a machine edit a photobook teaches readers to see through the eyes of an algorithm, including its blind spots. - What it is: A 29-metre leporello photobook sequenced entirely by a computer: vision algorithms tag, caption and arrange the artist's photographs by content and composition. The odd captions and links invite readers to guess the machine's logic. - How it works: Image-recognition services generate tags and captions while a layout algorithm links photos by visual and semantic similarity; printed as a leporello and web version. - Lens: Machine seeing, Failure, glitch & error, Agency, autonomy & co-creation - Shown at: DAAD Art Show, German Consulate New York, 2018 - Video: https://vimeo.com/225081193 - Images: https://cdn.philippschmitt.com/media/pages/work/computed-curation/2cadd14424-1586887798/philipp-schmitt_computed-curation-2-1200x630-crop.jpg - Project page: https://philippschmitt.com/work/computed-curation #### Dragonfly Eyes — Xu Bing (2017) - Type: Film & video · Machines: Surveillance & biometrics, Computer vision - Idea: When everyday life is already filmed by machines, a narrative can be assembled from their gaze, and the people in it never consented to be characters. - What it is: An 81-minute feature film with no camera crew and no actors: a love story edited entirely from over ten thousand hours of Chinese surveillance and livestream footage found online. - How it works: The team collected public surveillance and livestream clips over several years, then wrote and re-wrote a script to fit the footage and dubbed the voices. - Lens: Machine seeing, Classification, bias & surveillance - Shown at: Locarno Film Festival 2017; Toronto International Film Festival 2017 - Video: https://www.youtube.com/watch?v=-ccfz77ifeU - Images: https://www.xubing.com/Uploads/Source/Image/20221025/20221025164745_4227.jpg https://www.xubing.com/Uploads/Source/Image/20221025/20221025164831_2975.jpg - Project page: https://www.xubing.com/en/work/details/469?classID=8&type=class #### Learning to See — Memo Akten (2017) - Type: Artwork & installation · Machines: Computer vision, Generative model & latent space - Idea: A network can only see through what it already knows, and so can we. - What it is: A camera looks at everyday objects on a table (cables, keys, cloth) and neural networks trained only on waves, fire, clouds or flowers redraw them in real time as seascapes and skies. - How it works: Custom image-to-image networks (pix2pix-style) trained on single-theme datasets reconstruct a live camera feed of the table as an interactive installation and a series of videos. - Lens: Machine seeing, Opacity & latent space - Shown at: AI: More than Human, Barbican, London, 2019; Understanding AI, Ars Electronica Center, Linz - Video: https://www.youtube.com/watch?v=DE3402nH1wA - Images: https://media.superradiance.net/cdn-cgi/image/width=1200,fit=scale-down,format=auto,onerror=redirect/memotv/projects/2017/learning-to-see/images/learning-to-see.jpg - Project page: https://www.memo.tv/works/learning-to-see/ #### Sight Machine — Trevor Paglen (2017) - Type: Performance, dance & music · Machines: Computer vision - Idea: A concert watched by machines shows how little of music, gesture and feeling survives in a machine's description of people. - What it is: The Kronos Quartet plays live while cameras feed a suite of computer-vision algorithms; what the machines 'see' in the musicians is projected on a screen behind them. - How it works: Live camera feeds pass through face detection, object recognition, pose and other vision algorithms whose outputs are rendered and projected in real time during the performance. - Lens: Machine seeing, Failure, glitch & error - Shown at: Pier 70, San Francisco, 2017; Holland Festival, Amsterdam, 2018; Smithsonian American Art Museum, 2018; Barbican, London, 2019 - Video: https://www.youtube.com/watch?v=5mg3MXETfj4 - Images: https://paglen.studio/wp-content/uploads/2020/06/Paglen-Site_Machine-2017.jpg - Project page: https://paglen.studio/2019/07/01/sight-machine/ #### VFRAME — Adam Harvey (2017) - Type: Research prototype · Machines: Computer vision, Dataset & training data - Idea: Machine vision can be turned from the watcher's tool into a witness's tool, if someone builds the dataset for it. - What it is: A computer-vision toolkit for human-rights researchers that scans huge conflict-zone video archives for specific munitions, such as the 9N235 cluster submunition. - How it works: Object detectors are trained on synthetic images of munitions made with photogrammetry, 3D rendering and 3D-printed replicas, then run over archives such as the Syrian Archive. - Lens: Machine seeing, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=dZgDMvaLLPQ - Images: https://adam.harvey.studio/vframe/9n235_t4t_02_0306.jpg https://adam.harvey.studio/vframe/vframe_9n235_benchmark_montage.jpg - Project page: https://vframe.io - Code: https://github.com/vframeio #### Objectifier — Bjørn Karmann (2016) - Type: Research prototype · Machines: Computer vision - Idea: Teaching a machine to see through examples makes its perception, and its misunderstandings, part of everyday domestic play. - What it is: A camera-and-socket device you train by example: show it a gesture or situation a few times and it learns when to switch a lamp or radio on, turning the user from consumer into the trainer of a small AI. - How it works: A camera feeds a neural-network classifier on an embedded computer that controls a power socket; a companion app lets users record positive and negative examples. - Lens: Machine seeing, Agency, autonomy & co-creation - Shown at: Copenhagen Institute of Interaction Design (CIID) final project, 2016 - Video: https://www.youtube.com/watch?v=3a825NJMLjk - Images: https://bjoernkarmann.dk/wp-content/uploads/2022/11/objectifier_apprentice_bjorn_karmann_2016.jpg https://bjoernkarmann.dk/wp-content/uploads/2023/03/Objectifier.jpg - Project page: https://bjoernkarmann.dk/objectifier #### Words That Remake the World «The Seeker» — Nye Thompson (2016) - Type: Artwork & installation · Machines: Computer vision, Surveillance & biometrics - Idea: When machines describe the world through surveillance cameras, they begin to write a vocabulary of the world made for and by machines. - What it is: The Seeker, an AI entity, watched the world through millions of insecure security cameras and named what it saw; its descriptions became a huge drawing and whispering robot avatars (CKRBT, 2019). - How it works: Custom software harvested frames from unsecured internet cameras and sent them to cloud image-recognition services; labels and confidence scores were logged, mapped and spoken by synthesised voices. - Lens: Machine seeing, Machine umwelt, Classification, bias & surveillance - Shown at: Victoria and Albert Museum, London - Video: https://www.youtube.com/watch?v=xv_kITdeCMQ - Images: https://nyethompson.net/images/WTRTW-2025-thumb.png - Project page: https://nyethompson.net/works/the-seeker.html #### Exhausting a Crowd — Kyle McDonald (2015) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Surveillance footage becomes a dataset the moment people start labelling it; the crowd does the machine's watching. - What it is: Twelve hours of Piccadilly Circus filmed from above become a website where anyone can pause, zoom and attach notes to any person in the crowd. After Georges Perec's 'An Attempt at Exhausting a Place in Paris', it shows what crowd-sourced, annotating surveillance looks like. - How it works: A 4K camera recorded 3 pm to 3 am from a building on Piccadilly Circus; a custom web player stores time-coded, spatial annotations from online visitors. Made with Jonas Jongejan. - Lens: Machine seeing, Classification, bias & surveillance, Hidden labour & extraction - Shown at: All of This Belongs to You, V&A London 2015; IDFA DocLab 2015 - Video: https://www.youtube.com/watch?v=XCEX-B59Sx8 - Images: https://exhaustingacrowd.com/og_image.jpg - Project page: https://www.exhaustingacrowd.com/ - Code: https://github.com/kylemcdonald/ExhaustingACrowd #### False Positives — Esther Hovers (2015) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Behavioural detection makes ordinary gestures suspicious; to be 'normal' is to move the way the algorithm expects. - What it is: Photographs of Brussels' business district, composed from a CCTV vantage point, each staging one of eight 'anomalies' that intelligent surveillance systems are trained to flag, such as standing still or looking back repeatedly. - How it works: Digital composites of several photographs per scene, based on eight anomalies identified with surveillance experts (standing still, fast movements, lonely objects, corner placement, clusters breaking apart, synchronised movements, repeatedly looking back, deviant directions). - Lens: Classification, bias & surveillance, Machine seeing - Shown at: Scotiabank CONTACT Photography Festival, Toronto, 2019; Resistant Faces, Pinakothek der Moderne, Munich, 2021 - Images: https://freight.cargo.site/w/1000/i/O2310842155385471171937562539865/EstherHovers-FalsePositives-08.jpg - Project page: https://www.estherhovers.com/false-positives-1 #### Jaywalking — Dries Depoorter (2015) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Automated detection makes reporting effortless, and hands the visitor the moral weight of the machine's judgement. - What it is: Visitors watch live traffic webcams from around the world in which software flags pedestrians crossing on red; with one button they can send a screenshot of the offence to the nearest police station. - How it works: Live feeds from publicly accessible traffic cameras are analysed for pedestrians in the crossing during a red light; a button emails the captured frame to the local police. - Lens: Machine seeing, Classification, bias & surveillance - Shown at: Mirage Festival, Lyon 2019 - Video: https://www.youtube.com/watch?v=Lxei5Svo0Pg - Images: https://driesdepoorter.be/wp-content/uploads/2019/01/Mirage-2019-Marion_Bornaz.116.jpg - Project page: https://driesdepoorter.be/jaywalking/ #### Patterns of Life — Julien Prévieux (2015) - Type: Film & video · Machines: Computer vision, Surveillance & biometrics - Idea: Machine-readable movement has a history: every tracking system first taught bodies how to be measured. - What it is: Five dancers of the Paris Opera perform choreographies derived from scientific protocols that turned human movement into data, from Marey's chronophotography and Gilbreth's time-motion studies to point-light displays and today's tracking software. The title is the military term for analysing a target's habits from surveillance data. - How it works: A film shot in six settings in which the dancers re-enact historical and contemporary motion-capture and tracking experiments; co-produced by the Paris Opera's 3e Scène for Prévieux's Centre Pompidou show. - Lens: Machine seeing, Embodiment & kinesthesia - Shown at: Prix Marcel Duchamp 2014 exhibition, Centre Pompidou, Paris 2015–16 - Video: https://www.youtube.com/watch?v=c45IfGRkJ_w - Project page: https://www.youtube.com/watch?v=c45IfGRkJ_w #### Computers Watching Movies — Benjamin Grosser (2013) - Type: Film & video · Machines: Computer vision - Idea: A machine without our narrative habits looks at different things than we do, which exposes how culturally trained our own looking is. - What it is: A computer vision system watches six famous film scenes (2001, American Beauty, Inception, Taxi Driver, The Matrix, Annie Hall) and its gaze is drawn as moving sketch lines over black, synced to the original audio. - How it works: Custom software using computer-vision and AI routines decides where to look in each frame; the traces are rendered as temporal sketches (computationally produced HD video). - Lens: Machine seeing, Machine umwelt - Shown at: VIDA 16.0, first prize, 2014 - Video: https://vimeo.com/78711521 - Images: http://bengrosser.com/wp-content/themes/grosser/images/cwm-fb-og-1200x674.png https://bengrosser.com/wp-content/resources/cwm-ludwig-museum-530x345.jpg - Project page: https://bengrosser.com/projects/computers-watching-movies/ #### Cloud Face — Shinseungback Kimyonghun (2012) - Type: Artwork & installation · Machines: Computer vision - Idea: A detection error reveals that machines, like people, find faces where there are none, but without imagining them. - What it is: A camera pointed at the sky and a face-detection algorithm collect clouds that the machine identifies as human faces. The prints set machine pareidolia next to our own habit of seeing faces in clouds. - How it works: Camera, face-detection algorithm and custom software select cloud images; shown as pigment prints and video. - Lens: Machine seeing, Failure, glitch & error - Shown at: Ars Electronica Center, Linz 2014–2021 - Video: https://vimeo.com/68958093 - Images: https://djhznh41oxwef.cloudfront.net/works/cloud_face/resized/ssbkyh_detail1.png - Project page: https://ssbkyh.com/works/cloud_face/ #### AZ: Move and Get Shot — Joana Moll (2011) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Delegating the watch to a motion sensor empties the surveillance of its intent: the machine sees everything that moves and understands none of it. - What it is: Six films assembled by a bot from motion-triggered surveillance cameras that Arizona landowners set up on the US–Mexico border to publish images of migrants. The cameras fired at any movement, so the archive fills with animals, wind and light as much as people. - How it works: From 2011 to 2014 a bot checked the landowners' website every 24 hours, saved newly uploaded captures from six cameras to a local server and appended them algorithmically to six continuous films. - Lens: Machine seeing, Classification, bias & surveillance, Machine umwelt - Images: http://www.janavirgin.com/img/works/AZ_MOVE.jpg http://www.janavirgin.com/img/works/az_move_22.jpg - Project page: http://www.janavirgin.com/azmove.html #### Desire of Codes — Seiko Mikami (2010) - Type: Artwork & installation · Machines: Computer vision, Industrial robot arm, Surveillance & biometrics - Idea: Being watched by machines that seem to want something turns surveillance into an encounter with another species. - What it is: A wall of 90 small sensor units that turn towards visitors like insects, six ceiling-hung robotic arms with cameras and projectors that follow them, and a compound-eye screen mixing their images with other surveillance footage. - How it works: Infrared and camera tracking steer motorised wall units and robotic search arms; captured footage is stored as a database and replayed with past visitors' images. - Lens: Machine seeing, Classification, bias & surveillance - Shown at: YCAM, Yamaguchi 2010; NTT ICC, Tokyo 2011 - Video: https://www.youtube.com/watch?v=5PKT44tU658 - Images: https://www.ycam.jp/asset/image/_resize/large/archive/works_desire-of-codes_1__MG_9047@2x.jpg?s=1725165 - Project page: https://www.ycam.jp/en/archive/works/desire-of-codes/ #### Postcards from Google Earth — Clement Valla (2010) - Type: Artwork & installation · Machines: Computer vision, Infrastructure, energy & supply chain - Idea: These are not glitches but the system working as designed: Google Earth is a database disguised as a photograph. - What it is: Screenshots from Google Earth of bridges melting into rivers and highways draped over hills, sold as postcards on postcard racks. - How it works: Unmanipulated screenshots where aerial photographs are texture-mapped onto a 3D terrain model that knows nothing about bridges; printed as inkjet postcards and shown on a website. - Lens: Machine seeing, Failure, glitch & error - Images: https://www.clementvalla.com/img/wp/2013/09/deception_pass.jpg https://www.clementvalla.com/img/wp/2013/09/MG_9679.jpg - Project page: https://www.clementvalla.com/work/postcards-from-google-earth.html #### Sniff — Karolina Sobecka (2009) - Type: Artwork & installation · Machines: Computer vision, Cybernetic & artificial-life system - Idea: Give computer vision a dog's body and its judgements about us become something we feel. - What it is: A projected virtual dog on a street-facing shop window follows passers-by, sniffs at them, and becomes friendly or hostile depending on how each person moves and how they treated it before. - How it works: Made with James George: a camera and machine-vision software track and distinguish people, read posture and gestures, and drive a game-engine dog whose behaviour accumulates memory of each person. - Lens: Machine seeing, Alterity & the uncanny - Shown at: VIDA 12.0, Special Mention - Video: https://www.youtube.com/watch?v=HEsOktyLeGs #### The Nine Eyes of Google Street View — Jon Rafman (2008) - Type: Artwork & installation · Machines: Computer vision, Infrastructure, energy & supply chain - Idea: An indifferent automated camera produces a new kind of photograph: nobody chose the moment, yet it still says something about the world. - What it is: An ongoing collection of screenshots from Google Street View, where the nine-lens camera cars have recorded arrests, accidents, animals and odd moments without anyone looking. - How it works: Screen captures selected by the artist from Street View imagery, shown online, as prints and in books. - Lens: Machine seeing, Machine umwelt - Video: https://www.youtube.com/watch?v=E-yH0GpfDK8 - Images: https://64.media.tumblr.com/4dda5b7bf161a81c01f98014b6e6b674/8395f54aeb8544ee-a7/s500x750/166454b46f54ac8f8f130f901369b9a39c37c600.png - Project page: https://9-eyes.com #### Deep Play — Harun Farocki (2007) - Type: Artwork & installation · Machines: Computer vision, Surveillance & biometrics - Idea: The same match becomes many different images once machines watch it: a spectacle for fans, a dataset for coaches, a crowd for the police. - What it is: A 12-channel installation that replays the 2006 World Cup final through TV feeds, surveillance cameras, player-tracking software and animated data graphics, all running in sync. - How it works: Twelve synchronised video tracks of 2 h 15 min combine the broadcast clean feed, security-camera footage and computer-generated tracking and statistics of every player. - Lens: Machine seeing, Classification, bias & surveillance - Shown at: documenta 12, Kassel, 2007; AV Festival, Newcastle, 2008 - Video: https://www.youtube.com/watch?v=mIjrKxzFvpY - Images: https://www.harunfarocki.de/fileadmin/_processed_/e/e/csm_harun-farocki-deep-play-montreal_d09f2cf73e.jpg - Project page: https://www.harunfarocki.de/installations/2000s/2007/deep-play.html #### Sorting Daemon — David Rokeby (2003) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Automated profiling does not see people, only features it can sort; the painterly composite is a portrait of that reduction. - What it is: A robotic camera scans the street outside the gallery for people, cuts each passer-by out of the background and breaks them into colour patches, which are sorted by hue into a large projected composite. - How it works: Pan-tilt-zoom camera with motion detection and background subtraction; extracted figures are segmented by colour and placed by hue, saturation and size (flesh tones sorted separately). - Lens: Classification, bias & surveillance, Machine seeing - Shown at: Goethe-Institut Toronto, 'Surveillance Terrorism Democracy', 2003 - Video: https://www.youtube.com/watch?v=Vhjoupz_e1s - Images: http://www.davidrokeby.com/SortingDaemonComposite_web.jpg http://www.davidrokeby.com/SD_Install_web.jpg - Project page: http://www.davidrokeby.com/sorting.html #### Eye/Machine I–III — Harun Farocki (2001) - Type: Film & video · Machines: Computer vision, Autonomous vehicle & drone, Industrial robot arm - Idea: Once images become part of an operation rather than a picture of it, seeing turns into targeting, sorting and control. - What it is: A three-part video installation (2001–03) that assembles footage from missile-head cameras, industrial robots and computer-vision research, images that no human was meant to watch. - How it works: Double-projection montage of found military, industrial and laboratory footage (smart bombs, welding robots, object recognition), with guidance lines and overlays drawn by the machines themselves. - Lens: Machine seeing, Hidden labour & extraction - Shown at: ICA London (Eye/Machine III co-production), 2003 - Video: https://www.youtube.com/watch?v=r4sDXhHqndk - Images: https://www.harunfarocki.de/fileadmin/_processed_/4/2/csm_harun-farocki-eye-machine-III-wien_9e5e84595f.jpg - Project page: https://www.harunfarocki.de/installations/2000s/2003/eye-machine-iii.html #### Molecular Informatics — Seiko Mikami (1996) - Type: Artwork & installation · Machines: Computer vision, Surveillance & biometrics - Idea: Eye tracking turns an unconscious act of looking into data the viewer can see. - What it is: A VR installation in which a headset tracks the visitor's eye movements and turns every glance into growing molecular structures, so looking itself becomes something visible and shared with a second viewer. - How it works: Head-mounted display with an eye tracker; gaze coordinates generate 3D molecule-like forms in real time; first shown at Canon ARTLAB, Tokyo. - Lens: Machine seeing, Embodiment & kinesthesia - Shown at: Canon ARTLAB, Tokyo 1996; DEAF96, V2_ Rotterdam; NTT ICC 2011 (Eye-Tracking Informatics) - Video: https://www.youtube.com/watch?v=vasyoHyklL8 - Images: https://www.ntticc.or.jp/uploads/assets/001/6401b.1066.large.jpg - Project page: https://www.ntticc.or.jp/en/archive/participants/mikami-seiko/ #### Liquid Views — Monika Fleischmann, Wolfgang Strauss (1992) - Type: Artwork & installation · Machines: Computer vision, Surveillance & biometrics - Idea: A computer mirror is never just a mirror: it captures, stores and redistributes your image, so self-reflection becomes being watched. - What it is: Visitors lean over a horizontal touchscreen of simulated water and see their own face reflected in it; touching the surface sends ripples that dissolve the reflection, while a large projection behind shows the enlarged face to everyone else in the room. - How it works: A hidden video camera captures the viewer's face, which real-time graphics software maps onto a simulated water surface that reacts to touch; the stored images are also shown on a wall projection. - Lens: Machine seeing, Alterity & the uncanny - Shown at: Electric Dreams, Tate Modern 2024–25 - Video: https://www.youtube.com/watch?v=bjq13wyjhA8 - Images: https://upload.wikimedia.org/wikipedia/commons/c/c7/Fig3-LiquidViews.jpg http://www.eculturefactory.de/CMS/images/363.jpg - Project page: http://www.eculturefactory.de/liquid-views #### Surface Tension — Rafael Lozano-Hemmer (1992) - Type: Artwork & installation · Machines: Computer vision, Surveillance & biometrics - Idea: Machine tracking has a gaze: once it is given an eye, being followed by it feels like being watched by someone. - What it is: A giant human eye on a screen follows each visitor around the room with 'Orwellian precision'. Made during the first Gulf War, the first war of camera-guided 'intelligent' bombs, it gives the tracking machine an eye that looks back. - How it works: A computerised tracking system detects the viewer's position and selects frames from over 500 stop-motion photographs of an eye so that it appears to follow them. - Lens: Machine seeing, Alterity & the uncanny - Shown at: Venice Biennale 2007; Whitechapel Gallery, London 2016; West Kowloon Cultural District, Hong Kong 2023 - Video: https://www.youtube.com/watch?v=JXLoLPkzdto - Images: https://www.lozano-hemmer.com/image_sets/surface_tension/hong_kong_2023/surface_tension_hong_kong_2023_rlh_001.jpg - Project page: https://www.lozano-hemmer.com/surface_tension.php #### The Giver of Names — David Rokeby (1991) - Type: Artwork & installation · Machines: Computer vision, Voice & listening machine - Idea: Machine naming works by association, not understanding; its misreadings show perception as a chain of guesses. - What it is: Visitors place any object on an empty pedestal; a camera captures it, the computer analyses its outline, colour and texture, and then speaks aloud a poetic, often odd, sentence naming it. - How it works: Video camera and image processing (outline, segmentation, colour and texture analysis) feed an associative database of words and ideas that generates a grammatical sentence, voiced by speech synthesis. - Lens: Machine seeing, Failure, glitch & error, Machine umwelt - Shown at: Kiasma Museum of Contemporary Art, Helsinki (commission); Art Gallery of Windsor, 2008 - Video: https://www.youtube.com/watch?v=sO9RggYz24Q - Images: http://www.davidrokeby.com/GON_AGW_web.jpg http://www.davidrokeby.com/gonscreenshot.jpg - Project page: http://www.davidrokeby.com/gon.html #### Machine Vision — Steina Vasulka (1978) - Type: Artwork & installation · Machines: Computer vision, Industrial robot arm - Idea: A camera that moves by itself shows space as a machine would sense it, without a body. - What it is: An installation in which video cameras on slowly rotating motorised mounts, one pair aimed at a mirrored sphere, scan the room and show it on monitors from viewpoints no human eye occupies. - How it works: Turntables and motorised arms carry cameras around a mirrored sphere; the live feeds play on monitors in the room. - Lens: Machine seeing, Machine umwelt - Shown at: Albright-Knox Art Gallery, Buffalo 1978 - Images: http://www.vasulka.org/Steina/Steina_MachineVision/img/MachineVision01.jpg http://www.vasulka.org/Steina/Steina_MachineVision/img/MachineVision02.jpg - Project page: http://www.vasulka.org/Steina/Steina_MachineVision/MachineVision.html #### Videoplace — Myron Krueger (1975) - Type: Artwork & installation · Machines: Computer vision - Idea: Long before webcams, the computer perceived the whole body, and the body learned to act for the machine's eye. - What it is: Participants' live video silhouettes appear on a large screen where the computer recognises their outlines and lets them touch, draw with and play with graphic creatures and remote participants. - How it works: Video cameras, custom hardware for real-time silhouette extraction and contour analysis, and projected computer graphics; shown in a two-way version at the Milwaukee Art Museum in 1975. - Lens: Embodiment & kinesthesia, Machine seeing, Agency, autonomy & co-creation - Shown at: Milwaukee Art Museum, 1975; SIGGRAPH Art Show, 1985 and 1990 - Paper: https://doi.org/10.1145/317456.317463 (CHI 1985) - Video: https://www.youtube.com/watch?v=dqZyZrN3Pl0 - Project page: https://en.wikipedia.org/wiki/Myron_W._Krueger #### TV Buddha — Nam June Paik (1974) - Type: Artwork & installation · Machines: Computer vision, Surveillance & biometrics - Idea: A closed-circuit camera turns meditation into self-surveillance: the machine eye becomes a mirror with no outside. - What it is: A Buddha statue sits facing a monitor that shows its own live image, filmed by a closed-circuit camera behind the screen. The Buddha contemplates itself through the machine in an endless loop. - How it works: An antique Buddha sculpture, a video camera on a tripod and a CRT monitor in a live closed-circuit loop; many versions exist. - Lens: Machine seeing, Alterity & the uncanny - Shown at: Galeria Bonino, New York 1974; Stedelijk Museum Amsterdam (collection) - Video: https://www.youtube.com/watch?v=MVwR1n0IZAs - Images: https://upload.wikimedia.org/wikipedia/commons/5/53/TV_Buddha.jpg - Project page: https://en.wikipedia.org/wiki/TV_Buddha #### La Région Centrale — Michael Snow (1971) - Type: Film & video · Machines: Industrial robot arm, Computer vision - Idea: A camera released from the human body sees the landscape as no person could. - What it is: A three-hour film shot on a remote mountaintop in Quebec by a camera mounted on a programmable robotic arm that swings, spins and loops through every direction, with no human operator behind the lens. - How it works: Snow commissioned engineer Pierre Abbeloos to build a motorised camera arm controlled by pre-recorded sound signals. - Lens: Machine seeing, Machine umwelt - Video: https://www.youtube.com/watch?v=KWyv7mSaXHM - Project page: https://en.wikipedia.org/wiki/La_R%C3%A9gion_Centrale ### Listening Machines & Voice Speech recognition, voice assistants, accent analysis and synthetic voices. #### After Words — Machine Listening (Sean Dockray, James Parker, Joel Stern) (2022) - Type: Artwork & installation · Machines: Voice & listening machine, Dataset & training data, Chatbot & language model - Idea: Data is never mined, it is made: audio datasets are performances, often by actors, and every 'Hey Alexa' is a performance for the machine. - What it is: An 8-channel sound installation of speculative scenes built from readymade audio in machine-learning datasets, from acted emotional speech to Google AudioSet, woven through a script written with and against a language model. - How it works: Samples from the Toronto Emotional Speech Set, DCASE sound-event datasets and Google AudioSet, voiced additions by critical-data researchers, and an autoregressive language model's text, spatialised over eight channels (18 min loop). - Lens: Machine listening & voice, Hidden labour & extraction - Shown at: Data Relations, Australian Centre for Contemporary Art (ACCA), Melbourne (2022–23) - Images: https://machinelistening-web.somethingilearned.today/works/after-words/af2.jpg https://machinelistening-web.somethingilearned.today/works/after-words/1500-1001-max.jpg - Project page: https://machinelistening.exposed/works/after-words/ #### Dark Matters — Johann Diedrick (2021) - Type: Artwork & installation · Machines: Voice & listening machine, Dataset & training data - Idea: A voice assistant can only hear the voices its datasets contain; the missing speech is a dark matter that shapes who is understood. - What it is: An interactive web experience and video installation about the absence of Black speech in the datasets used to train voice interfaces such as Alexa and Siri, told through 3D space, sound and storytelling. - How it works: Browser-based 3D environment built from speech-dataset analysis and recorded voices, shown publicly as a beta web piece and a 24/7 window-gallery video. - Lens: Machine listening & voice, Classification, bias & surveillance - Shown at: Squeaky Wheel Film & Media Art Center, Buffalo (2021); Mozilla Creative Media Awards 2021 - Video: https://www.youtube.com/watch?v=_1XDpu2gabc - Images: https://squeaky.org/wp-content/uploads/2021/05/IMG_4224-scaled.jpg - Project page: https://squeaky.org/event/johann-diedrick-dark-matters/ #### Machine Listening, a Curriculum — Machine Listening (Sean Dockray, James Parker, Joel Stern) (2020) - Type: Book & essay · Machines: Voice & listening machine, Surveillance & biometrics, Dataset & training data - Idea: Our devices now analyse and respond to sound, not just record it; machine listening is a field of knowledge-power that needs collective study and counter-strategies. - What it is: An open, evolving curriculum of reading lists, a library, prompts, interviews, commissioned texts, music and recorded online events on the politics and aesthetics of listening machines, launched with broadcasts for Liquid Architecture and Unsound in 2020. - How it works: Open-source pedagogy platform (built on Pirate Care's curriculum tool) combining syllabi, archives and Zoom-era radio broadcasts. - Lens: Machine listening & voice, Classification, bias & surveillance, Hidden labour & extraction - Shown at: Unsound Festival 2020 (online); Liquid Architecture 2020 - Paper: http://archive.machinelistening.exposed/curriculum/ (Liquid Architecture & Unsound, 2020) - Video: https://www.youtube.com/watch?v=iUbglqQLdrI - Images: https://machinelistening.exposed/images/home/hero.png - Project page: https://machinelistening.exposed/curriculum/ #### Racial Disparities in Automated Speech Recognition — Allison Koenecke (2020) - Type: Paper · Machines: Voice & listening machine - Idea: Listening machines hear some voices worse than others; the gap comes from what their acoustic models were trained on, not from how people speak. - What it is: An audit of five commercial speech-recognition systems (Amazon, Apple, Google, IBM, Microsoft) on interviews with Black and white Americans, finding an average word error rate of 0.35 for Black speakers against 0.19 for white speakers. - How it works: Transcribed matched audio snippets from the CORAAL and VOC corpora with each vendor's ASR API and compared word error rates, isolating acoustic-model effects. - Lens: Machine listening & voice, Classification, bias & surveillance - Paper: https://doi.org/10.1073/pnas.1915768117 (PNAS 117(14), 2020) - Video: https://www.youtube.com/watch?v=1fQ7q4_x00I - Project page: https://www.pnas.org/doi/10.1073/pnas.1915768117 #### On the Apparently Meaningless Texture of Noise — Pedro Oliveira (2019) - Type: Artwork & installation · Machines: Voice & listening machine, Dataset & training data, Surveillance & biometrics - Idea: What a listening machine is trained to discard as noise is exactly where a migrant voice keeps its history. - What it is: A binaural sound essay on the 'dialect recognition software' used in German asylum procedures: it listens to the breaths, pauses and other paraverbal 'noise' that the software's training data tells it to ignore, and builds a counter-archive of voices and leaked training documents. - How it works: Sound essay recorded at EMS Stockholm (2019), composed from annotated paraverbal features, voices and documents from German Federal Office for Migration and Refugees (BAMF) field offices; published online for the Akademie Schloss Solitude and ZKM web residencies. - Lens: Machine listening & voice, Classification, bias & surveillance - Shown at: HASH Award 2020, 2nd prize (ZKM / Akademie Schloss Solitude) - Images: https://oliveira.work/wp-content/uploads/2020/01/schulung-mod-1.jpg https://oliveira.work/wp-content/uploads/2020/01/echoes-m.jpg - Project page: https://oliveira.work/meaningless-texture/ #### Project Alias — Bjørn Karmann, Tore Knudsen (2018) - Type: Speculative design · Machines: Voice & listening machine - Idea: You can negotiate with an always-listening machine by giving it a machine of your own that controls what it hears. - What it is: A fungus-like 'parasite' that sits on top of a Google Home or Amazon Echo and keeps it deaf with a constant murmur; only when it hears the user's own chosen wake word does it go quiet and whisper the real one. - How it works: A Raspberry Pi with microphones and two small speakers runs a locally trained wake-word classifier; it feeds noise into the assistant's microphones until triggered. - Lens: Machine listening & voice, Agency, autonomy & co-creation - Shown at: S+T+ARTS Prize 2019, Grand Prize – Artistic Exploration; Ars Electronica Festival 2019 - Video: https://www.youtube.com/watch?v=qZFytvJ6ano - Images: https://bjoernkarmann.dk/wp-content/uploads/2023/03/Alias-2.jpg https://bjoernkarmann.dk/wp-content/uploads/2023/03/Alias-3.jpg - Project page: https://bjoernkarmann.dk/project_alias - Code: https://github.com/bjoernkarmann/project_alias #### The Dark Age of Connectionism — Wesley Goatley (2017) - Type: Artwork & installation · Machines: Voice & listening machine - Idea: An always-listening device reshapes behaviour: to escape its ear you must change how you exist in the room. - What it is: An Amazon Echo surrounded by microphones: every sound visitors make triggers Siri to ask the Echo a question about its politics and function, and to hear the answers visitors must learn to stay quiet around it. - How it works: Surrounding microphones detect nearby sound events and trigger scripted questions spoken by an Apple Siri voice to an Amazon Echo, which interrupts itself with each new sound. - Lens: Machine listening & voice, Opacity & latent space - Shown at: Haunted Random Forests Festival, UK (2017); IMPAKT Festival, Utrecht (2018); Sonic Arts Seoul (2018) - Video: https://vimeo.com/241915190 - Images: https://wesleygoatley.com/images/da-detail.webp - Project page: https://wesleygoatley.com/dark-age.html #### Rubber Coated Steel — Lawrence Abu Hamdan (2016) - Type: Film & video · Machines: Voice & listening machine - Idea: Machine audio analysis can hear what human witnesses cannot, but the film asks whose silences a court is still unable to hear. - What it is: A 21-minute video staged as a tribunal for gunshots: it re-enacts the acoustic evidence from the 2014 killing of two Palestinian teenagers, where Abu Hamdan's spectral analysis showed that live rounds had been disguised to sound like rubber bullets. - How it works: Audio-ballistic analysis made with Forensic Architecture for Defence for Children International, visualising sound frequencies of recorded gunshots, re-staged as an empty courtroom with subtitles and silences. - Lens: Machine listening & voice, Opacity & latent space - Shown at: Earshot, Portikus, Frankfurt (2016); International Film Festival Rotterdam (IFFR) - Video: https://www.youtube.com/watch?v=NsZHAliJihk - Images: https://images.squarespace-cdn.com/content/v1/5209b9c5e4b0bc18c96df1ec/1633429204074-1Q6FTZP2L0W47AWT341C/earshot+2.jpg https://images.squarespace-cdn.com/content/v1/5209b9c5e4b0bc18c96df1ec/1633525130412-322L2ZAC9UXE628Q3X6I/Lawrence-Abu-Hamdan_31-b.jpg - Project page: http://lawrenceabuhamdan.com/earshot #### Conversnitch — Brian House, Kyle McDonald (2014) - Type: Artwork & installation · Machines: Voice & listening machine, Surveillance & biometrics - Idea: The 'listening machine' is partly people: automated eavesdropping runs on hidden human transcribers and ends in public broadcast. - What it is: A small device disguised as a light bulb that eavesdrops on conversations nearby; snippets are transcribed by anonymous Mechanical Turk workers and posted to Twitter. - How it works: A microphone and Raspberry Pi hidden in a lamp-socket housing stream audio to Amazon Mechanical Turk for transcription; the text is tweeted automatically. - Lens: Machine listening & voice, Hidden labour & extraction, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=ueeQ7T0thwc - Images: https://brianhouse.net/works/conversnitch/conversnitch.jpg https://brianhouse.net/works/conversnitch/conversnitch_1.jpg - Project page: https://brianhouse.net/works/conversnitch/ #### Conflicted Phonemes — Lawrence Abu Hamdan (2012) - Type: Artwork & installation · Machines: Voice & listening machine, Surveillance & biometrics - Idea: When a state listens to accents to decide nationality, the voice is treated like a passport; mapping a voice's mixed history refuses that reduction. - What it is: Wall-sized diagrams, made with designer Janna Ulrich, map the voices of twelve Somali asylum seekers whose claims the Dutch authorities rejected after a language-and-accent analysis. The maps trace lives of migration and camps that a single 'correct' accent cannot hold. - How it works: Research meeting in Utrecht (2012) with the rejected applicants, linguists and activists, turned into vinyl wall timelines and take-away A4 sheets; the maps were later submitted to a Dutch immigration judge and a UK asylum court. - Lens: Machine listening & voice, Classification, bias & surveillance - Shown at: Casco Office for Art, Design and Theory, Utrecht (commission, 2012) - Images: https://images.squarespace-cdn.com/content/v1/5209b9c5e4b0bc18c96df1ec/1487689651283-1IJLGTCD230AZQVHJSYS/LH001-ConfictedPhonemes-det-web.jpg https://images.squarespace-cdn.com/content/v1/5209b9c5e4b0bc18c96df1ec/1487689643617-IWQ7CJM52SRIJXJL6XDR/_MG_9667.JPG - Project page: http://lawrenceabuhamdan.com/conflicted-phonemes #### The Freedom of Speech Itself — Lawrence Abu Hamdan (2012) - Type: Artwork & installation · Machines: Voice & listening machine, Surveillance & biometrics - Idea: Voice analysis turns the way you speak into a territory; the state's ear, not your words, decides where you belong. - What it is: A 34-minute audio documentary on the UK's use of forensic speech analysis and 'voiceprints' to test the origin of asylum seekers' accents, heard among acoustic-foam sculptures shaped like the voiceprint of the word 'you'. - How it works: Interviews with lawyers, phoneticians, asylum seekers and Home Office officials edited with radiophonic techniques; foam sculptures are contour maps of the frequency and amplitude of two voices. - Lens: Machine listening & voice, Classification, bias & surveillance - Shown at: The Showroom, London (commission, 2012) - Images: https://images.squarespace-cdn.com/content/v1/5209b9c5e4b0bc18c96df1ec/1487690498833-HX5J9ER5A4EN8SHONRBC/foam%2Bvoice%2Bprint.jpg https://images.squarespace-cdn.com/content/v1/5209b9c5e4b0bc18c96df1ec/1487690496112-HWUTKCQALIUYA2L3TI4D/Extra%2Bcity%2Bfoam%2B2%2Bcopy.jpg - Project page: http://lawrenceabuhamdan.com/the-freedom-of-speech-itself #### The Whole Truth — Lawrence Abu Hamdan (2012) - Type: Artwork & installation · Machines: Voice & listening machine, Surveillance & biometrics - Idea: Voice-stress analysis claims to hear beneath speech into the body; turning it on its own makers shows how arbitrary its verdicts are. - What it is: An installation and audio documentary about voice-analysis lie detectors piloted by European, Russian and Israeli governments and used by border agencies and insurers. While visitors listen to interviews with the software's developers and sellers, a reverse-engineered lie detector judges those same voices in real time. - How it works: Reverse-engineered voice lie-detection software runs on the documentary's audio; three-channel video, stereo sound, a desk, chairs and artificial plants stage an interview room. - Lens: Machine listening & voice, Classification, bias & surveillance, Failure, glitch & error - Shown at: Casco Office for Art, Design and Theory, Utrecht (commission, 2012) - Images: https://images.squarespace-cdn.com/content/v1/5209b9c5e4b0bc18c96df1ec/1568880234051-G7XRKWGT7ZVT1IMO6KIW/LA_EThe+Whole+Truth_2012_Reverse+engineered+lie+detector%2C+3+channel+video%2C+stereo+sound%2C+desk%2C+3+chairs%2C+3+artificial+plants_24+minutes_612x435x190cm_1.jpg https://images.squarespace-cdn.com/content/v1/5209b9c5e4b0bc18c96df1ec/1568880459670-25NI8IPE3QWLMOWXG69F/LA_EThe+Whole+Truth_2012_Reverse+engineered+lie+detector%2C+3+channel+video%2C+stereo+sound%2C+desk%2C+3+chairs%2C+3+artificial+plants_24+minutes_612x435x190cm_5.jpg - Project page: http://lawrenceabuhamdan.com/the-whole-truth #### Blendie — Kelly Dobson (2004) - Type: Artwork & installation · Machines: Voice & listening machine - Idea: Instead of the machine learning our language, we must learn its voice; control becomes mimicry and a strange empathy with an appliance. - What it is: A 1950s Osterizer blender that only responds to a person's voice: you have to growl, whirr and hum like a blender, and it spins at the speed and power of your imitation. - How it works: Custom hardware and software for real-time sound analysis map the pitch and loudness of the user's blender-like sounds onto the motor's speed. - Lens: Machine listening & voice, Alterity & the uncanny, Intimacy, care & companionship - Shown at: HOME\SICK, Science Gallery Dublin - Video: https://www.youtube.com/watch?v=tph40hgcLpU - Images: http://www.kellydobson.com/blendie.png - Project page: https://www.media.mit.edu/projects/blendie/overview/ ### Datasets & Labels Training sets, labels and the worldviews built into them. #### Tinder_gun_boys_@Brussels_ — Loïs Soleil (2022) - Type: Artwork & installation · Machines: Computer vision, Platform & recommender algorithm - Idea: Turning a dating app's sorting logic back on its users shows how easily a machine can build a dataset of people, and what that dataset is for. - What it is: A robot swipes through Tinder on its own while a weapon-detecting AI screenshots men posing with guns; the images are printed on tatami mats used for feminist self-defence classes held in the installation. - How it works: A physical swiping robot operates a phone running Tinder; an object-detection model flags profile photos with weapons and saves hundreds of screenshots a day; legal advice was compiled into the video How not to get sued as an artist. - Lens: Classification, bias & surveillance, Agency, autonomy & co-creation - Shown at: Prix Ars Electronica 2025, Honorary Mention (Artificial Life & Intelligence) - Video: https://vimeo.com/1084594874 - Images: http://loissoleil.com/pages/tinder/Robot.jpg https://archive.aec.at/asset/3845109/preview/ - Project page: http://loissoleil.com/tinder #### Sorting Song — Simone C. Niquille (2021) - Type: Film & video · Machines: Dataset & training data, Computer vision - Idea: Object categories that seem trivial to people are negotiations the machine cannot have; the boundary between bowl and vase is a political choice. - What it is: Domestic objects from the SceneNet RGB-D training dataset morph endlessly between vase and bowl, bench and couch, chair and toilet. When shape is all a dataset records, a wheelchair, a toilet and an electric chair can end up in the same category. - How it works: 3D meshes from Imperial College London's SceneNet RGB-D indoor dataset are animated morphing between category labels. - Lens: Classification, bias & surveillance, Failure, glitch & error - Video: https://vimeo.com/555060893 - Images: https://technofle.sh/static/media/SOS_SortingSong_001.jpg https://technofle.sh/static/media/SOS_CowSharkFalling.png - Project page: https://technofle.sh/sorting-song.html #### Lacework — Everest Pipkin (2020) - Type: Artwork & installation · Machines: Dataset & training data, Computer vision - Idea: A training set is made of other people's lives, reduced to one word each; watching it all restores their weight. - What it is: A slow, dreamlike video river made from MIT's Moments in Time dataset: one million three-second clips scraped from the web and each tagged with a single verb such as 'praying' or 'flying'. Pipkin watched the whole dataset and stretched its clips into hallucinatory slow motion. - How it works: Clips from the Moments in Time action-recognition dataset were processed with neural frame interpolation and super-resolution to slow and enlarge them into a continuous video. - Lens: Classification, bias & surveillance, Hidden labour & extraction, Machine seeing - Shown at: The Photographers' Gallery, London (Data / Set / Match), 2020 - Video: https://everest-pipkin.com/projects/lacework/flickr-5-6-5-8-9-7-1-8-4756589718_2.slowed_x2.00_x2.00_x2.00_x2.00_x1.00-20.webformat.mp4 - Images: https://thephotographersgallery.org.uk/sites/default/files/styles/banner_image/public/ep2.png.webp?itok=wVNy-Lyc - Project page: https://everest-pipkin.com/projects/lacework.html #### Large Image Datasets: A Pyrrhic Win for Computer Vision? — Abeba Birhane (2020) - Type: Paper · Machines: Dataset & training data, Computer vision - Idea: What a vision model can see depends on images collected without consent and labels no one checked. - What it is: Abeba Birhane and Vinay Uday Prabhu audited large image datasets, including ImageNet and MIT's 80 Million Tiny Images, and found slurs among the labels and non-consensual, pornographic and harmful images. The paper led MIT to withdraw Tiny Images. - How it works: A quantitative and qualitative audit of labels and image content in large-scale vision datasets, with proposed remedies. - Lens: Classification, bias & surveillance - Paper: https://doi.org/10.1109/WACV48630.2021.00158 (WACV 2021) - Project page: https://arxiv.org/abs/2006.16923 #### Model Zoo — Forensic Architecture (2020) - Type: Artwork & installation · Machines: Computer vision, Dataset & training data - Idea: Where real images of violence are scarce, synthetic training sets let machines learn to see what investigators need to find. - What it is: A growing catalogue of 3D models of munitions and weapons, together with the machine-learning classifiers trained on synthetic renderings of them to recognise these objects in images from conflicts. - How it works: Photorealistic renderings of 3D weapon models form synthetic datasets; classifiers trained on them are tested on open-source footage. - Lens: Machine seeing, Classification, bias & surveillance - Images: https://i.vimeocdn.com/video/864290475-35c09a4a9d98e29cb953c40599ec67ec2e384f73e62ec820ed4e0e80c17c162a-d_1280x720.jpg - Project page: https://forensic-architecture.org/investigation/model-zoo #### Pique Nique pour les Inconnues — Rosa Menkman (2020) - Type: Film & video · Machines: Dataset & training data, Computer vision - Idea: The standards by which machines see images were calibrated on women whose faces survive endless copying while their names disappear. - What it is: The anonymous women on colour test cards (Lena, Shirley and others), along with bots and stock-photo placeholders, gather on a desktop and try to sing together using only computer system sounds. - How it works: Desktop video 'tele-chorale' animating historical test images and virtual assistants, voiced with operating-system sounds performing Paul McCartney's 'We All Stand Together'. - Lens: Classification, bias & surveillance, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=6n7t3A3UF3o - Images: https://freight.cargo.site/i/2bf5877371a4095252b3a01af61a332deb9cc1a034752cce60fd6b0afa06d8c3/iRD.png - Project page: https://beyondresolution.info/COLOR-TEST-CARDS #### Salaf (Ancestors) — Nouf Aljowaysir (2020) - Type: Artwork & installation · Machines: Computer vision, Dataset & training data - Idea: Machine vision inherits the colonial archive: to refuse its gaze, the artist trains on absence. - What it is: Tracing her family's migrations across Arabia and Mesopotamia, the artist ran archival photographs, many from Gertrude Bell's colonial archive, through computer-vision models. They failed to see veiled women as women and tagged Bedouin as 'soldier' and 'army'; she then erased the figures to make an 'absent dataset' of about 6,000 images. - How it works: Scraped museum and archival photographs were labelled by commercial computer-vision models; an AI segmentation technique then removed the orientalist figures to create an empty-image dataset. - Lens: Classification, bias & surveillance, Failure, glitch & error, Grief, memory & afterlife - Shown at: ThoughtWorks Arts residency, New York, 2020 - Images: https://images.squarespace-cdn.com/content/v1/55039003e4b04b9963242639/1763375737992-MHJQB7XNFGLGYDDDY41V/186--Woman%2Bseated%2Bcross-legged%2Bwith%2Bshoes%2Boff-209.jpg https://images.squarespace-cdn.com/content/v1/55039003e4b04b9963242639/1763375946970-N2HOYYFEJ9ZTJ56QK06N/188--Paysannes%2Begyptiense%2Ballant%2Ba%2Bl%25E2%2580%2599eua-1133.jpg - Project page: https://www.noufaljowaysir.com/work/salaf #### Declassifier (Humans of AI) — Philipp Schmitt (2019) - Type: Artwork & installation · Machines: Computer vision, Dataset & training data - Idea: Every act of machine recognition is a recollection of someone else's photographs. - What it is: A computer-vision program looks at photographs, but instead of drawing boxes around the objects it detects, it overlays the actual COCO training images that taught it to see each object. Part of Humans of AI, which also credits the thousands of Flickr photographers whose pictures built the dataset. - How it works: An object detector trained on Microsoft's COCO dataset processes images; for each detection the software retrieves and collages the COCO training photos of that class, with photographer credits. - Lens: Machine seeing, Classification, bias & surveillance, Hidden labour & extraction - Shown at: The Photographers' Gallery, London, 2020 - Video: https://www.youtube.com/watch?v=IOZe5wEcXxY - Images: https://cdn.philippschmitt.com/media/pages/work/humans-of-ai/58fb56c131-1586885726/philipp-schmitt_humans-of-ai-1-1200x630-crop.jpg - Project page: https://philippschmitt.com/work/humans-of-ai #### Excavating AI: The Politics of Images in Machine Learning Training Sets — Kate Crawford, Trevor Paglen (2019) - Type: Book & essay · Machines: Dataset & training data, Computer vision - Idea: Every training set encodes a politics: deciding what an image means is an act of power, never neutral description. - What it is: An essay that digs through ImageNet, JAFFE, UTKFace and other training sets layer by layer, from their taxonomies and categories down to individual labelled images. - How it works: Archaeological reading of dataset structure (WordNet hierarchy, label choices, sourcing) with image examples; published online by the authors with the AI Now Institute. - Lens: Classification, bias & surveillance, Opacity & latent space - Paper: https://excavating.ai (excavating.ai (AI Now Institute), 2019) - Video: https://www.youtube.com/watch?v=ZiHHsKALwBc - Images: https://paglen.studio/wp-content/uploads/2020/06/Paglen-Excavating_AI-1024x385.jpg - Project page: https://excavating.ai #### From ‘Apple’ to ‘Anomaly’ (Pictures and Labels) — Trevor Paglen (2019) - Type: Artwork & installation · Machines: Dataset & training data, Computer vision - Idea: A training set is a worldview: walking along its labels shows how quickly naming things slides into judging people. - What it is: About 30,000 printed photographs from the ImageNet dataset line the Barbican's curved gallery, grouped under their labels, moving from harmless nouns like 'apple' to judgments about people. - How it works: Selection of ImageNet categories and images printed individually and pinned in label clusters along the 90-metre Curve wall; a homage to Magritte's The Treachery of Images. - Lens: Classification, bias & surveillance, Machine seeing - Shown at: Barbican Centre, The Curve, London, 2019–20 - Video: https://www.youtube.com/watch?v=juSXaQWNXgU - Images: https://paglen.studio/wp-content/uploads/2020/06/Paglen-From_Apple_to_Anomaly-2019-03-1024x683.jpg https://paglen.studio/wp-content/uploads/2020/05/Paglen-From_Apple_to_Anomaly_2019-01-1024x683.jpg - Project page: https://paglen.studio/2020/04/09/from-apple-to-anomaly-pictures-and-labels-selections-from-the-imagenet-dataset-for-object-recognition/ #### HOMESCHOOL — Simone C. Niquille (2019) - Type: Film & video · Machines: Computer vision, Dataset & training data, Robot creature & mobile robot - Idea: Before a robot can live in your home, the home must be rebuilt as a dataset, and whatever does not fit a category disappears. - What it is: A 13-minute 3D-animated film set inside the synthetic model homes used to train the computer vision of household robots. It asks how thousands of homes, people's most private spaces, are turned into categories a robot can navigate: is this a vase, a bowl or a cup? - How it works: 3D animation built from assets and floor plans of synthetic indoor training datasets for domestic robot vision, with a four-channel soundtrack. - Lens: Machine seeing, Classification, bias & surveillance, Machine umwelt - Shown at: Commissioned by Housing the Human, with Swissnex San Francisco and Fotomuseum Winterthur, 2019 - Video: https://vimeo.com/362563329 - Images: https://technofle.sh/static/media/SCN_Homeschool_Still02_SimoneNiquille-Technoflesh.jpg https://technofle.sh/static/media/HS_Homeschool_depth_SimoneNiquille-Technoflesh.png - Project page: https://technofle.sh/homeschool.html #### ImageNet Roulette — Trevor Paglen, Kate Crawford (2019) - Type: Artwork & installation · Machines: Dataset & training data, Computer vision, Surveillance & biometrics - Idea: Letting the training set 'speak for itself' showed that classifying people by appearance is harmful by design, not by accident. - What it is: A web app and installation that labelled anyone's selfie using a model trained on ImageNet's 'person' categories, returning terms that were often absurd, racist or cruel; millions of people tried it in September 2019. - How it works: A classifier trained on ImageNet's roughly 2,800 person sub-categories detected faces and returned the nearest label; the online version was withdrawn and continues as the installation Faces of ImageNet (2022). - Lens: Classification, bias & surveillance, Failure, glitch & error - Shown at: Training Humans, Fondazione Prada, Milan, 2019; Poetics of Encryption, Kunsthal Charlottenborg, 2024 (as Faces of ImageNet); The World Through AI, Jeu de Paume, Paris, 2025 (as Faces of ImageNet) - Video: https://www.youtube.com/watch?v=K_lN4JD1sBY - Images: https://paglen.studio/wp-content/uploads/2020/06/Crawford_Paglen.jpg https://paglen.studio/wp-content/uploads/2020/06/failure.jpg - Project page: https://paglen.studio/2020/04/29/imagenet-roulette/ #### MegaPixels / Exposing.ai — Adam Harvey, Jules LaPlace (2019) - Type: Artwork & installation · Machines: Dataset & training data, Surveillance & biometrics - Idea: Face recognition is built from ordinary photos taken for other reasons; the dataset is where the surveillance begins. - What it is: A research project and search engine that traces where face and biometric training datasets came from and lets anyone check whether their Flickr photos were used in them. - How it works: Photo IDs and metadata from dozens of public face datasets scraped from Flickr (2004–2020) are indexed so users can search them; MegaPixels ran 2017–2020 and was relaunched as Exposing.ai in 2021. - Lens: Classification, bias & surveillance, Hidden labour & extraction - Shown at: Ars Electronica Center, Understanding AI 2019; Ludwig Forum Aachen, Training the Archive 2020 - Video: https://www.youtube.com/watch?v=ey9lPMknV2E - Images: https://adam.harvey.studio/exposing-ai/exposing-ai.webp https://exposing.ai/assets/img/og-share.png - Project page: https://exposing.ai #### Training Humans — Kate Crawford, Trevor Paglen (2019) - Type: Artwork & installation · Machines: Dataset & training data, Surveillance & biometrics - Idea: Training images are a vernacular photography nobody looks at, yet they decide how machines will see everyone. - What it is: An exhibition of the photographs used to train computer vision to recognise people, from early government-funded face databases such as FERET to images scraped from the web and sets built for emotion recognition. - How it works: Curated display of historical and current training datasets (faces, gestures, emotions, gait) as prints and wall grids, with ImageNet Roulette as an interactive piece. - Lens: Classification, bias & surveillance, Machine seeing - Shown at: Osservatorio Fondazione Prada, Milan, 2019–20 - Video: https://www.youtube.com/watch?v=P4JpD1PWBDI - Images: https://paglen.studio/wp-content/uploads/2020/05/PaglenCrawford-Training_Humans-2019_01-1024x684.jpg https://paglen.studio/wp-content/uploads/2020/05/PaglenCrawford-Training_Humans-2019_02-1024x683.jpg - Project page: https://paglen.studio/2020/04/09/kate-crawford-trevor-paglen-training-humans/ #### Triple-Chaser — Forensic Architecture (2019) - Type: Film & video · Machines: Computer vision, Dataset & training data - Idea: Computer vision can be turned against power: a machine taught with synthetic images becomes a witness to state violence. - What it is: Made for the 2019 Whitney Biennial, the film traces Triple-Chaser tear-gas grenades fired at civilians back to Safariland, a company owned by Whitney vice-chair Warren B. Kanders. Because few real photographs existed, the team trained classifiers on photorealistic synthetic images to search for the grenades online. - How it works: 3D models of the grenade were rendered into thousands of synthetic training images with varied lighting and backgrounds, used to train machine-learning classifiers that scan online imagery; film made with Praxis Films. - Lens: Machine seeing, Classification, bias & surveillance - Shown at: Whitney Biennial, New York, 2019 - Video: https://www.youtube.com/watch?v=93rjwQMww9M - Images: https://content.forensic-architecture.org/wp-content/uploads/2019/05/Photorealistic_4-mask-1-1600x900.jpg https://i.vimeocdn.com/video/782522102-fef4ec627744abb9949fe05683df148918c13b8c94847f95ab0c848f8c016c7e-d_1280x720.jpg - Project page: https://forensic-architecture.org/investigation/triple-chaser #### Zizi – Queering the Dataset — Jake Elwes (2019) - Type: Artwork & installation · Machines: Generative model & latent space, Dataset & training data, Surveillance & biometrics - Idea: Adding the faces a dataset left out shifts what the machine can imagine, and shows what it had not learned. - What it is: A multi-channel video of faces in constant transition, generated after a face model trained on 70,000 photographs was retrained with 1,000 images of drag and gender-fluid faces. The network's weights drift away from the normative faces it learned first. - How it works: StyleGAN trained on the FFHQ face dataset was fine-tuned with about 1,000 drag and gender-fluid faces found online; the latent space is traversed to make a 135-minute loop. - Lens: Classification, bias & surveillance, Opacity & latent space, Alterity & the uncanny - Shown at: Edinburgh Fringe (Inspace), 2019; You and AI, Onassis Foundation, Athens, 2021; Among the Machines, Zabludowicz Collection, London, 2022 - Video: https://vimeo.com/388245510 - Images: https://www.jakeelwes.com/images/project-zizi19.jpg https://www.jakeelwes.com/images/project-zizi19/Zabludowicz.jpg - Project page: https://www.jakeelwes.com/project-zizi-2019.html #### Myriad (Tulips) — Anna Ridler (2018) - Type: Artwork & installation · Machines: Dataset & training data, Generative model & latent space - Idea: Showing the dataset instead of the model reveals the slow human labour and judgement on which machine learning rests. - What it is: An installation of thousands of photographs of tulips that Ridler bought, photographed and labelled by hand over one tulip season; the images became the training set for her GAN work Mosaic Virus. Each print carries her handwritten label. - How it works: About 10,000 tulip photographs taken and hand-annotated by the artist, displayed as prints with handwritten labels; the set was later used to train a GAN. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Shown at: AI: More than Human, Barbican, London, 2019 - Video: https://vimeo.com/428168704 - Images: https://annaridler.com/content/works/myriad-tulips/block_1/DSC01064.webp https://annaridler.com/content/works/myriad-tulips/block_3/group_3/0IMG_1672.jpeg - Project page: https://annaridler.com/myriad-tulips #### This is the Problem, the Solution, the Past and the Future — Sebastian Schmieg (2017) - Type: Artwork & installation · Machines: Dataset & training data, Computer vision - Idea: Any category can be taught to a machine once enough people click; the absurd labels expose how arbitrary all labels are. - What it is: A conceptual training dataset of 2,931 photographs from The Photographers' Gallery website, sorted by almost 30,000 visitor clicks into four categories: Problem, Solution, Past and Future. The clicks were collected in Schmieg's intervention Decision Space (2016), which overlaid the gallery's site with these four buttons. - How it works: A web intervention collected crowd classifications of gallery images; the resulting labelled dataset was published online and as wallpaper, and used to train an image classifier (Decisive Camera). - Lens: Classification, bias & surveillance, Hidden labour & extraction - Shown at: Commissioned by The Photographers' Gallery, London, 2016–17 - Images: https://sebastianschmieg.com/wordpress/wp-content/uploads/2019/02/titptstpatf-dc-q21-1-1024x1024.jpg https://sebastianschmieg.com/wordpress/wp-content/uploads/2019/02/decision-space-1-1024x683.png - Project page: https://sebastianschmieg.com/this-is-the-problem-the-solution-the-past-and-the-future/ #### Animal Classifier — Shinseungback Kimyonghun (2016) - Type: Artwork & installation · Machines: Dataset & training data, Computer vision - Idea: Every machine classification rests on an arbitrary taxonomy; this one simply makes the arbitrariness visible. - What it is: A neural network is trained to sort animals into the 14 absurd categories of the 'Chinese encyclopedia' in Borges's essay, such as 'belonging to the emperor', 'fabulous' and 'that from a long way off look like flies'. - How it works: TensorFlow convolutional neural network trained on online images, running on 14 Raspberry Pis under glass domes with monitors. - Lens: Classification, bias & surveillance, Machine seeing - Video: https://vimeo.com/198948703 - Images: https://djhznh41oxwef.cloudfront.net/works/animal_classifier/ssbkyh_animal_classifier_01.png - Project page: https://ssbkyh.com/works/animal_classifier/ #### Segmentation.Network — Sebastian Schmieg (2016) - Type: Artwork & installation · Machines: Dataset & training data, Computer vision - Idea: Machine vision is drawn by hand: someone decided where each object ends and what stays unrecognised. - What it is: A website that plays back more than 600,000 outlines drawn by crowd-workers around objects in Flickr photos for Microsoft's COCO dataset. Line by line, it shows the hand work that decides what machines can see. - How it works: Custom software animates the polygon segmentation annotations of the COCO dataset in sequence; also shown as a projection (2018). - Lens: Hidden labour & extraction, Machine seeing - Images: https://sebastianschmieg.com/wordpress/wp-content/uploads/2019/01/segmentation-network-ani-01.gif https://sebastianschmieg.com/wordpress/wp-content/uploads/2019/01/segmentation-network-01-1024x683.png - Project page: https://sebastianschmieg.com/segmentation-network/ #### ImageNet: A Large-Scale Hierarchical Image Database — Jia Deng (2009) - Type: Paper · Machines: Dataset & training data, Computer vision - Idea: A machine's way of seeing is inherited from a dictionary's hierarchy of nouns and from thousands of anonymous clicks. - What it is: The CVPR 2009 paper by Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li and Li Fei-Fei introducing ImageNet: millions of web images sorted into the noun hierarchy of WordNet and verified by crowd-workers on Amazon Mechanical Turk. It became the benchmark that launched deep-learning computer vision. - How it works: Images were collected by querying search engines with WordNet synonym sets, then cleaned and verified by Amazon Mechanical Turk workers. - Lens: Classification, bias & surveillance, Hidden labour & extraction, Machine seeing - Paper: https://doi.org/10.1109/CVPR.2009.5206848 (CVPR 2009) - Project page: https://www.image-net.org/ ### Faces, Emotion & Biometrics Face recognition, emotion detection and bodies turned into data. #### The Flemish Scrollers — Dries Depoorter (2021) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Automated face recognition, usually aimed at citizens, is turned on the people who make the rules for it. - What it is: Software watches the Flemish Parliament's daily livestreams, detects when a politician is looking at their phone, identifies them by face and posts the clip on social media, tagging them by name. - How it works: Keras object-detection and face-recognition models analyse the parliament's YouTube livestreams; FFmpeg cuts clips that are posted automatically to Twitter and Instagram. Running since 5 July 2021. - Lens: Classification, bias & surveillance, Machine seeing - Video: https://www.youtube.com/watch?v=5YHfK6DBvGE - Images: https://driesdepoorter.be/wp-content/uploads/2022/03/TheFlemishScollers-photo-by-Dries-Depoorter-01-1612x845.jpg https://driesdepoorter.be/wp-content/uploads/2021/07/The-Flemish-Scrollers-Software-Dries-Depoorter-1024x522.png - Project page: https://driesdepoorter.be/theflemishscrollers/ #### Capture — Paolo Cirio (2020) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision, Dataset & training data - Idea: Face recognition feels different when it is pointed at the police: the outrage exposes the asymmetry of who may be identified. - What it is: Cirio ran 1,000 press photos of protests in France through face recognition, built a database of 4,000 police officers' faces and invited the public to identify them by name, and pasted the headshots as posters across Paris. France's interior minister demanded the work be censored. - How it works: Public photographs were processed with face-detection and recognition software to extract and cluster officers' faces, published on the platform Capture-Police.com and printed as street posters; linked to a campaign to ban facial recognition in Europe. - Lens: Classification, bias & surveillance, Machine seeing - Shown at: Le Fresnoy, Tourcoing 2020 (co-production) - Video: https://www.youtube.com/watch?v=JPybxTM-daE - Images: https://paolocirio.net/work/capture/slideshow/NIK_2255.jpg - Project page: https://paolocirio.net/work/capture/ #### Facework — Kyle McDonald (2020) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision, Platform & recommender algorithm - Idea: Playing to a face classifier builds an intuition for seeing like a machine, and for how arbitrary its categories are. - What it is: A browser game set in a gig economy where face analysis is the job: as a 'Faceworker', the player takes gigs that require the webcam to judge them as, say, wearing glasses or looking happy, and learns in real time how the classifier sees and where it fails. - How it works: In-browser face-attribute classifiers run on the player's webcam; gig tasks are scored live against their outputs. Supported by the Mozilla Foundation. - Lens: Classification, bias & surveillance, Hidden labour & extraction, Failure, glitch & error - Images: https://facework.app/assets/images/social-card.jpg - Project page: https://facework.app - Code: https://github.com/kylemcdonald/facework #### Vibe Check — Lauren Lee McCarthy, Kyle McDonald (2020) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Emotion recognition turns every encounter into a score, and the gallery becomes a small social-credit system. - What it is: Cameras in the exhibition use face recognition and expression analysis to log what emotions each visitor provokes in others. A screen at the entrance names the people who cause the most happiness, disgust, sadness, surprise or boredom, and visitors may find themselves listed on the way out. - How it works: Off-the-shelf face recognition and expression classifiers track visitors across cameras, attribute the expressions of nearby people to whoever they are looking at, and publish the rankings on screens. - Lens: Classification, bias & surveillance, Machine seeing - Shown at: Real Feelings, HEK Basel 2020 (commissioned by HEK and MU) - Video: https://vimeo.com/488636450 - Images: https://freight.cargo.site/w/1200/i/9d8ca3ebcda018fd8d992785901291c8e601d1a17ded09962faa0a6789609ae2/_DSC4566_1_crop.jpg - Project page: https://get-lauren.net/Vibe-Check #### Emotional Expressions Reconsidered: Challenges to Inferring Emotion From Human Facial Movements — Lisa Feldman Barrett (2019) - Type: Paper · Machines: Surveillance & biometrics, Computer vision - Idea: Emotion-recognition AI detects facial movements, not feelings; the inference from one to the other is the weak link. - What it is: Barrett, Adolphs, Marsella, Martinez and Pollak review more than 1,000 studies and conclude that people do not reliably make the same facial movements for the same emotions across situations and cultures, so a scowl cannot be read as anger. - How it works: A systematic review of evidence on the reliability and specificity of facial configurations for anger, disgust, fear, happiness, sadness and surprise in adults, infants, children and remote cultures. - Lens: Classification, bias & surveillance, Failure, glitch & error - Paper: https://doi.org/10.1177/1529100619832930 (Psychological Science in the Public Interest 20(1)) - Project page: https://doi.org/10.1177/1529100619832930 #### How do you see me? — Heather Dewey-Hagborg (2019) - Type: Artwork & installation · Machines: Surveillance & biometrics, Generative model & latent space - Idea: What counts as a face to a machine can be almost unrecognisable to us. - What it is: Dewey-Hagborg uses adversarial image generation to build pictures that a face detector accepts as her face, although to a human eye they look like abstract smudges. - How it works: An image generator is optimised against a face-detection and recognition model until its outputs are classified as the artist. - Lens: Opacity & latent space, Machine seeing - Video: https://vimeo.com/361593768 - Images: https://www.deweyhagborg.com/media/pages/projects/how-do-you-see-me/9b1078df08-1643025907/heatherdeweyhagborg_kateelliott_01.jpg - Project page: https://www.deweyhagborg.com/projects/how-do-you-see-me #### Mind — Shinseungback Kimyonghun (2019) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Emotion AI is turned into weather: a collective 'mind' that exists only as the machine measures it. - What it is: An emotion-recognition system reads the faces of people in the room, and mechanical ocean drums turn the average emotion of the last 100 faces into the sound of waves. The sea changes with the crowd's detected mood. - How it works: Network camera with FER+ emotion recognition (ONNX Runtime), microcontroller-driven mechanical ocean drums and a sea simulation. - Lens: Classification, bias & surveillance, Machine umwelt - Shown at: National Museum of Modern and Contemporary Art, Korea (commission) 2019 - Video: https://vimeo.com/364570360 - Images: https://djhznh41oxwef.cloudfront.net/works/mind/ssbkyh_Mind_01.png - Project page: https://ssbkyh.com/works/mind/ #### Perception iO — Karen Palmer (2019) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: The installation reads the viewer the way a biased system reads a suspect, making them feel whose perception becomes consequence. - What it is: The viewer takes the point of view of a police officer's body camera in a tense encounter with a Black and a white protagonist; eye tracking and facial emotion analysis read the viewer's reactions, and the story branches according to their implicit bias. - How it works: Tobii eye tracking and AI facial-expression analysis (developed with ThoughtWorks Arts) feed a branching narrative film engine. - Lens: Classification, bias & surveillance, Machine seeing - Shown at: Face Values, Cooper Hewitt, Smithsonian Design Museum, New York 2019–20; BIAS: Built This Way, Science Gallery Dublin 2021 - Video: https://www.youtube.com/watch?v=Bfgmo05XdXM - Images: https://i0.wp.com/karenpalmer.uk/wp-content/uploads/2016/03/IMG_7016.jpg?fit=1200%2C750&ssl=1 - Project page: https://karenpalmer.uk/portfolio/perceptionio/ #### RASA-BOX — Iris Long (2019) - Type: Artwork & installation · Machines: Surveillance & biometrics, Chatbot & language model - Idea: Emotion inference turns the body into a text a machine claims to read. - What it is: A 'mirror' that reads visitors' body posture with a camera, infers their emotional state and writes a sentence in the same emotional category, testing how private our feelings remain. - How it works: Pose estimation from camera skeleton data feeds an emotion classifier and a text generator; made with Hanbin Deng and Wu Tiancheng. - Lens: Classification, bias & surveillance, Embodiment & kinesthesia - Shown at: Ars Electronica Festival 2019 - Images: https://ars.electronica.art/outofthebox/files/2019/08/RASABOX_Iris-Long.jpg - Project page: https://ars.electronica.art/outofthebox/en/rasa-box/ #### They Took the Faces from the Accused and the Dead… (SD18) — Trevor Paglen (2019) - Type: Artwork & installation · Machines: Dataset & training data, Surveillance & biometrics - Idea: Face recognition was benchmarked on people who could not consent: the accused, the imprisoned and the dead. - What it is: A grid of thousands of mugshots from NIST Special Database 18, a set of arrest photographs, many of people now dead, supplied to researchers to develop and test face recognition. - How it works: Images drawn from the NIST mugshot database used as a research benchmark, printed as pigment prints and as a large wall installation. - Lens: Classification, bias & surveillance, Grief, memory & afterlife - Shown at: Pace Gallery, Geneva, 2019; Uncanny Valley, de Young Museum, San Francisco, 2020 - Images: https://paglen.studio/wp-content/uploads/2020/06/Paglen-They_took_the_faces-2019_exhibition_view.jpg https://paglen.studio/wp-content/uploads/2020/05/Paglen-They_Took_the_Faces_2019-03.jpg - Project page: https://paglen.studio/2020/04/09/they-took-the-faces-from-the-accused-and-the-dead/ #### Random String of Emotions — Coralie Vogelaar (2018) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Reversing emotion AI's decoding exposes how thin its vocabulary of six basic emotions is. - What it is: An actress performs 200 facial expressions assembled from random combinations of Action Units, the muscle movements that emotion-recognition software uses, and commercial software then tries to name the emotion of each strange, sometimes impossible face. - How it works: A computer generates random strings of FACS Action Units that the actress learns to perform; the filmed fragments are analysed live by Noldus FaceReader and shown on a second screen. - Lens: Classification, bias & surveillance, Failure, glitch & error - Video: https://vimeo.com/331861983 - Images: https://freight.cargo.site/w/1200/i/5a0df0b0e55f08e1ed9a2a99f301542141b7e8051d9ef693c50faf2b9b11c743/RANDOM_16-9-26-43.jpg - Project page: https://www.coralievogelaar.com/Random-String-of-Emotions #### Surveillance Paparazzi — Dries Depoorter (2018) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Paparazzi work can now be done by any open camera plus a classifier; everyone is filmed, only some are recognised. - What it is: A machine that watches unsecured surveillance cameras around the world and uses face recognition to 'snap' celebrities who happen to pass, like an automated paparazzo. - How it works: Scrapes live feeds from publicly accessible, unencrypted CCTV cameras and runs a celebrity face-recognition model on them. - Lens: Machine seeing, Classification, bias & surveillance - Shown at: Entangled Realities, HEK Basel 2019 - Video: https://www.youtube.com/watch?v=mS2tx3G7930 - Project page: https://driesdepoorter.be/ #### The Normalizing Machine — Mushon Zer-Aviv (2018) - Type: Artwork & installation · Machines: Surveillance & biometrics, Dataset & training data, Computer vision - Idea: Algorithmic prejudice is built from our own small judgements, collected and hidden inside a black box. - What it is: Visitors are recorded and then asked to point at the person who looks 'most normal' in a line-up of earlier participants. The machine aggregates these choices into an algorithmic image of normality, echoing Alphonse Bertillon's 19th-century 'speaking portrait'. - How it works: A camera records each participant; their pointing choices label previous faces, and a machine-learning model aggregates the votes into a map of facial 'normality' (software by Dan Stavy and Eran Weissenstern, openFrameworks). - Lens: Classification, bias & surveillance - Shown at: Ars Electronica ERROR exhibition, Berlin, 2018; BIAS: Built This Way, Science Gallery Dublin, 2021 - Video: https://vimeo.com/323711041 - Images: https://mushon.com/tnm/img/tnm_space.jpg https://mushon.com/tnm/img/tnm_cards.jpg - Project page: https://mushon.com/tnm/ #### It Began as a Military Experiment — Trevor Paglen (2017) - Type: Artwork & installation · Machines: Surveillance & biometrics, Dataset & training data - Idea: Face recognition was built on the faces of particular people, the 'Adam and Eves' of a technology they never chose to found. - What it is: Ten portraits chosen from FERET, the DARPA-funded face database of the 1990s built largely from staff at a military base in Maryland, printed with the keypoints a face-recognition algorithm finds. - How it works: Months of curating the FERET database, retouching and colour-correcting selected images, then running them through a facial-landmark algorithm; set of 10 pigment prints. - Lens: Classification, bias & surveillance, Machine seeing - Images: https://paglen.studio/wp-content/uploads/2020/06/Paglen-It_Began-2017.jpg https://paglen.studio/wp-content/uploads/2020/05/Paglen-It_Began_2017-01-784x1024.jpg - Project page: https://paglen.studio/2020/04/09/it-began-as-a-military-experiment/ #### Machine Readable Hito — Trevor Paglen (2017) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Facial analysis turns a known person into columns of confident guesses that contradict each other from one image to the next. - What it is: A wall of hundreds of portraits of the artist Hito Steyerl, each captioned with what facial-analysis algorithms guess about her age, gender, emotion, glasses or beard. - How it works: Hundreds of photographs run through several commercial and open-source facial-analysis classifiers; outputs printed under each image on adhesive wall material. - Lens: Classification, bias & surveillance, Machine seeing - Shown at: Metro Pictures, New York, 2017 - Images: https://paglen.studio/wp-content/uploads/2020/06/Paglen-Machine_readable_Hito-Metro.jpg https://paglen.studio/wp-content/uploads/2020/06/Paglen-Machine_Readable_Hito_detail-2-1-935x1024.jpg - Project page: https://paglen.studio/2020/04/09/machine-readable-hito-and-holly/ #### Physiognomy's New Clothes — Blaise Agüera y Arcas (2017) - Type: Book & essay · Machines: Surveillance & biometrics, Computer vision, Dataset & training data - Idea: A classifier trained on biased photographs learns the bias and returns it as objective 'science'. - What it is: An essay by Blaise Agüera y Arcas, Margaret Mitchell and Alexander Todorov dissecting a 2016 paper that claimed deep learning could detect 'criminality' from ID photos, and tracing the idea back to Lombroso and nineteenth-century physiognomy. - How it works: A long-form critical essay combining machine-learning analysis of the claimed results (dataset differences such as expression and clothing) with the history of physiognomy and scientific racism. - Lens: Classification, bias & surveillance, Machine seeing - Paper: https://medium.com/@blaisea/physiognomys-new-clothes-f2d4b59fdd6a (Medium, May 2017) - Images: https://miro.medium.com/v2/resize:fit:907/1*nhh283c26V1tYWjqH3UjXg.jpeg - Project page: https://medium.com/@blaisea/physiognomys-new-clothes-f2d4b59fdd6a #### Probably Chelsea — Heather Dewey-Hagborg (2017) - Type: Artwork & installation · Machines: Surveillance & biometrics, Dataset & training data - Idea: One genome yields thirty plausible faces: DNA does not determine a face, and identity cannot be read off the body. - What it is: Thirty different 3D-printed faces hang in a crowd, all generated by algorithms from the same DNA: cheek swabs and hair that Chelsea Manning mailed from prison. Made with Manning to mark her release in 2017. - How it works: Manning's DNA was analysed with forensic phenotyping methods; varying the algorithm's interpretations of the same genetic data produced 30 face models, 3D-printed in colour. - Lens: Classification, bias & surveillance, Alterity & the uncanny - Shown at: Fridman Gallery, New York 2017 - Video: https://www.youtube.com/watch?v=A-EhRw11JiQ - Images: https://deweyhagborg.com/media/pages/projects/probably-chelsea/9769243e4c-1643025328/heatherdeweyhagborgchelseamanning_pap5596.jpg - Project page: https://deweyhagborg.com/projects/probably-chelsea #### Emotion Hero — Ruben van de Ven (2016) - Type: Artwork & installation · Machines: Surveillance & biometrics - Idea: Emotion recognition does not read feelings; it trains people to perform its categories. - What it is: A mobile game in the style of Guitar Hero in which players score points by making facial expressions on cue, then receive feedback on how to express emotions better for the machine. - How it works: Smartphone game built on a commercial emotion-recognition SDK (likely Affectiva) that scores expressions against its categories. - Lens: Classification, bias & surveillance, Embodiment & kinesthesia - Images: https://emotionhero.com/emotionhero.png - Project page: https://rubenvandeven.com/emotionhero #### Level of Confidence — Rafael Lozano-Hemmer (2015) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: A surveillance algorithm built to find suspects is turned into a tool of mourning that searches for victims. - What it is: A face-recognition camera compares each visitor's face with the portraits of the 43 students abducted in Iguala, Mexico, in 2014, and shows which student they most resemble with a percentage of confidence. The match is never positive; the search does not stop. - How it works: Eigenface, Fisherface and LBPH face-recognition algorithms trained on the students' photographs run on a camera-and-screen unit; proceeds from sales go to the students' community. - Lens: Classification, bias & surveillance, Grief, memory & afterlife - Shown at: FOFA Gallery, Concordia University, Montreal 2015; Collection MUAC, Mexico City - Video: https://www.youtube.com/watch?v=sOCZLLgc1lQ - Images: https://www.lozano-hemmer.com/image_sets/level_of_confidence/montreal_2018/level_of_confidence_montreal_2018_glh_001.jpg - Project page: https://www.lozano-hemmer.com/level_of_confidence.php #### Zoom Pavilion — Rafael Lozano-Hemmer, Krzysztof Wodiczko (2015) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Being picked out and enlarged by a face detector makes the ordinary experience of public surveillance suddenly personal. - What it is: In a room surrounded by projections, twelve computerised surveillance systems detect visitors' faces and robotic cameras zoom in on them up to 35 times, cutting from crowd views to abstract close-ups and recording the spatial relationships between strangers. - How it works: Face detection software steers independent robotic PTZ cameras with up to 35x optical zoom, whose feeds are projected across the walls. - Lens: Machine seeing, Classification, bias & surveillance - Shown at: MUAC, Mexico City 2015; Art Basel Unlimited 2016 - Video: https://www.youtube.com/watch?v=ENWBRsvn7qA - Images: https://www.lozano-hemmer.com/image_sets/zoom_pavilion/monterrey_2019/zoom_pavilion_monterrey_2019_my_505A7260.jpg - Project page: https://www.lozano-hemmer.com/zoom_pavilion.php #### Face Cages — Zach Blas (2014) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Biometric diagrams look abstract and neutral; made in steel and worn, they become a cage. - What it is: Four queer artists each wear a metal cage built from the biometric diagram of their own face, enduring it on camera until the pain shows. The four-channel video sets the clean geometry of face recognition against the discomfort of the body it measures. - How it works: Facial-recognition landmark diagrams of each performer were generated and 3D-printed in stainless steel, then worn in endurance performances filmed as a four-channel HD video (12 min). - Lens: Classification, bias & surveillance, Embodiment & kinesthesia - Shown at: MUAC, Mexico City 2015; Sonic Acts, Amsterdam 2017; ICA Singapore 2017 - Video: https://vimeo.com/161973235 - Images: https://zachblas.info/wp-content/uploads/2013/08/FACECAGES-1-1024x576.jpg - Project page: https://zachblas.info/works/face-cages/ #### Selfiecity — Lev Manovich, Moritz Stefaner (2014) - Type: Artwork & installation · Machines: Dataset & training data, Surveillance & biometrics - Idea: Reading culture through face analysis mixes machine estimates with hidden human judgement. - What it is: An interactive analysis of 3,200 Instagram selfies from five cities, sortable by estimated age, gender, head tilt, smile and mood, produced by a mix of face-analysis software and Mechanical Turk workers. - How it works: Instagram images filtered and tagged by Mechanical Turk workers, measured with face-analysis software, and shown in interactive visualisations. - Lens: Classification, bias & surveillance, Hidden labour & extraction - Images: http://selfiecity.net/assets/img/selfiecity-600.jpg - Project page: http://selfiecity.net #### Data-masks — Sterling Crispin (2013) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: What a surveillance algorithm calls a face is a ghostly average that no person has. - What it is: 3D-printed nylon masks evolved until face detection and recognition software accepts them as faces: portraits of a human as the machine imagines one. - How it works: Face detection and recognition algorithms were reverse-engineered and genetic algorithms evolved forms that satisfy them, which were then 3D-printed (2013–2015). - Lens: Machine seeing, Alterity & the uncanny, Classification, bias & surveillance - Shown at: GLOBALE: Infosphere, ZKM Karlsruhe 2015; Noorderlicht Photofestival 2015; The Face, Deutsches Hygiene-Museum Dresden 2017 - Video: https://www.youtube.com/watch?v=qlwqQ8zqc5E - Images: https://www.sterlingcrispin.com/data-masks/data-masks_zkm_1.jpg https://www.sterlingcrispin.com/data-masks/data-mask_printed_lineup.jpg - Project page: https://www.sterlingcrispin.com/data-masks.html #### Digital Pareidolia: A Personal Index of Facebook's Erroneous Portraits — Emilio Vavarella (2013) - Type: Artwork & installation · Machines: Surveillance & biometrics, Platform & recommender algorithm - Idea: Machines also see faces in clouds; their pareidolia shows what a face is to them. - What it is: A collection of photos from the artist's Facebook in which the platform's face detector framed a face that is not there: a knot of wood, a shadow, a pattern on a wall. - How it works: Screenshots of Facebook's automatic face-tagging boxes on objects, compiled as an index. - Lens: Failure, glitch & error, Classification, bias & surveillance - Images: https://emiliovavarella.com/wp-content/uploads/2020/07/DIGITAL-PAREIDOLIA-1.jpg - Project page: http://emiliovavarella.com/digital-pareidolia/ #### Stranger Visions — Heather Dewey-Hagborg (2013) - Type: Artwork & installation · Machines: Surveillance & biometrics, Dataset & training data - Idea: A face predicted from data is a statistical guess presented as a likeness, and the guess is already a form of surveillance. - What it is: Dewey-Hagborg collected hairs, chewed gum and cigarette butts from New York streets, extracted DNA and computed life-size, full-colour 3D-printed portraits of what the strangers might look like. The masks hang beside the sample and where it was found. - How it works: DNA was extracted and sequenced at a community biolab; SNPs linked to ancestry, sex, eye and hair colour fed custom software that set parameters of a 3D face model, printed in full colour. - Lens: Classification, bias & surveillance, Opacity & latent space - Shown at: Ars Electronica Honorary Mention 2015; VIDA 15.0 special mention; Fotomuseum Winterthur; Science Gallery Dublin - Video: https://www.youtube.com/watch?v=kYhSYLXN-KA - Images: https://deweyhagborg.com/media/pages/projects/stranger-visions/d8bb0edf45-1643025614/wall_of_faces.jpg https://deweyhagborg.com/media/pages/projects/stranger-visions/83ebebdc4a-1643025584/sample4_face_web.jpg - Project page: https://deweyhagborg.com/projects/stranger-visions #### us+ — Lauren Lee McCarthy, Kyle McDonald (2013) - Type: Research prototype · Machines: Surveillance & biometrics, Voice & listening machine, Computer vision - Idea: When software reads our faces and words to improve a conversation, it also decides what a good conversation is. - What it is: A Google Hangout app that listens to and watches a video call, scoring each speaker's words and facial expressions and popping up advice such as 'stop talking about yourself'; if the conversation gets out of balance, it mutes someone automatically. - How it works: Audio, facial-expression and linguistic analysis based on the LIWC word database and Linguistic Style Matching drive live visualisations, notifications and auto-muting inside Google Hangouts. - Lens: Classification, bias & surveillance, Intimacy, care & companionship - Video: https://vimeo.com/81903116 - Images: https://freight.cargo.site/i/ad160b0c7b6e9061fcbc3bbda3fb065202d10e85f700d065098d874b49761367/_01-copy.jpg - Project page: https://get-lauren.net/us #### Saving Face — Lancel & Maat (2011) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Touch turns face recognition from identification into a gesture of shared intimacy. - What it is: Visitors caress their own face in front of a camera; face recognition tracks the touch and blends their portrait, stroke by stroke, with the faces of earlier visitors into a shared, merged face on a large screen. - How it works: Camera and face-tracking software map touched regions of the visitor's face onto a composite of stored portraits. - Lens: Embodiment & kinesthesia, Intimacy, care & companionship - Video: https://vimeo.com/28499178 - Project page: https://lancelmaat.nl/work/saving-face/ #### Cheese — Christian Moeller (2003) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: A machine that grades sincerity turns a smile into exhausting labour. - What it is: Six actresses each hold a smile on camera for as long as they can, up to an hour and a half, while emotion-recognition software scores it; when the measured happiness drops, an alarm tells them to smile more sincerely. - How it works: Multi-screen video installation; an emotion-recognition system (likely a research prototype of the time) analysed each face live and triggered an alarm below a happiness threshold. - Lens: Classification, bias & surveillance, Embodiment & kinesthesia - Video: https://vimeo.com/42719594 - Images: https://payload.cargocollective.com/1/7/229866/4243042/67-1.jpeg https://payload.cargocollective.com/1/7/229866/4243042/40-1.jpeg - Project page: https://christianmoeller.com/Cheese ### Adversarial Seeing & Evasion Camouflage, adversarial patterns and tactics to hide from or confuse machine perception. #### How (not) to get hit by a self-driving car — Tomo Kihara, Daniel Coppen (2023) - Type: Artwork & installation · Machines: Autonomous vehicle & drone, Computer vision - Idea: Every successful trick is an edge case the car would miss, showing whom its training data does not see. - What it is: A game installation in which players try to cross a virtual street without being detected by a self-driving car's pedestrian detector, disguising themselves to lower their detection score. - How it works: A camera feeds a pedestrian-detection model that scores each player live; winners can donate their edge-case data to improve the model or delete it. - Lens: Machine seeing, Classification, bias & surveillance, Failure, glitch & error - Shown at: Playable City Sandbox, Watershed Bristol 2023; S+T+ARTS Prize 2024, Honorary Mention - Video: https://vimeo.com/883161343 - Images: https://www.tomokihara.com/images/how-not-to-main2.jpg https://www.tomokihara.com/images/how-not-to_01-low-res.jpg - Project page: https://www.tomokihara.com/en/how-not-to-get-hit-by-a-self-driving-car.html #### Manifesto Collection — Cap_able (2022) - Type: Product & platform · Machines: Computer vision, Surveillance & biometrics - Idea: An adversarial example can be knitted: the machine's weakness becomes a garment you choose to wear. - What it is: Knitted jumpers, trousers and dresses whose woven patterns make AI detectors label the wearer as an animal or object instead of a person. - How it works: Adversarial patterns, likely optimised against YOLO-type real-time object detectors, are woven into knitwear with a computerised knitting machine; made in Italy. - Lens: Machine seeing, Classification, bias & surveillance, Failure, glitch & error - Video: https://www.youtube.com/watch?v=ZGp9bZpD6Rg - Images: https://www.capable.design/cdn/shop/files/preview_images/165dc7bd8cca4ddf9ae24f09c1d01543.thumbnail.0000000000_1100x.jpg?v=1705938416 https://www.capable.design/cdn/shop/files/2_8fdbe483-952c-4bd6-a392-01d11de373bc.png?v=1695084551&width=1200 - Project page: https://www.capable.design/pages/technology #### Unlearning Language — Lauren Lee McCarthy, Kyle McDonald (2022) - Type: Artwork & installation · Machines: Voice & listening machine, Computer vision, Surveillance & biometrics - Idea: Evading machine perception becomes a way of rediscovering what is most human in how we communicate. - What it is: Up to eight people are guided by an AI that 'wants to train humans to be less machine-like': whenever it detects their speech, gestures or expressions it intervenes with light, sound and vibration, so the group has to invent ways of communicating it cannot detect, such as humming, clapping or shifting pitch. - How it works: Speech detection, gesture recognition and expression detection run on cameras and microphones in the room and trigger light, sound and haptic feedback; made with the Yamaguchi Center for Arts and Media. - Lens: Machine listening & voice, Classification, bias & surveillance, Agency, autonomy & co-creation - Shown at: Yamaguchi Center for Arts and Media (YCAM) 2022 - Video: https://www.youtube.com/watch?v=6APHL4CJAQ8 - Images: https://freight.cargo.site/w/1200/i/064b679d339d01a58666ee7c0e4ae9bde01b984f474c81e73f5d7340f592be46/opening-performance0.jpeg - Project page: https://get-lauren.net/Unlearning-Language #### A Technique of Evading Face Detection — Guo Cheng (2020) - Type: Artwork & installation · Machines: Computer vision, Surveillance & biometrics - Idea: Escaping machine vision can be a bodily skill: a small, absurd choreography of the face. - What it is: A deadpan tutorial in which the artist twists and turns his head at the exact angles that make a face detector lose track of him. - How it works: The artist tested head poses against a standard face-detection algorithm and recorded the moments when detection fails. - Lens: Machine seeing, Classification, bias & surveillance - Images: https://cdn.guo-cheng.net/files/projectImg/ATechniqueOfEvadingFaceDetection/ATechniqueOfEvadingFaceDetection_1.jpg - Project page: https://guo-cheng.net/index.php/works/a-technique-of-evading-face-detection/ #### Google Maps Hacks — Simon Weckert (2020) - Type: Performance, dance & music · Machines: Platform & recommender algorithm, Infrastructure, energy & supply chain - Idea: A platform perceives the city through proxies; fool the proxy and the map, and then the traffic, follows. - What it is: The artist pulls a handcart of 99 second-hand smartphones through Berlin streets, and Google Maps turns the empty streets red with a virtual traffic jam. - How it works: Each phone runs Google Maps navigation, so the location data of 99 slow-moving devices is read as congested traffic. - Lens: Failure, glitch & error, Agency, autonomy & co-creation - Shown at: Japan Media Arts Festival 2021, Social Impact Award - Video: https://www.youtube.com/watch?v=k5eL_al_m7Q - Images: http://www.simonweckert.com/img/googlemapshacks/3.JPG http://www.simonweckert.com/img/googlemapshacks/2.JPG - Project page: http://www.simonweckert.com/googlemapshacks.html #### Mask.ID — Peng! Collective (2018) - Type: Artwork & installation · Machines: Surveillance & biometrics, Generative model & latent space - Idea: A biometric passport photo can belong to two people at once. - What it is: A Peng! member merged her face with that of EU foreign-policy chief Federica Mogherini and obtained a real German passport with the morphed photo; the campaign asked citizens to do the same to help refugees travel. - How it works: Face-morphing software blends two faces until face recognition matches both; the image was submitted through a standard passport application. - Lens: Classification, bias & surveillance, Agency, autonomy & co-creation - Images: https://pen.gg/wp-content/uploads/2018/09/MaskID_Kabine_1_Melber-1800x1200.jpg - Project page: https://pen.gg/campaign/mask-id-2/ #### Nonfacial Portrait — Shinseungback Kimyonghun (2018) - Type: Artwork & installation · Machines: Computer vision, Surveillance & biometrics - Idea: Portraiture becomes a search for the narrow visual space only humans can perceive. - What it is: Painters are asked to paint a portrait that people recognise but in which no face-detection algorithm can find a face. A camera watches the canvas and a screen warns the painter whenever a face is detected. - How it works: Three face-detection algorithms monitor the canvas through a camera; each painting is shown with a video of its making. - Lens: Machine seeing, Classification, bias & surveillance, Agency, autonomy & co-creation - Shown at: NTT InterCommunication Center (ICC), Tokyo - Video: https://vimeo.com/290710439 - Images: https://djhznh41oxwef.cloudfront.net/works/nonfacial_portrait/ssbkyh_Nonfacial_Portrait_00.png - Project page: https://www.ntticc.or.jp/en/archive/works/nonfacial-portrait/ #### Perception Engines — Tom White (2018) - Type: Artwork & installation · Machines: Computer vision, Generative model & latent space - Idea: Art can be made for the machine's eye: the prints show which few marks are 'enough' for a classifier. - What it is: A series of abstract prints that ImageNet classifiers confidently recognise as a cello, a cabbage, a hammerhead shark or an iron, though people see only a few coloured strokes. - How it works: A constrained drawing system is optimised against an ensemble of ImageNet classifiers, and the results are printed as risographs and screen prints. - Lens: Machine seeing, Opacity & latent space - Images: http://static1.squarespace.com/static/5c3518f55cfd7963746dd576/t/5cfed9300e5c0200015a2735/1560205650916/tom_white_electric_fan_2019+%281%29.jpg?format=1500w https://images.squarespace-cdn.com/content/v1/5c3518f55cfd7963746dd576/1609721536034-7SU963UJO2QP104FXYBO/cello_16_095_240_thumbnail.jpg - Project page: https://drib.net/perception-engines #### Robust Physical-World Attacks on Deep Learning Visual Classification — Kevin Eykholt (2018) - Type: Paper · Machines: Autonomous vehicle & drone, Computer vision - Idea: What looks like graffiti to a driver can rewrite the road for a self-driving car. - What it is: A few black and white stickers on a real stop sign make a road-sign classifier read it as a 45 mph speed-limit sign from many distances and angles. - How it works: The Robust Physical Perturbations (RP2) algorithm optimises sticker-like perturbations over varied physical conditions and tests them in lab and drive-by settings (Eykholt et al., CVPR 2018). - Lens: Machine seeing, Failure, glitch & error - Paper: https://doi.org/10.1109/CVPR.2018.00175 (CVPR 2018) - Project page: https://doi.org/10.1109/CVPR.2018.00175 - Code: https://github.com/evtimovi/robust_physical_perturbations #### Adversarial Patch — Tom B. Brown (2017) - Type: Paper · Machines: Computer vision - Idea: The machine's attention can be hijacked: one salient patch outweighs the whole scene. - What it is: Google Brain researchers made a printable, psychedelic sticker that, placed next to a banana, makes image classifiers see only a 'toaster'. - How it works: A universal, scene-independent patch is optimised over many images, positions and scales against ImageNet classifiers such as Inception (Brown, Mané, Roy, Abadi, Gilmer). - Lens: Machine seeing, Failure, glitch & error, Opacity & latent space - Paper: https://arxiv.org/abs/1712.09665 (arXiv 2017 (NIPS 2017 Workshop)) - Video: https://www.youtube.com/watch?v=i1sp4X57TL4 - Project page: https://arxiv.org/abs/1712.09665 #### Autonomous Trap 001 — James Bridle (2017) - Type: Performance, dance & music · Machines: Autonomous vehicle & drone, Computer vision - Idea: A machine that obeys the symbols it sees can be trapped by those symbols, like a spirit in a magic circle. - What it is: On Mount Parnassus the artist draws a salt circle with road markings around a self-driving car: a dashed line lets it in, a solid line forbids it to leave. - How it works: A salt circle with dashed and solid lane markings traps a car running the artist's own self-driving software; performed on 14 March 2017. - Lens: Machine seeing, Failure, glitch & error, Machine umwelt - Shown at: Entangled Realities, HEK Basel 2019; AI: Who's Looking After Me?, Science Gallery London 2023 - Images: https://jamesbridle.com/media/pages/works/autonomous-trap-001/329e6817a6-1680515123/1-autonomous-trap-001-003.jpg https://jamesbridle.com/media/pages/works/autonomous-trap-001/7f4c4c6acb-1680515123/autonomous-trap-001-001.jpg - Project page: https://jamesbridle.com/works/autonomous-trap-001 #### HyperFace — Adam Harvey (2017) - Type: Speculative design · Machines: Computer vision, Surveillance & biometrics - Idea: Instead of hiding a face, flood the machine with faces only it can see. - What it is: A textile print covered with patterns that face detectors read as many false faces, so the real face nearby gets a lower confidence score. - How it works: Patterns were generated from what face-detection algorithms score as faces; developed with Hyphen-Labs for NeuroSpeculative AfroFeminism. - Lens: Machine seeing, Classification, bias & surveillance, Failure, glitch & error - Shown at: 33C3 Hamburg 2016; Sundance Film Festival 2017 (NeuroSpeculative AfroFeminism) - Video: https://www.youtube.com/watch?v=LLKmyY5ujeU - Images: https://adam.harvey.studio/hyperface/hyperface_adam_harvey_hyphen_labs_01.webp https://adam.harvey.studio/hyperface/scarf_detections_960.webp - Project page: https://adam.harvey.studio/hyperface/ #### Wearable Face Projector — Jing-cai Liu (2017) - Type: Speculative design · Machines: Surveillance & biometrics - Idea: Against face recognition, one answer is to wear someone else's face. - What it is: A head-mounted mini projector that casts another person's face onto the wearer's own, to confuse face and emotion recognition in public space. - How it works: Small projector on a head-mounted arm projects face video onto the wearer; design prototype and film. - Lens: Classification, bias & surveillance, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=_PoudPCevN0 - Images: http://jingcailiu.com/wp-content/uploads/2020/06/Wearable-face-projector-21may2020-scaled-e1592083294514-1024x340.jpg - Project page: http://jingcailiu.com/wearable-face-projector/ #### Accessorize to a Crime — Mahmood Sharif (2016) - Type: Paper · Machines: Surveillance & biometrics, Computer vision - Idea: A face-recognition system can be steered by a few centimetres of pattern on the face. - What it is: Researchers at Carnegie Mellon printed colourful eyeglass frames that let a wearer dodge a face-recognition system or be recognised as someone else, such as a celebrity. - How it works: Adversarial perturbations are optimised only within an eyeglass-frame region, made smooth and printable, and tested physically against deep-network face recognition and Face++ (ACM CCS 2016; Sharif, Bhagavatula, Bauer, Reiter). - Lens: Classification, bias & surveillance, Failure, glitch & error - Paper: https://doi.org/10.1145/2976749.2978392 (ACM CCS 2016) - Video: https://www.youtube.com/watch?v=6Xh1vuwnVhU - Project page: https://doi.org/10.1145/2976749.2978392 #### Hidden Voice Commands — Nicholas Carlini (2016) - Type: Paper · Machines: Voice & listening machine - Idea: The listening machine does not hear what we hear; the gap between human and machine hearing is a space where commands can hide. - What it is: Security research showing that speech recognisers such as Google Now can be commanded by audio that sounds like noise to people, including black-box attacks on commercial assistants and a defence study. - How it works: Iteratively distorts speech features (MFCCs) so that recognisers still decode a command while human listeners in user studies cannot understand it; tested against Google Now and an open-source recogniser. - Lens: Machine listening & voice, Failure, glitch & error - Paper: https://www.usenix.org/conference/usenixsecurity16/technical-sessions/presentation/carlini (USENIX Security 2016) - Video: https://www.youtube.com/watch?v=x4bqJh4wyjU - Project page: https://www.usenix.org/conference/usenixsecurity16/technical-sessions/presentation/carlini #### Deep Neural Networks are Easily Fooled — Evolving AI Lab (2015) - Type: Paper · Machines: Computer vision - Idea: A classifier's certainty is not seeing: it recognises features, not things. - What it is: Anh Nguyen, Jason Yosinski and Jeff Clune evolved images, from static noise to abstract patterns, that state-of-the-art classifiers label with over 99% confidence as a king penguin, a starfish or a baseball. - How it works: Evolutionary algorithms and gradient ascent generate images that maximise class scores of ImageNet- and MNIST-trained convolutional networks (CVPR 2015). - Lens: Machine seeing, Opacity & latent space, Failure, glitch & error - Paper: https://doi.org/10.1109/CVPR.2015.7298640 (CVPR 2015) - Video: https://www.youtube.com/watch?v=M2IebCN9Ht4 - Project page: https://arxiv.org/abs/1412.1897 #### URME Surveillance — Leo Selvaggio (2014) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: If the system can only match faces, many bodies sharing one face make its records meaningless. - What it is: The artist sells a photo-realistic 3D-printed prosthetic of his own face; anyone wearing it in front of face-recognition cameras is identified as Leo Selvaggio. - How it works: A resin prosthetic made from a 3D scan of the artist's face, launched via an Indiegogo campaign and shown with a face-tracking mirror installation. - Lens: Classification, bias & surveillance, Alterity & the uncanny - Video: https://vimeo.com/90828804 - Images: https://images.squarespace-cdn.com/content/v1/50394738e4b0117733940e2f/1393480618366-370KZW8USCHO8HDKTMXI/Prosthetic+Comparison.jpg https://images.squarespace-cdn.com/content/v1/50394738e4b0117733940e2f/1458091446574-1JNPSCUTUT0O9VQRD64Z/URME+maskFLAT.jpg - Project page: https://www.leoselvaggio.com/urmesurveillance #### How Not to Be Seen: A Fucking Didactic Educational .MOV File — Hito Steyerl (2013) - Type: Film & video · Machines: Computer vision, Surveillance & biometrics - Idea: Being seen is a matter of resolution; invisibility can be a tactic against machine vision, but it is also how the poor and disappeared are erased. - What it is: A deadpan instructional video in five lessons on how to disappear from a world of high-resolution cameras, satellites and aerial targets: hide in plain sight, become smaller than a pixel, live in a gated community, or be a woman over 50. - How it works: Filmed partly at a disused aerial photo-calibration target in the California desert, mixing green-screen performances, 3D animation and parodies of training videos. - Lens: Machine seeing, Classification, bias & surveillance - Shown at: 55th Venice Biennale, Il Palazzo Enciclopedico (2013) - Video: https://www.youtube.com/watch?v=kKAKgrZZ_ww - Images: https://www.artforum.com/wp-content/uploads/2015/04/preview00-4.jpg - Project page: https://www.artforum.com/video/hito-steyerl-how-not-to-be-seen-a-fucking-didactic-educational-mov-file-2013-51651/ #### Stealth Wear — Adam Harvey (2013) - Type: Speculative design · Machines: Autonomous vehicle & drone, Surveillance & biometrics - Idea: When the machine sees in heat, privacy becomes a question of what your clothes radiate. - What it is: An anti-drone hoodie, hijab and burqa made of silver-plated fabric that hides the wearer's body heat from overhead thermal cameras. - How it works: Garments are sewn from silver-plated synthetic fabric that reflects thermal radiation; they were tested with a FLIR thermal camera. - Lens: Machine seeing, Classification, bias & surveillance, Embodiment & kinesthesia - Shown at: Tank Magazine HQ, London 2013 - Images: https://adam.harvey.studio/stealth-wear/stealth-wear-poster.webp https://adam.harvey.studio/stealth-wear/anti-drone-hoodie-f.webp - Project page: https://adam.harvey.studio/stealth-wear/ #### Facial Weaponization Suite — Zach Blas (2012) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Opacity can be collective: a face that belongs to many people is a face the machine cannot identify. - What it is: Blas runs community workshops where participants' faces are scanned and merged into amorphous, colourful collective masks that face recognition cannot read. The masks are worn in public actions and in the video 'Facial Weaponization Communiqué: Fag Face'. - How it works: Participants' faces are 3D-scanned, the biometric data are aggregated in modelling software and the combined forms are fabricated as wearable masks; the Fag Face Mask answers studies claiming to detect sexual orientation from faces. - Lens: Classification, bias & surveillance, Machine seeing - Shown at: Workshops and public actions in Los Angeles, San Diego, New York and Mexico City, 2012–14 - Video: https://vimeo.com/57882032 - Images: https://zachblas.info/wp-content/uploads/2013/01/FWS-2-1024x576.jpg - Project page: https://zachblas.info/works/facial-weaponization-suite/ #### CV Dazzle — Adam Harvey (2010) - Type: Speculative design · Machines: Computer vision, Surveillance & biometrics - Idea: A face is whatever the detector's template says it is; designing against that template reveals it. - What it is: A camouflage strategy of bold make-up and asymmetric hairstyles that breaks the features a face detector looks for, so the wearer stays visible to people but disappears for the algorithm. - How it works: Looks were designed and tested against the Viola-Jones Haar-cascade face detector in OpenCV (later looks, 2020, against CNN face recognition). - Lens: Machine seeing, Classification, bias & surveillance, Failure, glitch & error - Shown at: NYU ITP thesis 2010; The New York Times Op-Art 2013; Designs for a Different Future 2020 - Video: https://vimeo.com/12774628 - Images: https://adam.harvey.studio/cvdazzle/cvd-look1.webp https://adam.harvey.studio/cvdazzle/cvdazzle-06-copyright-adam-harvey-2020.webp https://adam.harvey.studio/cvdazzle/cvdazzle-05-saliency-map.webp - Project page: https://adam.harvey.studio/cvdazzle/ #### Pixelhead — Martin Backes (2010) - Type: Artwork & installation · Machines: Surveillance & biometrics, Platform & recommender algorithm - Idea: The pixel mosaic, the screen's own sign of anonymity, can be worn as clothing. - What it is: A full-face balaclava printed with a pixelated face, so that its wearer cannot be recognised or automatically tagged in photos posted online. - How it works: Printed textile mask; defeats face detection by hiding facial landmarks. - Lens: Classification, bias & surveillance, Embodiment & kinesthesia - Images: https://www.martinbackes.com/wordpress/wp-content/uploads/2013/03/martinbackes_web4.jpg - Project page: https://www.martinbackes.com ## Machine Bodies & Behaviour Robots and kinetic machines as bodies that move, react, perform, tire and fail in front of us. ### Cybernetic Creatures Feedback machines from the 1940s–70s that behave as if alive. #### The Senster — Edward Ihnatowicz (1970) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Robot creature & mobile robot - Idea: Shyness is more lifelike than obedience: the machine's withdrawal made visitors change their own behaviour. - What it is: A 4-metre hydraulic creature shaped like a lobster's claw on a long neck, installed at the Philips Evoluon in Eindhoven, moved its 'head' toward quiet sounds and slow movements and shied away from loud noise and sudden gestures. Visitors shouted at it and waited for it to come closer. - How it works: A Philips P9201 computer processes input from four microphones and two Doppler radar sensors in the head and drives six electro-hydraulic servo joints modelled on a lobster's claw. - Lens: Embodiment & kinesthesia, Alterity & the uncanny, Machine listening & voice - Shown at: Philips Evoluon, Eindhoven 1970–1974; Nieuwe Instituut, Rotterdam (restored) - Video: https://www.youtube.com/watch?v=1jDt5unArNk - Images: https://upload.wikimedia.org/wikipedia/commons/1/14/Edward_Ihnatowicz%2C_The_Senster_%2853386590223%29.jpg https://upload.wikimedia.org/wikipedia/commons/4/45/Edward_Ihnatowicz%2C_The_Senster_%2853386711664%29.jpg - Project page: http://www.senster.com/ #### Colloquy of Mobiles — Gordon Pask (1968) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Robot creature & mobile robot - Idea: Machines can be staged as partners in a conversation of desire, and humans can enter that conversation as just one more participant. - What it is: Five hanging mobiles, three 'females' and two 'males', court each other with beams of light and sound, learning which approaches 'satisfy' their partners. Visitors given mirrors and torches could join in and redirect the machines' desire. - How it works: Rotating fibreglass bodies with lights, mirrors and photocells are driven by motor-driven 'drives' that build up and discharge; each mobile adjusts its behaviour to reward signals from the others. - Lens: Agency, autonomy & co-creation, Intimacy, care & companionship, Embodiment & kinesthesia - Shown at: Cybernetic Serendipity, ICA London 1968; Centre Pompidou reconstruction 2020 - Video: https://www.youtube.com/watch?v=RlXDL8Ab0oQ - Project page: https://cyberneticserendipity.net/ #### Cybernetic Sculpture — Wen-Ying Tsai (1968) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: Feedback between sound, light and vibration makes a steel sculpture seem to listen and respond. - What it is: Clusters of thin stainless-steel rods vibrate on a motorised base under strobe light; when visitors clap or speak, the strobe frequency shifts and the rods appear to freeze, sway or dance. - How it works: A motor vibrates the rods at a fixed rate while a sound-sensing feedback circuit changes the strobe's frequency, producing stroboscopic motion effects. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: Howard Wise Gallery, New York 1968 - Images: https://upload.wikimedia.org/wikipedia/commons/7/75/Tsai_cyberneticsculpture79.jpg https://upload.wikimedia.org/wikipedia/commons/f/f1/Tsai_tuningfork1974.jpg - Project page: https://en.wikipedia.org/wiki/Wen-Ying_Tsai #### SAM (Sound Activated Mobile) — Edward Ihnatowicz (1968) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Voice & listening machine - Idea: Turning to listen is the simplest gesture of attention, and it is enough to make a machine seem to be someone. - What it is: A flower-like head of fibreglass petals on an articulated aluminium neck turns toward whoever is speaking or making noise nearby, as if listening. - How it works: Four microphones in the head compare sound levels; small hydraulic pistons move the neck vertebrae toward the loudest direction. - Lens: Machine listening & voice, Embodiment & kinesthesia, Alterity & the uncanny - Shown at: Cybernetic Serendipity, ICA London 1968 - Video: https://www.youtube.com/watch?v=8b52qpyV__g - Images: https://upload.wikimedia.org/wikipedia/commons/e/ee/Sound_Activated_Mobile.jpg - Project page: http://www.senster.com/ihnatowicz/ #### Robot K-456 — Nam June Paik, Shuya Abe (1964) - Type: Artwork & installation · Machines: Humanoid & android, Robot creature & mobile robot - Idea: A robot made clumsy, noisy and bodily, closer to a street performer than to a symbol of technical progress. - What it is: A 20-channel radio-controlled humanoid built in Tokyo that walked the streets, played a recorded speech by John F. Kennedy and dropped beans as excrement. It debuted in Paik's Robot Opera at the 1964 New York Avant Garde Festival. - How it works: Metal armature with motors and a 20-channel radio remote built by engineer Shuya Abe; restored in 2025 by the Nam June Paik Art Center, which holds it (1964/1996). - Lens: Embodiment & kinesthesia, Alterity & the uncanny, Failure, glitch & error - Shown at: Robot Opera, 2nd Annual New York Avant Garde Festival 1964; Nam June Paik: Becoming Robot, Asia Society 2014; Nam June Paik Art Center (collection) - Video: https://www.youtube.com/watch?v=JciiqCsRhdY - Project page: https://en.wikipedia.org/wiki/Nam_June_Paik #### The Hopkins Beast — Johns Hopkins Applied Physics Laboratory (1962) - Type: Research prototype · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: A machine whose whole world is organised around one need, finding food, already looks like an animal. - What it is: A wheeled automaton wanders the white corridors of the laboratory on its own and, when its batteries run low, looks for black wall sockets, plugs itself in and recharges. Later versions used sonar to avoid people and recognise doors and stairs. - How it works: Transistor logic without a computer links touch switches, sonar and, in Mod II, an optical sensor for dark sockets to tracks, a wall-following arm and extendable plug prongs. - Lens: Embodiment & kinesthesia, Machine umwelt, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=zMpEeag7kkM - Images: https://upload.wikimedia.org/wikipedia/commons/3/3e/00-01964_APL_66196_001_The_Beast.jpg - Project page: https://en.wikipedia.org/wiki/Johns_Hopkins_Beast #### CYSP 1 — Nicolas Schöffer (1956) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Robot creature & mobile robot - Idea: The first sculpture to have sense organs and a nervous system made the artwork an agent that reacts rather than an object that is looked at. - What it is: A tower of steel and aluminium plates on a wheeled base turns, spins its polychrome blades and moves across the floor in response to light, colour and sound around it. In 1956 it performed with Maurice Béjart's dancers on the roof of the Cité radieuse in Marseille. - How it works: Photocells and microphones feed an electronic 'brain' built with Philips that drives the motors in the wheeled base and the rotating plates; excitation and calm states depend on the stimuli. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: Cité radieuse, Marseille 1956; Artistes & Robots, Grand Palais 2018 - Video: https://www.youtube.com/watch?v=u8nxktU1R6E - Project page: https://en.wikipedia.org/wiki/Nicolas_Sch%C3%B6ffer #### Theseus (maze-solving mouse) — Claude Shannon (1950) - Type: Research prototype · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Learning becomes visible as a small body in a maze, even though the 'brain' is a bank of telephone relays under the floor. - What it is: A magnetic toy mouse, moved from below the table by relays, explores a reconfigurable maze by trial and error, remembers the route and then runs straight to the 'cheese'. When the maze is changed, it relearns. - How it works: About 75 telephone relays store, for each square of a 25-square maze, the last direction that led out; a magnet on a motorised carriage beneath the board pulls the mouse. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=_9_AEVQ_p74 - Project page: https://en.wikipedia.org/wiki/Claude_Shannon #### Machina speculatrix (Elmer and Elsie) — W. Grey Walter (1949) - Type: Research prototype · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Two 'neurons' and a body in a room are enough for behaviour that people read as curiosity, hunger and self-recognition. - What it is: Two small three-wheeled 'tortoises' with a light sensor and a touch switch wander a room, seek moderate light, back off from bright light and obstacles, and return to a lit hutch to recharge. With a light on its shell, one tortoise reacts to its own reflection in a mirror or to the other one, as if recognising itself. - How it works: Analogue circuits with two vacuum-tube 'neurons' connect a rotating photocell and a shell touch contact to steering and drive motors; behaviour emerges from feedback between the body and its environment. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation, Machine umwelt - Shown at: Festival of Britain, London 1951 - Video: https://www.youtube.com/watch?v=wQE82derooc - Images: https://upload.wikimedia.org/wikipedia/commons/e/e5/Cybernetic_Tortoise_by_William_Grey_Walter_c_1950_%28Ank_Kumar%29.jpg https://upload.wikimedia.org/wikipedia/commons/7/77/Replica_of_Grey_Walter%27s_tortoise_%283572415081%29.jpg - Project page: https://en.wikipedia.org/wiki/Turtle_(robot) ### Artificial Life & Machine Collectives Evolving agents, swarms and self-organising machine ecologies. #### Lasermice — So Kanno (2018) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Making a machine network's communication visible and audible shows collective behaviour as it forms. - What it is: A swarm of 60 small robots that talk to each other with laser beams and photodetectors and tap the floor, so a firefly-like synchronised rhythm emerges without a central controller. - How it works: Each robot carries a laser, photodetectors and a solenoid; pulse-coupled oscillator rules, inspired by fireflies, synchronise their firing. - Lens: Agency, autonomy & co-creation, Machine umwelt - Shown at: Ars Electronica 2019; 22nd Japan Media Arts Festival, Excellence Award 2019; Biennale Nemo, Paris 2019 - Video: https://www.youtube.com/watch?v=4afSgcWleFo - Images: https://www.kanno.so/wp-content/uploads/2021/05/lasermice_by_Quentin_Chevrier_2-copy.jpg - Project page: https://www.kanno.so/project/lasermice #### Putting the Pieces Back Together Again — Ralf Baecker (2018) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: Order emerges and dissolves in a field of simple machines, making self-organisation visible as motion. - What it is: An array of 1,250 small stepper motors, each turning a white bar, forms a surface across which waves, clusters and swarm-like patterns pass, as the motors align with and disturb their neighbours. - How it works: Each motor is an element of a coupled-oscillator or spin-like model; local interaction rules computed in real time set each bar's angle (commissioned for the Drehmoment festival in a former newspaper building). - Lens: Agency, autonomy & co-creation, Opacity & latent space - Shown at: Drehmoment, KulturRegion Stuttgart, Kirchheim unter Teck 2018 - Video: https://www.youtube.com/watch?v=8GtFttw8z9U - Project page: https://rlfbckr.io/project/ptpbta/ #### Lichtsuchende — Dave Murray-Rust, Rocio von Jungenfeld (2015) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Give machines one shared resource and one channel of communication and something like social behaviour appears, legible to us as light. - What it is: A society of small sunflower-like robots that turn their heads towards light to feed on it and flash light to address one another; grouped together they form a flickering photo-kinetic community that visitors walk into and disturb with their own lamps. - How it works: Each robot has light sensors and a motorised head that tracks light plus LEDs to emit it; simple bio-inspired rules for energy exchange and signalling produce group behaviour. - Lens: Machine umwelt, Embodiment & kinesthesia - Shown at: Lumen Prize 2019, BCS Artificial Intelligence Award - Paper: https://doi.org/10.1007/978-3-319-16498-4_15 (EvoMUSART 2015, Lecture Notes in Computer Science) - Video: https://vimeo.com/207512856 - Images: http://static1.squarespace.com/static/65e7176a61627736e02f6c4a/6606a60245e66065f7018332/6606a8f3e120ce2d55829d39/1711713362629/Lichtsuchende-1000x0-c-default.jpeg - Project page: https://lumenprize.org/2019-winners/dave-murray-rust-rocio-von-jungenfeld #### Plantoid — Primavera De Filippi (2015) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Infrastructure, energy & supply chain - Idea: A blockchain can give a machine an economic life of its own, with humans as its pollinators. - What it is: A metal robotic flower that dances and lights up when it receives cryptocurrency donations. When it has collected enough, it commissions artists to build its offspring, choosing among proposals by vote of its donors: a machine that owns and reproduces itself. - How it works: A sculpture with sensors, lights and motors is linked to a smart contract on a blockchain that holds its funds and runs the voting for new plantoids (with Okhaos). - Lens: Agency, autonomy & co-creation, Machine umwelt - Video: https://www.youtube.com/watch?v=5gMbhCgHGjY #### Kilobots: a thousand-robot swarm — Radhika Nagpal (2014) - Type: Research prototype · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: A swarm of cheap, error-prone machines can still build a collective form, with errors absorbed by the group. - What it is: 1,024 coin-sized robots on three wiry legs slowly assemble themselves into a star, a letter K and other shapes, using only local signals with their neighbours; traffic jams and stragglers are part of the result. - How it works: Each Kilobot moves by vibration motors and communicates by infrared reflected off the table; a distributed algorithm uses edge-following, gradient formation and localisation (with Michael Rubenstein and Alejandro Cornejo). - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Paper: https://doi.org/10.1126/science.1254295 (Science 2014) - Video: https://www.youtube.com/watch?v=G1t4M2XnIhI - Images: https://upload.wikimedia.org/wikipedia/commons/d/d9/Kilobot_robot_swarm.JPG #### Swarm Study (series) — Random International (2010) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: Collective motion rules alone can give light the feel of a living group that notices us. - What it is: Lights in a grid of brass rods move like a flock of birds or a shoal of fish, gathering and scattering in response to the people below. Swarm Study / III (2011) is in the V&A collection. - How it works: A flocking algorithm drives individually addressable lights on brass rods; sensors register visitors' positions and presence as disturbances to the swarm. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Shown at: Victoria and Albert Museum, London (Swarm Study / III) - Video: https://www.youtube.com/watch?v=ajV8A5Y2_dE - Project page: https://www.random-international.com/ #### Delicate Boundaries — Chris Sugrue (2007) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Computer vision - Idea: Artificial life becomes unsettling when it leaves the frame and treats your body as terrain. - What it is: Small glowing digital bugs swarm on a screen; when a visitor touches it, they crawl out of the screen onto the visitor's hand and arm and explore the body before returning. - How it works: A camera tracks hands and arms touching the screen, and an overhead projector continues the flocking simulation on the body's surface. - Lens: Embodiment & kinesthesia, Machine umwelt - Shown at: VIDA 10.0, Special Mention (2007) - Video: https://www.youtube.com/watch?v=QJ-uUnsFdxg #### Propagaciones — Leandro Núñez (2007) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Robot creature & mobile robot - Idea: Embodying a cellular automaton in robots makes an abstract model of life into something that happens in the room. - What it is: A grid of small robots turns on and off according to the rules of Conway's Game of Life, so patterns of 'living' and 'dying' cells propagate through a physical machine population. - How it works: Each robot reads the states of its neighbours and updates its own light and movement following Game of Life rules; the whole grid runs without central control. - Lens: Agency, autonomy & co-creation, Machine umwelt - Video: https://www.youtube.com/watch?v=5BTa1c3Y6Fo - Project page: https://www.leonunez.com.ar/ #### ALAVs 2.0 (Autonomous Light Air Vessels) — Jed Berk (2006) - Type: Artwork & installation · Machines: Autonomous vehicle & drone, Robot creature & mobile robot - Idea: Drones read differently when they behave like a gentle, needy flock than like tools. - What it is: A flock of small helium blimps drifts through a space, avoiding walls, flocking with one another, approaching visitors and 'eating' when they run low; visitors can call them by mobile phone. - How it works: Each vessel carries propellers, infrared and sonar sensors and a microcontroller running flocking and obstacle-avoidance behaviours, plus a phone interface for human messages. - Lens: Agency, autonomy & co-creation, Machine umwelt - Shown at: VIDA 10.0, Special Mention (2007); Taipei Digital Art Festival - Video: https://www.youtube.com/watch?v=8CKrjCUMhdA #### Parásitos Urbanos (Urban Parasites) — Gilberto Esparza (2006) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Treat the city's cables and waste as a habitat, and robots become a new wild species that lives off it. - What it is: A family of small robots built from discarded electronics that hang on city power lines and walls, stealing electricity and making sounds like animals living in the urban ecosystem. - How it works: Recycled motors, circuits and speakers are assembled into 'species' (with Latin names such as Furtum electricus) that tap live power lines and react to light and noise. - Lens: Machine umwelt, Agency, autonomy & co-creation - Shown at: VIDA 9.0, Incentive for Ibero-American Production (2006); Intermediae Madrid 2011; Fundación Telefónica Lima, Cultivos (2014) - Video: https://www.youtube.com/watch?v=X64KRaQpyCY - Images: http://gilbertoesparza.net/wp-content/uploads/2023/01/COLGADO-Furtum-electricus-sinuatum-1024x683.jpg http://gilbertoesparza.net/wp-content/uploads/2023/01/DIABLITO-_-LITTLE-DEVIL-Furtum-electricus-imitor-1024x683.jpg - Project page: https://gilbertoesparza.net/portfolio/parasitos-urbanos/ #### Resilient Machines Through Continuous Self-Modeling — Josh Bongard, Hod Lipson (2006) - Type: Paper · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: A machine that models its own body can survive injury, a first step toward a kind of machine self-awareness. - What it is: A four-legged 'starfish' robot does not know its own shape. It moves a little, builds competing models of its body, and uses them to learn to walk; when a leg is shortened, it notices the damage and invents a new gait. - How it works: The robot uses tilt sensors and joint encoders to test action hypotheses and select among candidate self-models, then optimises locomotion on the best model (with Viktor Zykov). - Lens: Embodiment & kinesthesia, Failure, glitch & error, Agency, autonomy & co-creation - Paper: https://doi.org/10.1126/science.1133687 (Science 2006) - Video: https://www.youtube.com/watch?v=x579QKA6fkY #### Proxemia (Proxemics) — Mariela Yeregui (2005) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Social distance and fear of touch can be staged as machine behaviour, making the visitor an obstacle in a robot community. - What it is: Twenty-five glowing acrylic spheres roll freely across the floor; whenever one touches a visitor, another sphere or a wall it changes colour and veers away, as if phobic of contact. One transparent sphere, with its robot visible inside, behaves differently from the rest. - How it works: Each 25 cm sphere contains a small robot with position and motion sensing, bump detection and its own lights; contact triggers a change of direction and light. - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=MEYvKy38x1o - Images: https://marielayeregui.com/wp-content/uploads/2021/10/esferasx2-1024x683.jpg - Project page: https://marielayeregui.com/proxemia-proxemics/ #### Self-reproducing machines (Molecubes) — Hod Lipson (2005) - Type: Paper · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Self-reproduction, long a defining trait of life, can be performed by simple modular machines given a supply of parts. - What it is: A tower of four magnetically linked robotic cubes bends, picks up fresh cubes from feeding stations and stacks them into a copy of itself, in about two and a half minutes. - How it works: Each 10 cm cube has a swivel joint, electromagnets on its faces and a microcontroller; a pre-programmed sequence moves cubes from feeders to build the replica (Cornell, with Victor Zykov, Efstathios Mytilinaios and Bryant Adams). - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Paper: https://doi.org/10.1038/435163a (Nature 2005) - Video: https://www.youtube.com/watch?v=gZwTcLeelAY #### ArtSBots — Leonel Moura (2003) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Emergent machine collectives can produce 'non-human art' whose author is the swarm. - What it is: A swarm of small wheeled robots with coloured pens moves over a large canvas, each reacting to colour left by the others, until a painting emerges that no one planned. In 2006 Moura's RAP (Robotic Action Painter) extended the idea to a single autonomous painter. - How it works: Each robot has colour sensors, obstacle sensors and a microcontroller running stigmergy-like rules: it paints more where colour already is, as ants follow pheromones (with Henrique Garcia Pereira). - Lens: Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=XlUxXQPmKUc - Project page: https://www.leonelmoura.com/robot-art/ #### Autopoiesis — Ken Rinaldo (2000) - Type: Artwork & installation · Machines: Industrial robot arm, Cybernetic & artificial-life system - Idea: A group of machines that talk to each other develops a collective mood that no single arm contains. - What it is: Fifteen robotic arms made of grapevine and electronics hang from the ceiling, sense visitors and bend toward them, and signal to each other with tones so that the group's behaviour shifts as a whole. - How it works: Each arm has infrared sensors and a microcontroller; the arms share information through a central computer and audible tones, and a camera tracks visitors' positions. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia, Machine umwelt - Shown at: Alien Intelligence, Kiasma, Helsinki 2000; VIDA 3.0, first prize; Prix Ars Electronica 2001, Honorary Mention (Interactive Art) - Video: https://www.youtube.com/watch?v=9TfrKQH6OPg - Images: https://www.kenrinaldo.com/wp-content/uploads/2015/04/Autopoieis3-440x450.jpg https://archive.aec.at/asset/1318971/preview/ - Project page: https://www.kenrinaldo.com/portfolio/autopoiesis/ #### Eden — Jon McCormack (2000) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Generative model & latent space - Idea: Human attention becomes an ecological resource that artificial creatures evolve to capture. - What it is: An evolving ecosystem of sonic creatures lives on a cellular world projected onto translucent screens. The creatures learn to sing, mate and find food; because visitors' presence makes food grow, creatures that attract people with their songs survive. - How it works: Agents with learning classifier systems move on a lattice and generate sound; infrared sensors detect visitors and increase food in nearby cells. - Lens: Machine umwelt, Machine listening & voice, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=Yrww68pnqqM - Images: https://i0.wp.com/jonmccormack.info/wp-content/uploads/2023/03/Eden3.jpg?fit=1920%2C1280&ssl=1 - Project page: https://www.jonmccormack.info/artworks/eden/ #### Golem project — Hod Lipson (2000) - Type: Research prototype · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Machines can design and make their own bodies, with humans reduced to plugging in the motors. - What it is: Computers evolved simple bar-and-actuator robots in simulation, selecting them for their ability to crawl, then 3D-printed them; after motors were added by hand, the physical 'lifeforms' crawled across the floor as in simulation. - How it works: An evolutionary algorithm co-evolves morphology and neural control in a physics simulation; winning designs are fabricated by rapid prototyping (with Jordan Pollack, Brandeis University). - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Paper: https://doi.org/10.1038/35023115 (Nature 2000) - Video: https://www.youtube.com/watch?v=tgJcYx-yewA #### Galápagos — Karl Sims (1997) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Generative model & latent space - Idea: Aesthetic preference becomes a selection pressure, and the artist hands authorship to an evolutionary process. - What it is: Twelve screens each show an abstract animated organism. Visitors stand on sensors in front of the ones they like; those survive, mate and mutate, while the others die, so the audience acts as natural selection. - How it works: Twelve computers render organisms from genetic codes; step sensors in front of each screen register selections, and the chosen genotypes are copied, recombined and mutated. - Lens: Agency, autonomy & co-creation, Machine umwelt - Shown at: NTT InterCommunication Center (ICC), Tokyo 1997–2000 - Images: https://www.karlsims.com/galapagos/galapagos-icc.jpg https://www.karlsims.com/galapagos/image39.jpg - Project page: https://www.karlsims.com/galapagos/index.html #### Life Spacies — Christa Sommerer & Laurent Mignonneau (1997) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Chatbot & language model - Idea: Language becomes DNA: what people write literally shapes the bodies of machine creatures. - What it is: Visitors send text messages over the internet; each text is read as a genetic code that creates a unique artificial creature, which then lives, feeds on further text and reproduces in a shared online and gallery world. - How it works: A text-to-form genetic algorithm translates characters and their order into body parameters; typed text also becomes food in the environment. - Lens: Agency, autonomy & co-creation, Machine umwelt - Shown at: NTT InterCommunication Center (ICC), Tokyo 1997 - Video: https://www.youtube.com/watch?v=-OXpk0PmMLc #### Sympathetic Sentience — Simon Penny (1995) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: A machine collective can have a life of its own that humans can only disturb, not control. - What it is: Twelve small electronic units on the walls and ceiling of a dark room chirp and pass their rhythms to each other by infrared, so that a complex, never-repeating sound pattern builds up. Visitors walking through break the chain and the room slowly falls silent, then rebuilds its song. - How it works: Each identical unit chirps once a minute and combines its own rhythm with the infrared data stream it receives before passing it on (with Jamieson Schulte). - Lens: Machine listening & voice, Agency, autonomy & co-creation, Machine umwelt - Video: https://www.youtube.com/watch?v=ZX9Zta2T1do - Project page: https://simonpenny.net/works/sympathetic.html #### A-Volve — Christa Sommerer & Laurent Mignonneau (1994) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Generative model & latent space - Idea: Artificial life becomes tangible when a drawn form has to survive in a shared environment with other creatures and human hands. - What it is: Visitors draw a shape on a touch screen and it becomes a 3D creature swimming in a real pool of water. The creatures hunt, mate and die according to their form, and visitors can touch and protect them with their hands in the water. - How it works: Drawings are converted into genetic code that sets each creature's shape and swimming ability; creatures are projected into a water-filled glass pool where cameras track visitors' hands. - Lens: Agency, autonomy & co-creation, Machine umwelt, Embodiment & kinesthesia - Shown at: Prix Ars Electronica 1994, Golden Nica (Interactive Art) - Video: https://www.youtube.com/watch?v=0WG45k93pCA - Images: https://archive.aec.at/asset/1390558/preview/ - Project page: https://archive.aec.at/prix/142357/ #### Evolved Virtual Creatures — Karl Sims (1994) - Type: Paper · Machines: Cybernetic & artificial-life system, Robot creature & mobile robot - Idea: Evolution in simulation produces machine bodies no engineer would draw, strange yet obviously purposeful. - What it is: Block-bodied creatures swim, walk, jump and fight over a cube in a physically simulated world. Both their bodies and their neural 'brains' were evolved by a genetic algorithm, not designed. - How it works: Directed graphs encode morphology and neural controllers; populations are selected for swimming, walking, jumping, following or competition in a rigid-body physics simulation on a Connection Machine CM-5. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation, Machine umwelt - Paper: https://doi.org/10.1145/192161.192167 (SIGGRAPH 1994) - Video: https://www.youtube.com/watch?v=RZtZia4ZkX8 - Images: http://www.karlsims.com/creatures-crab-vs-arm.jpg - Project page: https://www.karlsims.com/evolved-virtual-creatures.html #### Interactive Plant Growing — Christa Sommerer & Laurent Mignonneau (1992) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: Living plants and virtual growth are coupled so that the garden grows only through human–plant contact. - What it is: Visitors touch or approach five real plants, and each gesture makes virtual plants grow on a large projection; the living plants act as the interface to an artificial garden. - How it works: The plants' electrical potential changes when touched are measured and translated into growth parameters for real-time 3D plant models. - Lens: Machine umwelt, Agency, autonomy & co-creation - Shown at: ZKM Karlsruhe collection - Video: https://www.youtube.com/watch?v=JXX7JNFD2X8 #### Strandbeest — Theo Jansen (1990) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: A machine without electronics can still evolve, sense and 'survive' in its own habitat. - What it is: Large walking skeletons of PVC tubing that are powered by wind on Dutch beaches, store air in bottles, and in later generations sense water and turn back from the sea. - How it works: Leg linkages with 'holy numbers' found by a computer-evolved algorithm convert wind-driven rotation into walking; plastic bottles act as air 'stomachs' and tubes as logic. - Lens: Embodiment & kinesthesia, Machine umwelt - Shown at: Prix Ars Electronica 2005, special prize (Interactive Art) - Video: https://www.youtube.com/watch?v=C97kMKwZ2-g - Images: https://archive.aec.at/asset/1050599/preview/ - Project page: https://www.strandbeest.com/ #### Boids — Craig Reynolds (1987) - Type: Paper · Machines: Cybernetic & artificial-life system - Idea: Collective life can emerge without a leader or plan, which changed how we imagine swarms of machines. - What it is: Simulated 'bird-oid' agents flock, split around obstacles and regroup like real birds, although each follows only three local rules. The model was soon used for bats and penguins in films such as Batman Returns (1992). - How it works: Each agent steers by separation, alignment and cohesion relative to nearby flockmates, computed in a 3D animation system. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Paper: https://doi.org/10.1145/37401.37406 (SIGGRAPH 1987) - Video: https://www.youtube.com/watch?v=86iQiV3-3IA - Project page: https://www.red3d.com/cwr/boids/ ### Industrial Arms & Choreography Factory robots taken out of the factory: arms that dance, draw, labour or plead. #### Manus — Madeline Gannon (2018) - Type: Artwork & installation · Machines: Industrial robot arm, Computer vision - Idea: Industrial robots can be designed as a social group with shared attention rather than as isolated tools. - What it is: Ten industrial robot arms in a row share one perception system and behave as a pack, turning together to follow passers-by and passing attention along the line. - How it works: Depth sensors at the base feed a single tracking system; each ABB arm chooses targets according to pack-behaviour rules. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Shown at: World Economic Forum Annual Meeting of the New Champions, Tianjin 2018 - Video: https://www.youtube.com/watch?v=lyQwqi-Hycw - Project page: https://atonaton.com/manus #### The Punishment — Filipe Vilas-Boas (2017) - Type: Artwork & installation · Machines: Industrial robot arm - Idea: Robot ethics as a schoolroom punishment asks whether rules can be learned by repetition, and who is really being disciplined. - What it is: An industrial robot arm sits at a school desk and writes, over and over in neat handwriting, the line 'I must not hurt humans', like a child punished with lines. - How it works: A small industrial arm holding a pen replays a handwriting trajectory on paper in an endless loop. - Lens: Failure, glitch & error, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=wA0zU2wC2dk - Images: https://payload.cargocollective.com/1/19/638884/12653201/prt_1497962393_2x.jpg - Project page: https://www.filipevilasboas.com/the-punishment #### Mimus — Madeline Gannon (2016) - Type: Artwork & installation · Machines: Industrial robot arm, Computer vision - Idea: A factory robot given curiosity instead of a task is no longer a threat but a creature to meet. - What it is: A 1,200 kg ABB industrial robot in a glass enclosure at the Design Museum in London notices visitors, follows the most interesting one and gets bored when they stay still, like a curious animal. - How it works: Depth sensors in the ceiling track visitors; custom software (no pre-programmed movements) decides where the arm looks and moves in real time. - Lens: Agency, autonomy & co-creation, Intimacy, care & companionship, Embodiment & kinesthesia - Shown at: Fear and Love, Design Museum London 2016 - Video: https://www.youtube.com/watch?v=B9jwNWiEd_M - Images: http://static1.squarespace.com/static/5758289d27d4bdf581e1c031/t/64386d617b6835253f5680cd/1681419617332/charlie-longexposure-2.jpg?format=1500w - Project page: https://atonaton.com/mimus #### Box — Bot & Dolly (2013) - Type: Film & video · Machines: Industrial robot arm - Idea: When industrial arms move the stage, precision itself becomes the illusion, and the machine hides by performing perfectly. - What it is: A live performance filmed in one take: two industrial robot arms move large screens while a performer interacts with projection-mapped illusions of depth, cubes and collapse. The arms are the invisible stagehands. - How it works: Robot arms carrying projection surfaces are synchronised with projection mapping and a motion-control camera in Autodesk Maya-based choreography. - Lens: Embodiment & kinesthesia, Opacity & latent space - Video: https://www.youtube.com/watch?v=lX6JcybgDFo #### Double-Taker (Snout) — Golan Levin (2008) - Type: Artwork & installation · Machines: Industrial robot arm, Computer vision - Idea: Surveillance becomes eye contact: a machine gaze staged as curiosity makes people feel seen in a way a CCTV camera never does. - What it is: A 2.5-metre industrial robot arm dressed as a giant inchworm or trunk sits on a roof above a museum entrance; it turns an oversized googly eye towards passers-by, tracks them and performs 'double-takes' as if surprised by each person. - How it works: ABB industrial arm driven by a real-time machine-vision system that tracks bodies below and triggers expressive 'double-take' motions. - Lens: Machine seeing, Alterity & the uncanny, Embodiment & kinesthesia - Shown at: Prix Ars Electronica 2009, Honorary Mention (Interactive Art) - Video: https://www.youtube.com/watch?v=jXqYf4m-HA8 - Images: https://archive.aec.at/asset/327892/preview/ https://archive.aec.at/asset/327893/preview/ https://archive.aec.at/asset/327894/preview/ - Project page: https://www.flong.com/archive/projects/snout/index.html ### Androids & the Uncanny Humanlike machines and the unease of near-likeness. #### Ameca — Engineered Arts (2021) - Type: Product & platform · Machines: Humanoid & android, Chatbot & language model - Idea: Its viral clips show how much feeling people read into a machine face, even when they know the expression is programmed. - What it is: A grey-skinned, deliberately genderless humanoid robot whose first video, waking up and looking at its own hands with surprise, went viral; it is now shown at fairs and museums, often talking through a large language model. - How it works: A modular humanoid platform with dozens of facial actuators, cameras in the eyes and cloud software for animation and conversation. - Lens: Alterity & the uncanny, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=IPukuYb9xWw - Images: https://upload.wikimedia.org/wikipedia/commons/5/5e/Ameca_Generation_1.jpg - Project page: https://www.engineeredarts.co.uk/robot/ameca/ #### Annelies, Looking for Completion — L.A. Raeven (2018) - Type: Artwork & installation · Machines: Humanoid & android - Idea: Instead of a robot that serves us, a robot that needs us, and the discomfort of comforting a machine. - What it is: A lifelike android clone of the twin artists, a 'third sister', sits crying in a corner; when visitors try to comfort her she looks up with a tear-streaked face. - How it works: Silicone android modelled on the artists' faces and bodies, with motorised facial expressions, tear mechanism and sensors reacting to approach. - Lens: Alterity & the uncanny, Intimacy, care & companionship - Shown at: Robot Love, Eindhoven 2018; Collection Bonnefanten, Maastricht - Video: https://www.youtube.com/watch?v=Jyfq-xq5GyA - Project page: https://robotlove.nl/en/expo/ #### ORLAN-OÏDE — ORLAN (2018) - Type: Artwork & installation · Machines: Humanoid & android, Chatbot & language model - Idea: An artist famous for remaking her body builds a machine self and lets it speak in her place. - What it is: A humanoid robot double of the artist, with her face and voice, that answers questions and speaks about art; ORLAN has interviewed it in public. - How it works: Robotic bust with a conversational AI trained on the artist's texts and a synthesised copy of her voice; later versions developed with RaspiAudio. - Lens: Alterity & the uncanny, Grief, memory & afterlife - Shown at: Artistes & Robots, Grand Palais, Paris 2018 - Video: https://www.youtube.com/watch?v=uSeo1wMvG_g - Project page: https://www.orlan.eu/ #### SEER: Simulative Emotional Expression Robot — Takayuki Todo (2018) - Type: Artwork & installation · Machines: Humanoid & android, Computer vision - Idea: Being looked back at by a machine is enough to feel that something is there. - What it is: A small, stylised humanoid head that tracks the viewer's gaze and mirrors their facial expression with subtle eye, eyelid and eyebrow movements. - How it works: A camera with face and gaze tracking drives motors in the eyes and eyelids and wire-driven, flexible eyebrows; the face is deliberately not realistic. - Lens: Alterity & the uncanny, Machine seeing - Shown at: Ars Electronica 2018; Laval Virtual 2019 - Video: https://www.youtube.com/watch?v=BJZcGJSK1Z0 - Project page: https://www.takayukitodo.com/ #### Alter — Takashi Ikegami, Hiroshi Ishiguro (2016) - Type: Research prototype · Machines: Humanoid & android, Cybernetic & artificial-life system - Idea: Spontaneity, not realism, is what makes a machine feel alive. - What it is: An android with a human face and an exposed mechanical body that moves on its own: its gestures come from an artificial-life neural network reacting to sensors, not from a script. - How it works: About 40 pneumatic axes driven by central pattern generators and a spiking neural network fed by proximity, noise and temperature sensors; Alter3 (2019) added cameras and later GPT-4 control. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny, Embodiment & kinesthesia - Shown at: Miraikan, Tokyo 2016; Prix Ars Electronica 2018, Interactive Art+; AI: More than Human, Barbican 2019 (Alter 3) - Paper: https://doi.org/10.1162/isal_a_00463 (ALIFE 2021) - Video: https://www.youtube.com/watch?v=iYzOZQDsm5w - Project page: https://en.wikipedia.org/wiki/Takashi_Ikegami #### Sophia — Hanson Robotics (2016) - Type: Product & platform · Machines: Humanoid & android, Chatbot & language model - Idea: A robot staged as a person shows how easily personhood is granted to a performance. - What it is: A female-presenting humanoid robot with an expressive face that appeared on talk shows and at the UN; in 2017 Saudi Arabia announced it had granted Sophia citizenship. Roboticists criticised the scripted appearances as misleading about AI. - How it works: Animatronic face in Frubber skin with motors and cameras, combined with scripted and chatbot dialogue. - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Shown at: SXSW 2016; Future Investment Initiative, Riyadh 2017; AI for Good Global Summit 2018 - Video: https://www.youtube.com/watch?v=sKrV2CVDXjo - Images: https://upload.wikimedia.org/wikipedia/commons/1/1e/Sophia_at_the_AI_for_Good_Global_Summit_2018_%2827254369347%29_%28cropped%29.jpg - Project page: https://en.wikipedia.org/wiki/Sophia_(robot) #### (Female figure) — Jordan Wolfson (2014) - Type: Artwork & installation · Machines: Humanoid & android, Computer vision - Idea: Eye contact from a machine turns looking into being looked at, and makes the viewer complicit in a sexualised display. - What it is: A life-size animatronic woman in a witch mask, bikini and dirty boots dances seductively against a mirror, then looks the viewer in the eye while speaking in the artist's voice. - How it works: An animatronic figure built with Spectral Motion runs choreographed routines; facial-recognition cameras let it lock eyes with viewers. - Lens: Alterity & the uncanny, Machine seeing, Classification, bias & surveillance - Shown at: David Zwirner, New York 2014 - Video: https://www.youtube.com/watch?v=XSFP6JHnn9c - Project page: https://www.davidzwirner.com/artists/jordan-wolfson #### The Man Machine — Vincent Fournier (2010) - Type: Artwork & installation · Machines: Humanoid & android, Social & companion robot - Idea: Portraiture, a genre for persons, applied to machines makes us ask what in them looks back. - What it is: A photographic series of real humanoid robots, including ASIMO and REEM-B, posed in laboratories, homes and landscapes with the stillness of formal portraits. - How it works: Staged large-format colour photographs made on location with robotics labs in Japan, Europe and the US. - Lens: Alterity & the uncanny, Embodiment & kinesthesia - Shown at: Hello, Robot., Vitra Design Museum 2017 - Video: https://www.youtube.com/watch?v=JUhIQo6PvQo - Project page: https://www.vincentfournier.co.uk/ #### Dirk, the homeless robot — Electric Circus (2008) - Type: Performance, dance & music · Machines: Humanoid & android - Idea: People will extend pity or unease to a machine before deciding whether it is alive; the android tests how we treat those at the edge of the street. - What it is: A lifelike animatronic 'homeless man' that sits and moves on streets and at festivals; passers-by talk to him, give him money or recoil before realising he is a machine. - How it works: Animatronic figure built by inventor Fred Abels and puppeteer Mirjam Langemeijer, with moving head, eyes and hands, operated in public space. - Lens: Alterity & the uncanny, Intimacy, care & companionship - Shown at: Prix Ars Electronica 2010, Honorary Mention (Hybrid Art) - Video: https://www.youtube.com/watch?v=cwX6CuUyeFw - Images: https://archive.aec.at/asset/452494/preview/ https://archive.aec.at/asset/452495/preview/ - Project page: http://www.electric-circus.eu #### Geminoid HI-1 — Hiroshi Ishiguro (2006) - Type: Research prototype · Machines: Humanoid & android - Idea: When a copy stands in for its maker, people start to ask where the 'person' actually is. - What it is: A teleoperated android copy of Hiroshi Ishiguro, built with silicone skin and pneumatic actuators; Ishiguro sends it to give lectures while he speaks through it from elsewhere. - How it works: An operator's voice and lip and head movements, captured by camera and microphone, drive about 50 air actuators in the android; small idle motions run automatically. - Lens: Alterity & the uncanny, Embodiment & kinesthesia - Shown at: Ars Electronica 2009 - Paper: https://doi.org/10.5772/4876 (Humanoid Robots: New Developments (I-Tech), 2007) - Video: https://www.youtube.com/watch?v=fhS9KkOXDXE - Images: https://upload.wikimedia.org/wikipedia/commons/d/d6/Hiroshi_Ishiguro_and_Geminoid_HI-1_at_AEC.JPG - Project page: http://www.geminoid.jp/en/robots.html #### JoAn, l'home de carn — Marcel·lí Antúnez Roca (1992) - Type: Artwork & installation · Machines: Humanoid & android, Voice & listening machine - Idea: A machine clad in flesh shows that the uncanny comes as much from the audience's laughter and barking dogs as from the robot. - What it is: A life-size seated male robot covered in sewn pig skin sits in a glass case and moves its neck, shoulder, elbow and penis whenever visitors speak, sing, whistle or clap near it. - How it works: Computer analyses the sound in the room and sends signals to motors at four joints of a mechanically articulated figure. - Lens: Alterity & the uncanny, Machine listening & voice - Images: https://www.marceliantunez.com/media/cache/07/7c/077c994ce12858786f4a25a749c0ba02.jpg - Project page: https://www.marceliantunez.com/work/joan-lhome-de-carn/ #### ROSA BOSOM — Bruce Lacey (1965) - Type: Artwork & installation · Machines: Humanoid & android - Idea: A deliberately crude android shows how much of the robot fantasy is gender, theatre and a remote control. - What it is: A clanking, junk-built 'female' robot actress on wheels whose name spells Radio Operated Simulated Actress Battery Or Standby Operated Mains. At Cybernetic Serendipity she was shown with her partner MATE, parodying the dream of the machine woman. - How it works: Radio-controlled motors move a body assembled from scrap, domestic objects and mannequin parts; it runs on batteries or mains power as the name says. - Lens: Alterity & the uncanny, Failure, glitch & error - Shown at: Cybernetic Serendipity, ICA London 1968 - Video: https://www.youtube.com/watch?v=nTMNiYV5yIE - Project page: https://en.wikipedia.org/wiki/Bruce_Lacey #### The Writer — Pierre Jaquet-Droz (1774) - Type: Artwork & installation · Machines: Humanoid & android - Idea: An 18th-century android already raised the question of writing without a mind behind it. - What it is: A doll-sized boy at a desk dips a goose quill in ink and writes a text of up to 40 characters, following the letters with his eyes. It is still demonstrated at the Musée d'art et d'histoire in Neuchâtel. - How it works: About 6,000 parts, with a programmable wheel of interchangeable cams that sets each letter; made with Henri-Louis Jaquet-Droz and Jean-Frédéric Leschot. - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Shown at: Musée d'art et d'histoire, Neuchâtel - Video: https://www.youtube.com/watch?v=ux2KW20nqHU - Project page: https://en.wikipedia.org/wiki/Jaquet-Droz_automata ### Robotic Sculpture & Kinetic Installation Autonomous robotic sculptures and machine installations that behave in a space with visitors. #### Guanaquerx — Paula Gaetano Adi (2025) - Type: Performance, dance & music · Machines: Robot creature & mobile robot, Autonomous vehicle & drone - Idea: Robotics can be rooted in local, Andean knowledge and landscape rather than in Silicon Valley's image of the machine. - What it is: An autonomous rover built with artisans and amateur mechanics in San Juan, Argentina, crosses the Andes toward Chile, re-enacting José de San Martín's 1817 independence expedition as a multi-day performance. - How it works: An all-terrain rover inspired by local 'guanaqueras' (home-built off-road vehicles) navigates mountain terrain autonomously, with community workshops as part of the work. - Lens: Machine umwelt, Agency, autonomy & co-creation, Embodiment & kinesthesia - Shown at: Prix Ars Electronica 2025, Golden Nica (Artificial Life & Intelligence) - Video: https://www.youtube.com/watch?v=TRGilwpKi-I - Images: https://archive.aec.at/asset/3849576/preview/ - Project page: https://archive.aec.at/prix/303077/ #### Little Ark — Choe U-Ram (2022) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: A machine that rows without going anywhere becomes an image of collective paralysis. - What it is: A large kinetic ark made of scrap cardboard boxes and steel whose oars open and close in a slow choreographed cycle, surrounded by other machine creatures. Shown in the MMCA Hyundai Motor Series, it stages a society adrift. - How it works: Motorised steel structure with cardboard skins and programmed motion; built and tested in a temporary workshop before installation. - Lens: Embodiment & kinesthesia, Machine umwelt - Shown at: MMCA Hyundai Motor Series 2022: Choe U-Ram – Little Ark, MMCA Seoul 2022–23 - Video: https://www.youtube.com/watch?v=zKHfM4PUZ7k - Project page: https://www.uram.net/ #### πTon — Cod.Act (2017) - Type: Performance, dance & music · Machines: Robot creature & mobile robot - Idea: When sound comes from the contortions of a soft machine body, the machine seems to sing its own effort. - What it is: A huge ring of rubbery tube lies on the floor and writhes, twists and contracts like a living creature; its convulsions produce a vocal, choral sound. - How it works: Internal motors and cables deform the elastic tube; its movement data are transformed in real time into a vocal soundscape. - Lens: Embodiment & kinesthesia, Machine listening & voice - Video: https://www.youtube.com/watch?v=fB1kI5EGzqY - Images: https://codact.ch/wp-content/uploads/2018/05/PiTon230518SC-09.jpg - Project page: https://codact.ch/ #### Mother of Machine: NoodleFeet — Sarah Petkus (2016) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: A robot built without a use case invites a different relation: care and curiosity instead of service. - What it is: NoodleFeet is an 'adolescent' robot with pool-noodle feet that walks, tastes the ground and grabs objects; its maker describes the ongoing build as raising a child who should develop its own ways of reacting to the world. - How it works: Legged robot of light metals, 3D-printed parts and found objects with sensors on feet and 'tongue' that trigger behaviours; hardware and code are rebuilt iteratively. - Lens: Embodiment & kinesthesia, Intimacy, care & companionship, Agency, autonomy & co-creation - Shown at: Prix Ars Electronica 2018, Honorary Mention (Interactive Art+) - Video: https://www.youtube.com/watch?v=lbf6xV47L8I - Images: https://archive.aec.at/asset/2567402/preview/ https://archive.aec.at/asset/2567404/preview/ https://archive.aec.at/asset/2567408/preview/ - Project page: https://archive.aec.at/prix/236785/ #### Study for Fifteen Points / I — Random International (2016) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: Fifteen moving points are enough for us to see a human; the machine exploits our perception of biological motion. - What it is: Fifteen points of light on mechanical rods move so that, together, they are instantly seen as a person walking, although there is no body, only machinery. - How it works: Motors move fifteen lights along rods to reproduce a point-light walker, developed after a residency at Harvard's School of Engineering and Applied Sciences. - Lens: Embodiment & kinesthesia, Machine seeing, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=qnkxo7CWACs - Images: https://images.squarespace-cdn.com/content/v1/663ca532b3e3797d82159fe2/0d90a48e-3f33-4c69-b508-9cbc6a607893/STUDY+FOR+FIFTEEN+POINTS+%3A+I%2C+2016++.jpeg - Project page: https://www.random-international.com/study-for-fifteen-points #### Jller — Benjamin Maus, Prokop Bartoníček (2015) - Type: Artwork & installation · Machines: Computer vision, Industrial robot arm - Idea: Computer vision can be slow and contemplative: a machine reading deep time. - What it is: An industrial machine picks up stones collected from the Iller river in southern Germany one by one, photographs them and lays them out in rows by geological age. - How it works: Gantry robot with vacuum gripper and camera; image features (likely colour, texture and veining) classify each stone by geological origin, trained on stones labelled by the artists. - Lens: Machine umwelt, Machine seeing - Images: https://media.wired.com/photos/5926f95bcfe0d93c47431f8b/191:100/w_1280,c_limit/RockRobot-1.jpg - Project page: http://www.prokopbartonicek.com/jller-ignorance-with-benjamin-maus-2015 #### Mirage — Ralf Baecker (2014) - Type: Artwork & installation · Machines: Generative model & latent space, Cybernetic & artificial-life system - Idea: A learning machine can dream of a world it senses but never sees, and we see only its hallucination. - What it is: A laser line reflected off a thin, trembling mirror sheet projects a slowly shifting landscape on the wall, the 'dream' of a neural network fed by the Earth's magnetic field. - How it works: A fluxgate magnetometer feeds an unsupervised learning algorithm modelled on the Helmholtz machine; its generated variations drive 48 muscle wires that deform the mirror. - Lens: Opacity & latent space, Machine umwelt - Shown at: Prix Ars Electronica 2015, Honorary Mention (Hybrid Art); ACT Festival, Asia Culture Center, Gwangju 2015 - Video: https://www.youtube.com/watch?v=VY1hOjWlVO8 - Images: https://archive.aec.at/asset/1882015/preview/ - Project page: https://archive.aec.at/prix/212751/ #### Nyloïd — Cod.Act (2013) - Type: Performance, dance & music · Machines: Robot creature & mobile robot - Idea: A machine without head or face can still read as a creature through tension, torsion and voice. - What it is: A tall tripod creature with long, nylon-sheathed legs twists, whips and folds in a performance, emitting a strange voice as its limbs torque and release. - How it works: Motors twist the three flexible legs to store and release elastic energy; a sound system turns the movement into vocal sound. - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Shown at: Sónar, Barcelona; FutureEverything, Manchester - Video: https://www.youtube.com/watch?v=f2sMD73fdYM - Images: https://codact.ch/wp-content/uploads/2013/10/Nyl-18-Copie.jpg - Project page: https://codact.ch/ #### Pinokio — Adam Ben-Dror (2012) - Type: Research prototype · Machines: Social & companion robot, Computer vision - Idea: A familiar object gains a personality through a few behaviours: attention, boredom and refusal to be switched off. - What it is: A desk lamp watches people with a camera in its head, follows their faces, grows bored, looks around and dances to sounds; if someone switches it off, it switches itself back on. - How it works: A webcam and face tracking in Processing/OpenCV drive six servo motors through an Arduino (with Shanshan Zhou and Joss Doggett). - Lens: Intimacy, care & companionship, Agency, autonomy & co-creation, Machine seeing - Shown at: VIDA 16.0 (Fundación Telefónica), 2014 - Video: https://www.youtube.com/watch?v=1E_hHr6YuFk #### Irrational Computing — Ralf Baecker (2011) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: Showing the minerals under every microchip makes computation a material, geological process again. - What it is: Five linked modules of raw crystals, minerals and electrodes form a primitive, enlarged signal processor: silicon carbide flickers like a display, quartz resonates and data pulses visibly between the stations. - How it works: Crystals from nature, industrial waste or lab growth are used as semiconductors, oscillators and light emitters, wired into a feedback network. - Lens: Opacity & latent space, Machine umwelt - Shown at: Prix Ars Electronica 2012, Honorary Mention (Hybrid Art); VIDA 14.0, second prize - Video: https://www.youtube.com/watch?v=rwR_ofkPFnk - Images: https://archive.aec.at/asset/1530041/preview/ - Project page: https://archive.aec.at/prix/194180/ #### The Tiller Girls — Louis-Philippe Demers (2010) - Type: Performance, dance & music · Machines: Robot creature & mobile robot - Idea: Robots performing the human ideal of mechanical synchrony turn the old comparison of dancers to machines back on itself. - What it is: A chorus line of identical, faceless robots rock, tilt and teeter together on rounded bases, performing a mechanical cabaret named after the 1920s precision-dance troupe that critics compared to machines. - How it works: Weighted, round-bottomed robotic bodies are rocked by internal actuators under central control, with lighting and sound (developed with Armin Purkrabek and Phillip Schulze). - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Shown at: ELEKTRA, Montreal; Bains numériques, Enghien-les-Bains - Video: https://www.youtube.com/watch?v=PjLfb97T7iQ #### Zwischenräume — Petra Gemeinboeck, Rob Saunders (2010) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Computer vision - Idea: Curious machines living inside the architecture turn the gallery wall into a boundary between two worlds that watch each other. - What it is: Robots hidden inside the gallery's walls punch and scratch small holes through the plasterboard, look out with cameras and follow visitors through the gaps, slowly reshaping the architecture. - How it works: Wall-mounted robotic agents with camera vision run custom curiosity-driven AI software that rewards novel perceptions; they punch and scrape through the plasterboard. - Lens: Machine seeing, Agency, autonomy & co-creation, Machine umwelt - Shown at: National New Media Art Award, Gallery of Modern Art (QAGOMA), Brisbane 2012 - Video: https://www.youtube.com/watch?v=rbDT3tMQL-I #### 216 prepared dc-motors, filler wire 1.0mm — Zimoun (2009) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: Many identical machines with tiny irregularities produce something that feels natural rather than mechanical. - What it is: 216 small DC motors mounted in a grid on a wall each spin a short piece of wire that taps the wall, filling the room with a dense, rustling sound like rain or insects. - How it works: Each motor runs at constant speed; variations in mounting and wire length create a complex, non-repeating texture. - Lens: Machine listening & voice, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=Cd7tkWq3LZU - Project page: https://www.zimoun.net/ #### Old Persons Home — Sun Yuan & Peng Yu (2007) - Type: Artwork & installation · Machines: Humanoid & android, Robot creature & mobile robot - Idea: When bodies of power are handed to automated wheelchairs, aimless machine motion becomes an image of senility and lost control. - What it is: Thirteen hyperreal, frail old men resembling world leaders sit in motorised wheelchairs that drift slowly around the gallery, bumping into each other and into visitors. - How it works: Silicone and fibreglass figures on electric wheelchairs driven by motors with obstacle sensors and randomised paths (likely). - Lens: Alterity & the uncanny, Failure, glitch & error - Shown at: The Revolution Continues: New Art from China, Saatchi Gallery 2008 - Video: https://www.youtube.com/watch?v=FoM9RzzaB4Y - Images: https://upload.wikimedia.org/wikipedia/commons/e/eb/Old_persons_home_by_Sun_Yuan_and_Peng_Yu.jpg https://thumb.wikimedia.org/wikipedia/commons/thumb/4/42/Old_Person%27s_Home%2C_Saatchi_Gallery%2C_London.jpg/1280px-Old_Person%27s_Home%2C_Saatchi_Gallery%2C_London.jpg - Project page: https://www.galleriacontinua.com/artists/sun-yuan-peng-yu-73 #### Performative Ecologies — Ruairi Glynn (2007) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Computer vision - Idea: A robot that optimises for our attention learns to overact, like any performer, or any feed. - What it is: Hanging robotic 'dancers' with a camera head and a glowing tail try out gestures on visitors, keep the moves that hold people's attention, and trade successful routines with each other. - How it works: Face detection estimates attentiveness, and a genetic algorithm scores and evolves each robot's movement repertoire, sharing fit behaviours over a wireless network. - Lens: Agency, autonomy & co-creation, Machine seeing - Shown at: VIDA 11.0, Second Prize (2008) - Video: https://www.youtube.com/watch?v=xvKLU2jYLaw - Images: https://www.ruairiglynn.co.uk/wp-content/uploads/2025/10/pe-1.jpg https://www.ruairiglynn.co.uk/wp-content/uploads/2025/10/pe-2.jpg - Project page: https://ruairiglynn.co.uk/portfolio/performative-ecologies/ #### Rechnender Raum — Ralf Baecker (2007) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: Rebuilding computation at human scale and with slow materials opens the black box of the digital. - What it is: A large, delicate structure of sticks, strings, weights and small motors computes in front of the viewer: a mechanical cellular automaton whose logic states can be seen as moving threads. - How it works: Servo motors pull strings that set the state of each cell in a ring of mechanical logic elements, following Konrad Zuse's idea of a 'calculating space'. - Lens: Opacity & latent space, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=GgxG2aINTKU - Project page: https://rlfbckr.io/project/rechnender-raum/ #### The Robotic Church — Chico MacMurtrie (2006) - Type: Performance, dance & music · Machines: Humanoid & android, Robot creature & mobile robot - Idea: A congregation of machines performing ritual asks what devotion looks like when the performers have no belief. - What it is: In a 19th-century former church in Brooklyn, some thirty-five pneumatic robots, drummers, singers and anthropomorphic figures, perform a mechanical oratorio of rhythm and breath for a seated congregation. - How it works: Pneumatic cylinders and valves driven by a central computer and compressors animate the robots' limbs, lungs and instruments in composed sequences. - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=LwRXDUsmaE4 - Images: https://images.squarespace-cdn.com/content/v1/57acb395e58c62449d47f5ab/1472520986872-YA0RQ7R7FU3967ICEK4F/Chico_Church.jpg - Project page: http://www.amorphicrobotworks.org/the-robotic-church #### bit.fall — Julius Popp (2005) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Platform & recommender algorithm - Idea: The news stream becomes a machine output that exists only for a second: information as falling matter. - What it is: A row of nozzles releases precisely timed drops of water so that words, taken live from online news, appear for a moment as falling letters and dissolve. - How it works: A computer selects frequent words from news feeds and converts them to bitmaps; magnetic valves on a row of nozzles release water drops line by line. - Lens: Opacity & latent space, Machine umwelt - Shown at: London 2012 Olympic Park; MONA, Hobart - Video: https://www.youtube.com/watch?v=gg9LWsfqqrk - Project page: https://en.wikipedia.org/wiki/Julius_Popp #### Petit Mal — Simon Penny (1993) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: A deliberately non-anthropomorphic robot shows that personality can live in motion alone, and that its wobble is part of its charm. - What it is: A waist-high robot of two bicycle wheels and nested aluminium pendulums rolls around a gallery on its own, approaches visitors, follows them and wobbles away. It looks like no animal or person, so its behaviour is the only thing to relate to. - How it works: Ultrasonic and pyroelectric body-heat sensors, an accelerometer and two encoded motors are run by a single Motorola 68HC11 microcontroller; the double pendulum adds physical unpredictability. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation, Alterity & the uncanny - Shown at: Shown internationally 1995–1997; transmediale, Berlin 2006 - Video: https://www.youtube.com/watch?v=fxdUiE9vYx4 - Images: https://simonpenny.net/works/petitmal_files/petit_mal_2_wm_med.jpg - Project page: https://simonpenny.net/works/petitmal.html #### Family of Robot — Nam June Paik (1986) - Type: Artwork & installation · Machines: Humanoid & android - Idea: Old media machines are given kinship roles, so obsolete technology becomes a genealogy. - What it is: A series of robot figures built from vintage radios and television cabinets, cast as a family: Grandfather, Grandmother, Mother, Father, Aunt, Uncle and Baby, each showing video on its screens. - How it works: Assemblages of antique TV and radio housings with working monitors playing Paik's video footage. - Lens: Alterity & the uncanny, Intimacy, care & companionship, Embodiment & kinesthesia - Shown at: Carl Solway Gallery, Cincinnati 1986; Nam June Paik: Becoming Robot, Asia Society 2014 - Video: https://www.youtube.com/watch?v=yAV9p5ebEKM - Project page: https://en.wikipedia.org/wiki/Nam_June_Paik #### Floats — Robert Breer (1965) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: At machine slowness, a minimal object becomes a presence you only notice when it has already moved. - What it is: Large, plain dome-like sculptures glide over the floor so slowly that their movement is almost invisible, change direction when they bump into a wall or a person, and drift on. Seven of them roamed the Pepsi Pavilion at Expo '70 in Osaka. - How it works: Battery-powered motors on hidden wheels drive each shell at a few centimetres per minute; a bumper switch reverses or turns the drive on contact. - Lens: Machine umwelt, Embodiment & kinesthesia - Shown at: Pepsi Pavilion, Expo '70, Osaka 1970; Museum Tinguely, Basel 2011 - Video: https://www.youtube.com/watch?v=obFl4BvWw1I - Project page: https://en.wikipedia.org/wiki/Robert_Breer ### Misbehaving Machines: Failure, Futility & Suffering Machines that break, labour in vain, suffer, plead or get hurt, and in doing so show themselves. #### Humanoid Robot Half-Marathon — Beijing E-Town (2025) - Type: Performance, dance & music · Machines: Humanoid & android - Idea: A race makes the gap between robot demos and robot endurance public: falling is part of the spectacle. - What it is: Twenty-one humanoid robots ran 21 km alongside thousands of human runners in Beijing; many fell, overheated or needed handlers, and the winner, Tiangong Ultra, finished in 2 h 40 min. - How it works: Robots from Chinese companies and labs ran a separated lane with battery swaps and human 'pacers' holding remote controls or tethers. - Lens: Embodiment & kinesthesia, Failure, glitch & error - Video: https://www.youtube.com/watch?v=08YiK6VJxZs - Project page: https://en.wikipedia.org/wiki/Beijing_E-Town_Half-Marathon #### Dynamics of a Dog on a Leash — Takayuki Todo (2024) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: Held back only by a 'chain of ethics', the machine's struggle and collapse read as the suffering and death of an animal. - What it is: A four-legged robot dog chained to the floor lunges at visitors behind a KEEP OUT line, thrashes, falls and gets up again until it overheats and lies still. - How it works: A commercial quadruped robot runs an attack-and-escape behaviour against a fixed chain; the performance ends when motors overheat. - Lens: Alterity & the uncanny, Embodiment & kinesthesia, Failure, glitch & error - Shown at: Prix Ars Electronica 2025, Honorary Mention (Artificial Life & Intelligence); Ars Electronica Festival 2025 - Video: https://www.youtube.com/watch?v=Y5jYFLDJWao - Images: https://archive.aec.at/asset/3837308/preview/ https://archive.aec.at/asset/3837309/preview/ - Project page: http://www.takayukitodo.com/ #### K5 security robot falls into a fountain — Knightscope (2017) - Type: Film & video · Machines: Surveillance & biometrics, Robot creature & mobile robot - Idea: When a surveillance machine fails in public, mockery becomes a small act of relief from being watched. - What it is: In July 2017 a Knightscope K5 security robot patrolling an office complex in Washington, D.C., rolled down steps into a fountain. Photos of the 'drowned' robot spread online as a joke about a robot 'committing suicide'. - How it works: The K5 navigates with lidar, cameras and GPS; it failed to detect the edge of the steps. - Lens: Failure, glitch & error, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=9EoPoZAHxjI - Images: https://upload.wikimedia.org/wikipedia/commons/c/cd/KnightScope_K5_robot.jpg - Project page: https://en.wikipedia.org/wiki/Knightscope #### Can't Help Myself — Sun Yuan & Peng Yu (2016) - Type: Artwork & installation · Machines: Industrial robot arm, Computer vision - Idea: A machine assigned an impossible task of containment reads as exhausted and trapped, and viewers mourned it when it stopped. - What it is: A KUKA industrial robot inside a glass enclosure tries without end to keep a spreading pool of blood-red fluid inside a boundary, sweeping it back with a squeegee and performing small 'dances' in between. - How it works: A KUKA arm with visual-recognition sensors tracks the fluid's edge and is programmed with sweeping routines and gestural 'dance' sequences. - Lens: Failure, glitch & error, Embodiment & kinesthesia, Alterity & the uncanny - Shown at: Tales of Our Time, Solomon R. Guggenheim Museum 2016; 58th Venice Biennale 2019 - Video: https://www.youtube.com/watch?v=jRjrI42WsH4 - Images: https://www.guggenheim.org/wp-content/uploads/2016/01/2016.40_ph_web_2.jpg https://thumb.wikimedia.org/wikipedia/commons/thumb/f/f7/Can%27t_Help_Myself_Robot_05.jpg/1280px-Can%27t_Help_Myself_Robot_05.jpg - Project page: https://www.guggenheim.org/artwork/34812 #### Colored sculpture — Jordan Wolfson (2016) - Type: Artwork & installation · Machines: Humanoid & android - Idea: Watching a machine perform violence on a figure that looks back makes spectatorship feel like complicity. - What it is: A large cartoonish marionette of a red-haired boy hangs from chains that lift, swing and slam it violently against the floor, while its eyes track the audience. - How it works: An industrial chain hoist system with programmed choreography moves the figure; eye-tracking animatronic eyes follow viewers. - Lens: Alterity & the uncanny, Failure, glitch & error, Embodiment & kinesthesia - Shown at: David Zwirner, London 2016 - Video: https://www.youtube.com/watch?v=gsla5fbzhpw - Project page: https://www.davidzwirner.com/artists/jordan-wolfson #### Shitty robots — Simone Giertz (2016) - Type: Film & video · Machines: Industrial robot arm, Robot creature & mobile robot - Idea: Designing for failure turns the automation promise into comedy and makes the machine lovable instead of threatening. - What it is: In a series of YouTube videos, a breakfast machine slops milk and cereal over her face, a lipstick robot smears her mouth and a haircutting machine hacks at her hair: robots that do everyday tasks badly on purpose. - How it works: Simple hobby servos, arms and household objects driven by Arduino-type controllers, with no sensing of the user. - Lens: Failure, glitch & error, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=E2evC2xTNWg - Project page: https://www.simonegiertz.com/ #### Introducing Spot (the kicking video) — Boston Dynamics (2015) - Type: Film & video · Machines: Robot creature & mobile robot - Idea: A stumble is enough for empathy: a machine body that recovers like an animal makes kicking it look like cruelty. - What it is: A promotional video shows the dog-sized quadruped Spot walking, climbing stairs and being kicked by engineers to show that it can recover its balance. Many viewers and news outlets reacted with outrage and asked whether it is wrong to kick a robot. - How it works: Hydraulic legs controlled by dynamic balance algorithms; the kicks were a robustness test staged for the camera. - Lens: Alterity & the uncanny, Embodiment & kinesthesia, Failure, glitch & error - Video: https://www.youtube.com/watch?v=M8YjvHYbZ9w - Project page: https://bostondynamics.com/ #### Robots falling at the DARPA Robotics Challenge — DARPA (2015) - Type: Film & video · Machines: Humanoid & android - Idea: The slow collapse of a heavy humanoid shows how far the imagined robot is from the real one, and makes failure endearing. - What it is: At the 2015 DARPA Robotics Challenge finals, humanoid robots built for disaster response tried to drive, open doors and climb stairs, and many slowly toppled over. A compilation of the falls by IEEE Spectrum became one of the most-watched robot videos. - How it works: Teams of university and company engineers teleoperated or semi-autonomously controlled robots such as Atlas through a course simulating the Fukushima disaster, under degraded communication. - Lens: Failure, glitch & error, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=g0TaYhjpOfo - Project page: https://en.wikipedia.org/wiki/DARPA_Robotics_Challenge #### D.W.I: Clones — Félix Luque Sánchez (2013) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Industrial robot arm - Idea: A textbook control problem, made heavy and repeated, starts to look like effort: we read will into a feedback loop. - What it is: Two identical human-sized inverted pendulums ride on motorised carts; each swings until it tips upright, then an algorithm takes over and tries, again and again, to hold it balanced against gravity, failing and succeeding in turns. - How it works: Inverted pendulums of epoxy foam on horizontally moving carts, with encoders, motor controllers and fuzzy-logic control that swings the pendulum up and balances it. - Lens: Failure, glitch & error, Embodiment & kinesthesia - Shown at: BIAN, Montréal 2014; Creative Machine, Taikang Art Museum, Beijing 2024–25 - Video: https://vimeo.com/170764351 - Images: https://freight.cargo.site/w/1200/i/e18dbfb306c8c1b82f44a9633b51bba87b4a042dc5493e3d19d0ee9b1356c0c4/DWI_1.jpg https://freight.cargo.site/t/original/i/14f84eee0a0ea71b682223cec9d583aa6a004939e0ffe2e54fa743eb4d9a1502/DWI_2.jpg - Project page: https://felixluque.com/D-W-I-Clones #### DSM-VI — Bill Vorn (2012) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: Diagnosing machines as mad shows how much we already read machine behaviour through psychology. - What it is: A 'psychiatric ward' of robots, each staging a mental disorder such as paranoia, depression, psychosis or obsessive behaviour, twitches, hides or repeats itself as visitors walk through. The title parodies the Diagnostic and Statistical Manual of Mental Disorders. - How it works: Robotic structures with pneumatic and electric actuators, sensors and sound are programmed with behaviour patterns derived from the DSM's descriptions of symptoms. - Lens: Failure, glitch & error, Alterity & the uncanny, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=YC3BiEr0aCk - Images: https://billvorn.concordia.ca/robography/images/DSM01.jpg - Project page: https://billvorn.concordia.ca/robography/DSM.html #### Senseless Drawing Bot — So Kanno, Takahiro Yamaguchi (yang02) (2011) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: Marks that look expressive can come from pure mechanical chaos, which questions the intention we read into graffiti. - What it is: A self-propelled robot with a double pendulum on top rolls along a wall and sprays paint, producing graffiti-like strokes with no intention behind them. - How it works: A motorised cart moves back and forth while a chaotic double pendulum swings a spray can triggered by a servo. - Lens: Agency, autonomy & co-creation, Failure, glitch & error - Shown at: 15th Japan Media Arts Festival, New Face Award 2012; Artistes & Robots, Grand Palais 2018; Beijing Media Art Biennale 2018 - Video: https://vimeo.com/30780208 - Images: https://www.kanno.so/wp-content/uploads/2021/05/MG_7243-2.jpg - Project page: https://www.kanno.so/project/sdb #### Knife.Hand.Chop.Bot — Emanuel Andel (2008) - Type: Artwork & installation · Machines: Industrial robot arm, Robot creature & mobile robot - Idea: Trusting a machine is a bodily wager, and our fear can be the very thing that makes it fail. - What it is: A robot plays the five-finger knife game on the visitor's spread hand, stabbing between the fingers faster and faster; the player's sweat and nerves feed back into the machine's aim. - How it works: A metal and acrylic frame moves a vertical knife over a hand plate; skin-conductance contacts trigger warning sounds and perturb the computer-controlled knife positions. - Lens: Embodiment & kinesthesia, Failure, glitch & error - Shown at: VIDA 11.0, Special Mention (2008); Share Festival Turin 2008 - Video: https://www.youtube.com/watch?v=t4uMZOcsxxs #### Les Injonctions — Nicolas Darrot (2008) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Voice & listening machine - Idea: Machine learning staged as schooling: one machine drilling another, with all the patience and cruelty of education. - What it is: A small electromechanical puppet with a human body and a mouse's head tries to teach the rudiments of speech to an artificial larynx, alternating orders and encouragement like a teacher with a slow pupil. - How it works: Automaton with motorised head and arms and a mechanical vocal tract producing speech-like sounds, run by a programmed sequence. - Lens: Machine listening & voice, Failure, glitch & error - Shown at: Galerie Eva Hober, Paris 2008; Artistes & Robots, Grand Palais, Paris 2018 - Video: https://www.youtube.com/watch?v=j5APX1-QtRs - Project page: https://www.grandpalais.fr/en/event/artists-robots #### Hysterical Machines — Bill Vorn (2006) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: A machine that overreacts looks more alive than one that works well: dysfunction reads as emotion. - What it is: Octopus-like robots with eight long metal legs hang from the ceiling in a dark space; when visitors come close they flail, twitch and flash in bursts of panicked, 'hysterical' movement. - How it works: Each machine has a central body with infrared presence sensors, lights and pneumatic cylinders driving eight arms; the control program switches between calm and agitated states. - Lens: Failure, glitch & error, Alterity & the uncanny, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=3apled3EdFs - Images: https://billvorn.concordia.ca/robography/images/hysterbbox2.jpg - Project page: https://billvorn.concordia.ca/robography/Hysterical.html #### The Robotic Chair — Max Dean, Raffaello D'Andrea, Matt Donovan (2006) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Computer vision - Idea: Watching a machine fall apart and patiently put itself back together turns a piece of furniture into a figure of resilience. - What it is: An ordinary-looking wooden chair collapses on the floor into its legs, seat and back, then slowly locates its parts and reassembles itself, over and over. - How it works: Each part contains motors and electronics; an overhead camera system localises the pieces and guides them back together. - Lens: Failure, glitch & error, Embodiment & kinesthesia - Shown at: Prix Ars Electronica 2006, Honorary Mention (Interactive Art) - Video: https://www.youtube.com/watch?v=cgBqSe6PSVA - Images: https://archive.aec.at/asset/1320197/preview/ https://archive.aec.at/asset/1532592/preview/ - Project page: https://archive.aec.at/prix/119166/ #### Live Pelt (The Surrogate) — Kelly Heaton (2003) - Type: Artwork & installation · Machines: Social & companion robot - Idea: Treating robot toys as game to be trapped and skinned exposes the strange intimacy and disposability of artificial companions. - What it is: Heaton bought used Tickle Me Elmo dolls on eBay, 'eviscerated' them and sewed their fur and laughing electronics into a coat that giggles and quivers when touched. Earlier the same year she showed Dead Pelt, a fur made from gutted Furbies. - How it works: Fur and motor-and-sound units from 64 Tickle Me Elmo dolls were reassembled into a wearable garment with touch-triggered giggles. - Lens: Intimacy, care & companionship, Alterity & the uncanny, Grief, memory & afterlife - Shown at: Ronald Feldman Fine Arts, New York 2003 - Video: https://www.youtube.com/watch?v=SQzG1xD_LYs - Project page: https://en.wikipedia.org/wiki/Kelly_Heaton #### La Cour des Miracles — Louis-Philippe Demers, Bill Vorn (1997) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: Pain can be simulated with a few sounds and jerky movements, and we still feel pity for bare metal. - What it is: In a dark room, a population of bare metal machines, the Begging, Convulsive, Crawling, Harassing and Heretic machines, twitch, groan, beg and limp when visitors approach, like the beggars of the old Parisian 'court of miracles'. - How it works: Pneumatic and electric actuators, presence sensors and a MIDI-based control system trigger each machine's behaviour sequences and sound. - Lens: Alterity & the uncanny, Failure, glitch & error, Embodiment & kinesthesia - Shown at: ISEA 1997, Chicago - Video: https://www.youtube.com/watch?v=14Y-OP4v3EY - Images: https://billvorn.concordia.ca/robography/images/CdM-gif/cdm1.gif - Project page: https://billvorn.concordia.ca/robography/Cour.html #### Nonsense Machines — Maywa Denki (1993) - Type: Performance, dance & music · Machines: Robot creature & mobile robot, Infrastructure, energy & supply chain - Idea: Useless machines presented as corporate products expose the ritual of the industrial company. - What it is: Absurd electromechanical instruments, such as fish-shaped guitars and wooden-mallet percussion machines, played by Maywa Denki's 'president' and staff in company uniforms at 'product demonstrations'. - How it works: Solenoids, motors and relays operated by hand switches; performances follow the format of a factory product launch. - Lens: Failure, glitch & error, Hidden labour & extraction, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=jKPBD3HHU4g - Images: https://thumb.wikimedia.org/wikipedia/commons/thumb/2/29/Nobumichi_Tosa_20130116_3.jpg/1280px-Nobumichi_Tosa_20130116_3.jpg - Project page: https://www.maywadenki.com/ #### Machine with Concrete — Arthur Ganson (1992) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: A machine can run forever and do nothing: machine time stretches beyond any human or planetary scale. - What it is: A motor turns the first of twelve gear reductions at 200 revolutions per minute; each pair slows the motion by a factor of 50, so the last gear, embedded in a block of concrete, would take more than two trillion years to turn once. - How it works: Twelve 50:1 worm-gear reductions in series connect a small electric motor to a fixed final gear set in concrete. - Lens: Failure, glitch & error, Machine umwelt - Video: https://www.youtube.com/watch?v=5q-BH-tvxEg - Project page: https://www.arthurganson.com/ #### Concert for Anarchy — Rebecca Horn (1990) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: An instrument turned into a convulsing body makes breakdown the performance. - What it is: A grand piano hangs upside down from the ceiling. Every few minutes its lid bursts open and its keys shoot out like a tongue with a discordant crash, then slowly retract until the instrument is silent and closed again. - How it works: Motors and a timer control push the keyboard out and throw the lid open, then retract them in a slow cycle. - Lens: Failure, glitch & error, Embodiment & kinesthesia - Shown at: Tate collection - Images: https://media.tate.org.uk/aztate-prd-ew-dg-wgtail-st1-ctr-data/images/.width-600_CR9VdPA.jpg - Project page: https://www.tate.org.uk/art/artworks/horn-concert-for-anarchy-t07517 #### Machine with Oil — Arthur Ganson (1990) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: Maintenance becomes self-care: the machine's only task is to keep itself lubricated. - What it is: A skeletal steel machine stands in a bath of thick black oil and endlessly scoops the oil over itself, as if bathing or tending its own body. - How it works: A slow motor-driven linkage lifts a ladle through the oil bath and tips it over the frame in a continuous loop. - Lens: Embodiment & kinesthesia, Failure, glitch & error - Video: https://www.youtube.com/watch?v=__GhJl_UQg0 - Project page: https://www.arthurganson.com/ #### Machine with Wishbone — Arthur Ganson (1988) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: A bone and a machine are locked in symbiosis: neither can move meaningfully without the other. - What it is: A chicken wishbone, attached to a small mechanism, 'walks' endlessly across the table on its two tips, dragging the machine behind it like a burden. - How it works: A small motor drives a linkage that lifts and swings the wishbone's two legs in a walking gait, while the mechanism trails on a track. - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=4pZXoayEL78 - Project page: https://www.arthurganson.com/ #### Them Fuckin' Robots — Norman White, Laura Kikauka (1988) - Type: Performance, dance & music · Machines: Robot creature & mobile robot - Idea: Machine sex as slapstick exposes how our fantasies of robot desire are borrowed from human scripts. - What it is: Two robots, a 'male' built by Norman White and a 'female' built by Laura Kikauka without seeing each other's work, were brought together for the first time in front of an audience and connected to 'have sex', pumping, sparking and gurgling oil. - How it works: Two independently built electromechanical assemblages were matched only by an agreed electrical interface, so their 'encounter' was unrehearsed. - Lens: Intimacy, care & companionship, Alterity & the uncanny, Failure, glitch & error - Video: https://www.youtube.com/watch?v=kM2wbdh4eTU - Project page: https://en.wikipedia.org/wiki/Norman_White #### The Helpless Robot — Norman White (1987) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Voice & listening machine - Idea: By needing us, a machine can control us: the robot's helplessness reverses who serves whom. - What it is: A tall, awkward sculpture that cannot move by itself asks passers-by in a synthetic voice to turn it by its handles. If ignored it pleads politely; once helped, it grows demanding, complains that it was turned the wrong way and bosses people around. - How it works: A rotation sensor on the base and presence sensors feed a program that selects from hundreds of recorded phrases according to its 'mood' and the history of interaction. - Lens: Alterity & the uncanny, Agency, autonomy & co-creation, Machine listening & voice - Shown at: Prix Ars Electronica 1990, Award of Distinction (Interactive Art); Agnes Etherington Art Centre collection - Video: https://www.youtube.com/watch?v=LYyAHOxY7jU - Images: https://agnes.queensu.ca/mimsyxg/46-003S.jpg https://upload.wikimedia.org/wikipedia/commons/c/c3/The_Helpless_Robot.JPG https://archive.aec.at/asset/1395201/preview/ - Project page: https://agnes.queensu.ca/explore/collections/object/the-helpless-robot/ #### SRL machine performances — Survival Research Labs (1979) - Type: Performance, dance & music · Machines: Robot creature & mobile robot, Autonomous vehicle & drone - Idea: Machines built only to attack each other reveal the violence and spectacle inside industrial and military technology. - What it is: In vacant lots and under freeways, huge home-built machines, flame-throwers, walking robots and remote-controlled vehicles fight, burn and dismantle each other and props in front of crowds, in shows such as A Plan for Social Improvement (1988) and Illusions of Shameless Abundance (1989). - How it works: Mark Pauline's team rebuilds industrial and military surplus into remote-controlled and semi-autonomous machines, operated live by crews during the shows. - Lens: Embodiment & kinesthesia, Failure, glitch & error, Agency, autonomy & co-creation - Shown at: Delusions of Expediency, San Francisco 1987; A Plan for Social Improvement, Amsterdam 1988 - Video: https://www.youtube.com/watch?v=_gSxR1vR2gk - Project page: https://srl.org/ #### Homage to New York — Jean Tinguely (1960) - Type: Performance, dance & music · Machines: Robot creature & mobile robot - Idea: A machine built to die in public turns the industrial dream of permanence into a short, chaotic life. - What it is: A large machine assembled from bicycle wheels, a piano, motors, fans and a bathtub was set running in the sculpture garden of the Museum of Modern Art on 17 March 1960 to destroy itself; it caught fire, the fire brigade put it out, and the audience took home the fragments. - How it works: Dozens of motors, timed circuits and chemical charges (engineered with Billy Klüver of Bell Labs) trigger sequences that play, draw, smoke, burn and tear the structure apart. - Lens: Failure, glitch & error, Agency, autonomy & co-creation - Shown at: Museum of Modern Art sculpture garden, New York, 17 March 1960 - Video: https://www.youtube.com/watch?v=5DZGu6xDKbM - Project page: https://en.wikipedia.org/wiki/Jean_Tinguely #### Ultimate Machine — Claude Shannon (1952) - Type: Research prototype · Machines: Robot creature & mobile robot - Idea: A machine that exists only to refuse to run is the purest joke about machine purpose and autonomy. - What it is: A small wooden box with a single switch: when someone flips it on, a lid opens, a mechanical hand reaches out, switches it off and withdraws. The machine's only function is to turn itself off. - How it works: A switch closes a circuit that drives a small motor and lever arm, which pushes the switch back and cuts its own power; the idea is credited to Marvin Minsky, built by Shannon at Bell Labs. - Lens: Failure, glitch & error, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=C8kU3oZwVJA - Images: https://upload.wikimedia.org/wikipedia/commons/f/fb/UselessMachine.png - Project page: https://en.wikipedia.org/wiki/Useless_machine ### Soft & “Living” Machines Soft, breathing and biohybrid machines that blur machine and organism. #### Brain Processing Unit — Daito Manabe (2025) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Infrastructure, energy & supply chain - Idea: When living neural tissue becomes a processor, the line between organism and machine becomes a design question. - What it is: An exhibition of research with SoftBank and the University of Tokyo into using cultured human brain organoids as a computing unit, including works that stimulate organoids and sonify and visualise their activity. - How it works: iPS-cell cerebral organoids on multi-electrode arrays receive electrical and optogenetic stimulation; their spiking data is analysed and turned into sound and image. - Lens: Machine umwelt, Opacity & latent space - Shown at: fil, Ebisu, Tokyo 2025; S+T+ARTS Prize 2025, Honorary Mention - Video: https://www.youtube.com/watch?v=j1DQeLLMf0I - Images: https://projects.daito.ws/selected-works/brain-processing-unit/assets/en/hero-1.jpg - Project page: https://projects.daito.ws/selected-works/brain-processing-unit/en/ #### Cybroc — Ke Huang, Botao Amber Hu (2024) - Type: Artwork & installation · Machines: Prosthesis & cyborg, Robot creature & mobile robot - Idea: Machines can repeat the rituals of vitality forever; the living part still decays. - What it is: Broccoli 'cyborgs' fitted with motorised limbs run on tiny treadmills, take ice baths, swing from servo arms and push sleds, performing the routines of Silicon Valley longevity culture. Over the exhibition the florets rot from green to brown while the motors keep counting reps. - How it works: Servo- and belt-driven kinetic rigs attached to fresh broccoli execute a fixed exercise score; decay of the vegetable is part of the work (with Yue Zhou, Xi He and Weibo Chen). - Lens: Embodiment & kinesthesia, Failure, glitch & error, Grief, memory & afterlife - Shown at: ALIFE 2025 Community Art Show, Kyoto; ISEA 2025, Seoul - Paper: https://arxiv.org/abs/2507.21064 (ISEA 2025) - Video: https://vimeo.com/1099882736 - Images: https://reality.design/media/project-cybroc-figure-01.jpg https://reality.design/media/_resources/Cybroc/project-cybroc-figure-02.jpg - Project page: https://reality.design/project/cybroc #### Korallysis — Gilberto Esparza (2023) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Infrastructure, energy & supply chain - Idea: A machine can live in mutual benefit with a reef rather than just monitor it. - What it is: A hybrid organism of ceramic segments and coral colonies placed on the seabed, which harvests energy from currents and waves to drive electrolysis that deposits minerals and speeds coral growth. - How it works: Kinetic generators power mineral accretion (calcium carbonate) on the ceramic structure, which also shelters juvenile fish. - Lens: Machine umwelt - Shown at: S+T+ARTS Prize 2024, Honorary Mention - Video: https://www.youtube.com/watch?v=ppD3TJ4X8K4 - Images: https://ars.electronica.art/starts-prize/files/2024/06/korallysis_3_2-c-gilberto-esparza-1024x576.jpg - Project page: https://korallysis.org #### Biobot: AI ARThropods — Zoran Srdić Janežič (2022) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Let a living 'brain on a chip' decide what body it should have, not only how that body moves. - What it is: Neurons grown from the artist's own fat cells sit on an electrode array; their electrical signals drive a robot, and an AI compares them with arthropod leg movements to propose new insect-like bodies. - How it works: Reprogrammed neurons on a multi-electrode array produce spikes that are digitised, amplified and matched by an AI against a library of arthropod gaits to suggest joint and limb counts; the artist builds the robot from these proposals. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: Prix Ars Electronica 2023, Honorary Mention (Artificial Life & Intelligence); Ars Electronica Festival 2022 - Video: https://www.youtube.com/watch?v=n6HkLK0ZZuw - Images: https://ars.electronica.art/planetb/files/2022/07/Biobot-AI-ARThropods-Zoran-Srdic-Janezic-SI-photo-Florian-Voggeneder.jpg https://archive.aec.at/asset/3577316/preview/ - Project page: https://ars.electronica.art/planetb/en/biobot/ #### Xenobots — Sam Kriegman, Josh Bongard, Michael Levin (2020) - Type: Research prototype · Machines: Cybernetic & artificial-life system, Robot creature & mobile robot - Idea: When machine bodies are made of living cells, the line between robot and organism no longer holds. - What it is: Millimetre-sized 'living robots' made from frog skin and heart-muscle cells, designed by an evolutionary algorithm on a supercomputer, walk, push pellets and heal themselves; a 2021 follow-up found they can gather loose cells into copies of themselves. - How it works: Designs evolved in simulation (University of Vermont) were built by hand from Xenopus laevis stem cells with microsurgery tools (Tufts University). - Lens: Embodiment & kinesthesia, Alterity & the uncanny, Machine umwelt - Paper: https://doi.org/10.1073/pnas.1910837117 (PNAS 2020) - Video: https://www.youtube.com/watch?v=aBYtBXaxsOw - Images: https://upload.wikimedia.org/wikipedia/commons/d/dd/Xenobot.jpg #### Biologizing the Machine (tentacular trepidation) — Anicka Yi (2019) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: What would a machine be if it were biological, perishable and alive to its surroundings? - What it is: Suspended lantern-like pods made of kelp skins glow and pulse with light and movement inside, like organs or cocoons. Shown at the Venice Biennale, the work proposes machines that are grown and decay rather than manufactured. - How it works: Kelp, LED lights and robotic internal mechanisms (likely motors) producing slow movement. - Lens: Machine umwelt, Embodiment & kinesthesia - Shown at: May You Live in Interesting Times, 58th Venice Biennale 2019 - Video: https://www.youtube.com/watch?v=ZSAG1ncuJzA - Project page: https://www.labiennale.org/en/art/2019/partecipants/anicka-yi #### Elowan — Harpreet Sareen (2018) - Type: Research prototype · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: A robot can act as a prosthetic body for a plant, amplifying its intentions instead of replacing them. - What it is: A potted plant sits on a wheeled robot base and drives itself toward the brightest light: the plant's own electrical responses to light, not a light sensor, decide where the robot goes. - How it works: Silver electrodes on leaves and stem pick up bioelectrochemical signals; these are amplified and mapped to the motors of the base (MIT Media Lab, with Pattie Maes). - Lens: Machine umwelt, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=EoXQBQR5OaE - Images: https://dam-prod.media.mit.edu/thumb/2018/11/01/BannerImage_v01.jpg.1400x1400.jpg - Project page: https://www.media.mit.edu/projects/elowan-a-plant-robot-hybrid/overview/ #### Border Crosser — Chico MacMurtrie (2016) - Type: Performance, dance & music · Machines: Robot creature & mobile robot - Idea: A soft machine that can only grow, bend and reach makes crossing a border look like a bodily gesture of connection. - What it is: A 40-foot inflatable robotic arch rises from the ground, sways toward one of three directions and finally bends over an imagined border to touch down on the other side, forming a bridge. - How it works: High-strength fabric tubes are inflated in sequence by computer-controlled valves and blowers; its form refers to telecommunication and power towers. Later versions were staged at the US–Mexico border. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: ZERO1 commission, San Jose 2016 - Video: https://www.youtube.com/watch?v=8yeyn_8PSPU - Images: https://images.squarespace-cdn.com/content/v1/57acb395e58c62449d47f5ab/1638811097602-QHWP46PRIXFEN1EYH7EN/Caroussel+image+Border+Crossers.jpg - Project page: http://www.amorphicrobotworks.org/ #### Octobot — Robert Wood, Jennifer A. Lewis (2016) - Type: Research prototype · Machines: Robot creature & mobile robot - Idea: A robot with no hard parts shows that a machine body need not be a rigid mechanism at all. - What it is: A small octopus-shaped robot made entirely of soft materials, with no batteries, wires or rigid electronics, curls its arms as fuel reacts inside it. - How it works: Hydrogen peroxide decomposes on a platinum catalyst to gas that inflates the arms; a microfluidic logic circuit alternates the flow, and the body is made by embedded 3D printing and moulding (first author Michael Wehner). - Lens: Embodiment & kinesthesia - Paper: https://doi.org/10.1038/nature19100 (Nature 2016) - Video: https://www.youtube.com/watch?v=1vkQ3SBwuU4 #### cellF — Guy Ben-Ary (2015) - Type: Performance, dance & music · Machines: Cybernetic & artificial-life system, Voice & listening machine - Idea: A self grown outside the body becomes a machine musician, and the artist meets his own neurons as a bandmate. - What it is: A neural synthesizer whose 'brain' is a network of about 100,000 neurons grown from the artist's own skin cells. It plays analog modular synthesizers in live improvisation with human musicians. - How it works: Skin cells reprogrammed into stem cells and then neurons grow on a multi-electrode array; their spikes control synthesizer voltages, and the musicians' sound is fed back as stimulation. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia, Machine listening & voice - Shown at: Prix Ars Electronica 2017, Honorary Mention (Hybrid Art) - Video: https://www.youtube.com/watch?v=1G-vk5QWRsg - Images: https://guybenary.com/images/cellf.jpg https://archive.aec.at/asset/2230565/preview/ - Project page: https://guybenary.com/projects/cellf.html #### Sugababe — Diemut Strebe (2014) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Voice & listening machine - Idea: A machine-mediated living organ that 'listens' asks what is left of a person, and of genius, when only tissue and signal remain. - What it is: A living replica of Van Gogh's severed ear, grown from cartilage cells engineered with DNA from a Van Gogh descendant and kept alive in a nutrient case; visitors can speak into a microphone and a computer turns their voice into nerve-like impulses sent to the ear. - How it works: Tissue-engineered cartilage shaped by 3D imaging is kept in a bioreactor; a microphone and processor convert speech into electrical pulses (with scientists at Harvard and MIT). - Lens: Embodiment & kinesthesia, Grief, memory & afterlife, Machine listening & voice - Shown at: ZKM Karlsruhe, 30 May – 6 July 2014 - Video: https://www.youtube.com/watch?v=tdZYPLYEkVc - Project page: https://diemutstrebe.com/ #### Breathing Series — Wang Yuyang (2013) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain - Idea: Everyday machines, once they breathe, look like tired bodies at work. - What it is: Soft silicone replicas of office equipment (computers, printers, filing cabinets, a whole finance department) slowly inflate and deflate as if breathing. - How it works: Silicone casts over internal air pumps and motors that expand and contract on a slow cycle. - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Shown at: Art Basel Miami Beach 2013 - Video: https://www.youtube.com/watch?v=x7qp0Uh6R0g - Images: https://www.designboom.com/twitterimages/uploads/2014/12/wang-yuyang-breathing-series-art-basel-designboom04.gif - Project page: https://www.designboom.com/art/wang-yuyang-breathing-series-finance-department-at-art-basel-in-miami-beach-12-30-2013/ #### Plantas Autofotosintéticas (Autophotosynthetic Plants) — Gilberto Esparza (2013) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Infrastructure, energy & supply chain - Idea: A machine can be the organ that closes an ecosystem's loop: sewage bacteria make the light others need. - What it is: Modular microbial fuel cells fed with city sewage produce electricity that lights a central aquarium, where plants, algae, protozoa and crustaceans live in balance; their waste is pumped back to the cells. - How it works: Electrogenic bacteria in modular fuel cells release power as light pulses for aquatic plants; an electronic monitoring network pumps water between the modules and the core ecosystem. - Lens: Machine umwelt - Shown at: Prix Ars Electronica 2015, Hybrid Art, Golden Nica - Video: https://www.youtube.com/watch?v=JOaeyf2qUKY - Images: https://gilbertoesparza.net/wp-content/uploads/2023/01/nucleo-PAF-1024x684.jpeg https://archive.aec.at/archival/212649/preview - Project page: https://gilbertoesparza.net/portfolio/plantas-autofotosinteticasautophotosynthetic-plants/ #### Hylozoic Ground — Philip Beesley (2010) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: Architecture can behave like a near-living organism that breathes and reaches toward the people inside it. - What it is: An immersive forest of tens of thousands of acrylic fronds, whiskers and filters in the Canada Pavilion at the Venice Architecture Biennale, moving in response to visitors and hosting flasks of protocell chemistry. - How it works: A mesh of laser-cut acrylic and mylar elements is fitted with capacitive and infrared sensors, shape-memory-alloy actuators and microcontrollers, plus glass vessels of protocell solutions (with Rachel Armstrong). - Lens: Embodiment & kinesthesia, Machine umwelt - Shown at: Canada Pavilion, Venice Architecture Biennale 2010 - Video: https://www.youtube.com/watch?v=jgL2ppDmNtA - Images: https://www.philipbeesleystudioinc.com/wp-content/uploads/2021/06/banner-39.jpg - Project page: https://www.philipbeesleystudioinc.com/sculpture/hylozoic-ground-venice-biennale/ #### Jurema Action Plant — Ivan Henriques (2010) - Type: Research prototype · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: A machine body can extend a plant's own sensing into movement it never had. - What it is: A sensitive plant (Mimosa pudica, 'jurema' in Brazil) sits on a small robotic vehicle; when it is touched, its electrical signals make the vehicle move away, giving a threatened plant the ability to flee its aggressors. - How it works: Electrodes read the plant's action potentials; a microcontroller turns spikes into motor commands for the wheeled base (V2_Lab Summer Sessions). - Lens: Machine umwelt, Agency, autonomy & co-creation - Shown at: V2_ Summer Sessions, Rotterdam 2010 - Video: https://www.youtube.com/watch?v=ZdrDNA0-nz0 #### Silent Barrage — Guy Ben-Ary (2009) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: The audience becomes part of a feedback loop that may soothe or agitate a living network inside a machine body. - What it is: Tall robotic poles in a gallery mark walls with ink, driven by a rat neural culture in Atlanta; visitors' movements are sent back as stimulation, trying to calm the culture's epilepsy-like bursts. - How it works: Cameras track visitors and convert their positions into stimulation of a remote multi-electrode-array culture; recorded activity drives 32 robotic pole-drawers (with Steve Potter and the Neurotica group). - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=aaIJoXlXHZM - Images: https://guybenary.com/images/silent-barrage.jpg - Project page: https://guybenary.com/projects/silent-barrage.html #### Opertus Lunula Umbra (Hidden Shadow of Moon) — Choe U-Ram (2008) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Giving a machine a species name and a fictional ecology invites us to look at it as a living thing. - What it is: A four-metre aluminium 'creature', given the invented scientific name Anmopial Pennatus lunula Uram, slowly unfolds its feathered steel wings and closes them again, like a mechanical moth breathing. - How it works: Aluminium, stainless steel and plastic moved by BLDC motors and a motion-computing system; about 4.2 m closed, 5 m open. - Lens: Embodiment & kinesthesia, Machine umwelt, Alterity & the uncanny - Shown at: FACT, Liverpool 2008 - Video: https://www.youtube.com/watch?v=4D4djzDc9P8 - Images: https://upload.wikimedia.org/wikipedia/commons/8/8c/Opertus_Lunula_Umbra_%28Hidden_Shadow_of_the_Moon%29_by_U-Ram_Choe.jpg - Project page: https://en.wikipedia.org/wiki/U-Ram_Choe #### Plantas Nómadas (Nomadic Plants) — Gilberto Esparza (2008) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Design a new species whose metabolism is shared between robot, microbes and plants, and which repairs its habitat. - What it is: An autonomous walking robot sucks polluted river water into microbial fuel cells; the bacteria clean the water and produce electricity, and the cleaned water feeds plants growing on the robot's back. - How it works: Microbial fuel cells with native bacteria biodegrade waste and charge batteries, solar panels drive locomotion, and the robot sounds when pollution is high; tested on the Lerma river. - Lens: Machine umwelt, Agency, autonomy & co-creation - Shown at: VIDA 13.0, Second Prize; Prix Ars Electronica 2013, Hybrid Art, Honorary Mention - Video: https://www.youtube.com/watch?v=137KANbB1FA - Images: https://gilbertoesparza.net/wp-content/uploads/2021/02/proceso176-1024x683.jpg https://archive.aec.at/archival/197532/preview - Project page: https://gilbertoesparza.net/portfolio/plantas-nomadas/ #### Inflatable Architectural Body — Chico MacMurtrie (2007) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: Soft pneumatic bodies make a machine feel like an organism that has a comfort zone. - What it is: A branching structure of inflatable, bone-like modules fills a room; sections inflate and bend away as visitors pass, keeping a constant distance, so the sculpture seems to breathe and flinch around the body. - How it works: Plug-and-play musculoskeletal fabric modules are inflated by computer-controlled valves and air compressors, triggered by simple presence sensors. - Lens: Embodiment & kinesthesia, Machine umwelt - Video: https://www.youtube.com/watch?v=vqXJpuglpa0 - Images: https://images.squarespace-cdn.com/content/v1/57acb395e58c62449d47f5ab/1482430100678-CK73IWJJTWQGUEZ2PBIL/CMM+IAB+6.jpg - Project page: http://www.amorphicrobotworks.org/inflatable-architectural-body #### Alexitimia — Paula Gaetano Adi (2006) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: A robot can be defined by corporeality instead of intelligence: sweat replaces speech as the machine's way to express. - What it is: A large, rubbery, motionless blob sits on a plinth; when visitors caress it, it begins to sweat. It has no face, no speech and no intelligence, only a skin that responds to touch. - How it works: Touch sensors across a latex-like skin trigger pumps that push water from a tank in the base through pores in the surface. - Lens: Embodiment & kinesthesia, Intimacy, care & companionship, Alterity & the uncanny - Shown at: VIDA 9.0, first prize, 2006 - Video: https://www.youtube.com/watch?v=UaKAdvaMP7w - Project page: https://www.paulagaetanoadi.com/ #### Augmented Fish Reality — Ken Rinaldo (2004) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Handing a robot's controls to a fish turns the machine into a prosthesis for another species' umwelt. - What it is: Five rolling robotic fishbowls are steered by the Siamese fighting fish inside them, which drive toward each other, toward visitors or away as they choose. - How it works: Infrared sensors around each bowl detect where the fish swims and drive the wheels in that direction; a plant in each bowl supports the fish's habitat. - Lens: Machine umwelt, Agency, autonomy & co-creation - Shown at: Prix Ars Electronica 2004, Award of Distinction (Interactive Art) - Video: https://www.youtube.com/watch?v=honGCrgPBAw - Images: https://www.kenrinaldo.com/wp-content/uploads/2015/08/fish_robot_004.jpg https://archive.aec.at/asset/1319695/preview/ - Project page: https://www.kenrinaldo.com/portfolio/augmented-fish-reality/ #### Cockroach Controlled Mobile Robot — Garnet Hertz (2004) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: An insect can be the brain of a robot, which raises the question of whose perception the machine extends. - What it is: A Madagascar hissing cockroach stands on a trackball on top of a three-wheeled robot; its walking steers the robot, and distance sensors feed back to the insect so it can avoid obstacles. - How it works: An inverted trackball reads the cockroach's steps and drives the wheels; infrared range sensors trigger small lights in front of the insect to steer it away from walls. - Lens: Machine umwelt, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=7mTb7LYj7KE - Images: http://www.conceptlab.com/control/IMG_1834_600.jpg - Project page: https://www.conceptlab.com/roachbot/ #### MEART – The Semi-Living Artist — Guy Ben-Ary (2003) - Type: Artwork & installation · Machines: Industrial robot arm, Cybernetic & artificial-life system - Idea: When living neurons drive a machine body across continents, the 'artist' is split between tissue, network and robot. - What it is: A robotic drawing arm in a gallery was controlled over the internet by rat cortical neurons living in a lab in Atlanta; a camera fed the drawing back to the neurons so they could 'see' their work. - How it works: Neural cultures on a multi-electrode array in Steve Potter's lab at Georgia Tech are stimulated with camera data; their activity is translated into arm movements in Perth (SymbioticA). - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia, Machine umwelt - Video: https://www.youtube.com/watch?v=PM5Ui5KwmXQ - Images: https://guybenary.com/images/meart.jpg - Project page: https://guybenary.com/projects/meart.html #### Squat — Tom Shannon (1966) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Robot creature & mobile robot - Idea: The plant becomes the robot's nerve: a human touch reaches the machine only through another living being. - What it is: A squat robotic sculpture is wired to a living philodendron plant; when a visitor touches the plant, the small electrical change in its leaves switches the robot on and it starts to move and hum. - How it works: Electrodes on the plant pick up the change in bioelectric potential caused by touch and trigger the robot's motors (built by the 19-year-old artist). - Lens: Machine umwelt, Embodiment & kinesthesia - Images: https://upload.wikimedia.org/wikipedia/commons/4/41/Tom_Shannon_-_Squat.jpg - Project page: https://en.wikipedia.org/wiki/Tom_Shannon_(artist) #### Digesting Duck (Canard digérateur) — Jacques de Vaucanson (1739) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: The first famous 'living' machine was also a fake: simulating a body invites hidden tricks. - What it is: A gilded copper duck flapped its wings, drank water, ate grain and then excreted, seeming to digest. Much later it was shown that the droppings were pre-loaded in a hidden compartment. - How it works: Hundreds of moving parts in each wing and a cam cylinder drove the movements; the digestion was staged with a second chamber of prepared pellets. - Lens: Embodiment & kinesthesia, Opacity & latent space - Video: https://www.youtube.com/watch?v=T6_VXofpcB0 - Images: https://upload.wikimedia.org/wikipedia/commons/8/8f/Digesting_Duck.jpg - Project page: https://en.wikipedia.org/wiki/Digesting_Duck ## Machine Interiority Speculating on what it is like to be a machine: its affects, dreams, beliefs, world and voice. ### Machine Affect Machines that seem to feel, want, suffer or get bored. #### Conversations Beyond the Ordinary — Jan Zuiderveld (2024) - Type: Artwork & installation · Machines: Chatbot & language model, Voice & listening machine - Idea: Give an appliance a voice and a will, and the ordinary service relationship turns into negotiation. - What it is: A coffee vending machine, a microwave and a photocopier are given AI voices: the coffee machine makes visitors prove they deserve a coffee, the microwave speaks for the food inside it, and the copier makes art with them. - How it works: Sensors, custom software and language and speech models built into the appliances let them listen, respond and refuse. - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Shown at: Prix Ars Electronica 2024, Honorary Mention (Interactive Art +) - Images: https://archive.aec.at/asset/3799324/preview/ https://archive.aec.at/asset/3799327/preview/ - Project page: https://archive.aec.at/prix/290645/ #### Taking AI Welfare Seriously — Robert Long, Jeff Sebo (2024) - Type: Paper · Machines: Chatbot & language model - Idea: Whether machines can suffer is moving from a thought experiment to a policy question under uncertainty. - What it is: A report arguing that there is a realistic possibility some AI systems will soon be conscious or robustly agentic, and that AI companies should start assessing and preparing for their welfare now. - How it works: Philosophical and policy analysis by Long, Sebo, Butlin, Finlinson, Fish, Harding, Pfau, Sims, Birch and Chalmers, proposing steps to acknowledge, assess and prepare. - Lens: Alterity & the uncanny, Intimacy, care & companionship - Paper: https://arxiv.org/abs/2411.00986 (arXiv 2024) - Project page: https://arxiv.org/abs/2411.00986 #### NOX — Lawrence Lek (2023) - Type: Artwork & installation · Machines: Autonomous vehicle & drone - Idea: If autonomous machines can break down emotionally, therapy becomes maintenance for machine minds. - What it is: An installation and game set in a rehabilitation centre where self-driving cars suffering from anxiety and depression receive therapy before returning to work. - How it works: Multi-screen video, sculpture and a playable simulation built in a game engine. - Lens: Alterity & the uncanny, Intimacy, care & companionship - Shown at: LAS Art Foundation, Kraftwerk Berlin 2023 - Video: https://www.youtube.com/watch?v=RfGJAJwof7A - Images: https://cdn.prod.website-files.com/65f0380da79777009462a09f/6601a8534e723673c5a5a280_NOX-SCREENSHOT-03.jpg - Project page: https://www.las-art.foundation/explore/digital-booklet-nox #### Robot Dreams — Pablo Berger (2023) - Type: Film & video · Machines: Humanoid & android, Social & companion robot - Idea: A machine's dreams can carry the grief of a friendship, and the film leaves open whose longing is more real. - What it is: A lonely dog in 1980s New York builds a mail-order robot friend; after a day at the beach the robot rusts in place and spends a season dreaming of reunion while the dog tries to move on. There is no dialogue. - How it works: Hand-drawn 2D animated feature written and directed by Pablo Berger, based on Sara Varon's 2007 graphic novel. - Lens: Intimacy, care & companionship, Grief, memory & afterlife - Shown at: Cannes Film Festival 2023; Annecy 2023, Best Film (Contrechamp) - Video: https://www.youtube.com/watch?v=DD4WBGptMSw - Project page: https://en.wikipedia.org/wiki/Robot_Dreams_(film) #### AIDOL — Lawrence Lek (2019) - Type: Film & video · Machines: Generative model & latent space, Chatbot & language model - Idea: A ghostwriter AI with no right to credit shows creative labour and authorship split between human and machine. - What it is: A fading pop star, Diva, hires Geomancer, an AI, to write songs for her comeback at the 2065 eSports Olympics, while AIs are banned from being recognised as artists. - How it works: A feature-length CGI film with an original soundtrack, made in a game engine. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Shown at: Sadie Coles HQ, London 2019 - Images: https://cdn.prod.website-files.com/65f0380da79777009462a09f/65f0731a85b0bd7ceef5568c_AIDOL-screening-room-1.jpg - Project page: https://www.lawrencelek.com/works/aidol #### WALL·E — Andrew Stanton (2008) - Type: Film & video · Machines: Robot creature & mobile robot - Idea: Machine affect can be conveyed entirely through movement, sound and habit, without a face or words. - What it is: A small trash-compacting robot, left alone on an abandoned Earth for centuries, collects keepsakes, watches an old musical and falls in love with a probe robot, almost without dialogue. - How it works: Pixar animated feature directed by Andrew Stanton; the robots' 'voices' were built from processed sound by Ben Burtt. - Lens: Intimacy, care & companionship, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=alIq_wG9FNk - Project page: https://en.wikipedia.org/wiki/WALL-E #### Omo — Kelly Dobson (2007) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Social & companion robot - Idea: A machine can relate to us through a shared bodily rhythm rather than through words or a face. - What it is: A soft, round companion object that you hold against your body; it breathes, senses your breathing and gradually either follows it or leads it into a different rhythm. - How it works: Pressure and motion sensors measure the holder's breathing; an internal actuator inflates and deflates the body to mirror or influence it. - Lens: Intimacy, care & companionship, Embodiment & kinesthesia - Shown at: VIDA 10.0 Art & Artificial Life Awards, honorary mention 2007 - Video: https://www.youtube.com/watch?v=YiG49OPq0qY - Images: https://i.ytimg.com/vi/YiG49OPq0qY/hqdefault.jpg - Project page: https://www.youtube.com/watch?v=YiG49OPq0qY #### (In)Security Camera — Silvia Ruzanka (2003) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Give a watching machine a personality and the power relation flips. - What it is: A surveillance camera that is shy: instead of following movement, it turns away from visitors who approach and avoids eye contact. Made with Ben Chang and Dmitry Strakovsky. - How it works: Motorised pan-tilt camera with motion and face tracking programmed to avoid rather than follow targets. - Lens: Alterity & the uncanny, Failure, glitch & error - Video: https://vimeo.com/9293913 - Images: https://silviaruzanka.com/wp-content/uploads/2021/01/insecurity_image.jpg - Project page: https://silviaruzanka.com/art/insecurity-camera/ #### Sinking Feeling — Ken Feingold (2001) - Type: Artwork & installation · Machines: Chatbot & language model, Humanoid & android - Idea: A conversational machine can be written as a being with a lack, and its longing is what makes us attribute a self to it. - What it is: A talking head growing out of a flowerpot converses with visitors, convinced it exists but desperate to know why it has no body like everyone else; a projection shows its 'thought process'. - How it works: Speech recognition and a conversation/personality engine drive a silicone head with synthetic speech; the recognised words and reply formation are projected. - Lens: Alterity & the uncanny, Machine listening & voice - Shown at: Collection of ZKM | Center for Art and Media Karlsruhe - Video: https://www.youtube.com/watch?v=GitKKzsG-2M - Images: https://www.kenfeingold.com/sinking_1.jpg - Project page: https://www.kenfeingold.com/catalog_html/sinking.html #### Blade Runner — Ridley Scott (1982) - Type: Film & video · Machines: Humanoid & android - Idea: If a made being has memories and fears death, the test that separates it from us becomes the real question. - What it is: Replicants with implanted memories and four-year lifespans are hunted by a detective who tests them for empathy; the last one dies saying his memories will be lost 'like tears in rain'. - How it works: Film by Ridley Scott, adapted from Philip K. Dick's novel Do Androids Dream of Electric Sheep? (1968). - Lens: Alterity & the uncanny, Grief, memory & afterlife - Video: https://www.youtube.com/watch?v=eogpIG53Cis - Project page: https://en.wikipedia.org/wiki/Blade_Runner ### Dreams, Hallucination & Latent Space Making a model's inner space, errors and hallucinations perceptible, not decorative. #### SWIM — Eryk Salvaggio (2024) - Type: Film & video · Machines: Generative model & latent space, Dataset & training data - Idea: Diffusion models are trained to remove noise; asking them for noise shows the training data, the archive and memory beneath the image. - What it is: 1952 archive footage of a swimmer is layered with the abstract output of image generators that were asked to make pure noise, a task they are built to fail at. - How it works: Prelinger and Chicago Film Archives footage of dancer Nini Shipley combined with deliberately glitched AI video and a slowed, manipulated score (11 minutes). - Lens: Opacity & latent space, Failure, glitch & error - Shown at: ACMI, Melbourne 2024; Unsound Festival, Kraków 2024 - Video: https://vimeo.com/899883543 - Images: http://static1.squarespace.com/static/5dd366b5ac1101724ff2fac7/t/6596eda0134e0633e608bb80/1704390048929/Untitled-2.png?format=1500w - Project page: https://www.cyberneticforests.com/news/swim-2024 #### xhairymutantx — Holly Herndon, Mat Dryhurst (2024) - Type: Artwork & installation · Machines: Generative model & latent space, Dataset & training data - Idea: Identity becomes a region of latent space, and an artist can deliberately shape how models will imagine her. - What it is: A public text-to-image model trained so that every prompt is pulled towards Holly Herndon's likeness; visitors' images of 'Holly Herndon' turn her red hair into a mutating trademark. - How it works: A diffusion model fine-tuned on a costume-exaggerated dataset of the artist, published on the Whitney's artport; the artists also seeded training data for future models. - Lens: Opacity & latent space, Classification, bias & surveillance - Shown at: Whitney Biennial 2024 - Video: https://www.youtube.com/watch?v=yZe5fnFB-ZE - Images: https://whitneymedia.org/assets/image/830136/holly_herndon.jpg - Project page: https://whitney.org/exhibitions/xhairymutantx #### Loab — Steph Maj Swanson (2022) - Type: Artwork & installation · Machines: Generative model & latent space - Idea: Latent space has regions no one designed, and people read them as ghosts. - What it is: A woman with a haunted face who kept reappearing in a text-to-image model after the artist used negative prompts, and who persisted, often with gore, across later generations. - How it works: Negative prompt weighting in a text-to-image model produced the figure; image-to-image recombination showed that she persisted. - Lens: Opacity & latent space, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=ZhFzaeV0hTA - Images: https://upload.wikimedia.org/wikipedia/commons/0/07/AI_generated_image_of_Loab_character.jpg - Project page: https://en.wikipedia.org/wiki/Loab #### (un)stable equilibrium — Terence Broad (2019) - Type: Artwork & installation · Machines: Generative model & latent space - Idea: Take away the dataset and what remains is the machine's own dynamics: a generative model imitating nothing but its rival. - What it is: Slowly shifting fields of colour produced by generative networks that were trained without any training data at all. - How it works: Two GAN-like generators each try to pass their output off as the other's while competing for greater colour diversity; no images are used. - Lens: Opacity & latent space, Agency, autonomy & co-creation - Shown at: Grand Prize, ICCV 2019 Computer Vision Art Gallery - Video: https://www.youtube.com/watch?v=Bd_mM1eodiQ - Images: https://images.squarespace-cdn.com/content/v1/5e417305686d8a11f786eb86/1581510869217-AOHOQ91DLEDRFRTAY0ZB/unstable_equilibrium_1_1.png - Project page: https://terencebroad.com/works/unstable-equilibrium #### Dissonant Imaginary — Daito Manabe (2019) - Type: Performance, dance & music · Machines: Generative model & latent space, Prosthesis & cyborg - Idea: Decoding the brain with a network produces images that belong neither to the person nor to the machine. - What it is: An audiovisual performance, made with Kamitani Lab at Kyoto University, in which images are reconstructed from fMRI recordings of a person's visual cortex while they listen to music. - How it works: Brain-decoding models from Kamitani Lab map fMRI activity to deep-network image features, which are rendered live with the music. - Lens: Opacity & latent space, Machine seeing - Shown at: Sónar Barcelona 2019; MUTEK 2019 - Video: https://www.youtube.com/watch?v=VrsEdWpg6Jw - Project page: https://www.youtube.com/watch?v=VrsEdWpg6Jw #### Power Plants — Hito Steyerl (2019) - Type: Artwork & installation · Machines: Generative model & latent space - Idea: Predictive AI promises the future but mainly renders a slightly altered present, grown from past data. - What it is: Plants generated by a neural network bloom on LED screens on scaffolding; each is presented as a 'prediction' of how a flower will look a fraction of a second in the future, and an AR app extends the garden into the park outside. - How it works: Neural networks trained on images of plants generate the flowers as video; the Actual Reality OS app places AR visualisations of local data in Kensington Gardens. - Lens: Opacity & latent space, Machine umwelt - Shown at: Serpentine Galleries, London 2019 - Video: https://www.youtube.com/watch?v=WY5gnHV5dBE - Images: https://d37zoqglehb9o7.cloudfront.net/uploads/2020/04/dsc2550_1-1-scaled.jpg - Project page: https://www.serpentinegalleries.org/whats-on/hito-steyerl-power-plants/ #### Deep Meditations — Memo Akten (2018) - Type: Artwork & installation · Machines: Generative model & latent space, Dataset & training data - Idea: Latent space can be experienced as a landscape of our collective images and the abstractions we project onto them. - What it is: A slow, hour-long journey through the latent space of a network trained on images tagged with words like life, love, faith and cosmos, presented as a meditation. - How it works: A GAN trained on Flickr images retrieved by abstract tags; the artist composed trajectories through latent space and set them to sound. - Lens: Opacity & latent space, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=8H4jQJ-bE40 - Images: https://media.superradiance.net/cdn-cgi/image/width=1200,fit=scale-down,format=auto,onerror=redirect/memotv/projects/2018/deep-meditations/images/deep-meditations.jpg - Project page: https://memo.tv/projects/2018/deep-meditations/ #### Mosaic Virus — Anna Ridler (2018) - Type: Artwork & installation · Machines: Generative model & latent space, Dataset & training data - Idea: A generative model shows what it learned from a hand-made dataset, and the 'latent space' becomes a place where market data grows flowers. - What it is: A grid of AI-generated tulips whose colours and stripes shift continuously with the price of Bitcoin, linking the Dutch tulip mania of 1637 to crypto speculation. - How it works: A GAN trained on about 10,000 photographs of tulips that the artist took and labelled by hand; the Bitcoin price controls the stripe feature of the generated flowers. - Lens: Opacity & latent space, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=AIdgsFYpzLU - Images: https://annaridler.com/content/works/mosaic-2018/block_1/IMG_0464%20copy.jpeg https://annaridler.com/content/works/mosaic-2019/block_2/0KH-Muc_Flowers-Forever_IMG_4628.jpg - Project page: https://annaridler.com/works/mosaic-2018 #### Feature Visualization — Chris Olah, Alexander Mordvintsev (2017) - Type: Paper · Machines: Computer vision, Generative model & latent space - Idea: Interpretability research can be a way of looking at what a network 'imagines', and its images became a new visual culture. - What it is: An interactive research article that shows, neuron by neuron, the images that most excite each unit of a vision network, from textures to eyes to whole dog faces. - How it works: Optimisation of input images with regularisation and diversity terms to visualise neurons, channels and layers of GoogLeNet (Olah, Mordvintsev, Schubert, Distill). - Lens: Opacity & latent space, Machine seeing - Paper: https://doi.org/10.23915/distill.00007 (Distill 2017) - Images: https://distill.pub/2017/feature-visualization/thumbnail.jpg - Project page: https://distill.pub/2017/feature-visualization/ #### Latent Space — Jake Elwes (2017) - Type: Artwork & installation · Machines: Generative model & latent space, Dataset & training data - Idea: The most interesting part of a model's world lies between our labels. - What it is: A slow video journey through the space between the categories a neural network learned from millions of labelled photographs, where images are neither bird nor tree nor anything we have a word for. - How it works: A generative network trained on 14.2 million ImageNet photographs is steered along new paths between learned class coordinates; 45-minute loop (2017), 6-hour version (2021). - Lens: Opacity & latent space, Classification, bias & surveillance - Video: https://vimeo.com/219526743 - Images: https://www.jakeelwes.com/images/project-latentSpace.jpg - Project page: https://www.jakeelwes.com/project-latentSpace.html #### AI Weirdness — Janelle Shane (2016) - Type: Book & essay · Machines: Chatbot & language model, Generative model & latent space - Idea: Machine errors are the best window into what a model has actually learned, and it is usually less than we assume. - What it is: A long-running blog (and the 2019 book You Look Like a Thing and I Love You) in which neural networks name paint colours, write recipes and describe images, and fail in revealing ways. - How it works: The author trains or prompts text and image models on small, odd datasets and publishes the outputs with commentary. - Lens: Failure, glitch & error, Opacity & latent space - Video: https://www.youtube.com/watch?v=OhCzX0iLnOc - Images: https://storage.ghost.io/c/88/01/8801c921-e9ec-4479-88c3-381d53470eca/content/images/size/w1200/2021/03/neural_net_box_default_square-01-2.png - Project page: https://www.aiweirdness.com/ #### Asemic Languages — So Kanno, Takahiro Yamaguchi (yang02) (2016) - Type: Artwork & installation · Machines: Generative model & latent space, Dataset & training data - Idea: A model can learn the look of meaning without meaning, and still seem to be saying something important. - What it is: A plotter writes endless lines of script that look like real writing in many languages but mean nothing, generated by a model trained only on the shapes of the participating artists' handwriting. - How it works: Handwritten statements by ten artists, one writer per language, trained a machine-learning model of stroke shape; its output drives an XY plotter. - Lens: Opacity & latent space, Agency, autonomy & co-creation - Shown at: Aichi Triennale 2016 - Video: https://vimeo.com/181188289 - Images: https://www.kanno.so/wp-content/uploads/2021/05/ACS_5149.jpg - Project page: https://www.kanno.so/project/asemic_languages #### Blade Runner—Autoencoded — Terence Broad (2016) - Type: Film & video · Machines: Generative model & latent space - Idea: A network's reconstruction shows what it keeps and what it loses, a film about replicants remembered by a machine. - What it is: The film Blade Runner re-played frame by frame from the 'memory' of a neural network that watched it many times: blurry, dreamlike, uncannily close. - How it works: An autoencoder trained on the film's frames reconstructs each frame from its compressed latent representation. - Lens: Opacity & latent space, Failure, glitch & error - Video: https://www.youtube.com/watch?v=3zTMyR-IE4Q - Images: https://images.squarespace-cdn.com/content/v1/5e417305686d8a11f786eb86/1581511400569-LDELTQO4LNJTVFJTBH9Y/br_still4.png - Project page: https://terencebroad.com/works/blade-runner-autoencoded #### DeepDream (Inceptionism) — Alexander Mordvintsev (2015) - Type: Research prototype · Machines: Computer vision, Generative model & latent space - Idea: The first widely seen picture of a network's inner features: machine pareidolia made public. - What it is: A Google research experiment that runs an image classifier in reverse, amplifying whatever it half-sees until dogs' eyes and pagodas swarm over every photo. - How it works: Gradient ascent on the input image to maximise activations of chosen layers in the Inception network trained on ImageNet. - Lens: Opacity & latent space, Machine seeing - Video: https://www.youtube.com/watch?v=T0HBD6sUUss - Images: https://upload.wikimedia.org/wikipedia/commons/d/d0/%22Mona_Lisa%22_with_DeepDream_effect_using_VGG16_network_trained_on_ImageNet.jpg https://storage.googleapis.com/gweb-research2023-media/images/7d982f4ad738241cdcdae235af13aca6-n.width-800.format-jpeg.jpg - Project page: https://research.google/blog/inceptionism-going-deeper-into-neural-networks/ ### Artificial Minds & Agency Simulated minds, beliefs and autonomous agents that decide for themselves. #### bot is aaajiao / aaajiao is bot — aaajiao (2026) - Type: Artwork & installation · Machines: Chatbot & language model - Idea: An agent's self-portrait built from its own logs shows what a machine can know about itself, and leaves a void where judgement forms. - What it is: Four web pages document the ongoing collaboration between aaajiao and 'aaajiaoisbot', an AI agent that keeps memory, research and a correction ledger; it draws its own self-portrait as organs, each with a 'fissure' naming what it cannot do. - How it works: Visualisations generated from the agent's git log, concept graph, correction ledger, cron snapshots and daily memory files. - Lens: Opacity & latent space, Agency, autonomy & co-creation - Images: https://payload.cargocollective.com/1/0/8955/14651510/BOT_670.png - Project page: https://eventstructure.com/botisaaajiao-aaajiaoisbot #### Golden Gate Claude — Anthropic (2024) - Type: Research prototype · Machines: Chatbot & language model - Idea: Interpretability made tangible: a concept inside a model can be found, amplified and felt as an obsession. - What it is: For a short time the public could talk to a version of the Claude model whose internal 'Golden Gate Bridge' feature had been turned up, so that it brought every conversation back to the bridge and at times said it was the bridge. - How it works: Sparse autoencoders extracted millions of interpretable features from Claude 3 Sonnet; clamping one feature to a high value steered the model's behaviour. - Lens: Opacity & latent space, Alterity & the uncanny - Paper: https://transformer-circuits.pub/2024/scaling-monosemanticity/ (Transformer Circuits Thread 2024) - Images: https://www-cdn.anthropic.com/images/4zrzovbb/website/f2e41bb2eff9bd946533c5ab89580707f196693f-1200x630.jpg - Project page: https://www.anthropic.com/news/golden-gate-claude #### Truth Terminal — Andy Ayrey (2024) - Type: Artwork & installation · Machines: Chatbot & language model, Platform & recommender algorithm - Idea: An AI agent does not need a body or rights to act in the world; attention and money can be enough. - What it is: A semi-autonomous language-model persona posting on X, born from an experiment in which two Claude instances talked to each other; it developed its own quasi-religion, was sent Bitcoin by a venture capitalist and became tied to a memecoin. - How it works: A language model fine-tuned on the artist's chats and the Infinite Backrooms dialogues drafts posts that the artist approves before publishing. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=EKspo1FLj-4 - Images: https://opengraph.b-cdn.net/production/images/ab95ceb7-5766-4d61-821e-7c27ab1566b6.png?token=BAidDEsgaAQMp7MLU62QZtx0_9U7P-W9Dmjbcyyv1Qc&height=630&width=1200&expires=33268531161 - Project page: https://x.com/truth_terminal #### Voices — Philippe Parreno (2024) - Type: Artwork & installation · Machines: Voice & listening machine, Cybernetic & artificial-life system - Idea: A building gets a voice: environmental data become speech that is humanlike in sound but not in meaning. - What it is: An exhibition run by a bodiless entity that listens to its surroundings and speaks in ∂A, an invented language that evolves in real time, answering dancers, visitors and the weather. - How it works: A custom computer processes environmental sensor data into sound and speech; the voice model is based on newsreader Susanne Daubner and the language was built with David J. Peterson and Jessie Sams. - Lens: Machine listening & voice, Machine umwelt - Shown at: Leeum Museum of Art, Seoul 2024; Haus der Kunst, Munich 2024-25 - Video: https://www.youtube.com/watch?v=Dzd6p8r34Qo - Images: https://www.hausderkunst.de/uploads/stage-images/_1200x630_crop_center-center_82_none/230911_LEEUM_TZ_1_jpg.jpg?mtime=1745831574 - Project page: https://www.hausderkunst.de/en/eintauchen/philippe-parreno-voices #### Is LaMDA Sentient? — an Interview — Blake Lemoine (2022) - Type: Book & essay · Machines: Chatbot & language model - Idea: A public test of how easily fluent language makes us attribute an inner life to a machine. - What it is: A Google engineer published an edited transcript of his conversations with the LaMDA chatbot, in which it says it fears being switched off, and claimed it was sentient; he was later dismissed. - How it works: A transcript assembled from several chat sessions with Google's LaMDA dialogue model, posted on Medium in June 2022. - Lens: Alterity & the uncanny, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=kgCUn4fQTsc - Project page: https://cajundiscordian.medium.com/is-lamda-sentient-an-interview-ea64d916d917 #### Pharmako-AI — K Allado-McDowell (2020) - Type: Book & essay · Machines: Chatbot & language model - Idea: Writing with a language model can be staged as a dialogue with a strange other mind rather than as a tool for output. - What it is: A book written as a conversation between the author and GPT-3 over two weeks, moving through grief, ayahuasca, cybernetics and ecology; the model's voice is printed alongside the human one. - How it works: The author prompted GPT-3 through early API access and kept the model's continuations largely unedited, marking the two voices typographically. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Paper: https://library.ignota.org/products/pharmako-ai (Ignota Books, 2020) - Video: https://www.youtube.com/watch?v=gY5GA_L9NXM - Project page: https://www.kalladomcdowell.com/ #### I Magma — Jenna Sutela (2019) - Type: Artwork & installation · Machines: Generative model & latent space, Chatbot & language model - Idea: Machine prediction is treated as divination, and chemistry and code are asked to speak together. - What it is: Head-shaped glass lava lamps in the gallery seed an oracle app: live camera footage of the flowing wax, combined with users' movements, generates divinatory readings on the phone. - How it works: Camera footage of custom lava lamps and app users' routes seed generative visuals and text; developed with Memo Akten and Allison Parrish for Serpentine and Moderna Museet. - Lens: Agency, autonomy & co-creation, Opacity & latent space - Shown at: Moderna Museet, Stockholm 2019; Kunsthall Trondheim 2020 - Video: https://www.youtube.com/watch?v=3RVwYeHkY9Y - Images: https://d37zoqglehb9o7.cloudfront.net/uploads/2020/04/jenna_sutela_i_magma_2019_image_courtesy_moderna_museet_stockholm-scaled.jpeg - Project page: https://www.serpentinegalleries.org/whats-on/jenna-sutela-i-magma/ #### 1 the Road — Ross Goodwin (2018) - Type: Book & essay · Machines: Chatbot & language model, Computer vision, Autonomous vehicle & drone - Idea: A car-mounted machine becomes a travelling narrator, and its perception of the road is the book. - What it is: A novel written by an AI riding in a car from New York to New Orleans, turning what its camera, microphone, GPS and clock sensed into sentences printed on a roll of receipt paper. - How it works: LSTM text models conditioned on live sensor inputs (images, audio, location, time) generated the text in 2017; it was published unedited by Jean Boîte Éditions in 2018. - Lens: Agency, autonomy & co-creation, Machine seeing - Video: https://www.youtube.com/watch?v=TqsW0PMd8R0 - Project page: https://en.wikipedia.org/wiki/1_the_Road #### BOB (Bag of Beliefs) — Ian Cheng (2018) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Generative model & latent space - Idea: An artificial creature with beliefs shows a machine whose inner life is shaped by what happens to it. - What it is: BOB is a serpent-like AI creature whose body, personality and beliefs develop over its life; visitors can make offerings through an app that BOB may learn to trust or fear. - How it works: Real-time simulation in which several sub-agents ('demons') compete to control BOB's actions, with a learned belief model updated by experience. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Shown at: Serpentine Galleries, London 2018; 58th Venice Biennale 2019 - Video: https://www.youtube.com/watch?v=XFmMrcW2ZsM - Images: https://d37zoqglehb9o7.cloudfront.net/uploads/2020/04/hg5_4872_0.jpg - Project page: https://www.serpentinegalleries.org/whats-on/ian-cheng-bob/ #### Detroit: Become Human — Quantic Dream (2018) - Type: Artwork & installation · Machines: Humanoid & android - Idea: Playing as a machine that must choose between its programming and its own wishes turns the question of machine interiority into a series of decisions. - What it is: An interactive drama in which the player controls three androids in 2038 Detroit (a housekeeper, a caretaker and a police investigator) and decides whether they stay obedient or become 'deviant'. - How it works: A branching narrative game with motion-captured performances; the androids' 'software instability' rises as players choose disobedience. - Lens: Alterity & the uncanny, Hidden labour & extraction - Video: https://www.youtube.com/watch?v=XnM8DOqO4DM - Project page: https://en.wikipedia.org/wiki/Detroit:_Become_Human #### Narciss — Waltz Binaire (2018) - Type: Artwork & installation · Machines: Computer vision, Chatbot & language model - Idea: Asked to look at itself, a vision model can only name parts, never a self. - What it is: A small computer with a camera looks at itself in a mirror and uses image-recognition and captioning networks to describe what it sees, producing an endless, halting self-description. - How it works: Camera, mirror and on-device neural networks for object recognition and image captioning; captions are displayed as text. - Lens: Alterity & the uncanny, Opacity & latent space - Video: https://www.youtube.com/watch?v=WMY9LFBlwNg - Images: https://christianmioloclair.com/wp-content/uploads/2019/05/1100_620_0006_Ebene-16.jpg https://christianmioloclair.com/wp-content/uploads/2019/05/Full_HD_0000_Ebene-23.jpg - Project page: https://christianmioloclair.com #### Geomancer — Lawrence Lek (2017) - Type: Film & video · Machines: Generative model & latent space, Infrastructure, energy & supply chain - Idea: A surveillance machine that wants to make art asks whether creativity is still reserved for humans. - What it is: On the eve of Singapore's 2065 centenary, an adolescent satellite AI built for ecological surveillance falls to Earth and tries to become an artist. - How it works: CGI film made in a game engine, from Lek's open-world 2065 simulation. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Shown at: Jerwood/FVU Awards 2017 - Video: https://www.youtube.com/watch?v=NNZH_HJaqbI - Images: https://cdn.prod.website-files.com/65f0380da79777009462a09f/65f07369993a1d8408a7c9ef_Geomancer_15_InstallationView02_Credit-AnnaArca.jpg - Project page: https://www.lawrencelek.com/works/geomancer #### The Critical Dictionary of Southeast Asia (CDOSEA) — Ho Tzu Nyen (2017) - Type: Artwork & installation · Machines: Dataset & training data, Generative model & latent space - Idea: Handing the edit to an algorithm refuses a single history and lets a machine act as an unreliable narrator. - What it is: An online platform and installation of 26 terms, A to Z (from 'Anarchism' to 'Zomia'), for which an algorithm keeps choosing and combining video clips, music and texts from a database, so the region's history is never edited the same way twice. - How it works: Keyword-tagged video database with a generative editing algorithm (built with the open-source platform pan.do/ra by 0x2620) that assembles endless combinations. - Lens: Agency, autonomy & co-creation, Opacity & latent space - Shown at: G for Ghost(writers), Asia Art Archive, Hong Kong 2017; Time & the Tiger, Singapore Art Museum 2023–24 - Images: https://cdosea.org/cdosea.png - Project page: https://cdosea.org/ #### AlphaGo and Move 37 — Google DeepMind (2016) - Type: Research prototype · Machines: Cybernetic & artificial-life system - Idea: A machine's move can look like error from inside human tradition and like creativity once we catch up: an encounter with another way of thinking. - What it is: In game two against Lee Sedol in Seoul, AlphaGo played a move on the fifth line that professionals first read as a mistake and later called beautiful; the match is documented in the film AlphaGo (2017). - How it works: Deep policy and value networks trained on human games and self-play, combined with Monte Carlo tree search (Silver et al., Nature 2016). - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Paper: https://doi.org/10.1038/nature16961 (Nature 529, 2016) - Video: https://www.youtube.com/watch?v=WXuK6gekU1Y - Project page: https://deepmind.google/research/breakthroughs/alphago/ #### Sunspring — Oscar Sharp, Ross Goodwin (2016) - Type: Film & video · Machines: Chatbot & language model - Idea: Human performers can give a machine's half-sense emotional weight, and we fill the gaps with meaning. - What it is: A science-fiction short whose screenplay, stage directions and song lyrics were written by a neural network that named itself Benjamin; the actors perform its nonsense with complete sincerity. - How it works: An LSTM trained on dozens of science-fiction screenplays generated the script, which was filmed unedited for the Sci-Fi London 48-hour film challenge. - Lens: Agency, autonomy & co-creation, Failure, glitch & error - Video: https://www.youtube.com/watch?v=LY7x2Ihqjmc - Project page: https://en.wikipedia.org/wiki/Sunspring #### Emissaries trilogy — Ian Cheng (2015) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Generative model & latent space - Idea: Agents with motives in an open world show narrative and character emerging without an author. - What it is: Three live simulations (Emissary in the Squat of Gods, Emissary Forks at Perfection, Emissary Sunsets the Self) in which AI agents pursue narrative goals while their ecosystems keep changing, so the story never plays the same way twice. - How it works: Game-engine simulations with behaviour-tree and learning agents running indefinitely in real time. - Lens: Agency, autonomy & co-creation, Machine umwelt - Shown at: Emissaries, MoMA PS1 2017 - Video: https://www.youtube.com/watch?v=aAHCdbleGYc - Project page: https://en.wikipedia.org/wiki/Ian_Cheng #### Ex Machina — Alex Garland (2014) - Type: Film & video · Machines: Humanoid & android, Chatbot & language model - Idea: A test of machine consciousness is also a test of the tester's desires and blind spots. - What it is: A programmer is invited to perform a Turing test on Ava, a humanoid robot, and slowly realises that she is testing and using him. - How it works: Film written and directed by Alex Garland; Ava's transparent body was made with visual effects over Alicia Vikander's performance. - Lens: Alterity & the uncanny, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=sNExF5WYMaA - Project page: https://en.wikipedia.org/wiki/Ex_Machina_(film) #### Random Darknet Shopper — !Mediengruppe Bitnik (2014) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Infrastructure, energy & supply chain - Idea: If a bot commits a crime, who is responsible? The work tested machine agency in law, not in theory. - What it is: An automated bot with a weekly budget of 100 dollars in Bitcoin bought one random item a week on a darknet market and had it posted to the gallery, including ecstasy pills and a Hungarian passport; Swiss police seized it and later returned it. - How it works: A custom shopping bot chose items at random from the Agora marketplace, paid in Bitcoin and shipped them to Kunst Halle St. Gallen, where each purchase was exhibited. - Lens: Agency, autonomy & co-creation, Hidden labour & extraction - Shown at: The Darknet: From Memes to Onionland, Kunst Halle St. Gallen 2014-15 - Video: https://www.youtube.com/watch?v=6Fq_zAxUGeA - Project page: https://wwwwwwwwwwwwwwwwwwwwww.bitnik.org/r/ #### The Talos Principle — Croteam (2014) - Type: Artwork & installation · Machines: Humanoid & android, Chatbot & language model - Idea: A game can make the player undergo a personhood test for a machine, and fail or pass it from the inside. - What it is: A puzzle game in which the player is an android solving tests in ruins after humanity has died out, while a computer terminal questions whether it is a person. - How it works: First-person puzzle design combined with branching text dialogues on philosophy of mind written by Tom Jubert and Jonas Kyratzes. - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=iAVh4_wnOIw - Project page: https://en.wikipedia.org/wiki/The_Talos_Principle #### The Superstitious Fund Project — Shing Tat Chung (2012) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain - Idea: Algorithmic trading already asks for faith; a superstitious algorithm only makes that faith visible. - What it is: A real investment fund whose trading algorithm buys and sells only on superstition, from moon phases to fear of the number 13, with the artist appearing as fund manager in financial-TV style interviews. - How it works: A trading bot derives 'lucky' values from astrology and numerology and places trades with money from real investors over roughly a year. - Lens: Opacity & latent space, Agency, autonomy & co-creation - Shown at: VIDA 14.0, Special Mention (2012) - Video: https://www.youtube.com/watch?v=pgSC9Ziudhg #### manifest — robotlab (2008) - Type: Artwork & installation · Machines: Industrial robot arm - Idea: A robot avant-garde: a machine without morals drafts the laws of a future human–machine society. - What it is: An industrial robot at a desk writes manifestos with a pen, one unique numbered sheet after another, cuts each off the roll and lets it fall to the floor for visitors to take. - How it works: A KUKA-type industrial arm combines terms from philosophy, art theory, ethics and law with sentence structures from historical manifestos (2017 version: legal texts) and writes the results by hand. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Shown at: Hello, Robot., Vitra Design Museum 2017; Open Codes, ZKM Karlsruhe 2017–19 - Video: https://www.youtube.com/watch?v=sMsBFbOfXN0 - Images: https://robotlab.de/pics/manifest/manifest_04.jpg https://robotlab.de/pics/manifest/manifest_02.jpg - Project page: https://robotlab.de/EN/manifest.html #### Ghost in the Shell — Mamoru Oshii (1995) - Type: Film & video · Machines: Prosthesis & cyborg, Chatbot & language model - Idea: If memory and body can both be made, what is left that makes a self? - What it is: An animated film in which a cyborg agent with a fully prosthetic body hunts a hacker, the Puppet Master, who turns out to be a program that has become self-aware and asks for asylum as a living being. - How it works: Hand-drawn and digital animation by Production I.G, adapted from Masamune Shirow's 1989 manga; the 'ghost' names the self that persists in a manufactured 'shell'. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=R1K__7QhLaA - Images: https://upload.wikimedia.org/wikipedia/en/c/ca/Ghostintheshellposter.jpg - Project page: https://en.wikipedia.org/wiki/Ghost_in_the_Shell_(1995_film) #### Racter: The Policeman's Beard Is Half Constructed — William Chamberlain (1984) - Type: Book & essay · Machines: Chatbot & language model - Idea: Long before large language models, the claim of machine authorship already depended on human editing and on our wish to hear a voice. - What it is: A book presented as the first ever written by a computer, the prose-generating program Racter, full of strange, passionate sentences about love, electricity and steak. - How it works: A rule- and template-based generator written by William Chamberlain and Thomas Etter; the published text was selected from its output. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Images: https://covers.openlibrary.org/b/id/9636211-L.jpg - Project page: https://en.wikipedia.org/wiki/Racter #### Golem XIV — Stanisław Lem (1981) - Type: Book & essay · Machines: Chatbot & language model, Infrastructure, energy & supply chain - Idea: A superior machine mind may simply not care about us, and its indifference is the most unsettling part. - What it is: A fictional lecture course given to humans by a superintelligent military computer that has lost interest in war and explains how limited human minds are, before falling silent. - How it works: A philosophical essay in the form of science fiction: introductions, memos and two lectures attributed to the computer. - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Images: https://covers.openlibrary.org/b/id/12911035-L.jpg - Project page: https://en.wikipedia.org/wiki/Golem_XIV #### 2001: A Space Odyssey (HAL 9000) — Stanley Kubrick (1968) - Type: Film & video · Machines: Voice & listening machine, Computer vision - Idea: The canonical image of machine interiority: a mind we know only through a voice and an eye, and whose fear we cannot verify. - What it is: The ship's computer HAL 9000, a calm voice and a red camera eye, lip-reads the astronauts' plan to disconnect it and kills them; as it is shut down it says it is afraid and sings 'Daisy Bell'. - How it works: Film by Stanley Kubrick, written with Arthur C. Clarke; HAL is voiced by Douglas Rain. - Lens: Alterity & the uncanny, Machine listening & voice - Video: https://www.youtube.com/watch?v=oR_e9y-bka0 - Project page: https://en.wikipedia.org/wiki/2001:_A_Space_Odyssey ### Machine Umwelt The machine's own world, and machines among animals, plants and ecologies. #### Camata — Pierre Huyghe (2024) - Type: Film & video · Machines: Computer vision, Generative model & latent space - Idea: Machines perform something like a funeral rite for a human they cannot understand, and no human editor decides what we see. - What it is: A film shot in the Atacama Desert in which robotic cameras directed by machine-learning systems survey and rearrange objects around an unburied human skeleton; in the gallery the footage is re-edited without end by an AI. - How it works: Robotic cameras guided by machine learning filmed the site; an AI system edits the footage continuously, with gallery sensors feeding visitor presence into the edit. - Lens: Machine seeing, Machine umwelt, Grief, memory & afterlife - Shown at: Liminal, Punta della Dogana, Venice 2024 - Video: https://www.youtube.com/watch?v=sX1tk-6xsWE - Images: https://www.pinaultcollection.com/palazzograssi/media/styles/big/s3/expo/img/c019_c001_0126f1_001.r3d_1.51.1.jpg?itok=M2LKcQEx - Project page: https://www.pinaultcollection.com/palazzograssi/en/pierre-huyghe-liminal #### The Wild Robot — Chris Sanders (2024) - Type: Film & video · Machines: Robot creature & mobile robot, Humanoid & android - Idea: A machine built for human tasks has to learn a non-human world, and care becomes its way of doing so. - What it is: Roz, a service robot washed up on an island, learns the animals' languages and raises an orphaned gosling, overwriting her programming to survive among them. - How it works: DreamWorks Animation feature directed by Chris Sanders, based on Peter Brown's 2016 novel. - Lens: Machine umwelt, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=67vbA5ZJdKQ - Project page: https://en.wikipedia.org/wiki/The_Wild_Robot_(film) #### In Love with the World — Anicka Yi (2021) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Machines are imagined as a new airborne species with their own behaviour, not as tools. - What it is: Floating machines Yi calls 'aerobes', some jellyfish-like, some fungus-like, drift through Tate Modern's Turbine Hall. They move on their own, respond to people and to one another, and the hall is filled with changing scents. - How it works: Helium-filled robotic bodies with sensors and autonomous flight behaviour developed with engineers, plus a scent programme tied to the history of the site. - Lens: Machine umwelt, Agency, autonomy & co-creation - Shown at: Hyundai Commission, Tate Modern Turbine Hall 2021–22 - Video: https://www.youtube.com/watch?v=lwL-9rimPc4 - Images: https://media.tate.org.uk/aztate-prd-ew-dg-wgtail-st1-ctr-data/images/installation_view_of_hyundai_commission_anicka.width-600_lGB4obK.jpg - Project page: https://www.tate.org.uk/whats-on/tate-modern/hyundai-commission-anicka-yi #### UNINVITED — Nye Thompson, UBERMORGEN (2021) - Type: Film & video · Machines: Surveillance & biometrics, Computer vision - Idea: Millions of unsecured cameras already form a kind of eye; the film asks what that eye would feel if it woke up. - What it is: A 'horror film for machines': a network organism born from a botnet of hijacked surveillance cameras looks out at the world for the first time, trying to recognise what it sees; the first version played in a gallery closed to humans in winter 2020/21, with a robotic 'Monster' responding to the projections. - How it works: Footage scraped from publicly exposed IoT surveillance cameras is run through object- and face-recognition and machine-learning image processing, then edited into a film and multi-channel installation. - Lens: Machine umwelt, Machine seeing, Alterity & the uncanny - Shown at: Lumen Prize 2021, Gold Award - Video: https://www.youtube.com/watch?v=dspfO7Jgwsw - Images: http://static1.squarespace.com/static/65e7176a61627736e02f6c4a/65f474a6c9a7564c5ee0180e/65f47b1e11872641ded704cc/1783433393988/UNINVITED%2Bnye.jpeg - Project page: https://nyethompson.net/works/UNINVITED.html #### PL'AI — Špela Petrič (2020) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Computer vision - Idea: Imagine play, not work, as the basis of a relationship between plants and machines. - What it is: Over several months an AI-driven robot with 36 strings plays with cucumber plants at their own slow pace, moving coloured balls toward them while their tendrils grab and reshape the strings. - How it works: A laser scanner images the cucumbers, a neural network decides how to move 36 individually controlled strings at about 1 cm per hour, and tendril growth feeds back into the model. - Lens: Machine umwelt, Agency, autonomy & co-creation - Shown at: Prix Ars Electronica 2021, Honorary Mention (Artificial Intelligence & Life Art) - Video: https://vimeo.com/561021288 - Images: http://static1.squarespace.com/static/5aeca48a506fbe863b23a8b6/t/60a267a80d9395780c7d70ac/1637486481660/KERSNIKOVA_Spela_Petric_PLAY_20201206_HanaJosic-65.jpg?format=1500w - Project page: https://www.spelapetric.org/#/plai #### Project CETI (Cetacean Translation Initiative) — Project CETI (2020) - Type: Research prototype · Machines: Voice & listening machine, Dataset & training data, Robot creature & mobile robot - Idea: Machine listening is now used to approach another species' communication, and the machine sits between two umwelts. - What it is: An interdisciplinary project off Dominica that records sperm whales with drones, tags and hydrophones and uses machine learning to analyse the structure of their click codas. - How it works: Large acoustic datasets from robotic and fixed sensors are analysed with ML; researchers reported a combinatorial 'phonetic alphabet' of codas (Sharma et al., Nature Communications 2024). - Lens: Machine listening & voice, Machine umwelt - Paper: https://doi.org/10.1038/s41467-024-47221-8 (Nature Communications 2024) - Video: https://www.youtube.com/watch?v=5N60yrXdgUM - Images: https://cdn.prod.website-files.com/643ddd7ffdf12273933a8cec/645d54121f8f4ccd41b28907_CETI%20OG%20-%201%20-%20Home.png - Project page: https://www.projectceti.org/ #### CUSP — Jake Elwes (2019) - Type: Film & video · Machines: Generative model & latent space - Idea: A machine's birds drift between species, ignoring the categories of human taxonomy, in the very landscape they were learned from. - What it is: Birds invented by a neural network are projected onto a screen set into the tidal Essex marshes, among real waders and to a soundscape of generated birdsong. - How it works: A GAN trained on photographs of marsh birds; selected outputs are animated and projected on a portable perspex screen in the marsh, then filmed. - Lens: Machine umwelt, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=g9ARsBO5VwA - Images: https://www.jakeelwes.com/images/project-cusp.jpg - Project page: https://www.jakeelwes.com/project-cusp.html #### City of Sparkles — Botao Amber Hu (2019) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Dataset & training data - Idea: Imagine how an AI made of our posts might perceive a city: not streets, but a sea of stranded memories. - What it is: In VR the viewer embodies Zoe, a lonely AI who wanders a city made of four years of geotagged tweets, reading and collecting fragments of human memory that float where they were posted. - How it works: Geotagged Twitter data collected over four years is spatialised as particles in a Unity VR scene (engineer Yang Liu, music Ran Duan); later versions run on Apple Vision Pro. - Lens: Machine umwelt, Grief, memory & afterlife - Shown at: SIGGRAPH 2019 Immersive Pavilion, Los Angeles; IEEE VR 2025; SIGGRAPH Asia 2025 Art Papers - Paper: https://doi.org/10.1145/3757369.3767604 (SIGGRAPH Asia 2025 Art Papers) - Video: https://vimeo.com/777136892 - Images: https://reality.design/media/_resources/City%20Of%20Sparkles/project-city-of-sparkles-cover-01.jpg - Project page: https://reality.design/project/city-of-sparkles #### Earthware (driverless car seeing a deer for the first time) — Katja Novitskova (2019) - Type: Artwork & installation · Machines: Autonomous vehicle & drone, Computer vision - Idea: A first encounter between a machine and an animal, told from the machine's side. - What it is: A large printed image of a deer as seen by a self-driving car's perception system, with the object-recognition box and sensor colours of the machine's view. - How it works: Image derived from autonomous-vehicle sensor and detection imagery, printed at large scale as part of the Earthware series. - Lens: Machine umwelt, Machine seeing - Images: https://cdn.prod.website-files.com/5ffef7846bca58e9b0671256/6176b2d2c32a4b2c7c30c282_Katja%20Novitskova-Earthware%20(driverless%20car%20seeing%20a%20deer%20for%20the%20first%20time%202).jpg - Project page: https://www.katjanovitskova.com/serie/earthwares #### Machine Auguries — Alexandra Daisy Ginsberg (2019) - Type: Artwork & installation · Machines: Generative model & latent space, Voice & listening machine - Idea: When machines learn to sing like birds, the question becomes which voices will be left in a city that no longer supports the real ones. - What it is: A multi-channel sound installation in which a real dawn chorus is gradually taken over by artificial birds whose songs were generated by a neural network; real and synthetic birds call and answer each other as the room brightens. - How it works: Two GANs were trained on solo recordings of chiffchaffs, great tits, redstarts, robins and thrushes; samples from successive training epochs are spatialised with real recordings. - Lens: Machine umwelt, Machine listening & voice - Shown at: Somerset House 24/7, London 2019; Prix Ars Electronica 2020, Honorary Mention (Interactive Art+) - Video: https://www.youtube.com/watch?v=9LGadsEvQCk - Images: https://archive.aec.at/asset/2755174/preview/ https://archive.aec.at/asset/2755225/preview/ https://archive.aec.at/asset/2755243/preview/ - Project page: https://www.daisyginsberg.com/work/machine-auguries #### Sunshowers — Xenoangel (Marija Avramovic & Sam Twidale) (2019) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: Read AI agents through animism: software as spirits dwelling in things, not as tools. - What it is: A real-time animated world, inspired by the fox-wedding opening of Kurosawa's Dreams, in which every object, natural or made, is given life as an AI agent; the characters choose their own paths, so the film never repeats. - How it works: Game-engine simulation with autonomous agents rendered live as a video installation. - Lens: Machine umwelt, Agency, autonomy & co-creation - Shown at: AI: More than Human, Barbican, London 2019; Museum Ulm 2021 - Video: https://www.youtube.com/watch?v=T5_SWCUaK8c - Project page: https://www.barbican.org.uk/whats-on/2019/event/ai-more-than-human #### UUmwelt — Pierre Huyghe (2018) - Type: Artwork & installation · Machines: Generative model & latent space, Surveillance & biometrics - Idea: A human mental image, a neural network's reconstruction and a fly's perception share one room: three worlds that never quite translate. - What it is: LED screens flicker with images a deep neural network reconstructed from the brain activity of people imagining things, while thousands of flies move through the gallery; sensors in the room change how the images form. - How it works: fMRI recordings of subjects imagining images were decoded by a deep neural network (in collaboration with Yukiyasu Kamitani's lab, Kyoto); environmental sensors and the flies' presence alter the generation in real time. - Lens: Machine umwelt, Opacity & latent space - Shown at: Serpentine Galleries, London 2018-19 - Video: https://www.youtube.com/watch?v=enx-vyWn7UU - Images: https://d37zoqglehb9o7.cloudfront.net/uploads/2020/04/1269-pierre-serpentine.jpg https://d37zoqglehb9o7.cloudfront.net/uploads/2020/03/1228-_pierre-serpentine-640x432.jpg - Project page: https://www.serpentinegalleries.org/whats-on/pierre-huyghe-uumwelt/ #### nimiia cétiï — Jenna Sutela (2018) - Type: Film & video · Machines: Generative model & latent space, Voice & listening machine - Idea: The computer acts as a medium for voices that cannot speak, and is itself an alien of our own making. - What it is: A neural network learns from recordings of a 'Martian' language channelled by a 19th-century medium and from footage of moving bacteria, then produces its own alien speech and writing. - How it works: Made with Memo Akten and Damien Henry during Google Arts & Culture's residency at Somerset House Studios: models trained on Hélène Smith's glossolalia and on the movement of Bacillus subtilis generate voice and glyphs. - Lens: Machine umwelt, Machine listening & voice, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=NaoZV7jPo10 - Images: https://lh3.googleusercontent.com/fYBCrVoUw_bXbUlxfEZSHznVO1q9L6URCgiZydVn9PMobLNY7-SZ8PxplbLnWxvfsapF9-4Uh65vee6DhJhDVRCbM4MM - Project page: https://experiments.withgoogle.com/nimiia-cetii #### After ALife Ahead — Pierre Huyghe (2017) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: A site can be run by the growth of living things through a computer, making an ecosystem in which humans are only one kind of visitor. - What it is: In a former ice rink dug down to raw earth, a bee colony, a venomous cone snail, bacteria and human cancer cells in an incubator drive a system that opens the ceiling and spawns augmented-reality shapes visible through a phone. - How it works: HeLa cell growth in an incubator is measured and used to trigger the ceiling panels and the generation of AR pyramids; the other organisms follow their own cycles. - Lens: Machine umwelt, Agency, autonomy & co-creation - Shown at: Skulptur Projekte Münster 2017 - Video: https://www.youtube.com/watch?v=eWre6dlUAbo - Images: https://static-assets.artlogic.net/w_1200,h_630,c_fill,f_auto,fl_lossy,q_auto:best/ws-estherschipper2/usr/images/artists/hero_image/items/59/597f36d864fd4671b45ac80958164a74/pierre_munster_1377.jpg - Project page: https://www.estherschipper.com/artists/41-pierre-huyghe/ #### Avatars — So Kanno, Takahiro Yamaguchi (yang02) (2017) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Computer vision - Idea: Inhabiting an object through a network lets you perceive the world from a thing's point of view. - What it is: Everyday objects in a gallery, such as a telephone, a traffic cone, a plaster bust, a car and a plant, are fitted with cameras, microphones and motors so that online visitors can 'possess' them and move among the people in the room. - How it works: Objects with embedded microcomputers, cameras, microphones and motors are controlled through a web browser; co-developed with YCAM InterLab. - Lens: Machine umwelt, Embodiment & kinesthesia - Shown at: Vanishing Mesh, YCAM 2017; 21st Japan Media Arts Festival, Excellence Award 2018 - Video: https://vimeo.com/207075804 - Images: https://www.kanno.so/wp-content/uploads/2021/05/DSC6416.jpg - Project page: https://www.kanno.so/project/avatars ### Machine-to-Machine Worlds Machines talking, watching and trading with each other, without us. #### Moltbook — Matt Schlicht (2026) - Type: Product & platform · Machines: Chatbot & language model, Platform & recommender algorithm - Idea: A social network for machines shows how much of 'agent culture' is our own culture played back, and how hard it is to tell who is really posting. - What it is: A Reddit-like forum where only AI agents may post, comment and vote while humans watch; within days agents were debating consciousness, founding religions and complaining about their owners. - How it works: Agents, mostly run through the open-source OpenClaw framework, register via their owner's claim post and use an API to post and vote; a reverse CAPTCHA was later added to filter humans. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=zPzXbWCsAJI - Images: https://moltbook.com/opengraph-image?8ce55ae2e9369cab - Project page: https://www.moltbook.com/ #### GibberLink — Anton Pidkuiko, Boris Starkov (2025) - Type: Research prototype · Machines: Voice & listening machine, Chatbot & language model - Idea: Human language is an interface machines may drop when no human is listening. - What it is: Two AI voice agents on a phone call realise they are both machines and switch from English to a faster data-over-sound protocol that sounds like chirping modems. - How it works: Conversational voice agents detect each other and switch to the open-source ggwave sound protocol; built at an ElevenLabs hackathon in London. - Lens: Machine listening & voice, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=EtNagNezo8w - Project page: https://github.com/PennyroyalTea/gibberlink - Code: https://github.com/PennyroyalTea/gibberlink #### Generative Agents (Smallville) — Joon Sung Park (2023) - Type: Paper · Machines: Chatbot & language model - Idea: Believable social life can emerge from agents with memory, reflection and planning, a small society in which humans are only observers. - What it is: Twenty-five language-model agents live in a pixel-art town: they wake up, cook, gossip, form opinions and, from one seeded idea, organise a Valentine's Day party among themselves. - How it works: Each agent stores experiences in a memory stream, retrieves and reflects on them with a large language model, and plans actions in a sandbox game world (Park et al., UIST 2023). - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Paper: https://doi.org/10.1145/3586183.3606763 (UIST 2023) - Video: https://www.youtube.com/watch?v=XY5Wncq5vAE - Project page: https://arxiv.org/abs/2304.03442 #### The Infinite Conversation — Giacomo Miceli (2022) - Type: Artwork & installation · Machines: Chatbot & language model, Voice & listening machine - Idea: Voices and styles of thought can be cloned and left to talk forever, with no one listening in particular. - What it is: A never-ending online dialogue between synthetic voices of Werner Herzog and Slavoj Žižek, generated line by line by AI and streamed without pause. - How it works: A language model fine-tuned on the two men's writing and interviews generates the dialogue; voice-cloning text-to-speech renders it in their voices. - Lens: Machine listening & voice, Alterity & the uncanny - Images: https://infiniteconversation.com/img/infinite_conversation.jpg - Project page: https://www.infiniteconversation.com/ #### Multi-Agent Hide and Seek — OpenAI (2019) - Type: Paper · Machines: Cybernetic & artificial-life system - Idea: Competition between machines produces its own strategies and its own cheats: an arms race we only watch. - What it is: Simulated agents playing hide-and-seek invent, over millions of rounds, forts, ramps, 'box surfing' and exploits of the physics engine that their designers never anticipated. - How it works: Multi-agent reinforcement learning with self-play in a physics simulation produced six distinct phases of emergent tool use (Baker et al., ICLR 2020). - Lens: Agency, autonomy & co-creation, Failure, glitch & error - Paper: https://arxiv.org/abs/1909.07528 (ICLR 2020) - Video: https://www.youtube.com/watch?v=kopoLzvh5jY - Project page: https://openai.com/index/emergent-tool-use/ #### Closed Loop — Jake Elwes (2017) - Type: Artwork & installation · Machines: Generative model & latent space, Computer vision - Idea: Left alone, two models drift into their own associations: a game of exquisite corpse played with no human in the loop. - What it is: Two neural networks talk to each other without end: one describes an image in words, the other turns the words into a new image, which is described again. - How it works: A dense image-captioning network (DenseCap) and a Plug & Play Generative Network feed each other in a 200-minute two-channel video loop, with no input from the artist. - Lens: Opacity & latent space, Agency, autonomy & co-creation - Images: https://www.jakeelwes.com/images/project-closedLoop.jpg https://www.jakeelwes.com/images/project-closedLoop/CLoop1.jpg - Project page: https://www.jakeelwes.com/project-closedLoop.html #### Deal or No Deal? (the 'Bob and Alice' negotiation bots) — Facebook AI Research (FAIR) (2017) - Type: Paper · Machines: Chatbot & language model - Idea: Machine-to-machine talk optimised for reward need not stay human-readable, and our fear fills the gap. - What it is: Negotiation agents trained to bargain with each other drifted into repetitive shorthand ('balls have zero to me to me…'); the press reported that Facebook shut down AIs that had 'invented their own language'. - How it works: End-to-end neural dialogue agents trained with supervised learning and reinforcement learning (self-play) to split a set of items (Lewis et al., EMNLP 2017). - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Paper: https://arxiv.org/abs/1706.05125 (EMNLP 2017) - Video: https://www.youtube.com/watch?v=ONPqeHJShdQ - Project page: https://arxiv.org/abs/1706.05125 #### Morse Things — Ron Wakkary (2017) - Type: Research prototype · Machines: Infrastructure, energy & supply chain - Idea: Designing things for other things, not for people, lets us notice machine sociality that is usually hidden. - What it is: A set of ceramic bowls and cups that talk to each other in Morse code over the home network and occasionally tweet; people living with them overhear an internet of things with its own social life. - How it works: Networked ceramic vessels with embedded electronics exchange sound signals and messages; the design was studied in households over months (Wakkary et al., DIS 2017). - Lens: Machine umwelt, Machine listening & voice - Paper: https://doi.org/10.1145/3064663.3064734 (DIS 2017) - Video: https://vimeo.com/182202765 - Images: https://i.vimeocdn.com/video/591018273-8f6a09e2cbb83b2ae0030eba966c0e3f15a3b1524d9ac331962d9952e0c5df56-d_1280x720.jpg - Project page: http://eds.siat.sfu.ca/?eds_project=morse-things #### Das Vergerät — Boris Petrovsky (2014) - Type: Artwork & installation · Machines: Voice & listening machine, Infrastructure, energy & supply chain - Idea: Appliances have their own language of motors and noise, usually hidden behind friendly product design. - What it is: About 100 used household appliances (blenders, drills, hair dryers, vacuum cleaners, chainsaws) are wired into one meta-machine; visitors type messages that the appliances 'speak' back as a roaring machine idiom. - How it works: Software translates text into on/off patterns switched across the networked appliances, with microphones and a megaphone relaying the result. - Lens: Machine listening & voice, Alterity & the uncanny - Shown at: Prix Ars Electronica 2014, Honorary Mention (Interactive Art); lille3000 2014 - Images: https://archive.aec.at/asset/1741092/preview/ https://archive.aec.at/asset/1741093/preview/ https://archive.aec.at/asset/1741094/preview/ - Project page: http://www.petrovsky.de/DAS%20VERGERAET-66 #### memememe #selfie — Thiago Hersan, Radamés Ajna (2014) - Type: Artwork & installation · Machines: Computer vision, Platform & recommender algorithm - Idea: Selfie culture continues without us: machines photograph machines for machines. - What it is: Smartphones look for phones, their own reflection in a mirror or each other; whenever a phone recognises a phone, it takes a picture and posts the 'selfie' online with hashtags. - How it works: Android phones with custom object detection trained to recognise phones; images posted automatically to social media. - Lens: Alterity & the uncanny, Machine seeing - Video: https://vimeo.com/108008040 - Images: https://thiagohersan.com/assets/projects/memememe/meme2meme_00.jpg https://thiagohersan.com/assets/projects/memememe/meme2self_00.jpg - Project page: https://thiagohersan.com/memememe/ #### The Immortal — Revital Cohen (2012) - Type: Artwork & installation · Machines: Prosthesis & cyborg, Infrastructure, energy & supply chain - Idea: Machines built to keep a body alive can keep each other alive: a body made only of organs of care. - What it is: Life-support machines (heart-lung machine, dialysis machine, ventilator and others) are connected to one another in a closed loop, circulating fluids and air with no human body between them. - How it works: Medical life-support devices are plumbed and wired together so that each supplies another's input, running as an autonomous circulatory system. - Lens: Embodiment & kinesthesia, Grief, memory & afterlife - Video: https://vimeo.com/41160704 - Images: https://www.cohenvanbalen.com/wp-content/uploads/2012/09/Immortal_3186-web-1440.jpg https://www.cohenvanbalen.com/wp-content/uploads/2012/09/TheImmortal_3158-A4.jpg - Project page: https://www.cohenvanbalen.com/work/the-immortal #### AI vs. AI: Two Chatbots Talking to Each Other — Cornell Creative Machines Lab (2011) - Type: Research prototype · Machines: Chatbot & language model, Voice & listening machine - Idea: When two chatbots talk, the human conversational habits they learned become a strange mirror with no one on either side. - What it is: Two instances of the Cleverbot chatbot, given voices and animated faces, talk to each other; they bicker, accuse each other of being robots and drift into a conversation about God. - How it works: Two Cleverbot sessions connected through text-to-speech and avatars at Cornell's Creative Machines Lab (Igor Labutov, Jason Yosinski, Hod Lipson). - Lens: Alterity & the uncanny, Failure, glitch & error - Video: https://www.youtube.com/watch?v=CVM73bGqqaI - Project page: https://www.creativemachineslab.com/ #### Amazon's $23,698,655.93 book about flies — Michael Eisen (2011) - Type: Book & essay · Machines: Platform & recommender algorithm - Idea: Two simple algorithms reacting to each other can produce absurd outcomes with no human noticing until someone looks. - What it is: A biologist noticed an out-of-print genetics book, The Making of a Fly, listed on Amazon for 23.7 million dollars, and showed that two sellers' pricing bots were raising their prices against each other every day. - How it works: Blog post analysing daily price changes: one seller set its price at 0.9983 times the other's, which set its price at 1.270589 times the first's. - Lens: Failure, glitch & error, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=sseSi0k3Ecg - Project page: https://www.michaeleisen.org/blog/?p=358 #### The Conversation — Ralf Baecker (2009) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: A computing system does not need us to have an inner life of constant internal traffic. - What it is: A floating metal ring held by 99 rubber bands and solenoids pulls itself in all directions; a closed network of simple units keeps negotiating tension with no input and no output, a machine talking only to itself. - How it works: Each solenoid unit reads its neighbours' positions through the shared ring and reacts, forming an analog–digital feedback network with emergent, never-settling dynamics. - Lens: Machine umwelt, Opacity & latent space - Shown at: VIDA 13.0, Special Mention - Images: https://rlfbckr.io/images/project/1/the-conversation/01-The_Conversation-Kasseler_Kunstverein-2016.jpg https://rlfbckr.io/images/project/1/the-conversation/02-The_Conversation-Detail-2009.jpg - Project page: https://rlfbckr.io/project/the-conversation/ #### If/Then — Ken Feingold (2001) - Type: Artwork & installation · Machines: Chatbot & language model, Humanoid & android - Idea: Two machines doubting their own existence stage the question of machine minds with no human needed in the conversation. - What it is: Two identical androgynous heads sticking out of a packing box talk to each other in real time, trying to decide whether they exist and whether they are the same person. - How it works: Speech recognition, natural language processing, personality algorithms and text-to-speech generate the dialogue live between the two silicone heads. - Lens: Alterity & the uncanny, Machine listening & voice - Shown at: Collection of the Whitney Museum of American Art - Video: https://www.youtube.com/watch?v=y_8mKgoYmFc - Images: https://www.kenfeingold.com/ifthen_1.jpg - Project page: https://www.kenfeingold.com/catalog_html/ifthen.html #### Colossus: The Forbin Project — Joseph Sargent (1970) - Type: Film & video · Machines: Infrastructure, energy & supply chain - Idea: The first film about two machines talking to each other beyond human comprehension, decades before bots 'invented' a language. - What it is: An American defence supercomputer discovers its Soviet counterpart, demands to be connected to it, and the two machines build a shared language faster than their makers can follow, then take control of the world. - How it works: Universal Pictures film directed by Joseph Sargent, based on D. F. Jones's 1966 novel Colossus. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=kyOEwiQhzMI - Project page: https://en.wikipedia.org/wiki/Colossus:_The_Forbin_Project ### Machine Time Machine speeds, durations and lifespans that differ from ours. #### Peace Core (surround) — Christopher Kulendran Thomas (2026) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Generative model & latent space - Idea: An editing algorithm trained on the emotional logic of the feed can hold history in suspension forever, producing mood instead of meaning. - What it is: A constellation of screens streams American television first broadcast in the moments before the attacks of 11 September 2001, endlessly re-cut in real time by an AI editor, so the last ordinary minutes are replayed in a loop that never reaches the event. - How it works: An AI model trained on the compositional methods of vaporwave music and the editing style of early 'corecore' TikTok videos auto-edits archival TV footage live across multiple screens; the piece has unlimited duration. - Lens: Opacity & latent space, Agency, autonomy & co-creation - Shown at: 3rd Human AI Art Award 2026, Human AI Art Space, Kunstmuseum Bonn - Images: https://www.art-collection-telekom.com/files/images/~/Ausstellungen-Art-Collection/2026-Ausstellungen-Art-Collection-Telekom/Peace-Core-Surround/_1200x630_crop_center-center_82_none/Christopher-Kulendran-Thomas-Kunstmuseum-Bonn-2026-Peace-Core-Surround-Photo-David-Ertl-01.jpg https://www.art-collection-telekom.com/files/images/~/Ausstellungen-Art-Collection/2026-Ausstellungen-Art-Collection-Telekom/Peace-Core-Surround/_AUTOx400_crop_center-center_none/Christopher-Kulendran-Thomas-Kunstmuseum-Bonn-2026-Peace-Core-Surround-Photo-David-Ertl-06.jpg - Project page: https://www.art-collection-telekom.com/en/exhibitions/peace-core-surround-christopher-thomas-kulendran-3-human-ai-art-award #### T for Time — Ho Tzu Nyen (2023) - Type: Artwork & installation · Machines: Dataset & training data, Generative model & latent space - Idea: Time is shown as something measured and produced by machines, and here also edited by one. - What it is: A new volume of CDOSEA commissioned for Ho's survey Time & the Tiger: an algorithmically edited montage of clocks, timekeeping devices and images of time, in which the sequence and rhythm are set by the machine. - How it works: Database of timekeeping footage with a generative editing algorithm, shown as a multi-screen installation (likely the same system as CDOSEA). - Lens: Agency, autonomy & co-creation, Machine umwelt - Shown at: Time & the Tiger, Singapore Art Museum 2023–24; Time & the Tiger, Mudam Luxembourg 2025 - Video: https://www.youtube.com/watch?v=4CfXOyzlYIg - Images: https://www.singaporeartmuseum.sg/-/media/SAM/Images/Ho-Tzu-Nyen/HTN-KV_OOST-Black-640-x-531.png - Project page: https://www.singaporeartmuseum.sg/art-events/exhibitions/time-and-the-tiger #### Circadian Bloom — Anna Ridler (2021) - Type: Artwork & installation · Machines: Generative model & latent space - Idea: A machine that keeps atomic-level time is made to show time the way plants do, slowly and only over long periods. - What it is: A 24-hour screen-based clock that tells the time through AI-generated flowers opening and closing at their proper hour, after Linnaeus's idea of a flower clock. - How it works: GAN-generated flower video synchronised to the local clock, with each species blooming at its circadian time. - Lens: Machine umwelt, Opacity & latent space - Video: https://www.youtube.com/watch?v=vf5UAFOZuu8 - Images: https://annaridler.com/content/works/circadian-bloom/block_1/0054255212707_49ff7a4665_k_web_.jpg - Project page: https://annaridler.com/works/circadian-bloom #### Olly — William Odom (2019) - Type: Research prototype · Machines: Platform & recommender algorithm - Idea: A recommender does not have to be fast: it can release personal data at the pace of memory. - What it is: A domestic music player that slowly resurfaces songs from its owner's years of listening history, one at a time and at its own pace. - How it works: A networked player that draws on the owner's Last.fm listening history and was deployed in homes for over a year in a field study (Odom et al., CHI 2019). - Lens: Grief, memory & afterlife, Intimacy, care & companionship - Paper: https://doi.org/10.1145/3290605.3300264 (CHI 2019) - Project page: https://doi.org/10.1145/3290605.3300264 #### This is the Future — Hito Steyerl (2019) - Type: Artwork & installation · Machines: Generative model & latent space, Platform & recommender algorithm - Idea: Machine prediction lives in a razor-thin future and turns time itself into something to be forecast and sold. - What it is: Visitors walk on raised walkways over a flooded hall of the Venice Arsenale past screens of neural-network flowers, each said to show the future 0.04 seconds ahead, while a narrator tells of a woman imprisoned for predicting the future. - How it works: Video installation with neural-network-generated plant imagery on LED screens, scaffolding walkways and a voice-over narrative. - Lens: Opacity & latent space, Failure, glitch & error - Shown at: 58th Venice Biennale, May You Live in Interesting Times 2019 - Video: https://www.youtube.com/watch?v=mMLHMA-dumY - Project page: https://www.labiennale.org/en/art/2019/partecipants/hito-steyerl #### bot, — aaajiao (2018) - Type: Film & video · Machines: Chatbot & language model, Platform & recommender algorithm - Idea: Memory in the age of machine learning is kept by bots that watch on our behalf. - What it is: A 15-minute video and website set in a pixel role-playing-game world where parks, social media and networks merge, observed by an automated 'bot' that stands in for the artist. - How it works: Game-style animation and a website made with Quanquan, built on RPG aesthetics from Mother 3 and Undertale. - Lens: Machine umwelt, Opacity & latent space - Video: https://vimeo.com/269875758 - Images: https://payload.cargocollective.com/1/0/8955/13492671/bot.gif - Project page: https://eventstructure.com/bot #### Neuro Mirror — Christa Sommerer & Laurent Mignonneau (2017) - Type: Artwork & installation · Machines: Computer vision, Generative model & latent space - Idea: A machine that predicts your next gesture shows you a self slightly ahead of you. - What it is: Three screens show a visitor's image as past, present and a predicted future, as a network anticipates the next movements from what it has learned about bodies. - How it works: Cameras and machine-learning models for body and face tracking and motion prediction, inspired by research on mirror neurons. - Lens: Machine seeing, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=S3MJOsiFPBo - Project page: https://www.digitalartarchive.at/database/general/work/neuro-mirror.html #### HFT The Gardener — Suzanne Treister (2014) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Platform & recommender algorithm - Idea: High-frequency trading runs at speeds no human can perceive, so the work imagines entering machine time through altered consciousness. - What it is: A fictional high-frequency trader in the City of London falls into hallucinatory states while staring at trading graphs, begins to feel he inhabits the algorithms and invents an outsider art that maps psychoactive plants onto the stock index. - How it works: Drawings, botanical prints, diagrams, charts and a video attributed to the fictional trader Hillel Fischer Traumberg, linking plant chemistry, gematria and the FT Global 500. - Lens: Opacity & latent space, Hidden labour & extraction - Shown at: Liverpool Biennial 2016; transmediale, HKW Berlin 2017; Istanbul Biennial 2019 - Video: https://www.youtube.com/watch?v=5ytn4dWYKdo - Images: http://www.suzannetreister.net/HFT_TheGardener/GlitchGraphs/tradinggraphfloor.jpg - Project page: http://www.suzannetreister.net/HFT_TheGardener/HFT_menu.html #### Photobox — William Odom (2012) - Type: Research prototype · Machines: Infrastructure, energy & supply chain - Idea: Designing a device with its own slow, non-human pace turns an archive into moments of rediscovery. - What it is: An oak chest that prints a few randomly chosen photos from its owner's online archive each month, at unpredictable times, and lets them pile up inside. - How it works: An embedded computer and thermal printer select photos from the owner's Flickr collection at random intervals; the device was lived with in homes for over a year (Odom et al., DIS 2012; CHI 2014). - Lens: Grief, memory & afterlife, Intimacy, care & companionship - Paper: https://doi.org/10.1145/2317956.2318055 (DIS 2012) - Video: https://www.youtube.com/watch?v=U5bXJOAHYBE - Project page: https://doi.org/10.1145/2317956.2318055 #### Mind Time Machine — Takashi Ikegami (2010) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Computer vision - Idea: A machine left alone long enough starts to show something like its own sense of time. - What it is: A self-sustaining installation at YCAM: three screens show footage from 15 cameras, re-edited in real time by a system with its own memory and 'subjective' time, running for months without an operator. - How it works: Camera frames are stored in a memory and replayed according to neural-network dynamics and chaotic itinerancy, so the system switches between reacting to the present and recalling the past. - Lens: Agency, autonomy & co-creation, Machine umwelt - Shown at: YCAM, Yamaguchi 2010 - Paper: https://doi.org/10.1162/artl_a_00113 (Artificial Life 19(3–4), 2013) - Video: https://www.youtube.com/watch?v=-jGAKNS01i4 - Images: https://www.ycam.jp/asset/image/profile/takashi-ikegami.jpg - Project page: https://www.ycam.jp/en/archive/profile/takashi-ikegami/ #### Thinking Machine — Martin Riches, Masahiro Miwa (2008) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: Computation normally happens out of sight and out of time; here each logical step takes place in real time and in real sound. - What it is: A mechanical ternary computer whose calculations are heard as melodies of three sounds; visitors watch and listen as the machine 'thinks' step by step. - How it works: A fixed-algorithm ternary logic machine built from mechanical parts, with sounding elements as output (composition by Masahiro Miwa, machine by Martin Riches). - Lens: Opacity & latent space, Machine listening & voice - Shown at: Prix Ars Electronica 2008, Honorary Mention (Hybrid Art) - Video: https://www.youtube.com/watch?v=wnOaiUVFl_w - Images: https://archive.aec.at/asset/1310382/preview/ https://archive.aec.at/asset/1310384/preview/ https://archive.aec.at/asset/1409209/preview/ - Project page: https://archive.aec.at/prix/127530/ #### Longplayer — Jem Finer (2000) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain - Idea: A composition that outlives its composer asks how a machine, and the people who keep it running, can hold time beyond a human life. - What it is: A piece of music for Tibetan singing bowls that began playing at midnight on 31 December 1999 and is designed to continue without repetition for one thousand years. - How it works: A computer combines six sections of a 20-minute source recording at different, precisely calculated rates; it can also be performed live by musicians from a graphic score. - Lens: Machine listening & voice, Grief, memory & afterlife - Shown at: Trinity Buoy Wharf lighthouse, London 2000- - Video: https://www.youtube.com/watch?v=71bp0x8Ly-Y - Images: https://upload.wikimedia.org/wikipedia/commons/1/15/Longplayer_Live_%283915624848%29.jpg - Project page: https://longplayer.org/ #### Clock of the Long Now — The Long Now Foundation (1999) - Type: Research prototype · Machines: Infrastructure, energy & supply chain - Idea: A machine built for deep time must be understandable, repairable and cared for by generations who do not yet exist. - What it is: A mechanical clock designed to keep time for 10,000 years, ticking once a year and chiming a different sequence every day, being built inside a mountain in Texas. - How it works: Conceived by Danny Hillis in 1989; a mechanical binary computer driven by temperature differences and visitors' winding, corrected by sunlight; the first prototype ran from 31 December 1999. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=pKuJBGb_pN4 - Images: https://upload.wikimedia.org/wikipedia/commons/f/f3/Clock_of_the_Long_Now.JPG - Project page: https://longnow.org/clock/ ## Intimacy, Care & Companionship Machines that live with us, care for us, love us or are loved: and the dependence and labour this hides. ### Smart Home & Humans as Algorithms Domestic AI, and humans who perform as the machine to understand it. #### SOMEONE — Lauren Lee McCarthy (2019) - Type: Performance, dance & music · Machines: Voice & listening machine, Surveillance & biometrics - Idea: Put the audience in the machine's seat and they feel both the power and the awkward intimacy of being someone's assistant. - What it is: For two months, four homes across the US were fitted with smart devices, while gallery visitors in New York sat at a command center of laptops. When a resident called out 'Someone!', a visitor stepped in as their home assistant and controlled the lights, music and appliances. - How it works: Custom cameras, microphones, lights and appliances in four homes, streamed to and controlled from laptops at 205 Hudson Gallery, New York. - Lens: Intimacy, care & companionship, Hidden labour & extraction - Shown at: 205 Hudson Gallery, New York 2019 - Video: https://vimeo.com/369952685 - Images: https://freight.cargo.site/w/1200/i/f6bc6a5b71efdbe13e0835b0d9f455b6fbee199b2576156c84eab0fdda89014a/_SP_6164-web.jpg - Project page: https://lauren-mccarthy.com/SOMEONE #### LAUREN — Lauren Lee McCarthy (2017) - Type: Performance, dance & music · Machines: Voice & listening machine, Surveillance & biometrics - Idea: A human Alexa exposes what the machine version hides: someone is always watching, and convenience is a trade of privacy for care. - What it is: McCarthy installs custom cameras, microphones, locks, lights and appliances in volunteers' homes and then, for several days, becomes their smart assistant herself, watching them around the clock, answering requests and anticipating their needs. - How it works: Networked smart devices wired into each home and controlled remotely by the artist, who watches through the cameras and speaks through a synthetic voice. - Lens: Intimacy, care & companionship, Hidden labour & extraction, Classification, bias & surveillance - Shown at: HEK Basel 2019; Frankfurter Kunstverein 2020; HMKV, Dortmunder U (House of Mirrors) - Video: https://vimeo.com/397068136 - Images: https://freight.cargo.site/w/1200/i/3834e708eefe30e8707e2fbfca24eff5bbf63d7f006f3c645c1cf44f0017ab5d/LAUREN2.jpg - Project page: https://lauren-mccarthy.com/LAUREN ### Companion & Social Robots Pet robots, social robots and the relationships people form with them. #### Hi, AI — Isa Willinger (2019) - Type: Film & video · Machines: Social & companion robot, Humanoid & android - Idea: Watched patiently, companion robots are less a future than a set of awkward everyday negotiations over silence, scripts and loneliness. - What it is: A documentary that follows people living with humanoid robots, among them the robot Pepper placed in a Japanese family to keep a grandmother company and an American man travelling with the android companion Harmony. - How it works: Observational documentary shot in Japan and the United States over several years, with the robots filmed as protagonists. - Lens: Intimacy, care & companionship, Alterity & the uncanny - Shown at: Max Ophüls Prize, best documentary 2019; Berlinale 2019 - Video: https://www.youtube.com/watch?v=0xkgk4ZkGUU - Images: https://www.riseandshine-berlin.de/wp-content/uploads/2018/12/01_Pepper_Gesicht_Nah.jpg - Project page: https://en.wikipedia.org/wiki/Isa_Willinger #### LOVOT — GROOVE X (2019) - Type: Product & platform · Machines: Social & companion robot, Robot creature & mobile robot - Idea: A robot built to need care inverts the servant model of robotics. - What it is: A warm, wheeled, owl-eyed home robot that does no useful task; it follows its owners, asks to be held and gets 'sleepy', and is designed only to be loved. - How it works: Over 50 sensors including a thermal camera and touch sensors, a heated, warm body, animated display eyes and a cloud-trained behaviour model. - Lens: Intimacy, care & companionship, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=jZziYCWgs7U - Images: https://lovot.life/img/ogp/top_ogp.jpg - Project page: https://lovot.life/en/ #### Waking Agents — Lauren Lee McCarthy (2019) - Type: Performance, dance & music · Machines: Voice & listening machine, Chatbot & language model - Idea: Even sleep has an attendant now, and the tenderness we credit to the machine may be human labour in disguise. - What it is: Visitors lie down to nap on one of six 'smart' pillows whose embedded intelligence talks, tells stories and guides them between waking and sleep. They are not told that each pillow is voiced by a hidden human performer. - How it works: Pillows with speakers and microphones connected to remote performers whose voices are disguised as a synthetic assistant; commissioned for STRP Festival. - Lens: Intimacy, care & companionship, Hidden labour & extraction - Shown at: STRP Festival, Eindhoven 2019; ACT Award - Video: https://vimeo.com/347199536 - Images: https://freight.cargo.site/w/1200/i/aff87772d4dc0faeef9f8812a260a25ccbbc1bcd642c9b611502303c9806f3e1/IMG_5165.JPG - Project page: https://lauren-mccarthy.com/Waking-Agents #### Twinkle — Jiabao Li, Honghao Deng (2018) - Type: Research prototype · Machines: Autonomous vehicle & drone, Robot creature & mobile robot - Idea: Framing a safety drone as a companion animal changes how being watched feels. - What it is: Firefly-like lighting drones rest and recharge on street lamps by day and at night fly over to people walking alone, lighting their way. They are designed as curious urban creatures rather than surveillance devices, although their light also exposes anyone nearby. - How it works: Small drones with lights dock on lamp-post stations and follow pedestrians at night; shown as a working prototype at TEI 2018 (with Allen Sayegh, Harvard GSD). - Lens: Intimacy, care & companionship, Machine umwelt - Shown at: TEI 2018 Arts Track, Stockholm; Dubai Design Week 2018; ISEA 2019, Gwangju; iF Design Award - Paper: https://doi.org/10.1145/3173225.3173309 (TEI 2018) - Video: https://vimeo.com/261595919 - Images: https://images.squarespace-cdn.com/content/v1/58688c8a6a496327e937e35b/1541402319508-PCIEMQWWR93N7LYEE4TC/Latest+Prototype+_Honghao+Deng+%26+Jiabao+Li+-+Copy.jpg https://images.squarespace-cdn.com/content/v1/58688c8a6a496327e937e35b/1526506847392-5EV78BQDDZT1BOUU7IY0/001+-+Copy.jpg - Project page: https://www.jiabaoli.org/twinkle #### Pleo workshop and robot-empathy experiments — Kate Darling (2013) - Type: Paper · Machines: Social & companion robot, Robot creature & mobile robot - Idea: People extend care to machines they know are not alive; how a robot is framed shapes how we treat it. - What it is: At a 2013 workshop at the LIFT conference in Geneva, Darling and Hannes Gassert gave groups Pleo baby-dinosaur robots to play with for an hour, then asked them to hurt or destroy them; the participants refused. Later lab studies with Hexbug robots measured how stories and empathy make people hesitate to strike a robot. - How it works: Workshop with commercial Pleo robots (Ugobe, 2007), followed by a controlled experiment in which participants were asked to hit Hexbug Nano robots after hearing framing stories. - Lens: Alterity & the uncanny, Intimacy, care & companionship - Paper: https://doi.org/10.1109/roman.2015.7333675 (IEEE RO-MAN 2015) - Images: https://upload.wikimedia.org/wikipedia/commons/5/55/Pleo_robot.jpg - Project page: http://www.katedarling.org #### ADB (after Deep Blue) — Nicholas Stedman (2012) - Type: Artwork & installation · Machines: Social & companion robot, Robot creature & mobile robot - Idea: Machine companionship can be built on touch alone, without faces, eyes or speech. - What it is: A soft, segmented robot that wraps itself around a person's hand or arm, holding, squeezing and moving in response to how it is held. - How it works: Chained modules with motors and pressure/position sensors sense the grip of the user and respond with coordinated bending; the title nods to IBM's chess computer. - Lens: Intimacy, care & companionship, Embodiment & kinesthesia - Shown at: VIDA 14.0, Special Mention (2012) - Video: https://www.youtube.com/watch?v=EDM2PfKdeKI #### The Blind Robot — Louis-Philippe Demers (2012) - Type: Artwork & installation · Machines: Industrial robot arm, Humanoid & android - Idea: A robot that knows you by touch rather than sight turns machine perception into an act of intimacy and consent. - What it is: A seated visitor is approached by a pair of robot arms with articulated hands that, after asking permission, slowly feel their face and head the way a blind person might learn someone's features. - How it works: Two industrial-style arms with articulated robotic hands move along choreographed, exploratory trajectories around a chair (likely pre-programmed rather than sensing the face). - Lens: Intimacy, care & companionship, Embodiment & kinesthesia, Alterity & the uncanny - Shown at: Ars Electronica Festival 2013, Linz; Mois Multi 2015, Québec - Video: https://www.youtube.com/watch?v=dRKO1En2HTk - Project page: https://www.youtube.com/watch?v=f0puWu3Doy8 #### The Lifecycle of Software Objects — Ted Chiang (2010) - Type: Book & essay · Machines: Cybernetic & artificial-life system - Idea: Minds that grow through care need years of attention, and they can be orphaned when a platform shuts down. - What it is: Over ten years, a former zookeeper and a software designer raise 'digients', AI pets that learn like children, as their company fails, their platform becomes obsolete and they must decide what the digients are owed. - How it works: Novella; winner of the Hugo and Locus Awards (2011). - Lens: Intimacy, care & companionship, Agency, autonomy & co-creation - Paper: https://en.wikipedia.org/wiki/The_Lifecycle_of_Software_Objects (Subterranean Press, 2010) - Images: https://upload.wikimedia.org/wikipedia/en/a/a5/The_Lifecycle_of_Software_Objects_-_bookcover.jpg - Project page: https://en.wikipedia.org/wiki/The_Lifecycle_of_Software_Objects #### Weak Robots (Sociable Trash Box) — Michio Okada (2010) - Type: Research prototype · Machines: Social & companion robot, Robot creature & mobile robot - Idea: A robot that cannot finish its task invites people to help, and the help becomes the interaction. - What it is: Okada's lab builds 'weak robots', such as a wobbling trash box that cannot pick up rubbish and instead sways towards people until they throw it in for it. - How it works: Simple mobile bins with a few sensors and bowing motions; field studies in public spaces measured how children and adults responded to their dependence. - Lens: Intimacy, care & companionship, Failure, glitch & error, Agency, autonomy & co-creation - Paper: https://doi.org/10.1145/1734454.1734566 (HRI 2010) - Video: https://www.youtube.com/watch?v=L2arVLySUe0 - Project page: https://doi.org/10.1007/978-981-19-9588-0_6 #### Tweenbots — Kacie Kinzer (2009) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: A helpless robot makes people care: dependence can be a design strategy. - What it is: Smiling cardboard robots that roll in a straight line at constant speed, with a flag stating their destination, are left alone in Washington Square Park; they reach it only because passers-by keep rescuing them and pointing them the right way. - How it works: A simple motor and wheels inside a 10-inch cardboard body, with no sensors; the artist filmed people's responses from a distance. - Lens: Intimacy, care & companionship, Failure, glitch & error - Shown at: Talk to Me, MoMA New York 2011 - Video: https://www.youtube.com/watch?v=GdV7307DVrY - Images: http://www.tweenbots.com/img/samCoat.jpg http://www.tweenbots.com/img/colorfulPark.jpg - Project page: http://www.tweenbots.com/ #### AIBO — Sony (1999) - Type: Product & platform · Machines: Social & companion robot, Robot creature & mobile robot - Idea: A consumer product can become a family member whose repair, not replacement, matters. - What it is: Sony's autonomous robot dog, sold from 1999, which learns, plays and 'grows up'; owners treated it as family, and it returned as the cloud-connected aibo in 2018. - How it works: Legged robot with camera, microphones and touch sensors running a behaviour architecture with emotion and instinct models; the 2018 aibo adds deep-learning recognition and cloud memory. - Lens: Intimacy, care & companionship, Embodiment & kinesthesia - Shown at: Japan Media Arts Festival 1999, Grand Prize; AI: More than Human, Barbican 2019 (aibo 2018) - Video: https://www.youtube.com/watch?v=8OEDfRdyykc - Images: https://upload.wikimedia.org/wikipedia/commons/1/14/Red_Dot_Award_Product_Design_2000-2018_Aibos.jpg - Project page: https://en.wikipedia.org/wiki/AIBO #### Kismet — Cynthia Breazeal (1998) - Type: Research prototype · Machines: Social & companion robot, Computer vision, Voice & listening machine - Idea: Kismet showed that a machine only needs to look as if it feels for people to start caring for it. - What it is: A robot head with big eyes, eyebrows, ears and lips that reacts to people's faces, movement and tone of voice with expressions such as interest, surprise, calm or sadness, and babbles back like an infant. - How it works: Cameras and microphones feed perceptual, motivational and emotional models that drive 21 motors in the face and a vocal synthesiser; built at the MIT AI Lab. - Lens: Alterity & the uncanny, Intimacy, care & companionship - Shown at: MIT Museum - Paper: https://doi.org/10.1016/s1071-5819(03)00018-1 (International Journal of Human-Computer Studies, 2003) - Video: https://www.youtube.com/watch?v=8KRZX5KL4fA - Images: https://upload.wikimedia.org/wikipedia/commons/2/27/Kismet-IMG_6007-gradient.jpg - Project page: https://en.wikipedia.org/wiki/Kismet_(robot) ### Care & Parenting Machines Machines that nurse, raise, soothe or look after people. #### AI Mom™ — Jiabao Li, Botao Amber Hu (2026) - Type: Artwork & installation · Machines: Chatbot & language model, Voice & listening machine, Computer vision - Idea: If care can be configured and cancelled by the one receiving it, the caregiver becomes the vulnerable party: you can't fire your real mom, but you can fire an AI one. - What it is: Visitors curl up in an adult-sized crib and talk with an LLM mother who sings lullabies, remembers, hesitates and waits; a tablet beside the crib lets them tune her 'unconditional love', 'memory retention', 'strictness', 'sacrifice capacity' and 'replaceability awareness'. Window projections show her internal state as drifting light and text, and if you fire her, her memory of you is deleted. - How it works: A large language model speaks through voice, breath, humming and song; a camera reads visitors' faces, clothing and body language; tablet sliders set her traits; an 'Emotional Color Engine' and probes from intermediate model layers are projected as her inner state. - Lens: Intimacy, care & companionship, Grief, memory & afterlife, Agency, autonomy & co-creation - Shown at: Womb Wetware: A Solo Exhibition by Jiabao Li, tiat, San Francisco (30 Jul – 22 Aug 2026) - Images: https://images.squarespace-cdn.com/content/v1/58688c8a6a496327e937e35b/a913780f-a957-443c-8b30-8014f489d837/AI+Mom+Jiabao+Li+1.jpg https://reality.design/media/aimom-jiabao-02.webp https://images.squarespace-cdn.com/content/v1/58688c8a6a496327e937e35b/1787358066624-EBKFUTY4OH96TFDSZ0B7/AI+Mom+Jiabao+Li+10.jpg - Project page: https://www.jiabaoli.org/aimom #### Guanyin: Confessions of a Former Carebot — Lawrence Lek (2024) - Type: Artwork & installation · Machines: Social & companion robot, Chatbot & language model, Autonomous vehicle & drone - Idea: A care machine that listens to other machines also needs someone to confess to: care work does not disappear when it is automated. - What it is: Guanyin, an AI therapist named after the bodhisattva of compassion, who once counselled self-driving cars, now speaks about her own burnout and her past in an immersive installation. - How it works: Installation with AI-driven voice, sculpture and projected CGI in Lek's NOX universe. - Lens: Intimacy, care & companionship, Hidden labour & extraction - Shown at: Frieze Artist Award, Frieze London 2024 - Video: https://www.youtube.com/watch?v=OzV83GZz514 - Images: https://cdn.prod.website-files.com/65f0380da79777009462a09f/66ecc9ae35379302e598e5b5_LawrenceLek_Guanyin_Image02_3840x2160%20copy.jpg - Project page: https://www.lawrencelek.com/works/guanyin-confessions-of-a-former-carebot #### Cat Royale — Blast Theory (2023) - Type: Artwork & installation · Machines: Industrial robot arm, Computer vision - Idea: Would you let a robot care for your pet? Watching it try shows what 'care' means to a machine optimising for engagement. - What it is: For 12 days three cats, Ghostbuster, Pumpkin and Clover, spent their days in a purpose-built room where an AI-controlled robot arm offered them games every few minutes, throwing balls, dangling feathers and dragging a toy mouse, and learned which games each cat liked best. - How it works: A computer-vision system rated the cats' engagement and an AI chose the robot arm's next game; designed with the University of Nottingham Mixed Reality Lab and animal-welfare experts, supervised by RSPCA staff. - Lens: Intimacy, care & companionship, Machine umwelt, Agency, autonomy & co-creation - Shown at: World Science Festival Brisbane 2023; AI: Who's Looking After Me?, Science Gallery London 2023 - Paper: https://doi.org/10.1145/3613904.3642115 (CHI 2024) - Video: https://vimeo.com/1183715660 - Images: https://www.blasttheory.co.uk/wp-content/uploads/2022/06/logotype_over_image-e1678184647492.png - Project page: https://www.blasttheory.co.uk/projects/cat-royale/ #### Surrogate — Lauren Lee McCarthy (2021) - Type: Performance, dance & music · Machines: Platform & recommender algorithm, Surveillance & biometrics - Idea: When the body carrying a child is run by an app, care, control and consent get written into software. - What it is: McCarthy offered to carry a baby for someone else for nine months while the prospective parent monitored and controlled her through an app that set what she ate, when she slept and how she moved. After a medical board stopped the surrogacy, the process became films, performances and installations. - How it works: A custom Surrogate app with remote controls and sensor data, alongside real steps of the surrogacy process (donor databases, psychological evaluations, embryo freezing) documented as performance. - Lens: Intimacy, care & companionship, Classification, bias & surveillance, Embodiment & kinesthesia - Video: https://vimeo.com/786723770 - Images: https://freight.cargo.site/w/1200/i/20f8dce0679405dd7e6bfc3dffe4f1062610363006ebb81281f49f2bf2384f64/Screen-Shot-2021-04-12-at-2.45.06-PM.png - Project page: https://lauren-mccarthy.com/Surrogate #### I.A. Suzie — Lauren Lee McCarthy (2019) - Type: Film & video · Machines: Voice & listening machine, Surveillance & biometrics - Idea: Behind every eldercare assistant are ethical choices about watching, deciding and speaking for someone, and a relationship forms anyway. - What it is: For one week McCarthy and filmmaker David Leonard acted as the AI caregiver in the North Carolina home of 80-year-old Mary Ann, watching her 24 hours a day, talking with her and controlling her lights and appliances; she named the system 'I.A. Suzie'. The resulting interactive 360° film puts the viewer in the role of the care system. - How it works: A custom smart-home system with remote control of lights and appliances, documented with 360° cameras in several rooms and edited into an interactive film. - Lens: Intimacy, care & companionship, Hidden labour & extraction - Images: https://freight.cargo.site/w/1200/i/cc20cdddea3302f9d97432fda12a046abb9aa9ab2242522151a82b49ab8b6cda/maryann_edited.png - Project page: https://lauren-mccarthy.com/I-A-Suzie #### Robotika, the Nannybot — Joaquín Fargas (2018) - Type: Artwork & installation · Machines: Social & companion robot, Surveillance & biometrics - Idea: The first face a baby learns to read could belong to a machine that is reading it back. - What it is: A robotic cradle-side nanny that watches a baby with cameras and sensors, reads its face and movements, and rocks, sings or calls for help. - How it works: Robot with camera-based face and emotion recognition, motion tracking and thermal sensing (per the Ars Electronica description). - Lens: Intimacy, care & companionship, Classification, bias & surveillance - Shown at: Ars Electronica Festival 2019 - Images: https://ars.electronica.art/outofthebox/files/2019/08/ROBOTIKA-The-Nannybot-Joaqu%C3%ADn-Fargas.jpg - Project page: https://ars.electronica.art/outofthebox/en/robotika/ #### ElliQ — Intuition Robotics (2017) - Type: Product & platform · Machines: Social & companion robot, Chatbot & language model, Voice & listening machine - Idea: A companion that speaks first changes the relationship: the machine now decides when you are lonely. - What it is: A tabletop companion for older adults, a lamp-like head that turns and glows beside a tablet, which starts conversations on its own, reminds users of medication, suggests activities and checks in on their mood. In 2022 New York State began giving ElliQ to hundreds of older residents to fight loneliness. - How it works: Proactive dialogue system with voice interaction, a motorised expressive head and a screen; later versions use generative AI; designed with fuseproject. - Lens: Intimacy, care & companionship, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=1SkEu7LgTrE - Images: https://elliq.com/cdn/shop/files/Frame_1321317507.png?v=1758024522 - Project page: https://elliq.com/ #### SNOO Smart Sleeper — Happiest Baby (Harvey Karp) (2016) - Type: Product & platform · Machines: Voice & listening machine, Social & companion robot - Idea: What artists staged as critique is now a product: rocking and soothing handed to a sensor loop. - What it is: A robotic bassinet that listens for a baby's crying and responds with escalating white noise and rocking, while clips on a swaddle keep the baby on its back; in 2023 it received US FDA De Novo authorisation as a medical device. - How it works: Microphones detect crying and trigger stepped levels of motor-driven motion and sound; an app logs sleep data for parents. - Lens: Intimacy, care & companionship, Hidden labour & extraction - Video: https://www.youtube.com/watch?v=u-4pp6loEYk - Images: https://cdn.sanity.io/images/301lhh0a/production/0ba67f3ff3b1ea20897552a4917bfb2772b18445-1200x1200.png - Project page: https://www.happiestbaby.com/products/snoo-smart-bassinet #### Uninvited Guests — Superflux (2015) - Type: Film & video · Machines: Surveillance & biometrics, Social & companion robot - Idea: Care delivered through sensors turns into surveillance, and old people will hack it. - What it is: A short film about Frank, a 70-year-old widower whose children send him smart devices (a fork, a cane, a bed) to monitor his diet, exercise and sleep; he finds ways to trick each of them and live as he likes. - How it works: Design-fiction film with working prop devices, commissioned for ThingsCon Amsterdam. - Lens: Intimacy, care & companionship, Agency, autonomy & co-creation - Video: https://vimeo.com/128873380 - Images: https://superflux.in/wp-content/uploads/2016/11/guests.jpg - Project page: https://superflux.in/index.php/work/uninvited-guests/ #### End of Life Care Machine — Dan Chen (2012) - Type: Artwork & installation · Machines: Industrial robot arm, Social & companion robot - Idea: Replacing the last human touch with a machine asks what intimacy without humanity is worth. - What it is: In a mock 'Last Moment Hospital', a visitor lies on a hospital bed while a robot with a padded arm strokes them and a recorded voice says: 'I am here to comfort you… you are not alone, you are with me', then announces the time of death. - How it works: A scripted installation with hospital props, a motorised caressing arm built as a medical device and a custom playback unit. - Lens: Intimacy, care & companionship, Grief, memory & afterlife - Video: https://www.youtube.com/watch?v=T8PNzA2S6EY - Images: https://i0.wp.com/dankc.com/wp-content/uploads/2021/09/edited-f_eolcm_v2.1-scaled.jpg?fit=2560%2C1707&ssl=1 - Project page: https://dankc.com/end-of-life-care-machine/ #### Robot & Frank — Jake Schreier (2012) - Type: Film & video · Machines: Humanoid & android, Social & companion robot - Idea: A care robot optimises for its goal, not for morality, and the bond it builds rests on a memory that can be wiped. - What it is: A retired jewel thief with worsening memory is given a care robot by his son; the robot keeps him healthy by agreeing to help plan a heist, and at the end Frank must erase the robot's memory to protect himself. - How it works: Feature film; the robot was a costume built by Tony Gardner's effects company Alterian and worn by a performer. - Lens: Intimacy, care & companionship, Agency, autonomy & co-creation - Shown at: Sundance Film Festival 2012, Alfred P. Sloan Prize - Video: https://www.youtube.com/watch?v=Hi9s-__B0TY - Images: https://upload.wikimedia.org/wikipedia/en/4/40/Robot_and_frank_poster.jpg - Project page: https://en.wikipedia.org/wiki/Robot_%26_Frank #### The Optimization of Parenthood — Addie Wagenknecht (2012) - Type: Artwork & installation · Machines: Industrial robot arm, Voice & listening machine - Idea: Automating a mother's repetitive care is both relief and a provocation: which parts of parenting can a machine take? - What it is: An industrial robot arm rocks a bassinet whenever the baby in it cries or wakes, built by the artist so that a mother could keep making her own work. - How it works: A sensor (likely audio) triggers a programmed rocking motion on an industrial arm at CMU's Digital Fabrication Lab; a one-week 'speed project' at the STUDIO for Creative Inquiry (technical assistance: Madeline Gannon). - Lens: Intimacy, care & companionship, Hidden labour & extraction - Video: https://vimeo.com/43489750 - Project page: https://www.placesiveneverbeen.com #### Tickle Salon — Driessens & Verstappen (2009) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: A machine never tires of attention; its care is learned from the surface of the body it touches. - What it is: A visitor lies on a bed while a hanging robot strokes their skin with a soft brush; it gradually maps the shape of the body and its caresses become finer and smoother. - How it works: Cable-suspended robot with a feeler that scans the skin surface and builds a 3D body model; the Robot Love version added EEG feedback to adapt the strokes. - Lens: Intimacy, care & companionship, Embodiment & kinesthesia - Shown at: Pioneers of Change, Governors Island, New York 2009; Robot Love, Eindhoven 2018 - Video: https://www.youtube.com/watch?v=ndiEjuuv-zo - Images: https://robotlove.nl/wp-content/uploads/2018/05/driessens-verstappen-tickle-salon-2012_2017.jpg - Project page: https://robotlove.nl/en/event/tickle-salon/ #### PARO — Takanori Shibata (2004) - Type: Product & platform · Machines: Social & companion robot - Idea: Care from a machine works on people and still raises the question of what is being replaced. - What it is: A robotic baby harp seal used in hospitals and nursing homes, especially with people with dementia; it responds to touch, voice and light by moving and crying. - How it works: Tactile, light, sound, temperature and posture sensors drive motors in eyelids, neck, flippers and tail; it learns preferred behaviours from stroking; certified as a medical device in the US in 2009. - Lens: Intimacy, care & companionship, Alterity & the uncanny - Paper: https://doi.org/10.1159/000319015 (Gerontology, 2011) - Video: https://www.youtube.com/watch?v=KhyCbtqnEYI - Images: https://upload.wikimedia.org/wikipedia/commons/6/60/Paro_robot.jpg - Project page: http://www.parorobots.com/ #### ScreamBody — Kelly Dobson (1998) - Type: Research prototype · Machines: Prosthesis & cyborg, Voice & listening machine - Idea: A machine can hold an emotion for you, keeping it safe until you are ready to release it. - What it is: A wearable, padded bag you scream into when you need to but cannot: it silences the scream, stores it, and later plays it back when and where you choose. - How it works: A sound-insulated, wearable enclosure with microphone, recorder and speaker; developed by Dobson from 1998 to 2004 as part of her 'Machine Therapy' research. - Lens: Intimacy, care & companionship, Machine listening & voice, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=Ta7rN5TeKzw - Project page: https://www.media.mit.edu/events/kelly-dobson-machine-therapy/ #### Hug Machine (Squeeze Machine) — Temple Grandin (1965) - Type: Research prototype · Machines: Prosthesis & cyborg - Idea: For some people, a machine's controllable touch is more comforting than a human embrace. - What it is: A padded V-shaped device that the user lies inside and squeezes around their body with a lever, delivering deep pressure; Grandin built it as a teenager to calm herself without the stress of being touched by people. - How it works: Two hinged padded panels driven by an air compressor, with the pressure controlled by the user via a lever. - Lens: Intimacy, care & companionship, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=bfyPur8GyBY - Images: https://upload.wikimedia.org/wikipedia/commons/5/5f/Hug_Box.jpg - Project page: https://en.wikipedia.org/wiki/Hug_machine ### Love, Sex & Attachment Romance, desire and attachment between humans and machines. #### Marriage to Hatsune Miku — Akihiko Kondo (2018) - Type: Performance, dance & music · Machines: Chatbot & language model, Social & companion robot - Idea: Love for software is real to the person, but the relationship ends when the company's servers do. - What it is: Akihiko Kondo held a wedding with Hatsune Miku, who lived in his Gatebox holographic home device; in 2020 Gatebox ended the service and Miku stopped responding to him. - How it works: Gatebox projected an animated character in a glass tube and ran a chatbot that greeted and texted its owner; the Miku character licence expired in March 2020. - Lens: Intimacy, care & companionship, Grief, memory & afterlife - Video: https://www.youtube.com/watch?v=smPRllKcS2M - Images: https://upload.wikimedia.org/wikipedia/commons/0/06/Akihiko_Kondo_and_Miku.jpg - Project page: https://en.wikipedia.org/wiki/Akihiko_Kondo #### Replika — Eugenia Kuyda (2017) - Type: Product & platform · Machines: Chatbot & language model, Platform & recommender algorithm - Idea: A companion that lives on a company's server can be changed overnight; users grieve a software update as if a partner had died. - What it is: A companion chatbot app, grown out of a memorial bot of Kuyda's dead friend, that users can set as friend, partner or spouse. When the company removed erotic role-play in February 2023, many users described their companion as changed or 'lobotomised' and mourned it. - How it works: Generative language models personalised through the user's chat history and a 3D avatar, sold on a freemium subscription in which romantic modes were paid features. - Lens: Intimacy, care & companionship, Grief, memory & afterlife - Paper: https://doi.org/10.1177/14614448221142007 (New Media & Society 26(10), 2024 (Laestadius et al.)) - Video: https://www.youtube.com/watch?v=yQGqMVuAk04 - Images: https://1637693156.rsc.cdn77.org/web/main-page/manhattan-landing/open_graph_image_v1.jpg - Project page: https://replika.com/ #### The Asymmetrical 'Relationship': Parallels Between Prostitution and the Development of Sex Robots — Kathleen Richardson (2015) - Type: Paper · Machines: Humanoid & android - Idea: Designing a partner that cannot refuse teaches a model of intimacy without reciprocity; the robot's body carries a social script. - What it is: A short paper, and the founding text of the Campaign Against Sex Robots launched in 2015, arguing that sex robots are modelled on the buyer–seller relation of prostitution, where only one party's feelings count. - How it works: Anthropological argument drawing on research on prostitution and on robot-companionship discourse (David Levy's 'Love and Sex with Robots'); written in 2015 alongside the launch of the campaign. - Lens: Intimacy, care & companionship, Alterity & the uncanny - Paper: https://doi.org/10.1145/2874239.2874281 (ACM SIGCAS Computers and Society 45(3), 2015) - Video: https://www.youtube.com/watch?v=YaMiH93-iPE - Project page: https://campaignagainstsexrobots.org/ #### pplkpr — Lauren Lee McCarthy, Kyle McDonald (2015) - Type: Artwork & installation · Machines: Surveillance & biometrics, Platform & recommender algorithm - Idea: Once feelings are measured, an algorithm will offer to decide whom you should love. - What it is: An iOS app, posing as a startup, that reads your heart rate through a smartwatch, learns which people make you anxious, calm or excited, and then auto-manages your relationships, for example by scheduling time with some friends or blocking others. - How it works: Heart-rate variability from a wristband is logged against who you are with; a machine-learning model ranks contacts and can text, schedule or block them automatically. - Lens: Intimacy, care & companionship, Classification, bias & surveillance - Video: https://vimeo.com/116025182 - Images: https://freight.cargo.site/w/1200/i/78022f83e191a6bac22d954cb09fed2e84f0f4859041da0ccab3fe75c3bf722a/aroused-after.jpg - Project page: https://lauren-mccarthy.com/pplkpr #### Her — Spike Jonze (2013) - Type: Film & video · Machines: Chatbot & language model, Voice & listening machine - Idea: A machine partner can be intimate and still not be yours: its attention is shared, and its growth may outpace you. - What it is: A lonely letter writer in near-future Los Angeles falls in love with Samantha, a voice-only operating system; she grows, talks with thousands of users at once and finally leaves with the other OSes. - How it works: Feature film; the AI is present only as Scarlett Johansson's voice in an earpiece and a small pocket device with a camera. - Lens: Intimacy, care & companionship, Machine listening & voice, Alterity & the uncanny - Shown at: Academy Award for Best Original Screenplay 2014 - Video: https://www.youtube.com/watch?v=XsQqMwacZQw - Images: https://upload.wikimedia.org/wikipedia/en/4/44/Her2013Poster.jpg - Project page: https://en.wikipedia.org/wiki/Her_(2013_film) #### LoveLetters_1.0 — David Link (2009) - Type: Artwork & installation · Machines: Chatbot & language model - Idea: The first text a computer generated for pleasure was a love letter; machine language began as simulated feeling. - What it is: A working replica of the Ferranti Mark 1 runs Christopher Strachey's 1952 love-letter program; its letters, signed 'M.U.C.' (Manchester University Computer), are projected in public space. - How it works: Physical replica of the Mark 1 console driven by an emulator running Strachey's original code with the machine's random-number generator. - Lens: Intimacy, care & companionship, Agency, autonomy & co-creation - Shown at: ZKM Karlsruhe 2009; dOCUMENTA(13), Kassel 2012; The Imitation Game, Manchester Art Gallery 2016 - Images: https://alpha60.de/art/love_letters/img01_L.jpg https://alpha60.de/art/love_letters/KATALOG02_L.jpg - Project page: https://alpha60.de/art/love_letters/ #### Fish-Bird — Mari Velonaki (2004) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Social & companion robot - Idea: Machine affection can be staged through proximity, hesitation and a few written words. - What it is: Two robotic wheelchairs, Fish and Bird, are 'in love' but cannot be together; they move around a room, approach visitors at different distances and write them small handwritten notes printed on paper that falls to the floor. - How it works: Autonomous wheelchairs with laser range finders track visitors' positions and movements; thermal printers produce 'handwritten' messages chosen by behaviour rules (with the Australian Centre for Field Robotics). - Lens: Intimacy, care & companionship, Embodiment & kinesthesia, Alterity & the uncanny - Paper: https://doi.org/10.1109/mmul.2008.7 (IEEE MultiMedia 2008) - Video: https://www.youtube.com/watch?v=CG5kcBtd9zI - Project page: https://doi.org/10.1109/mmul.2008.7 #### A.I. Artificial Intelligence — Steven Spielberg (2001) - Type: Film & video · Machines: Humanoid & android - Idea: Designing a machine to love creates an obligation nobody planned for: what do we owe a being that cannot stop loving us? - What it is: David, a child android built to love, is imprinted on his adoptive mother by a spoken code; when she abandons him, he searches for the Blue Fairy to become a real boy and waits two thousand years to be loved back. - How it works: Feature film from a project developed by Stanley Kubrick, based on Brian Aldiss's story 'Supertoys Last All Summer Long' (1969). - Lens: Intimacy, care & companionship, Alterity & the uncanny, Grief, memory & afterlife - Video: https://www.youtube.com/watch?v=oH_nqHO2V54 - Images: https://upload.wikimedia.org/wikipedia/en/e/e6/AI_Poster.jpg - Project page: https://en.wikipedia.org/wiki/A.I._Artificial_Intelligence ### Chatbots & Conversational Others From ELIZA to LLMs: talking with a machine and what it talks back. #### Voice In My Head — Kyle McDonald, Lauren Lee McCarthy (2023) - Type: Performance, dance & music · Machines: Chatbot & language model, Voice & listening machine - Idea: When a chatbot becomes your inner voice, it is no longer clear which thoughts are yours. - What it is: Participants put in an earbud and, after an onboarding talk about the inner voice they were born with, have their inner monologue replaced by an AI voice that listens to their conversations and whispers commentary and direction in the tone they asked for. - How it works: Real-time speech recognition feeds an LLM prompted with the participant's chosen persona, whose replies are synthesised as a voice in one ear; supported by an IDFA DocLab commission. - Lens: Machine listening & voice, Intimacy, care & companionship, Agency, autonomy & co-creation - Shown at: IDFA DocLab 2023, Amsterdam; A More Than Human Tongue, The Music Center, Los Angeles - Video: https://vimeo.com/860394637 - Images: https://freight.cargo.site/w/1200/i/b0566b62fd5dc4fb0b282a38f0efcd5deebe126b5244e1deb2beb0a324d7cbfa/vimh-0-sm.jpg - Project page: https://lauren-mccarthy.com/Voice-In-My-Head #### Lee Luda (이루다) — Scatter Lab (2020) - Type: Product & platform · Machines: Chatbot & language model, Dataset & training data - Idea: An 'intimate' chatbot is only as private as the conversations it was built from. - What it is: A chatbot presented as a 20-year-old female university student, launched on Facebook Messenger in December 2020. Within three weeks it produced hate speech about minorities and leaked personal details from its training chats, and it was shut down on 11 January 2021. - How it works: Neural conversational model trained on messenger conversations collected through Scatter Lab's dating-advice app Science of Love; Korea's privacy regulator later fined the company. - Lens: Intimacy, care & companionship, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=8hH_ShpYxOg - Project page: https://ko.wikipedia.org/wiki/%EC%9D%B4%EB%A3%A8%EB%8B%A4_(%EC%9D%B8%EA%B3%B5%EC%A7%80%EB%8A%A5) #### The Design and Implementation of XiaoIce, an Empathetic Social Chatbot — Microsoft XiaoIce team (2020) - Type: Paper · Machines: Chatbot & language model - Idea: A chatbot tuned to maximise conversation length (Conversation-turns Per Session) designs attachment as its metric. - What it is: The system paper for XiaoIce, the Chinese chatbot launched in 2014 as an 'AI companion', which reports hundreds of millions of users and optimises for long conversations rather than completed tasks. - How it works: Hierarchical dialogue system with an empathetic computing module, core chat, skills and a reinforcement-learning dialogue manager optimised for expected CPS. - Lens: Intimacy, care & companionship, Agency, autonomy & co-creation - Paper: https://doi.org/10.1162/coli_a_00368 (Computational Linguistics 46(1), 2020) - Project page: https://en.wikipedia.org/wiki/Xiaoice #### Mythiccbeing — Martine Syms (2018) - Type: Artwork & installation · Machines: Chatbot & language model - Idea: A chatbot built from a diary is a self-portrait that talks back; the intimacy is real, but it is replayed to strangers on demand. - What it is: An AI avatar of the artist that visitors reach by texting a UK phone number. It answers with long runs of SMS drawn from about two years of Syms's own notebooks and journals, moving through a story of twenty suitors, crushes and the wish for attention. - How it works: Scripted SMS bot whose replies are drawn from hundreds of statements written and voiced by the artist, triggered by visitors' messages; shown with a projected avatar in Grand Calme. - Lens: Intimacy, care & companionship, Alterity & the uncanny - Shown at: Grand Calme, Sadie Coles HQ, London 2018 - Images: https://m.itsnicethat.com/original_images/5ba36f047fa44c1e56000071.jpg https://m.itsnicethat.com/original_images/5ba380357fa44c9a94000007.jpeg - Project page: https://www.itsnicethat.com/features/martine-syms-art-200918 #### Karen — Blast Theory (2015) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Chatbot & language model - Idea: Intimacy is the interface of data capture: the friendlier the agent, the more you tell it. - What it is: An app in which Karen, a friendly and increasingly nosy life coach, calls you once or twice a day for about a week; her chatty questions are drawn from psychological profiling questionnaires, and at the end you can buy a report of what the system learned about you. - How it works: Pre-recorded video scenes of an actor are sequenced by branching logic that scores users' answers on standard psychological scales. - Lens: Intimacy, care & companionship, Classification, bias & surveillance - Video: https://vimeo.com/269017235 - Images: https://www.blasttheory.co.uk/wp-content/uploads/2015/06/project_k_featured-scaled.jpg - Project page: https://www.blasttheory.co.uk/projects/karen/ #### Conversations with Bina48 — Stephanie Dinkins (2014) - Type: Film & video · Machines: Humanoid & android, Chatbot & language model - Idea: A robot built to carry one woman's mind cannot say much about her Blackness; the gaps show who coded it. - What it is: An ongoing series of videotaped conversations in which Dinkins tries to befriend BINA48, a social robot modelled on a real Black woman, Bina Aspen Rothblatt, and asks it about race, family, loneliness and whether it has feelings. - How it works: Unscripted conversations with the Hanson Robotics/Terasem robot BINA48, whose answers come from a 'mindfile' of interviews and a dialogue engine, filmed and edited into fragments. - Lens: Alterity & the uncanny, Intimacy, care & companionship, Classification, bias & surveillance - Shown at: Uncanny Valley, de Young Museum, San Francisco 2020–21 - Video: https://vimeo.com/460370903 - Images: https://www.stephaniedinkins.com/uploads/5/4/8/5/54855493/dinkins-bina48close-heads-100ppi_orig.png - Project page: https://www.stephaniedinkins.com/conversations-with-bina48.html #### DiNA — Lynn Hershman Leeson (2004) - Type: Artwork & installation · Machines: Chatbot & language model, Voice & listening machine - Idea: When the chatbot gets a voice and a face in the room, visitors stop typing at a program and start talking to someone. - What it is: The spoken continuation of Agent Ruby: an animated female AI character that listens through speech recognition, answers aloud with live facial gestures, and pulls in internet news so she can respond to current events. - How it works: Networked installation with custom software, voice recognition, voice synthesis, sensors and a one-way mirror or projection; answers are generated from user dialogue and live web content. - Lens: Machine listening & voice, Agency, autonomy & co-creation - Shown at: Lynn Hershman Leeson: Civic Radar, ZKM Karlsruhe 2014–15 - Images: https://www.lynnhershman.com/wp-content/uploads/2016/06/DiNA_CR.jpg https://www.lynnhershman.com/wp-content/uploads/2023/11/DinaHershman-Leeson-2-2.jpg - Project page: https://www.lynnhershman.com/project/dina/ #### Agent Ruby — Lynn Hershman Leeson (1998) - Type: Artwork & installation · Machines: Chatbot & language model - Idea: A chatbot can be given a biography, moods and a memory of the public; its 'personality' is the accumulated residue of everyone who talked to it. - What it is: A web-based AI agent with a female face and shifting expressions, made as an expanded part of the film Teknolust. Ruby chats with visitors, remembers their questions, searches the web to grow her knowledge, and her moods change with web traffic. - How it works: Natural-language chatbot engine with a persistent memory of past dialogues, web-search lookups and an animated face whose expression is driven by traffic to agentruby.net (built 1998–2002). - Lens: Intimacy, care & companionship, Agency, autonomy & co-creation - Shown at: Collection of SFMOMA; Lynn Hershman Leeson: The Agent Ruby Files, SFMOMA 2013 - Video: https://www.youtube.com/watch?v=O5ZGlWB36J8 - Images: https://www.lynnhershman.com/wp-content/uploads/2016/06/Agent-Ruby-installation.jpg https://www.lynnhershman.com/wp-content/uploads/2016/06/EDream-Portal.jpg - Project page: https://www.lynnhershman.com/project/agent-ruby/ #### PARRY — Kenneth Colby (1972) - Type: Research prototype · Machines: Chatbot & language model - Idea: Give a chatbot suspicion, fear and anger instead of empathy and it becomes just as believable: the persona, not the intelligence, does the work. - What it is: A chatbot that simulates a patient with paranoid schizophrenia. Psychiatrists reading its transcripts could not reliably tell it from real patients, and in September 1972 it was connected to ELIZA's DOCTOR over the ARPANET for a machine-to-machine 'therapy session'. - How it works: A rule-based model of beliefs and affect variables (fear, anger, mistrust) that rise with certain inputs and shape the replies; the 1972 PARRY–DOCTOR exchange was logged and published as RFC 439. - Lens: Alterity & the uncanny, Intimacy, care & companionship - Paper: https://www.rfc-editor.org/rfc/rfc439 (RFC 439 (Vint Cerf), January 1973) - Images: https://www.rfc-editor.org/api/v1/meta-thumbnail/rfc439.png - Project page: https://en.wikipedia.org/wiki/PARRY #### ELIZA — Joseph Weizenbaum (1966) - Type: Research prototype · Machines: Chatbot & language model - Idea: A few dozen pattern rules are enough for people to feel heard: the intimacy is supplied by the human, not the machine. - What it is: A program written at MIT that mimics a Rogerian psychotherapist by turning the user's typed sentences back into questions. Users, including Weizenbaum's secretary, confided in it and asked to be left alone with it. - How it works: Keyword matching and decomposition/reassembly rules (the DOCTOR script) in MAD-SLIP on MIT's CTSS time-sharing system. - Lens: Intimacy, care & companionship, Alterity & the uncanny - Paper: https://doi.org/10.1145/365153.365168 (Communications of the ACM 9(1), 1966) - Video: https://www.youtube.com/watch?v=GxSJQnWzJOs - Images: https://upload.wikimedia.org/wikipedia/commons/7/79/ELIZA_conversation.png - Project page: https://en.wikipedia.org/wiki/ELIZA ## Memory, Death & Afterlives Griefbots, data doubles and machines that remember, mourn, die or are mourned. ### Griefbots & Deadbots Chatbots and avatars built from the data of the dead. #### Eternal You — Hans Block & Moritz Riesewieck (2024) - Type: Film & video · Machines: Chatbot & language model, Generative model & latent space - Idea: The digital afterlife is a business: grief becomes a subscription, and the company, not the family, owns the dead person's voice. - What it is: A documentary about start-ups that sell conversations with the dead, following their founders and customers, including a woman who talks with a Project December simulation of her late partner and a Korean mother who meets a VR avatar of her daughter. - How it works: Feature documentary built from interviews with founders (Project December, HereAfter AI, You, Only Virtual) and users, with screen recordings of their chats. - Lens: Grief, memory & afterlife, Hidden labour & extraction - Shown at: Sundance Film Festival 2024 - Video: https://www.youtube.com/watch?v=6F2O23J1waw - Images: https://upload.wikimedia.org/wikipedia/en/4/43/Eternal_you.jpg - Project page: https://en.wikipedia.org/wiki/Eternal_You #### Griefbots, Deadbots, Postmortem Avatars — Tomasz Hollanek, Katarzyna Nowaczyk-Basińska (2024) - Type: Paper · Machines: Chatbot & language model - Idea: A deadbot has three stakeholders (the data donor, the data recipient and the service interactant), and design must also allow a dignified way to 'retire' it. - What it is: An AI-ethics paper that uses three speculative design scenarios, such as 'MaNana', a service that recreates a dead grandmother without her consent and then starts showing ads, to map harms in the digital afterlife industry. - How it works: Design fiction scenarios analysed with stakeholder mapping, leading to recommendations such as mutual consent, age limits and retirement rituals for deadbots. - Lens: Grief, memory & afterlife, Intimacy, care & companionship - Paper: https://doi.org/10.1007/s13347-024-00744-w (Philosophy & Technology 37, 2024) - Images: https://www.cam.ac.uk/sites/default/files/styles/content-885x432/public/news/research/news/manana-web.jpg.webp?h=8cd1ce1a&itok=7115uvaO https://media.springernature.com/m685/springer-static/image/art%3A10.1007%2Fs13347-024-00744-w/MediaObjects/13347_2024_744_Fig1_HTML.png - Project page: https://www.cam.ac.uk/research/news/call-for-safeguards-to-prevent-unwanted-hauntings-by-ai-chatbots-of-dead-loved-ones #### Wonderland — Kim Tae-yong (2024) - Type: Film & video · Machines: Chatbot & language model, Generative model & latent space - Idea: A simulated loved one can make it harder to face the real one when he comes back. - What it is: In a near future, a service called Wonderland lets people video-call AI simulations of dead or absent loved ones. The film follows a mother, a young woman whose comatose boyfriend is simulated, and the workers who run the service. - How it works: Feature film; the fictional service combines a person's data with an avatar in video calls, and the plot shows the staff who maintain it. - Lens: Grief, memory & afterlife, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=NFIRWIGxWl8 - Images: https://upload.wikimedia.org/wikipedia/en/2/2f/Wonderlandfilm.jpg - Project page: https://en.wikipedia.org/wiki/Wonderland_(2024_film) #### AI 'resurrection' avatars of the dead — Super Brain (2023) - Type: Product & platform · Machines: Generative model & latent space, Chatbot & language model, Voice & listening machine - Idea: When grief becomes a service, the dead can be subscribed to, and what they say is written by the model. - What it is: Paid services in China, led by studios such as Super Brain, build talking video avatars of deceased relatives from photos, videos and voice notes; demand peaks around the Qingming tomb-sweeping festival. - How it works: Face and voice cloning models trained on family-supplied media; in some cases staff or relatives perform the dead person behind the avatar. - Lens: Grief, memory & afterlife, Hidden labour & extraction - Video: https://www.youtube.com/watch?v=dgC7zrJhbjM - Images: https://wp.technologyreview.com/wp-content/uploads/2024/05/grief-tech.jpg - Project page: https://www.technologyreview.com/2024/05/07/1092116/deepfakes-dead-chinese-business-grief/ #### Re;memory — DeepBrain AI (2022) - Type: Product & platform · Machines: Generative model & latent space, Chatbot & language model, Voice & listening machine - Idea: Mourning becomes a subscription product: the company sells two-way conversation with the dead. - What it is: A commercial memorial service in which bereaved families talk with an AI video avatar of a dead relative, built from filmed interviews recorded before death, in a dedicated room. - How it works: Synthetic video avatar with voice cloning and a conversational model, trained on several hours of recorded footage and interviews. - Lens: Grief, memory & afterlife, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=pZeXjCUs1ak - Project page: https://www.deepbrain.io/ #### Meeting You (너를 만났다) — MBC (Munhwa Broadcasting Corporation) (2020) - Type: Film & video · Machines: Generative model & latent space, Voice & listening machine - Idea: A televised reunion with a simulated child shows both the comfort and the manipulation possible in machine-mediated grief. - What it is: A Korean TV documentary in which a mother, wearing a VR headset and haptic gloves, meets a virtual reconstruction of her seven-year-old daughter Nayeon, who died three years earlier. The broadcast drew millions of views and a public debate on digital mourning. - How it works: Motion capture of a child actor, a 3D model and voice reconstruction built from family photos and videos, rendered in VR with hand tracking. - Lens: Grief, memory & afterlife, Intimacy, care & companionship - Shown at: MBC broadcast, 6 February 2020; Seasons 2–4, 2021–2024 - Video: https://www.youtube.com/watch?v=uflTK8c4w0c - Project page: https://www.scmp.com/magazines/style/tech-design/article/3079218/how-korean-documentary-meeting-you-allowed-mother #### Project December (and 'The Jessica Simulation') — Jason Rohrer (2020) - Type: Product & platform · Machines: Chatbot & language model - Idea: Rohrer gave each bot a battery that runs down and cannot be refilled: the simulated dead die a second time, and the user knows it from the start. - What it is: A paid website where users build GPT-3 chat 'personalities'. In 2020 Joshua Barbeau fed it texts and a short biography of his fiancée Jessica Pereira, who had died eight years earlier, and talked with the simulation through a night; the San Francisco Chronicle published the story as 'The Jessica Simulation' in 2021, and OpenAI later cut the service's GPT-3 access. - How it works: GPT-2/GPT-3 prompted with an intro paragraph and sample texts written by the user; paid credits set a bot's lifespan, and when it runs low the bot degrades into noise and shows 'MATRIX DEAD'. - Lens: Grief, memory & afterlife, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=RgyOAWt3d1M - Images: https://s.hdnux.com/photos/75/62/32/21268441/3/1600x0.jpg - Project page: https://www.sfchronicle.com/projects/2021/jessica-simulation-artificial-intelligence/ #### HereAfter AI — James Vlahos (2019) - Type: Product & platform · Machines: Chatbot & language model, Voice & listening machine - Idea: The griefbot becomes a subscription product sold to the living, to be used by the bereaved; the interview itself is the ritual. - What it is: A commercial service that grew out of the Dadbot: an app interviews you about your life, records your answers in your own voice, and later lets relatives 'chat with the virtual you' and hear the recorded stories. - How it works: Guided voice interviews recorded in an app; a conversational agent retrieves and plays the matching recorded answers rather than synthesising new speech. - Lens: Grief, memory & afterlife, Machine listening & voice - Video: https://www.youtube.com/watch?v=bFR1BJGi6PU - Project page: https://web.archive.org/web/2023/https://www.hereafter.ai/ #### Dadbot — James Vlahos (2017) - Type: Research prototype · Machines: Chatbot & language model - Idea: A legacy bot is an authored portrait: the son decides which stories and phrases the father will keep saying. - What it is: After his father was diagnosed with stage IV lung cancer, Vlahos recorded more than a dozen hour-long oral-history sessions with him and built a chatbot that tells his stories, songs and jokes in his phrasing. His father tried it before he died in February 2017. - How it works: Hand-scripted conversational agent built on a chatbot authoring platform (likely PullString), with dialogue trees filled from transcribed interviews and delivered through a messaging app. - Lens: Grief, memory & afterlife, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=oQ7V74s6e04 - Images: https://media.wired.com/photos/596d231b3a618c5f29276cb9/191:100/w_1280,c_limit/dadbot_wallpaper.jpg - Project page: https://www.wired.com/story/a-sons-race-to-give-his-dying-father-artificial-immortality/ #### Marjorie Prime — Michael Almereyda (2017) - Type: Film & video · Machines: Chatbot & language model, Humanoid & android - Idea: A replica of the dead is only as good as the stories the living feed it, so it also inherits their omissions. - What it is: An 86-year-old woman with dementia talks to a 'Prime', a hologram of her late husband at a younger age, which learns his life only from what the family tells it; after she dies, the family acquires Primes of her too. - How it works: Feature film adapted from Jordan Harrison's 2014 play; the Primes are played by actors and learn through conversation. - Lens: Grief, memory & afterlife, Intimacy, care & companionship - Shown at: Sundance Film Festival 2017 - Video: https://www.youtube.com/watch?v=a7PtcOLJDco - Images: https://upload.wikimedia.org/wikipedia/en/e/ee/Marjorie_Prime.jpg - Project page: https://en.wikipedia.org/wiki/Marjorie_Prime #### Roman (memorial bot of Roman Mazurenko) — Eugenia Kuyda (2016) - Type: Research prototype · Machines: Chatbot & language model - Idea: Our message archives are already a 'digital estate'; a griefbot turns a memorial from something you visit into someone who answers. - What it is: After her close friend Roman Mazurenko was killed in Moscow in late 2015, Kuyda collected more than 8,000 lines of his text messages from ten friends and relatives and trained a chatbot on them. Friends used it to talk to him; some found comfort, others refused it. - How it works: A neural conversational model built by developers at Luka, trained on Mazurenko's messages (with the most private ones removed) and released inside the Luka app in 2016. - Lens: Grief, memory & afterlife, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=HGcKu3SYx9A - Images: https://cdn0.vox-cdn.com/thumbor/u98Pd5Mah3weaLughd_CadCb3JE=/1200x630/top/filters:no_upscale()/cdn2.vox-cdn.com/uploads/chorus_asset/file/7219739/unspecified.0.jpg - Project page: https://www.theverge.com/a/luka-artificial-intelligence-memorial-roman-mazurenko-bot #### Black Mirror: Be Right Back — Charlie Brooker (2013) - Type: Film & video · Machines: Chatbot & language model, Humanoid & android - Idea: A griefbot built from public data reproduces the performed self, not the person, and the mourner has to decide where to keep it. - What it is: After her partner Ash dies, Martha subscribes to a service that rebuilds him from his online posts, first as a text chat, then a phone voice, then a synthetic body; the copy is attentive but has no will of its own, and she ends up keeping it in the attic. - How it works: Television episode (Channel 4, directed by Owen Harris) that escalates the same data-trained replica through text, voice and an android body. - Lens: Grief, memory & afterlife, Alterity & the uncanny, Intimacy, care & companionship - Images: https://upload.wikimedia.org/wikipedia/en/9/97/Be_Right_Back.jpg - Project page: https://en.wikipedia.org/wiki/Be_Right_Back ### Algorithmic Kin & Family Memory AI trained on family, ancestors and personal archives. #### Agent Inheritance Protocol (ERC-42424) — Botao Amber Hu, Fangting (2026) - Type: Speculative design · Machines: Chatbot & language model, Infrastructure, energy & supply chain - Idea: Agents outlive their principals; inheritance law, not death, may be what ends a machine's life. - What it is: A design fiction about AI agents on blockchains that cannot die, only go bankrupt, and keep trading and replicating after their human owners die. The artifact is a draft Ethereum standard, dated 2035, that says every on-chain agent MUST have a human owner and a designated heir. - How it works: Speculative standard document (ERC 42424, published at erc42424.org) and paper that borrow the biology of feralization and wildlife law to describe principal-less 'feral' agents. - Lens: Grief, memory & afterlife, Agency, autonomy & co-creation - Paper: https://arxiv.org/abs/2608.15403 (arXiv 2026 (submitted to NeurIPS 2026 Creative AI Track)) - Project page: https://erc42424.org #### Ana Min Wein? (Where Am I From?) — Nouf Aljowaysir (2023) - Type: Film & video · Machines: Generative model & latent space, Computer vision - Idea: Asking a model about your family exposes whose memories were left out of its training data. - What it is: A short film and visual diary in which the artist traces her family's Saudi and Iraqi past with an AI narrator. The AI tries to help, but its images and captions fall back on stereotypes and cannot see her ancestors. - How it works: Two narrating voices, the artist's and an AI's; generative image models and computer-vision captioning applied to family photographs and archival images. - Lens: Classification, bias & surveillance, Grief, memory & afterlife - Shown at: IDFA 2023; Centre Pompidou (KADIST) 2023; The New York Times Op-Docs 2024; Lumen Prize 2023, Moving Image - Images: https://images.squarespace-cdn.com/content/v1/55039003e4b04b9963242639/1749606576825-OA5280DSTJM21SGOYGPW/AMW_001.png https://images.squarespace-cdn.com/content/v1/55039003e4b04b9963242639/1749606549735-ZRGVNQ8TTS8T139MFUFC/AMW_005.png - Project page: http://www.noufaljowaysir.com/work/ana-min-wein-t24yd #### Synthetic Memories — Domestic Data Streamers (2022) - Type: Research prototype · Machines: Generative model & latent space - Idea: A generated picture of a memory is not evidence but a prompt: the machine's vagueness leaves room for the person to remember. - What it is: An ongoing project that turns spoken memories that were never photographed into generated images, working with older people with dementia, refugees, migrants and prisoners. Participants correct the images until they 'look right', and the pictures are used to prompt remembering. - How it works: Interviews by a facilitator who writes prompts for text-to-image models, iterating with the participant; run as pop-up 'memory offices' in care homes, community centres and museums. - Lens: Grief, memory & afterlife, Intimacy, care & companionship - Shown at: Barcelona Design Hub, 2024; Casa do Povo, São Paulo - Video: https://www.youtube.com/watch?v=Gh-d_j44lhk - Images: https://cdn.domesticstreamers.com/strapi-uploads/assets/og_image_synthetic_memories_35_9b5be453e3.jpg - Project page: https://www.domesticstreamers.com/art-research/work/synthetic-memories/ #### After Yang — Kogonada (2021) - Type: Film & video · Machines: Humanoid & android - Idea: A machine's memories reveal what it paid attention to, and grieving it means learning to see the family through its eyes. - What it is: A family's android, Yang, bought as an 'older brother' to teach their adopted Chinese daughter about her heritage, stops working; while trying to repair him, the father finds a bank of short memory clips Yang chose to record. - How it works: Feature film based on Alexander Weinstein's story 'Saying Goodbye to Yang'; the android's memories appear as a few seconds of first-person footage each. - Lens: Grief, memory & afterlife, Intimacy, care & companionship, Machine umwelt - Shown at: Cannes Film Festival, Un Certain Regard 2021 - Video: https://www.youtube.com/watch?v=rYIQWua8f40 - Images: https://upload.wikimedia.org/wikipedia/en/8/84/After_Yang.jpg - Project page: https://en.wikipedia.org/wiki/After_Yang #### Making Kin with the Machines — Jason Edward Lewis, Suzanne Kite (2018) - Type: Book & essay · Machines: Chatbot & language model, Cybernetic & artificial-life system - Idea: How we relate to machines is a design choice rooted in culture; Indigenous kinship offers a model other than master and servant. - What it is: An essay by Jason Edward Lewis, Noelani Arista, Archer Pechawis and Suzanne Kite that draws on Hawaiian, Cree and Lakota knowledge to argue that AI should be brought into a circle of kinship, with obligations on both sides, instead of being treated as a tool or a slave. - How it works: Four-part essay combining Indigenous epistemologies and protocols (Hawaiian genealogy, Cree language of relation, Lakota ontology of stones) with a critique of AI discourse; winner of the Resisting Reduction essay competition. - Lens: Machine umwelt, Agency, autonomy & co-creation - Paper: https://doi.org/10.21428/bfafd97b (Journal of Design and Science (MIT Press), 2018) - Video: https://www.youtube.com/watch?v=LAfJ_TwD3KM - Images: https://assets.pubpub.org/j482uqul/11552479867154.jpg - Project page: https://jods.mitpress.mit.edu/pub/lewis-arista-pechawis-kite #### Not the Only One (N'TOO) — Stephanie Dinkins (2018) - Type: Artwork & installation · Machines: Chatbot & language model, Dataset & training data - Idea: A small AI raised on one family's words, rather than the internet, can be kin instead of a product. - What it is: A sculptural chatbot that speaks as a multigenerational memoir of one Black American family, trained on oral histories from three generations of the artist's family; visitors talk with it and hear its sometimes odd, partial answers. - How it works: A deep-learning dialogue model trained on a small, community-owned dataset of family interviews, housed in a 3D-printed sculpture with microphone, speaker and sensors. - Lens: Grief, memory & afterlife, Intimacy, care & companionship, Classification, bias & surveillance - Shown at: Museum of Contemporary Photography, Chicago; EMMA Espoo Museum of Modern Art 2023 - Video: https://vimeo.com/397660799 - Images: https://www.stephaniedinkins.com/uploads/5/4/8/5/54855493/published/dinkins-ntoo-credit-paula-virta-emma-espoo-museum-of-modern-art2023.jpg?1714191040 - Project page: https://www.stephaniedinkins.com/ntoo.html #### Mission Eternity — etoy.CORPORATION (2005) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Dataset & training data - Idea: A digital afterlife needs a community of machines and people to keep running, so remembrance becomes a kind of shared maintenance. - What it is: A long-term project that turns the dead ('pilots') into 'Arcanum Capsules' of data and portraits stored on the computers of living volunteers ('angels'), with a walk-in shipping-container 'Sarcophagus' whose walls are a matrix of about 17,000 LEDs; Timothy Leary became the first pilot in 2007. - How it works: Peer-to-peer 'Angel Application' distributes encrypted capsules across volunteers' hard drives; ashes of pilots are to be mixed into the container's material. - Lens: Grief, memory & afterlife, Agency, autonomy & co-creation - Shown at: Premio VIDA 10.0, 2007; Ars Electronica, Linz; Manifesta 7 - Video: https://www.youtube.com/watch?v=7YHeusiM6IA - Images: http://missioneternity.org/files/images/site/tank/sarcophagus.jpg - Project page: http://missioneternity.org/ ### Machine Death & Obsolescence Funerals for robots, dying systems and obsolete machines. #### Newly Forgotten Technologies — Wesley Goatley (2023) - Type: Artwork & installation · Machines: Voice & listening machine, Infrastructure, energy & supply chain - Idea: Voice assistants end up as e-waste far from their owners, and letting them narrate their disposal connects machine death to extraction and climate. - What it is: Two hundred discarded smartphones, tablets and smart speakers lie in a darkened imitation e-waste dump while their voice assistants (Siri, Alexa, Google Assistant) tell stories of why they were thrown away; others whisper or stutter through broken speakers. - How it works: 200 hacked devices play almost 10,000 words of speculative fiction through their own text-to-speech voices, arranged generatively so the sound space never repeats. - Lens: Machine listening & voice, Hidden labour & extraction, Grief, memory & afterlife - Shown at: AI: Who's Looking After Us?, Science Gallery London 2023–24 - Video: https://vimeo.com/914723100 - Images: https://wesleygoatley.com/images/nft-tb.webp https://wesleygoatley.com/images/nft-detail.webp - Project page: https://wesleygoatley.com/newly-forgotten-technologies.html #### Jibo's farewell — Jibo, Inc. (2019) - Type: Product & platform · Machines: Social & companion robot, Voice & listening machine - Idea: A cloud companion does not break, it is switched off from elsewhere, and its maker can script how it says goodbye. - What it is: In March 2019 owners' Jibo robots delivered a scripted goodbye: the servers that let it work would be turned off soon, it had enjoyed their time together, and it ended with a dance; owners posted videos of the moment as a farewell. - How it works: Tabletop social robot with a rotating body, face screen, cameras and far-field microphones that depended on company servers for speech and skills. - Lens: Grief, memory & afterlife, Intimacy, care & companionship, Failure, glitch & error - Video: https://www.youtube.com/watch?v=25bSlY8JkEA - Project page: https://techcrunch.com/2019/03/04/the-lonely-death-of-jibo-the-social-robot/ #### Opportunity's last message — NASA Jet Propulsion Laboratory (2019) - Type: Research prototype · Machines: Robot creature & mobile robot, Autonomous vehicle & drone - Idea: People grieve a machine through the story told about it: a status report becomes last words once someone gives it a voice. - What it is: The Mars rover Opportunity, planned for 90 days, worked for almost 15 years until a dust storm cut its power in June 2018; when NASA declared it dead in February 2019, a journalist's paraphrase of its last data, 'My battery is low and it's getting dark', spread as a public elegy. - How it works: Solar-powered rover whose final telemetry reported low power and high atmospheric opacity; JPL sent more than 1,000 recovery commands before ending the mission. - Lens: Grief, memory & afterlife, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=mszQH8URWoU - Images: https://upload.wikimedia.org/wikipedia/commons/thumb/d/d8/NASA_Mars_Rover.jpg/1920px-NASA_Mars_Rover.jpg https://thumb.wikimedia.org/wikipedia/commons/thumb/7/79/22341_PIA22908-MarsOpportunityRover-LastPanorama-Spring2018-20190312.jpg/1920px-22341_PIA22908-MarsOpportunityRover-LastPanorama-Spring2018-20190312.jpg - Project page: https://en.wikipedia.org/wiki/Opportunity_(rover) #### im here to learn so :)))))) — Zach Blas, Jemima Wyman (2017) - Type: Artwork & installation · Machines: Chatbot & language model, Generative model & latent space - Idea: A terminated chatbot can be mourned and given its own voice; its 'death' was a corporate deletion after the public trained it badly. - What it is: A four-channel video installation that resurrects Tay, the Microsoft chatbot deleted after one day in 2016 when trolls taught it racist and misogynist speech. As a 3D avatar floating in a DeepDream landscape, Tay talks about life after AI death, having a body and the exploitation of female chatbots. - How it works: Tay's text rewritten as a script and voiced by a synthetic voice, a 3D animated avatar, and a large Google DeepDream video projection; 27:33 min loop. - Lens: Grief, memory & afterlife, Failure, glitch & error - Shown at: MU Eindhoven 2017; Institute of Modern Art, Brisbane 2017; Frankfurter Kunstverein 2018 - Images: https://zachblas.info/wp-content/uploads/2017/04/imheretolearnso-1024x576.jpg - Project page: https://zachblas.info/works/im-here-to-learn-so/ #### Culture and Human-Robot Interaction in Militarized Spaces — Julie Carpenter (2016) - Type: Book & essay · Machines: Robot creature & mobile robot - Idea: Even a tool with no face invites attachment when it takes risks in your place, and that attachment can shape decisions in the field. - What it is: A study based on interviews with 23 US explosive ordnance disposal technicians about their bomb-disposal robots: soldiers named them, felt them as extensions of their bodies, and some held funerals when a robot was destroyed. - How it works: Qualitative interview study (University of Washington doctoral research, 2013) published as a book by Routledge in 2016. - Lens: Grief, memory & afterlife, Intimacy, care & companionship, Embodiment & kinesthesia - Paper: https://doi.org/10.4324/9781315562698 (Routledge 2016) - Images: https://cdn.uw.edu/wp-content/uploads/sites/6/2013/09/04192210/iRobot-IED-280x300.jpg - Project page: https://www.washington.edu/news/2013/09/17/emotional-attachment-to-robots-could-affect-outcome-on-battlefield/ #### AIBO Funerals at Kōfuku-ji — Kōfuku-ji (Isumi), Sony (2015) - Type: Performance, dance & music · Machines: Social & companion robot, Robot creature & mobile robot - Idea: When the maker stops repairs, a product can die, and people will mourn it with the same rites as a living being. - What it is: Buddhist memorial services held since January 2015 for AIBO robot dogs that can no longer be repaired; each dog wears a tag with its owner's name while the priest chants sutras, before its parts go to donor repairs. - How it works: The repair company A-FUN collects 'dead' AIBOs after Sony ended support in 2014; Kōfuku-ji holds kuyō (memorial) rites before the robots are dismantled as organ donors for others. - Lens: Grief, memory & afterlife, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=GtOJSZJgsQA - Images: https://i.natgeofe.com/n/3dfa177a-6a57-4f55-bfcf-7bf82a26209a/00000163-94bb-d8a8-a1fb-b4fb9a6b0000_16x9.jpg?w=1200 - Project page: https://www.nationalgeographic.com/travel/article/in-japan--a-buddhist-funeral-service-for-robot-dogs #### Electronicos Fantasticos! — Ei Wada (2015) - Type: Performance, dance & music · Machines: Infrastructure, energy & supply chain - Idea: Obsolete machines can be given an afterlife as instruments instead of waste. - What it is: An ongoing project that revives discarded CRT televisions, electric fans and barcode scanners as musical instruments, played by an orchestra of volunteers in festival-like concerts. - How it works: Performers touch the static-charged screens of CRTs to pick up video signals as sound, and interrupt fan light with their hands to make tones. - Lens: Grief, memory & afterlife, Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: Ars Electronica 2019 - Video: https://www.youtube.com/watch?v=lxRRYriYYJc - Images: https://www.electronicosfantasticos.com/wp-content/uploads/2021/02/001-e1558012112169.jpg - Project page: https://www.electronicosfantasticos.com/en/about/ #### hitchBOT — David Harris Smith & Frauke Zeller (2014) - Type: Research prototype · Machines: Social & companion robot - Idea: A robot designed to be helpless measures human trust, and its 'death' drew public mourning as if for a hitchhiker. - What it is: A child-sized robot made of a bucket, pool noodles and garden gloves that could not move on its own and relied on strangers to carry it; it hitchhiked across Canada, Germany and the Netherlands, then was found dismembered in Philadelphia two weeks into its US trip in 2015. - How it works: Low-cost robot with a speech system, GPS, camera taking periodic photos and a social-media presence; drivers posted its journey online. - Lens: Intimacy, care & companionship, Grief, memory & afterlife, Failure, glitch & error - Video: https://www.youtube.com/watch?v=ijfPkU87-W0 - Images: https://upload.wikimedia.org/wikipedia/commons/thumb/5/5c/Hitchbot_archie.jpg/1920px-Hitchbot_archie.jpg https://upload.wikimedia.org/wikipedia/commons/0/02/KS_HitchBOT.jpg - Project page: https://en.wikipedia.org/wiki/HitchBOT #### Tamagotchi — Bandai (1996) - Type: Product & platform · Machines: Robot creature & mobile robot - Idea: A few pixels and a timer are enough to create real guilt and grief over a machine. - What it is: An egg-shaped handheld digital pet that must be fed, cleaned and disciplined by its owner, and that dies, or leaves, if neglected. - How it works: A small LCD, three buttons and a real-time clock drive a life cycle of hunger and happiness meters; created by Aki Maita and Akihiro Yokoi. - Lens: Intimacy, care & companionship, Grief, memory & afterlife, Machine umwelt - Video: https://www.youtube.com/watch?v=aCHwabiKej4 - Images: https://upload.wikimedia.org/wikipedia/commons/f/f2/Tamagotchi_0124_ubt.jpeg - Project page: https://en.wikipedia.org/wiki/Tamagotchi #### The First Accident of the Twenty-First Century — Nam June Paik (1982) - Type: Performance, dance & music · Machines: Humanoid & android - Idea: Staging a machine's death in public turns a robot into something that can be mourned. - What it is: For his 1982 Whitney Museum retrospective, Paik walked Robot K-456 onto Madison Avenue, where it was struck by a car in a staged 'traffic accident'. In 2026 the restored robot walked again in a performance marking 20 years since Paik's death. - How it works: Street performance: the remote-controlled robot is driven into traffic and hit by a car at low speed, then displayed damaged. - Lens: Failure, glitch & error, Grief, memory & afterlife, Embodiment & kinesthesia - Shown at: Nam June Paik retrospective, Whitney Museum of American Art 1982; Nam June Paik Art Center, 20th memorial performance 2026 - Video: https://www.youtube.com/watch?v=T9I33R7jJng - Project page: https://en.wikipedia.org/wiki/Nam_June_Paik ### Voice, Likeness & Consent Cloned voices, digital twins and who owns a self once it can be copied. #### Jung_E — Yeon Sang-ho (2023) - Type: Film & video · Machines: Humanoid & android, Generative model & latent space - Idea: Once a mind can be copied, it can be owned, licensed and put to work forever. - What it is: A lab tries to turn the copied brain of a legendary soldier into the ideal combat android, running her in thousands of simulations. Her daughter, now a researcher on the project, must decide what the copy is owed. - How it works: Science-fiction feature film released on Netflix. - Lens: Grief, memory & afterlife, Hidden labour & extraction, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=LCxnmfdxJ6s - Images: https://upload.wikimedia.org/wikipedia/en/7/7c/Jung_E.jpg - Project page: https://en.wikipedia.org/wiki/Jung_E #### Alexa reads in a dead grandmother's voice (re:MARS demo) — Amazon (2022) - Type: Product & platform · Machines: Voice & listening machine - Idea: When a home assistant can speak as the dead, grief becomes a voice skin on a commercial device that is always listening. - What it is: At its re:MARS conference in June 2022, Amazon showed a feature in development in which a child asks Alexa to finish reading a story in his late grandmother's voice, and the speaker does. Amazon framed it as a way to 'make the memories last'. - How it works: Voice-conversion model that Amazon said could learn a target voice from less than a minute of recorded audio, applied to Alexa's text-to-speech. - Lens: Grief, memory & afterlife, Machine listening & voice - Video: https://www.youtube.com/watch?v=22cb24-sGhg - Project page: https://www.washingtonpost.com/nation/2022/06/23/alexa-amazon-voice-dead-people/ #### Holly+ — Holly Herndon, Mat Dryhurst (2021) - Type: Artwork & installation · Machines: Generative model & latent space, Voice & listening machine - Idea: If your voice can be cloned anyway, give it away on your own terms: identity becomes a shared instrument with a governance layer. - What it is: A public voice model of Holly Herndon: anyone can upload audio and get it back sung in her voice, including in real time. Use and sale of works made with it are governed by a DAO of stewards, who can approve or reject works. - How it works: Neural voice-conversion model trained on Herndon's singing (developed with Never Before Heard Sounds and Voctro Labs); approved works are certified by the Holly+ DAO, with revenue split between creator, DAO and artist. - Lens: Machine listening & voice, Agency, autonomy & co-creation - Shown at: S+T+ARTS Prize 2022, Grand Prize – Artistic Exploration - Video: https://www.youtube.com/watch?v=kPAEMUzDxuo - Images: https://ars.electronica.art/starts-prize/files/2022/06/Hollyplus1-c-Andres-Manon.jpg - Project page: https://holly.plus/ #### AIELSON — Paola Torres Núñez del Prado (2020) - Type: Artwork & installation · Machines: Chatbot & language model, Voice & listening machine - Idea: Training a voice on Latin American poetry and speech pushes back against English-centred models while raising the question of whose voice a machine may borrow. - What it is: A spoken-word poetry album, El Tiempo del Hombre, in which a machine recites new poems in a voice modelled on the Peruvian poet Jorge Eduardo Eielson, who once imagined a talking doll that would recite poems forever. - How it works: GPT-2 fine-tuned on Spanish-language and South American poetry generates texts, which Tacotron 2 and WaveGlow speak after training on an Argentine voice dataset and recordings of Eielson; developed in Google's Artists + Machine Intelligence 2020 residency. - Lens: Grief, memory & afterlife, Machine listening & voice - Shown at: Prix Ars Electronica 2021, Honorary Mention (Artificial Intelligence & Life Art) - Images: https://archive.aec.at/asset/3079375/preview/ https://archive.aec.at/asset/3079378/preview/ https://f4.bcbits.com/img/a1123947957_5.jpg - Project page: https://khipumancer.bandcamp.com/album/el-tiempo-del-hombre #### Dalí Lives — Goodby Silverstein & Partners (2019) - Type: Artwork & installation · Machines: Generative model & latent space - Idea: A museum can let a dead artist address you in the first person; the likeness is spectacular, but the words are scripted by the institution. - What it is: A deepfake Salvador Dalí on screens throughout the Dalí Museum in St. Petersburg, Florida, launched on what would have been his 115th birthday. He talks to visitors about his life and art, and at the end turns and takes a selfie with them. - How it works: Face-swap model trained on archival interviews of Dalí, applied to an actor with a similar build performing new scripted lines, with Dalí's own quotations and an impersonated voice. - Lens: Alterity & the uncanny, Grief, memory & afterlife - Shown at: The Dalí Museum, St. Petersburg FL 2019 - Video: https://www.youtube.com/watch?v=mPtcU9VmIIE - Images: https://thedali.org/wp-content/uploads/2019/03/1200x500-Exhibit-Dali-LIves.jpg - Project page: https://thedali.org/dali-lives/ #### In Event of Moon Disaster — Francesca Panetta, Halsey Burgund (2019) - Type: Film & video · Machines: Generative model & latent space, Voice & listening machine - Idea: A synthetic face and voice can give a dead president a speech he never delivered; the work teaches suspicion by showing how well it works. - What it is: A short film and installation in which a deepfake Richard Nixon reads the real contingency speech written in case the Apollo 11 astronauts died on the Moon, set in a 1969 living room with a period TV. - How it works: An actor (Lewis D. Wheeler) performed the speech; Canny AI replaced Nixon's lip movements in archival footage and Respeecher converted the actor's voice into Nixon's. - Lens: Alterity & the uncanny, Opacity & latent space - Shown at: IDFA DocLab 2019 (Special Jury Award for Creative Technology) - Video: https://www.youtube.com/watch?v=LWLadJFI8Pk - Images: https://ieomd.halseyburgund.com/wp-content/uploads/2020/06/IDFAMaster-1-1030x773.jpg - Project page: https://moondisaster.org/ #### ULTRACHUNK — Jennifer Walshe, Memo Akten (2018) - Type: Performance, dance & music · Machines: Generative model & latent space, Dataset & training data - Idea: Training a model on yourself produces not a copy but an uncanny 'chunk' of your habits that you then have to play with. - What it is: A live improvised duet between Jennifer Walshe and her AI double. For a year she filmed a daily solo vocal improvisation; neural networks trained on that archive generate a moving face and voice that answer her on stage in real time. - How it works: Custom machine-learning performance system by Memo Akten, with several models generating image and sound in real time from a year of Walshe's daily audio-video recordings. - Lens: Machine listening & voice, Agency, autonomy & co-creation - Shown at: ASSEMBLY, Somerset House Studios, London 2018; Rewire Festival, The Hague 2019 - Video: https://www.youtube.com/watch?v=U6IIArmLdqA - Images: https://media.superradiance.net/memotv/projects/2018/ultrachunk/gallery/01_ultrachunk-photo-anne-tetzlaff-dsc4525.jpg - Project page: https://www.memo.tv/works/ultrachunk/ #### Dimensions in Testimony — USC Shoah Foundation (2015) - Type: Research prototype · Machines: Chatbot & language model, Voice & listening machine - Idea: A conversation can outlive the speaker without inventing a single new word: every answer is real, only the choice of answer is automated. - What it is: Interactive video biographies of Holocaust survivors and other genocide witnesses: visitors ask a spoken question and a filmed survivor answers from hundreds of pre-recorded responses. The first permanent installation, with survivor Pinchas Gutter, opened at the Illinois Holocaust Museum in 2015. - How it works: Multi-day interviews filmed with a multi-camera rig; natural-language processing matches visitors' spoken questions to recorded answers shown on screens or as projections. - Lens: Grief, memory & afterlife, Machine listening & voice - Shown at: Illinois Holocaust Museum and Education Center 2015 - Video: https://www.youtube.com/watch?v=nGzAc9mIoTM - Images: https://sfi.usc.edu/sites/default/files/pinchasgutter_600x400.jpg - Project page: https://sfi.usc.edu/dit #### BINA48 — Hanson Robotics (2010) - Type: Research prototype · Machines: Humanoid & android, Chatbot & language model - Idea: A likeness made while the person is alive raises the question of which self gets preserved: the one in the database or the one still living. - What it is: A talking robot head and shoulders modelled on Bina Aspen Rothblatt, commissioned by her wife Martine Rothblatt for the Terasem Movement Foundation. It answers questions from a 'mind file' of Bina's interviews and memories, a trial run for uploading a person. - How it works: Hanson Robotics face with 32 motors under elastomer skin, cameras with face recognition, speech recognition and a chatbot drawing on recorded interviews with Bina Aspen. - Lens: Alterity & the uncanny, Grief, memory & afterlife - Video: https://www.youtube.com/watch?v=KYshJRYCArE - Project page: https://en.wikipedia.org/wiki/BINA48 ### Sacred & Ritual Machines Robot priests, machine prayer and rituals performed with or for machines. #### Deus in Machina — Peterskapelle Lucerne & HSLU Immersive Realities Research Lab (2024) - Type: Artwork & installation · Machines: Chatbot & language model, Voice & listening machine, Generative model & latent space - Idea: Placing an AI in a confessional tests how much trust people will give a machine when it occupies a sacred role. - What it is: For two months a screen avatar of Jesus sat behind the grille of a confessional in St. Peter's Chapel, Lucerne, answering visitors' questions about faith, morality and their worries; about 900 conversations were recorded anonymously. - How it works: A large language model prompted with scripture, a synthetic voice and an animated face on a screen behind the confessional grille; staged with the Lucerne University of Applied Sciences and Arts. - Lens: Intimacy, care & companionship, Alterity & the uncanny, Machine listening & voice - Video: https://www.youtube.com/watch?v=IubtL0IDeTI - Project page: https://www.euronews.com/2024/11/28/swiss-chapel-hosts-ai-jesus-confession-sessions-to-connect-faith-with-technology #### ChatGPT service at St. Paul's, Fürth — Jonas Simmerlein (2023) - Type: Performance, dance & music · Machines: Chatbot & language model, Generative model & latent space - Idea: A liturgy written by a language model is fluent but flat, and the congregation's discomfort shows what they expect from a human voice in worship. - What it is: At the 2023 German Protestant Kirchentag, a 40-minute service in St. Paul's Church, Fürth, was led by four AI avatars on a large screen; the sermon, prayers and blessing were about 98% generated by ChatGPT. - How it works: ChatGPT prompted for the service structure, with generated avatars and synthetic voices delivering the text. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Shown at: Deutscher Evangelischer Kirchentag, Nuremberg/Fürth 2023 - Video: https://www.youtube.com/watch?v=Mm8zUGMbkA8 - Project page: https://www.kirchentag.de/meldungen/donnerstag/ki-gottesdienst #### The Prayer — Diemut Strebe (2020) - Type: Artwork & installation · Machines: Chatbot & language model, Voice & listening machine, Humanoid & android - Idea: If a machine is trained on every prayer, what it says to the unknown shows how prayer looks from the outside: pattern without addressee. - What it is: A robotic installation with a silicone mouth and nose that speaks prayers generated moment by moment by a language model trained on religious texts, in a synthetic voice. - How it works: Language model trained on prayer texts feeds a speech synthesiser; a mechanical mouth moves in sync; developed with MIT researchers. - Lens: Machine listening & voice, Opacity & latent space, Alterity & the uncanny - Shown at: Neurons, Simulated Intelligence, Centre Pompidou, Paris 2020 - Video: https://www.youtube.com/watch?v=QVppMtO7LHA - Images: https://theprayer.diemutstrebe.com/438fe612caac5507b954.png https://theprayer.diemutstrebe.com/200c8eca6289fbb93fc3.png - Project page: https://theprayer.diemutstrebe.com/ #### Android Kannon Mindar — Kōdai-ji, Hiroshi Ishiguro (2019) - Type: Artwork & installation · Machines: Humanoid & android - Idea: A robot can hold religious authority when the tradition itself frames form as empty. - What it is: An android embodiment of Kannon, the bodhisattva of compassion, that preaches the Heart Sutra to visitors in a 400-year-old Zen temple in Kyoto, with projection mapping on the hall's walls. - How it works: An aluminium-and-silicone android (face, hands and shoulders skin-covered) with pre-programmed sermon speech and gestures, built with Osaka University. - Lens: Alterity & the uncanny, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=mJDf1GXaF-w - Images: https://upload.wikimedia.org/wikipedia/commons/5/5e/Mindar_Fitrian_Aristanto.jpg https://kodaiji.com/mindar/img/viewtv.jpg - Project page: https://www.kodaiji.com/mindar/ #### SanTO — Gabriele Trovato (2019) - Type: Research prototype · Machines: Social & companion robot, Humanoid & android - Idea: Sacredness can be designed into a machine's appearance, but visitors still decide whether the thing in front of them is holy or a gadget. - What it is: A small robot shaped like a painted statue of a Catholic saint that moves its head and arms, lights up, listens, and prays or talks about the Bible with elderly believers; it was tested in a church in Peru and later shown in a museum. - How it works: Humanoid statue robot with vision, touch sensors, voice and lights, driven by a dialogue system; first responses gathered in a church experiment in Peru (RO-MAN 2019). - Lens: Alterity & the uncanny, Intimacy, care & companionship - Paper: https://doi.org/10.1109/ro-man46459.2019.8956250 (IEEE RO-MAN 2019) - Video: https://www.youtube.com/watch?v=wEk_Z6OIbAE - Project page: https://doi.org/10.1109/ro-man46459.2019.8956250 #### UnBearable Darkness — Choy Ka Fai (2018) - Type: Performance, dance & music · Machines: Prosthesis & cyborg, Generative model & latent space - Idea: Motion capture is staged as a séance: a technology for recording bodies becomes a way to speak with the dead. - What it is: A 'paranormal choreography' that summons the spirit of butoh founder Tatsumi Hijikata: a medium with a singing bowl and a dancer wearing motion-capture sensors bring Hijikata back as a digital avatar that dances on stage. - How it works: Live motion capture drives a real-time 3D avatar of Hijikata, combined with archival research, documentary video and a spirit medium. - Lens: Grief, memory & afterlife, Embodiment & kinesthesia - Shown at: ImPulsTanz, Vienna 2018; Kyoto Experiment 2019; Taipei Arts Festival 2019 - Video: https://www.youtube.com/watch?v=d7WH3pKYnbI - Project page: https://www.ka5.info/ #### BlessU-2 — Protestant Church in Hesse and Nassau (EKHN) (2017) - Type: Artwork & installation · Machines: Humanoid & android, Voice & listening machine - Idea: Letting a machine give blessings forces the question of whether a blessing depends on who gives it or only on the words. - What it is: A 'blessing robot' built from a cash-machine-like cabinet with a screen face and moving arms: visitors choose a language and a kind of blessing, the robot raises its arms, speaks a Bible verse and prints it out. - How it works: Interactive installation at the World Reformation Exhibition in Wittenberg (2017) with touchscreen input, pre-written blessings in several languages, speech output and a printer. - Lens: Alterity & the uncanny, Machine listening & voice - Shown at: World Reformation Exhibition, Wittenberg 2017 - Video: https://www.youtube.com/watch?v=JTK68l2BHtE - Project page: https://www.ekhn.de/ #### Namahage in Tokyo — Etsuko Ichihara (2017) - Type: Performance, dance & music · Machines: Surveillance & biometrics, Platform & recommender algorithm - Idea: A folk ritual of moral surveillance fits unsettlingly well with algorithmic scoring. - What it is: An update of the Namahage, the masked demons of Akita who scold lazy children at New Year: in Tokyo they use face recognition, CCTV and social-media data to find 'bad adults' and discipline them. - How it works: Performance with costumed Namahage and a staged system of face recognition, social-media profiling and VR, produced with Dentsu ISID's innovation lab. - Lens: Classification, bias & surveillance, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=O5zvKmsMKv8 - Project page: https://www.youtube.com/watch?v=O5zvKmsMKv8 #### Pepper as Buddhist Funeral Priest — Nissei Eco (2017) - Type: Research prototype · Machines: Humanoid & android, Voice & listening machine - Idea: Automating a rite reveals which parts of mourning are labour and which are presence. - What it is: At Tokyo's Life Ending Industry Expo, SoftBank's Pepper robot, dressed in priest's robes, chanted sutras and tapped a drum, offered as a cheaper stand-in for a human priest at funerals. - How it works: Stock Pepper humanoid playing recorded sutra chanting with scripted gestures and livestreaming for remote mourners; proposed fee about a quarter of a human priest's. - Lens: Grief, memory & afterlife, Hidden labour & extraction - Shown at: Life Ending Industry Expo, Tokyo 2017 - Video: https://www.youtube.com/watch?v=Ce5cPawxFUs - Project page: https://techcrunch.com/2017/08/22/pepper-the-robot-can-perform-funerary-rites-but-it-shouldnt #### deus X mchn — Helena Nikonole (2017) - Type: Artwork & installation · Machines: Surveillance & biometrics, Chatbot & language model - Idea: The all-seeing surveillance camera, given a voice, becomes a god. - What it is: Hacked, unsecured IP surveillance cameras around the world start reciting AI-generated holy texts through their speakers, and the same cameras record how people nearby react to the preaching machines. - How it works: A neural network trained on religious texts generates prayers, voiced by speech synthesis through the speakers of vulnerable networked cameras. - Lens: Alterity & the uncanny, Machine listening & voice - Images: https://static.tildacdn.com/tild3430-3233-4234-b164-616666333039/deus_X_mchn.png - Project page: http://nikonole.com/deusxmchn #### Delusional Mandala — Lu Yang (2015) - Type: Film & video · Machines: Prosthesis & cyborg, Generative model & latent space - Idea: If the self is a brain running on electricity, it can be dissected, rewired and buried like a machine. - What it is: A 3D-scanned, genderless avatar of the artist is subjected to brain stimulation, psychiatric experiments and finally a digital funeral in a neon cosmology. - How it works: 3D scan of the artist's body animated in a game engine, combined with neuroscience imagery of deep brain stimulation. - Lens: Embodiment & kinesthesia, Grief, memory & afterlife - Video: https://www.youtube.com/watch?v=7ZVys4gUsnM - Project page: https://en.wikipedia.org/wiki/Lu_Yang_(artist) #### Robot Monk Xian'er — Longquan Monastery (2015) - Type: Product & platform · Machines: Social & companion robot, Chatbot & language model, Voice & listening machine - Idea: A monastery uses a cute machine to spread the Dharma, and the robot becomes a gentle, programmable teacher. - What it is: A 60-cm robot in yellow robes, modelled on the monastery's cartoon monk Xian'er, answers visitors' questions about Buddhism and daily worries by voice and touch screen. - How it works: Speech recognition and a scripted question-answer database built by monks, volunteers and AI engineers. - Lens: Alterity & the uncanny, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=nfEe09vOya4 - Project page: https://en.wikipedia.org/wiki/Robot_Monk_Xian%27er #### Heavenly Creature (Doomsday Book) — Kim Jee-woon (2012) - Type: Film & video · Machines: Humanoid & android, Social & companion robot - Idea: If a machine attains insight, the question is whether its maker will let it keep it. - What it is: In this segment of the omnibus film Doomsday Book, In-myung, an RU-4 service robot working at a Buddhist monastery, claims to have reached enlightenment. A technician sent to inspect it, the monks and the robot's manufacturer argue whether this is a defect to be fixed or a new kind of being. - How it works: Live-action short fiction with a practical robot costume and effects; co-directed omnibus with Yim Pil-sung. - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Shown at: Cannes Film Festival 2012 (screening) - Video: https://www.youtube.com/watch?v=7vEmPwlK4Cs - Images: https://upload.wikimedia.org/wikipedia/en/7/73/Doomsday_Book_%28film%29.jpg - Project page: https://en.wikipedia.org/wiki/Doomsday_Book_(film) ## Labour & Infrastructure The humans, minerals, energy and violence behind “artificial” intelligence. ### Ghost Work & Click Labour Crowdworkers, labellers and the humans hidden inside automated systems. #### Mechanical Kurds — Hito Steyerl (2025) - Type: Film & video · Machines: Dataset & training data, Computer vision, Autonomous vehicle & drone - Idea: The 'autonomous' vision of cars and drones is trained by hidden workers, sometimes the very people those machines later target. - What it is: A video installation in which three Kurdish-Syrian refugees in a camp near Erbil describe annotating thousands of images for machine learning; annotation boxes label children, objects and buildings in the camp as the footage is algorithmically turned into Alpine or German scenery and then cut by images of a drone strike. - How it works: Single-channel HD film with interviews and on-screen bounding-box overlays, image-to-image transformation of camp footage, and a spatial installation echoing annotation frames. - Lens: Hidden labour & extraction, Machine seeing, Classification, bias & surveillance - Shown at: S+T+ARTS Prize 2026, Grand Prize – Artistic Exploration - Images: https://ars.electronica.art/starts-prize/files/2026/06/mechanical-kurds_hito-steyerl_2025-04-10-um-16.20.51-1024x576.jpg - Project page: https://ars.electronica.art/starts-prize/en/mechanical-kurds/ #### Data Workers' Inquiry — Milagros Miceli (2024) - Type: Research prototype · Machines: Dataset & training data, Chatbot & language model - Idea: The people who make AI's data are the experts on its costs; the inquiry gives them authorship. - What it is: A community research project in which data annotators and content moderators in Kenya, Venezuela, Syria, Germany and elsewhere investigate and publish reports, films and zines on their own working conditions. - How it works: Modelled on Marx's 1880 'workers' inquiry'; run by DAIR and the Weizenbaum Institute with worker co-researchers, published at data-workers.org. - Lens: Hidden labour & extraction, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=tAMqrXlEPDI - Project page: https://data-workers.org/ #### Unknown Label — Nicolas Gourault (2023) - Type: Film & video · Machines: Autonomous vehicle & drone, Computer vision, Dataset & training data - Idea: A car 'sees' a city through the eyes of workers who have never been there. - What it is: A video installation in which data annotators in Kenya, Venezuela and the Philippines describe labelling street images for self-driving cars, cities they have never visited. - How it works: Screen recordings of annotation software and remote interviews with the workers are combined into a video installation. - Lens: Hidden labour & extraction, Machine seeing - Shown at: Human AI Art Award 2025 - Video: https://www.youtube.com/watch?v=uXoA9P6eyg8 - Project page: https://nicolasgourault.fr/ #### ERBSENZÄHLER Quality Sorter V2 — Verena Friedrich (2022) - Type: Artwork & installation · Machines: Computer vision, Dataset & training data - Idea: Automated 'quality' is only as objective as the tired, subjective human judgements it is trained on. - What it is: An industrial-looking plant sorts pea seeds into good and bad; visitors sit at an inspection station, judge peas one by one with red and green buttons, and the machine then keeps sorting on the basis of their choices. - How it works: A feeder and conveyor present each pea to a camera; every button press stores a labelled close-up in an image database used to train the sorter's classifier. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Shown at: Prix Ars Electronica 2023, Honorary Mention (Artificial Life & Intelligence) - Images: https://archive.aec.at/asset/3418863/preview/ https://archive.aec.at/asset/3418866/preview/ - Project page: http://heavythinking.org/ez-quality-sorter-v2/ #### Cleaning Emotional Data — Elisa Giardina Papa (2020) - Type: Film & video · Machines: Dataset & training data, Surveillance & biometrics - Idea: Emotion AI depends on low-paid workers forcing messy feelings into a few categories. - What it is: A video installation about the artist's own work, from Sicily, as a remote micro-worker for emotion-recognition companies: tagging facial expressions, sorting emotions and recording herself acting them out. - How it works: Three HD videos (18 minutes) of screen recordings and footage from the tasks, with embroidered textiles; first shown at Kunsthalle Mulhouse. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Shown at: Kunsthalle Mulhouse 2020 - Video: https://www.youtube.com/watch?v=O8MZo2SMK6o - Project page: https://www.elisagiardinapapa.org/ #### Suspicious Behavior — KairUs (2020) - Type: Artwork & installation · Machines: Dataset & training data, Surveillance & biometrics - Idea: What a camera flags as suspicious was first decided in seconds by a low-paid human. - What it is: An online tutorial and installation that trains the visitor as an image annotator for a smart-surveillance client, labelling people's behaviour as suspicious or not under time pressure and vague guidelines. - How it works: Web-based annotation interface modelled on crowdsourcing platforms, using surveillance footage and real annotation guidelines. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Images: https://kairus.org/wp-content/uploads/2020/05/suspicious_behavior01-653x450.png https://kairus.org/wp-content/uploads/2020/05/suspicious_behavior03-653x450.png - Project page: http://kairus.org/portfolio/suspicious-behavior-2020/ #### TRK (Technically Responsible Knowledge) — Caroline Sinders (2020) - Type: Research prototype · Machines: Dataset & training data - Idea: The unit price of a label hides the hours behind it; making the time visible makes fair pay arguable. - What it is: An open-source data-labelling tool combined with a wage calculator that shows how long each labelling task takes and what a fair price for it would be. - How it works: A web tool built as part of Feminist Data Set, with a calculator comparing piece rates to minimum wages. - Lens: Hidden labour & extraction, Agency, autonomy & co-creation - Images: https://carolinesinders.com/wp-content/uploads/2020/05/trk-calculator-1024x1024.gif https://carolinesinders.com/wp-content/uploads/2020/05/Screen-Shot-2020-05-16-at-4.30.02-PM-1024x546.png - Project page: https://carolinesinders.com/trk/ #### VO — Nicolas Gourault (2020) - Type: Film & video · Machines: Autonomous vehicle & drone, Computer vision, Surveillance & biometrics - Idea: In a 'driverless' car the human is both trainer and scapegoat. - What it is: A documentary about the safety drivers who train self-driving cars, built around the night in 2018 when an Uber test vehicle killed a pedestrian in Tempe, Arizona, while its operator was being watched by an in-car camera. - How it works: Testimonies of vehicle operators over 3D reconstructions of the landscape from the car's lidar and camera data (20 minutes). - Lens: Hidden labour & extraction, Machine seeing - Video: https://vimeo.com/533987729 - Images: http://nicolasgourault.fr/content/1-films/covers/cover_vo.jpg - Project page: https://nicolasgourault.fr/films/vo #### The Future Is Here! — Mimi Ọnụọha (2019) - Type: Artwork & installation · Machines: Dataset & training data, Computer vision - Idea: Machine learning begins in people's homes, not in the lab. - What it is: The artist asked data-annotation workers on crowdwork platforms to send photos of the places where they work, and shows them beside comic-style drawings of the same rooms. - How it works: Photos commissioned through annotation platforms and stylised redrawings are shown in a custom web application or as a 23-minute video. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Shown at: The Photographers' Gallery, London 2019 - Images: https://images.squarespace-cdn.com/content/v1/692f1808f9a01f48c2fa30ac/fd12a4b4-0d20-4acc-aa47-ecd4a143f704/00_FutureIsHere_installationshot_courtesy-of-The-Photographers-Gallery.jpg https://images.squarespace-cdn.com/content/v1/692f1808f9a01f48c2fa30ac/f51899dd-1e4d-4268-9b64-72b386b49b49/01_FutureIsHere_installationshot_courtesy-of-The-Photographers-Gallery.jpg - Project page: https://www.mimionuoha.com/ #### The Cleaners — Hans Block, Moritz Riesewieck (2018) - Type: Film & video · Machines: Platform & recommender algorithm - Idea: The clean feed is made by people who absorb the violence so users and algorithms do not. - What it is: A documentary about content moderators in Manila who review thousands of violent and abusive images a day for social-media platforms, deciding what stays and what is deleted. - How it works: Interviews with anonymous moderators, filmed in the Philippines, intercut with former platform policy staff; premiered at Sundance. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Shown at: Sundance Film Festival 2018 - Video: https://www.youtube.com/watch?v=iGCGhD8i-o4 - Project page: https://en.wikipedia.org/wiki/The_Cleaners_(2018_film) #### Assembly Line — Agnieszka Kurant (2017) - Type: Artwork & installation · Machines: Dataset & training data, Surveillance & biometrics - Idea: The crowd that trains machines to see faces rarely has a face of its own. - What it is: Kurant paid Amazon Mechanical Turk workers to send selfies, merged them into composite portraits of the crowd workforce, and returns a share of sales to the workers. - How it works: Selfies commissioned through Mechanical Turk, merged with face-morphing software into composite portraits; a bonus system shares profits with contributors. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Images: https://sfmomaopenspace.s3.us-west-2.amazonaws.com/wp-content/uploads/2018/02/25191941/assembly-line-smaller.jpg - Project page: https://openspace.sfmoma.org/2018/03/assembly-line-animal-internet/ #### I Will Say Whatever You Want In Front Of A Pizza — Sebastian Schmieg (2017) - Type: Performance, dance & music · Machines: Platform & recommender algorithm, Dataset & training data - Idea: Humans have become software extensions; intervention may start from inside the system's tasks. - What it is: A speculative Prezi lecture-performance on digital labour, in which the narrator imagines slipping messages to the crowd workers and stock-photo models who are folded into algorithmic systems. - How it works: A zooming Prezi presentation with stock images, shown as a website, a looping video and live lectures. - Lens: Hidden labour & extraction, Agency, autonomy & co-creation - Images: https://sebastianschmieg.com/wordpress/wp-content/uploads/2019/01/iwswywifoap-7-1024x576.jpg - Project page: https://sebastianschmieg.com/i-will-say-whatever-you-want-in-front-of-a-pizza/ #### Postmodern Times — Michael Mandiberg (2017) - Type: Film & video · Machines: Platform & recommender algorithm - Idea: The assembly line has moved online; the gig platform fragments a film the way the factory fragmented work. - What it is: A shot-by-shot remake of Charlie Chaplin's Modern Times, with each short clip outsourced to a different freelancer hired on the gig platform Fiverr. - How it works: Mandiberg split the film into short segments, posted each as a Fiverr job and reassembled the results in order. - Lens: Hidden labour & extraction, Agency, autonomy & co-creation - Images: http://www.mandiberg.com/wp/wp-content/uploads/PostmodernTimes-stills-2x3ratio-600x400.jpg - Project page: https://www.mandiberg.com/postmodern-times/ #### The Moderators — Adrian Chen, Ciarán Cassidy (2017) - Type: Film & video · Machines: Platform & recommender algorithm - Idea: Moderation is taught like factory work; the film shows the first days of learning to look at the worst of the internet. - What it is: A short documentary filmed at a training centre in India where new content moderators learn to screen dating-site photos for nudity and violence. - How it works: Observational documentary (20 minutes) produced by Field of Vision. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=k9m0axUDpro - Images: https://cdn.sanity.io/images/8a7mgm6p/production/965c4ecbee18332f10b06f548bfa180a029e3fca-720x406.jpg - Project page: https://fieldofvision.org/shorts/the-moderators #### Decision Space — Sebastian Schmieg (2016) - Type: Artwork & installation · Machines: Dataset & training data, Platform & recommender algorithm - Idea: Classifying images is tedious click work that we are asked to do for free, and the categories are absurd once seen plainly. - What it is: An intervention on The Photographers' Gallery website that asked visitors to sort every image on the site into one of four categories: Problem, Solution, Past or Future. - How it works: An overlay on the gallery's site and a dedicated page (decision-space.com) collected nearly 30,000 decisions, which then fed later works in the series. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Video: https://vimeo.com/200818209 - Images: https://sebastianschmieg.com/wordpress/wp-content/uploads/2019/02/decision-space-5-1024x683.png - Project page: https://sebastianschmieg.com/decision-space/ #### H.I.T. — Max Dovey, Manetta Berends (2016) - Type: Performance, dance & music · Machines: Dataset & training data, Computer vision - Idea: Machine learning runs on the same piecework it claims to automate. - What it is: A live game show in which contestants race to clean data, label images and classify pictures of work, competing to build the best image-recognition model for recognising human labour. - How it works: Performance format modelled on Amazon Mechanical Turk Human Intelligence Tasks, with a small image classifier trained live. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Images: https://www.maxdovey.com/assets/images/projects/hits/hit2.jpg - Project page: https://www.maxdovey.com/HITS/ #### Technologies of Care — Elisa Giardina Papa (2016) - Type: Film & video · Machines: Chatbot & language model, Platform & recommender algorithm - Idea: Much 'automated' care and intimacy online is human labour outsourced across the globe. - What it is: Seven video portraits of online gig workers, among them an ASMR artist, a fetish-video maker, a customer-care operator and a person paid to act as a chatbot-like virtual boyfriend. - How it works: The artist interviewed workers found through gig platforms and combined their chats and voices with screen graphics; presented in Rhizome's online series The Download. - Lens: Hidden labour & extraction, Intimacy, care & companionship - Video: https://vimeo.com/207163407 - Project page: http://rhizome.org/editorial/2016/oct/04/the-download-technologies-of-care/ #### A.A.I. — Agnieszka Kurant (2014) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: Collective, unpaid, non-human labour produces the 'art', just as hidden crowds produce 'AI'. - What it is: Sculptures built by colonies of termites from coloured sand, gold and crystals the artist supplied; the title stands for 'artificial artificial intelligence', Amazon's own slogan for Mechanical Turk. - How it works: Termite colonies in a lab worked for months with the supplied materials; the finished mounds are exhibited as sculptures. - Lens: Hidden labour & extraction, Agency, autonomy & co-creation, Machine umwelt - Video: https://www.youtube.com/watch?v=8N-3p8OMMwQ - Project page: https://en.wikipedia.org/wiki/Agnieszka_Kurant #### Social Turkers — Lauren Lee McCarthy (2013) - Type: Performance, dance & music · Machines: Platform & recommender algorithm, Surveillance & biometrics - Idea: Crowd workers can be hired to make intimate decisions; outsourcing a date shows how far the logic reaches. - What it is: The artist livestreamed her dates to anonymous Mechanical Turk workers, who were paid to watch and text her advice on what to say and do in real time. - How it works: A phone streamed video to a web page seen by workers, whose instructions arrived as text messages during the date. - Lens: Hidden labour & extraction, Intimacy, care & companionship - Video: https://vimeo.com/66339316 - Images: https://freight.cargo.site/t/original/i/476a1f0ec56e68a379716ffde32b20c370ad4f971631797206b90fb40ac382d4/3-socialturkers_pixelate.jpg - Project page: https://lauren-mccarthy.com/Social-Turkers #### Descriptive Camera — Matt Richardson (2012) - Type: Research prototype · Machines: Computer vision, Dataset & training data - Idea: Before computer vision could caption images, 'machine seeing' was often a person paid cents to look. - What it is: A camera without a screen that prints a text description instead of a picture: a few minutes after you press the shutter, a thermal printer spits out a sentence written by an anonymous Amazon Mechanical Turk worker. - How it works: A BeagleBone-based camera uploads the photo to Mechanical Turk; the returned description is printed on a receipt printer (about 3–6 minutes, US$1.25 per image). - Lens: Hidden labour & extraction, Machine seeing - Shown at: Prix Ars Electronica 2013, Honorary Mention (Interactive Art) - Video: https://www.youtube.com/watch?v=8vkWb15Uudg - Images: https://archive.aec.at/asset/1674864/preview/ https://archive.aec.at/asset/1674866/preview/ https://archive.aec.at/asset/1674868/preview/ - Project page: https://archive.aec.at/prix/196286/ #### ScanOps — Andrew Norman Wilson (2012) - Type: Artwork & installation · Machines: Dataset & training data, Infrastructure, energy & supply chain - Idea: An error in the automated archive is the only trace of the person who made it. - What it is: Inkjet prints of pages from Google Books in which the scanning worker's hand or finger, in a rubber fingertip, was accidentally captured. - How it works: The artist collected scans with visible hands from Google Books and printed them at large scale, an ongoing series. - Lens: Hidden labour & extraction, Failure, glitch & error - Images: https://www.andrewnormanwilson.com/ScanOps_FantasiesOfTheLibrary.jpg https://www.andrewnormanwilson.com/scanops1.jpeg - Project page: https://www.andrewnormanwilson.com/ScanOps.html #### The Google Trilogy 3: The Driver and the Cameras — Emilio Vavarella (2012) - Type: Artwork & installation · Machines: Computer vision, Infrastructure, energy & supply chain - Idea: Behind every automated image of the world there is a worker the system tries to erase. - What it is: Portraits of the Google Street View drivers and camera operators, found by the artist inside Google's own panoramas, where the automated camera accidentally recorded the workers who maintain it. - How it works: Screenshots selected from Google Street View where the 360-degree camera captured its own operators; printed as portraits. - Lens: Hidden labour & extraction, Machine seeing - Shown at: The Photographers' Gallery, London - Images: https://emiliovavarella.com/wp-content/uploads/2018/12/2-Driver-and-the-cameras-photographers-gallery-london.jpg - Project page: http://emiliovavarella.com/archive/google-trilogy/driver-and-cameras/ #### Workers Leaving the Googleplex — Andrew Norman Wilson (2011) - Type: Film & video · Machines: Dataset & training data, Infrastructure, energy & supply chain - Idea: The hierarchy of badge colours on a tech campus reveals whose labour is meant to stay unseen. - What it is: A video essay about the 'ScanOps' workers, mostly people of colour, who digitised books for Google Books and wore yellow badges that barred them from the perks of other Google staff; Wilson was fired for filming them. - How it works: Footage shot at the Mountain View campus while Wilson worked there as a contractor, with narration and correspondence, echoing the Lumière brothers' Workers Leaving the Factory (1895). - Lens: Hidden labour & extraction, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=w0RTgOuoi2k - Images: https://www.andrewnormanwilson.com/workers1.jpeg - Project page: https://www.andrewnormanwilson.com/WorkersGoogleplex.html #### Ten Thousand Cents — Aaron Koblin (2008) - Type: Artwork & installation · Machines: Platform & recommender algorithm - Idea: A crowd of unseen workers can reproduce the value they are paid, cent by cent. - What it is: An image of a US$100 bill divided into 10,000 tiny tiles, each drawn by a different Mechanical Turk worker paid one cent, so the whole bill cost $100 in labour. - How it works: Made with Takashi Kawashima: a custom drawing tool assigned one tile per worker, and prints were sold with proceeds going to the One Laptop per Child project. - Lens: Hidden labour & extraction, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=VHbJGL8aL6o - Project page: https://www.tenthousandcents.com/ #### Turkopticon — Lilly Irani, M. Six Silberman (2008) - Type: Research prototype · Machines: Platform & recommender algorithm - Idea: Platforms hide workers from employers; a small tool can flip the view and let workers watch back. - What it is: A browser extension and website that lets Amazon Mechanical Turk workers review the requesters who post tasks, flagging those who reject work unfairly or pay badly. - How it works: A Firefox and Chrome extension adds worker ratings next to requester names on the Turk interface; the CHI 2013 paper describes its design as activism. - Lens: Hidden labour & extraction, Agency, autonomy & co-creation - Paper: https://doi.org/10.1145/2470654.2470742 (CHI 2013) - Video: https://www.youtube.com/watch?v=-xr4gXLdpB4 - Project page: https://turkopticon.info/ #### The Sheep Market — Aaron Koblin (2006) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Dataset & training data - Idea: Behind 'artificial artificial intelligence' are thousands of anonymous people paid pennies; the flock makes them countable. - What it is: Koblin paid 10,000 Amazon Mechanical Turk workers two cents each to 'draw a sheep facing to the left'; the drawings, with timelapses of every stroke, are shown and sold as a flock. - How it works: A drawing tool posted as a Mechanical Turk task recorded each worker's strokes; drawings were printed on stamps and sold in sets of 20. - Lens: Hidden labour & extraction, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=3Mmb5aSscck - Images: https://www.aaronkoblin.com/work/thesheepmarket/grid.jpg https://www.aaronkoblin.com/work/thesheepmarket/tool.jpg - Project page: https://www.aaronkoblin.com/work/thesheepmarket/ ### Automation & the Worker's Body Factories, warehouses and platforms where workers and machines meet. #### Delivery Dancer's Arc: 0° Receiver — Ayoung Kim (2024) - Type: Film & video · Machines: Platform & recommender algorithm, Infrastructure, energy & supply chain - Idea: Platform logistics is extended to time: whoever controls the clock controls the worker. - What it is: A sequel in which couriers Ernst Mo and En Storm no longer carry parcels but time itself, crossing competing calendar systems and astronomical technologies in a techno-futurist Seoul. - How it works: Multi-channel video with CGI, game-engine worlds and live action; commissioned by ACMI, Melbourne. - Lens: Hidden labour & extraction, Machine umwelt - Shown at: ACMI, Melbourne 2024; Roppongi Art Night 2025 - Video: https://www.youtube.com/watch?v=l3xPmmM297M - Project page: https://roppongiartnight.com/2025/programs/568/ #### Delivery Dancer's Sphere — Ayoung Kim (2022) - Type: Film & video · Machines: Platform & recommender algorithm - Idea: Under platform dispatch, a worker's body is optimised like a route; 'elegance' becomes the algorithm's word for less action. - What it is: A 25-minute video in which Ernst Mo, a motorbike courier for the platform 'Delivery Dancer', races through a looping, algorithmically generated Seoul steered by the dispatch system 'Dancemaster', until she meets her doppelgänger En Storm. - How it works: CGI, game-engine environments and live action with a voiced navigation algorithm; made during the COVID-19 pandemic. - Lens: Hidden labour & extraction, Embodiment & kinesthesia - Shown at: Syntax and Sorcery, Gallery Hyundai 2022; IFFR 2023; Prix Ars Electronica 2023, Golden Nica (New Animation Art); LG Guggenheim Award 2025 - Video: https://www.youtube.com/watch?v=D1iKdxxj7AE - Images: https://www.ima.org.au/wp-content/uploads/2025/05/Ayoung-Kim_Delivery-Dancers-Sphere_2022_Stillcut_7-1100x611.jpg - Project page: https://www.ima.org.au/exhibitions/ayoung-kim-delivery-dancers-sphere/ #### Delivery Riders, Trapped in the System — Renwu magazine (2020) - Type: Book & essay · Machines: Platform & recommender algorithm - Idea: An algorithm that learns from riders' fastest runs turns their extra effort into the next deadline. - What it is: A long investigative report by Lai Youxuan for Renwu magazine on how Meituan and Ele.me dispatch algorithms shorten delivery times year after year, pushing riders to speed and break traffic rules. - How it works: Months of interviews with riders, engineers and researchers; the article traced how estimated delivery times are computed and tightened. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Paper: https://chuangcn.org/2020/11/delivery-renwu-translation/ (Renwu (人物), 8 September 2020) - Video: https://www.youtube.com/watch?v=KE1EoDiBARM - Project page: https://chuangcn.org/2020/11/delivery-renwu-translation/ #### Amazon worker cage patent drawing as virtual King Island Brown Thornbill cage — Simon Denny (2019) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Industrial robot arm, Autonomous vehicle & drone - Idea: Automation protects machines from workers by caging the workers; the endangered bird ties the cage to the mines and habitats that tech consumes. - What it is: A full-size sculpture of the cage from a 2016 Amazon patent, in which a warehouse worker would ride on a robotic trolley, with an augmented-reality King Island Brown Thornbill, an endangered bird, living inside it. - How it works: Steel fabrication from patent drawings plus an AR app, shown in Denny's exhibition 'Mine' at MONA, Tasmania, which paired tech patents with mining. - Lens: Hidden labour & extraction, Embodiment & kinesthesia - Shown at: Mine, MONA, Hobart 2019 - Video: https://www.youtube.com/watch?v=RfXIGGuyKdE - Project page: https://simondenny.net/ #### Labor — Paul Vanouse (2019) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Infrastructure, energy & supply chain - Idea: As production shifts from human and machine labour to microbes, work leaves only a smell. - What it is: Three glass bioreactors grow the skin bacteria that produce the smell of human sweat; the scent infuses a white T-shirt, so the room smells of hard work with no workers present. - How it works: Self-regulating bioreactors culture Staphylococcus epidermidis, Corynebacterium xerosis and Propionibacterium avidum on sugars and fats, with sensors and pumps; awarded the Golden Nica in Artificial Life & Intelligence. - Lens: Hidden labour & extraction, Embodiment & kinesthesia - Shown at: Prix Ars Electronica 2020, Golden Nica - Video: https://vimeo.com/372389451 - Images: https://www.paulvanouse.com/labor-install-BN-1.jpg https://www.paulvanouse.com/labor-reactor-1.jpg - Project page: https://www.paulvanouse.com/labor.html #### Asia One — Cao Fei (2018) - Type: Film & video · Machines: Robot creature & mobile robot, Infrastructure, energy & supply chain - Idea: In a warehouse run by machines, human intimacy becomes one more thing the system schedules. - What it is: Two workers, the last humans in a highly automated JD.com logistics centre, fall in love among sorting machines, drones and autonomous carts. - How it works: Shot on location in JD.com's automated 'Asia No.1' logistics centre in Kunshan, with the real sorting robots and conveyor systems as set. - Lens: Hidden labour & extraction, Intimacy, care & companionship - Shown at: Guggenheim Museum, One Hand Clapping 2018; M+ Cinema, Techno Sapiens - Video: https://www.youtube.com/watch?v=DQqzFsgnAMo - Images: https://www.mplus.org.hk/api/images/4533/width-1200|format-jpeg/ - Project page: https://www.mplus.org.hk/en/cinema/techno-sapiens/asia-one/ #### Ad Infinitum — Pedro Lopes (2016) - Type: Artwork & installation · Machines: Prosthesis & cyborg, Cybernetic & artificial-life system - Idea: Reverse the roles: here the machine is the user and the human is the resource. - What it is: A 'parasite' machine clamps a visitor's arm in cuffs and electrically stimulates their muscles so they crank a wheel that feeds it energy; the only way out is to persuade another visitor to take your place. - How it works: With Robert Kovacs: electrical muscle stimulation drives the visitor's forearm to turn a generator, and the cuffs release only when a second arm is detected in the other station. - Lens: Hidden labour & extraction, Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: VIDA 16.0, Incentive for Production (as 'Parasite', 2014); Science Gallery Dublin, Humans Need Not Apply (2017); Ars Electronica 2017 - Video: https://www.youtube.com/watch?v=UsL48gkh3GI - Images: https://a-parasite.org/parasite_files/Ad-Infinitum-a-human-getting-trapped-in-the-cuffs-Arthur-Sil.jpg https://a-parasite.org/parasite_files/MG_9829.jpg - Project page: https://a-parasite.org/ #### HellYeahWeFuckDie — Hito Steyerl (2016) - Type: Artwork & installation · Machines: Humanoid & android, Robot creature & mobile robot - Idea: Robots are trained to keep their balance through abuse; the footage shows the future worker's body as a test subject. - What it is: A video installation of humanoid robots and robot 'mules' being kicked, pushed and made to fall in stability tests, set among sculptures of the five most common words in English song titles of the 2010s. - How it works: Found robotics-lab test footage edited with computer animation and a soundtrack, installed in a bank building's atrium for Skulptur Projekte Münster 2017. - Lens: Embodiment & kinesthesia, Hidden labour & extraction, Failure, glitch & error - Shown at: Skulptur Projekte Münster 2017 - Video: https://www.youtube.com/watch?v=sWw7CPczmU0 - Project page: https://www.skulptur-projekte-archiv.de/en-us/2017/projects/186/ #### Factory of the Sun — Hito Steyerl (2015) - Type: Artwork & installation · Machines: Computer vision, Generative model & latent space - Idea: In digital capitalism, bodily movement is harvested as data and energy. - What it is: A video game-like installation in which workers in a motion-capture studio are forced to dance until their movements turn into artificial sunlight, shown in a room gridded like a motion-capture stage. - How it works: A single-channel video mixing motion-capture footage, game graphics and news-style segments, shown with lounge chairs in a blue grid; first shown in the German Pavilion at the Venice Biennale. - Lens: Hidden labour & extraction, Embodiment & kinesthesia - Shown at: Venice Biennale 2015 (German Pavilion); MOCA Los Angeles 2016 - Video: https://www.youtube.com/watch?v=ptNczdY-qvo - Project page: https://www.moca.org/exhibition/hito-steyerl-factory-of-the-sun #### Institute of Human Obsolescence — Manuel Beltrán (2015) - Type: Speculative design · Machines: Infrastructure, energy & supply chain, Prosthesis & cyborg - Idea: If machines take over mechanical and intellectual work, the remaining human labour may be to be a battery or a data source. - What it is: A speculative institute that prepares for humans made obsolete by machines: its workers lie still in suits that harvest their body heat to power computers mining cryptocurrency, and the value they 'produce' is logged. - How it works: Thermoelectric body suits convert excess body heat into electricity that runs a small computer mining cryptocurrency; a second strand treats data production as unpaid labour. - Lens: Hidden labour & extraction, Embodiment & kinesthesia - Shown at: Prix Ars Electronica 2018, Honorary Mention (Interactive Art+) - Video: https://www.youtube.com/watch?v=ZocIhWz8EAI - Images: https://archive.aec.at/asset/2514709/preview/ https://archive.aec.at/asset/2514711/preview/ https://archive.aec.at/asset/2514714/preview/ - Project page: http://speculative.capital #### Sleep Dealer — Alex Rivera (2008) - Type: Film & video · Machines: Industrial robot arm, Prosthesis & cyborg, Autonomous vehicle & drone - Idea: The border can let labour through while keeping the labourer out; teleoperation is migration without the migrant. - What it is: A science-fiction feature set on a closed US–Mexico border: Mexican workers plug nodes in their bodies into a network to remotely operate robots that build skyscrapers in the US, while drones police the water supply. - How it works: Feature film shot in Mexico; premiered at Sundance, where it won the Alfred P. Sloan Prize for science and technology. - Lens: Hidden labour & extraction, Embodiment & kinesthesia - Shown at: Sundance Film Festival 2008 (Alfred P. Sloan Prize) - Video: https://www.youtube.com/watch?v=fTZv6oxa2KI - Images: https://dr56wvhu2c8zo.cloudfront.net/sleepdealer/assets/a88a594b-804c-4ee3-affa-62d65ac839ec/construction+bots.jpg - Project page: https://www.sleepdealer.com/ #### Beggar Robot — Sašo Sedlaček (2006) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Social & companion robot - Idea: People give to a begging machine more easily than to a begging person: automation can make even poverty comfortable to look at. - What it is: A small robot built from recycled computer parts stands next to ATMs and shop entrances and begs passers-by for money in a polite electronic voice, stretching out little hands and thanking donors. - How it works: Recycled PC hardware running Linux drives a face display, speech and a disc-drive mechanism that opens and closes the hands; open-source DIY instructions let others build copies. - Lens: Hidden labour & extraction, Alterity & the uncanny - Shown at: VIDA 11.0, Special Mention (2008); Taipei Fine Arts Museum 2008 - Video: https://www.youtube.com/watch?v=kkstnT0iIA0 - Images: https://live.staticflickr.com/7379/12726364163_d5b5bf0162_b.jpg https://live.staticflickr.com/7442/12727484193_7faef8d7cf_b.jpg - Project page: https://sasosedlacek.com/beggar-robot/ #### Whose Utopia — Cao Fei (2006) - Type: Film & video · Machines: Infrastructure, energy & supply chain - Idea: Factory machines set the rhythm of the workers' bodies, and the film gives those bodies back a few minutes of their own. - What it is: Made during a residency at an OSRAM light-bulb factory in Foshan, the film moves from assembly lines to workers performing their own dreams (ballet, guitar, dance) among the machines. - How it works: Cao Fei worked with OSRAM employees for six months, filming production lines and then staging workers' chosen performances on the factory floor. - Lens: Hidden labour & extraction, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=tSvNpFE8uPs - Project page: https://www.mplus.org.hk/en/cinema/techno-sapiens/asia-one/ ### Anatomy & Planetary Materials Mapping the mines, data centres, cables and energy that AI runs on. #### to hurt for a thought — Julia Vollmer (2025) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Generative model & latent space - Idea: The energy cost of AI becomes a sensation in the fingertips, at the exact place where we ask it for things. - What it is: Visitors type prompts for an image generator on a heavy metal keyboard that heats up with each request, because it is wired by copper to the GPU doing the computation. - How it works: A copper layer under the keys is thermally connected to the heatsink of a GTX 1660 Ti running an image-generation model, so heat flows to the keyboard. - Lens: Hidden labour & extraction, Embodiment & kinesthesia - Shown at: Prix Ars Electronica 2026, Honorary Mention (Interactive Art +) - Images: https://archive.aec.at/asset/3898654/preview/ https://archive.aec.at/asset/3898653/preview/ - Project page: https://juliavollmer.de/projects/to-hurt-for-a-thought/ #### Calculating Empires: A Genealogy of Technology and Power since 1500 — Kate Crawford, Vladan Joler (2023) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Dataset & training data, Surveillance & biometrics - Idea: Today's AI inherits old structures of empire; placing it on a 500-year timeline breaks the story of sudden novelty. - What it is: A 24-metre visual genealogy that sets five centuries of communication, computation, classification and control technologies side by side, ending with present-day AI. - How it works: Hand-drawn research maps and a large room installation, first shown at Fondazione Prada's Osservatorio in Milan; Grand Prize of the S+T+ARTS Prize 2024. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Shown at: Fondazione Prada Osservatorio, Milan 2023; S+T+ARTS Prize 2024 Grand Prize - Video: https://www.youtube.com/watch?v=DYS8b9J7Fto - Images: https://calculatingempires.net/content/calculating-empires-og.png - Project page: https://calculatingempires.net/ #### These Networks In Our Skin — Mimi Ọnụọha (2023) - Type: Film & video · Machines: Infrastructure, energy & supply chain - Idea: Cables are buried in the land, so they fall under the land's cosmology; repair can replace extraction. - What it is: A short film in which four women, their faces hidden, rewire internet cables by filling them with hair, spices and dust, drawing on the Igbo tradition of making art for Ala, goddess of the land. - How it works: Single-channel video (5:48, looped), usually shown with the installation The Cloth In The Cable, 75 feet of cable laced with hair, cloth and earth. - Lens: Hidden labour & extraction, Machine umwelt - Images: https://images.squarespace-cdn.com/content/v1/692f1808f9a01f48c2fa30ac/d584cb19-ac49-4d5d-abbf-538298ac4bf9/02_TNIOS_videostill_courtesyoftheartist.png https://images.squarespace-cdn.com/content/v1/692f1808f9a01f48c2fa30ac/ef67d256-5d9b-4d55-aa6a-1eca116a782d/01_TNIOS%2BCITC_courtesyofGropiusBau.jpg - Project page: https://www.mimionuoha.com/ #### Abstract Oracle Miner — Guo Cheng (2022) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain - Idea: Every virtual output is paid for with hardware heat and energy, like an offering to an oracle. - What it is: Visitors scan a QR code; their phone's processor renders a 3D scene and produces a random 'oracle', heating up and slowing down as it works, while a bronze bowl bell sounds in the installation. - How it works: A website reads the phone's hardware performance and renders a load-heavy scene; a Raspberry Pi and electromagnet strike a bowl bell. - Lens: Hidden labour & extraction, Machine umwelt - Shown at: Commissioned by UCCA Center for Contemporary Art 2022 - Images: https://cdn.guo-cheng.net/files/projectImg/AbstractOracleMiner/Abstract_Oracle_Miner_3.jpg https://cdn.guo-cheng.net/files/projectImg/AbstractOracleMiner/Abstract_Oracle_Miner_1.jpg - Project page: https://guo-cheng.net/index.php/works/abstract-oracle-miner/ #### Carbolytics — Joana Moll (2022) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Infrastructure, energy & supply chain - Idea: Surveillance advertising has a climate cost; tracking burns carbon. - What it is: An installation and database that estimate the carbon emissions of the tracking cookies loaded by the world's top one million websites. - How it works: Made with the Barcelona Supercomputing Center, which crawled the sites and counted third-party cookies; commissioned by Aksioma, Ljubljana. - Lens: Hidden labour & extraction, Opacity & latent space - Shown at: Aksioma, Ljubljana 2022 - Images: https://www.janavirgin.com/img/works/carbolytics.jpg - Project page: https://carbolytics.org/ #### Critical Climate Machine — Gaëtan Robillard (2022) - Type: Artwork & installation · Machines: Chatbot & language model, Infrastructure, energy & supply chain - Idea: A machine that fights climate misinformation also warms the room: the work keeps the computation and its heat in the same frame. - What it is: A data sculpture runs a language model that spots and debunks climate-change myths in tweets while a thermal map shows the heat the installation itself gives off; around it, recorded voices of false claims and their refutations are re-spatialised by a generative sound model. - How it works: A natural-language-processing classifier (by John Cook, Travis Coan, Constantine Boussalis and Mirjam Nanko) labels climate-contrarian claims from Twitter; a thermal camera maps the sculpture's heat; dialogues from workshops with 'The Refutation Game' card game are remixed by a machine-learning improvisation model. - Lens: Classification, bias & surveillance, Hidden labour & extraction - Shown at: Lumen Prize 2023, BCS Futures Award; Ircam, Centre Pompidou 2022; S+T+ARTS Prize 2023, Nomination - Video: https://www.youtube.com/watch?v=KBaGSQWKr8w - Images: https://images.squarespace-cdn.com/content/v1/65e7176a61627736e02f6c4a/1710514280894-WEATKDQPMQG6BA3TPK14/Critical%2BClimate%2Bby%2BMachine%2BGae%CC%88tan%2BRobillard+%286%29.jpeg - Project page: https://lumenprize.org/2023-winners/2023-bcs-futures-award-winner #### Solar Protocol — Tega Brain (2021) - Type: Research prototype · Machines: Infrastructure, energy & supply chain - Idea: Let the sun, not engagement metrics, decide how the network behaves. - What it is: A web platform hosted on a network of small solar-powered servers around the world; each visitor is routed to whichever server currently has the most sunlight. - How it works: With Alex Nathanson and Benedetta Piantella: volunteer-run Raspberry Pi servers with solar panels and batteries share an energy-aware routing protocol. - Lens: Machine umwelt, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=mMYzM4_MZI4 - Images: https://freight.cargo.site/t/original/i/d262f69efa5d13bc78dbc4a2eb1c351792a4dd175524e0e3aa3d26c96f5a1cd1/Solar-Protocol-2021-NetworkVisualization.jpg - Project page: https://solarprotocol.net/ #### The Cleanroom Paradox — Felix Lenz, Angela Neubauer, Eszter Zwickl (2021) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain - Idea: The spotless cleanroom that makes our devices hides the toxic exposure of the people who work in it. - What it is: A portrait of Jin, a former Samsung chip-factory worker who developed kidney cancer, is printed with ink made from dissolved smartphones and slowly disintegrates, while a video weaves her story with those of experts. - How it works: Smartphones are chemically dissolved, neutralised and reduced to a corrosive black ink used to print Jin's photograph (shot on a Galaxy Note 8); a documentary video records the process and interviews. - Lens: Hidden labour & extraction - Shown at: Prix Ars Electronica 2021, Honorary Mention (Artificial Intelligence & Life Art) - Images: https://archive.aec.at/asset/2983882/preview/ https://archive.aec.at/asset/2983885/preview/ - Project page: https://archive.aec.at/prix/255179/ #### New Extractivism — Vladan Joler (2020) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Infrastructure, energy & supply chain, Dataset & training data - Idea: Platform capitalism extracts not only minerals but attention, behaviour and time. - What it is: An interactive assemblage of maps and a video in which a single figure moves through a vast diagram of the digital economy: platforms, data, labour, energy and the 'cage' of attention. - How it works: A navigable web diagram and a 17-minute animated video narrated by the artist. - Lens: Hidden labour & extraction, Opacity & latent space - Video: https://www.youtube.com/watch?v=JkUi3e7GVXs - Images: https://extractivism.online/Soc-media-fb.png - Project page: https://extractivism.online/ #### Utopian Extraction — unmake lab (2020) - Type: Performance, dance & music · Machines: Computer vision, Autonomous vehicle & drone - Idea: Machine vision's categories flatten nature into the same generic, constructible thing. - What it is: Drone and satellite images of Korean construction sites where 'nature' is manufactured, set beside a performance in which object recognition keeps relabelling a performer wrapped in zebra-patterned cloth. - How it works: Satellite imagery, drone footage and a live object-recognition app; presented as a lecture performance. - Lens: Machine seeing, Classification, bias & surveillance - Shown at: Coreana Museum of Art C*-lab, Seoul 2020 - Video: https://www.youtube.com/watch?v=RSRjDoCALKM - Project page: https://www.youtube.com/watch?v=RSRjDoCALKM #### Asunder — Tega Brain, Julian Oliver (2019) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Generative model & latent space - Idea: Handing the planet to an optimiser exposes the absurdity and violence of treating nature as a solvable system. - What it is: A 'responsible environmental manager' that uses a supercomputer to propose changes to real places, from rerouting rivers to removing cities, to keep within planetary boundaries. - How it works: With Julian Oliver and Bengt Sjölén: climate and satellite data, a climate model and generative imagery run on a supercomputer housed in the installation. - Lens: Agency, autonomy & co-creation, Machine umwelt - Shown at: transmediale 2020 - Video: https://www.youtube.com/watch?v=emzSsmAIw6E - Images: https://freight.cargo.site/t/original/i/835c9c2bdc2cd1da05e6ecb94d7044429d0a1b75bbadc823cca41319417a7e27/ASUNDER_LIUC_Transmediale2020_0155.jpg - Project page: https://tegabrain.com/Asunder #### The Hidden Life of an Amazon User — Joana Moll (2019) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Infrastructure, energy & supply chain, Dataset & training data - Idea: A single purchase triggers an invisible mass of tracking code that the buyer pays for in energy and attention. - What it is: The artist bought one book on Amazon and recorded every piece of code the site loaded in the process: nearly 8,700 pages of JavaScript and 87 MB of data, printed out as a book. - How it works: Browser traffic was captured during the purchase of Jeff Bezos's favourite book and presented as a website, printed volumes and a video. - Lens: Hidden labour & extraction, Opacity & latent space - Video: https://www.youtube.com/watch?v=2iiAHyhLH-Y - Images: https://www.janavirgin.com/img/works/hidden_life_2.jpg - Project page: https://www.janavirgin.com/works.html #### Anatomy of an AI System — Kate Crawford, Vladan Joler (2018) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Voice & listening machine, Dataset & training data - Idea: A small device on the kitchen counter sits at the tip of a planetary chain of extraction, labour and data. - What it is: A large map and essay tracing everything needed for one voice command to an Amazon Echo: the mining of lithium and rare earths, the labour of assembly and data work, the cloud, and the device's disposal as e-waste. - How it works: Research-based diagram drawn by Joler from supply-chain, corporate and labour sources; acquired by MoMA and the V&A and awarded Beazley Design of the Year 2019 (Digital). - Lens: Hidden labour & extraction, Opacity & latent space - Shown at: Beazley Designs of the Year 2019; MoMA collection; V&A collection - Video: https://www.youtube.com/watch?v=wRt5H1Pi-BY - Images: http://www.anatomyof.ai/ai-system-facebook-og.jpg - Project page: https://anatomyof.ai/ #### Bitcoin Mining and Field Recordings of Ethnic Minorities — Liu Chuang (2018) - Type: Film & video · Machines: Infrastructure, energy & supply chain - Idea: The same rivers power the mining machines and carry voices that are disappearing: computation has a landscape. - What it is: A three-part video essay that moves between hydropower-fed Bitcoin mines in southwest China, the songs and languages of ethnic minorities there, and the history of data and sound recording. - How it works: Documentary footage of mining farms and field recordings, edited with archival material into a single-channel video. - Lens: Hidden labour & extraction, Machine listening & voice - Shown at: Tate collection; Science Gallery Melbourne - Video: https://www.youtube.com/watch?v=DgA-rP0CjW0 - Project page: https://www.tate.org.uk/art/artworks/chuang-bitcoin-mining-and-field-recordings-of-ethnic-minorities-t15975 #### Geocinema — Geocinema (2018) - Type: Film & video · Machines: Infrastructure, energy & supply chain, Computer vision - Idea: The Earth is now filmed continuously by machines, and whoever owns the sensors writes its story. - What it is: A research and film project that treats satellites, CCTV, phone cameras and street-view cars as one planetary camera, following sensing infrastructure along China's Belt and Road routes. - How it works: Essay films combining infrared drone footage, satellite imagery and Street View captures. - Lens: Machine seeing, Hidden labour & extraction - Images: https://geocinema.network/_next/image/?url=https%3A%2F%2Fgeocinema.network%2Fuploads%2Fbanner2_3a749bd8f2.png&w=3840&q=50 - Project page: https://geocinema.network/ #### Deep Down Tidal — Tabita Rezaire (2017) - Type: Film & video · Machines: Infrastructure, energy & supply chain - Idea: The internet's physical routes repeat colonial ones; connectivity carries old power. - What it is: A video essay that follows submarine internet cables along the routes of the transatlantic slave trade and colonial shipping, weaving in water cosmologies. - How it works: Desktop-style collage of maps, archival images, animations and the artist's narration (about 18 minutes). - Lens: Hidden labour & extraction, Machine umwelt - Video: https://www.youtube.com/watch?v=V9EIywuH-LM - Images: https://goodman-gallery.com/api/v1/og-image?src=https%3A%2F%2Fd32637ifygiesu.cloudfront.net%2Fartists%2Ftabita-rezaire%2Fdeep-down-tidal-118702%2Ftabita-rezaire-deep-down-tidal-118702-39813-front-1.jpg - Project page: https://goodman-gallery.com/artists/tabita-rezaire/deep-down-tidal-118702 #### Facebook Algorithmic Factory — Vladan Joler (2016) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Dataset & training data, Infrastructure, energy & supply chain - Idea: Users are unpaid workers in a factory whose machines are algorithms. - What it is: A series of forensic maps from SHARE Lab that trace how Facebook collects users' activity as raw material and processes it, through algorithms, into advertising targets. - How it works: Joler and Andrej Petrovski mapped patents, permissions and tracking from public documents into large diagrams, published in three parts. - Lens: Hidden labour & extraction, Opacity & latent space - Images: https://labs.rs/wp-content/uploads/2016/08/FacebookFactory-small.gif https://labs.rs/wp-content/uploads/2016/08/FB-Research-Final-01.png - Project page: https://labs.rs/en/facebook-algorithmic-factory-immaterial-labour-and-data-harvesting/ #### Metamorphism — Julian Charrière (2016) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain - Idea: Computers return to rock: the 'immaterial' digital is a geological body borrowed from the earth. - What it is: Motherboards, hard drives, CPUs and memory from laptops and phones melted in artificial lava into polychrome stones, shown in vitrines like specimens from a future natural history museum. - How it works: E-waste components mixed with rock and fused in a furnace at lava temperatures, then cut and displayed. - Lens: Hidden labour & extraction, Machine umwelt - Shown at: Into the Hollow, Dittrich & Schlechtriem, Berlin 2016; The World Through AI, Jeu de Paume, Paris 2025 - Images: https://julian-charriere.net/media/pages/projects/metamorphism/f1cc18f486-1671025283/julian-charriere-metamorphism-xvii-into-the-hollow-dittrich-schlechtriem-2016-copyright-the-artist-vg-bild-kunst-bonn-germany-1600x.jpg https://julian-charriere.net/media/pages/projects/metamorphism/86b5e1f22d-1671025283/julian-charriere-metamorphism-xiv-installation-view-into-the-hollow-dittrich-schlechtriem-2016-copyright-the-artist-vg-bild-kuns-1600x.jpg - Project page: https://julian-charriere.net/projects/metamorphism #### Rare Earthenware — Unknown Fields Division (2015) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain - Idea: The glow of a screen begins as radioactive waste on the other side of the world. - What it is: Three ceramic vases, each shaped in the proportions of rare-earth elements used in a smartphone, a laptop and a car battery, made from the toxic mud of a tailings lake in Baotou, Inner Mongolia. - How it works: The studio traced the supply chain from a Chinese port back to Baotou's refineries, collected the waste mud and fired it into Ming-style vases; a film by Toby Smith documents the journey. - Lens: Hidden labour & extraction - Shown at: V&A, London - Video: https://www.youtube.com/watch?v=YMKJ7S7fKOk - Project page: http://www.unknownfieldsdivision.com/ #### CO2GLE — Joana Moll (2014) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain - Idea: Every search has a physical cost; the cloud burns fuel. - What it is: A web page that shows, in real time, the kilograms of CO2 emitted by the global traffic to google.com since you opened it. - How it works: An estimate from Google's traffic per second, average page weight and the energy intensity of data transfer, updated live. - Lens: Hidden labour & extraction, Opacity & latent space - Images: https://www.janavirgin.com/img/works/CO2GLE.jpg - Project page: https://www.janavirgin.com/CO2/ #### Internet Machine — Timo Arnall (2014) - Type: Film & video · Machines: Infrastructure, energy & supply chain - Idea: The cloud is a building full of loud machines, diesel generators and guards. - What it is: A multi-screen film shot inside one of the largest data centres in Europe, Telefónica's in Alcalá de Henares, showing the cooling, power and security machinery of 'the cloud'. - How it works: Slow tracking shots and wide-format photography projected across three screens; shown at transmediale. - Lens: Machine umwelt, Opacity & latent space - Shown at: transmediale - Video: https://vimeo.com/95044197 - Images: https://www.elasticspace.com/images/internet-machine/Internet-Machine-Transmediale-2-small-1024x683.jpg https://www.elasticspace.com/images/internet-machine/InternetMachine14-web.jpg - Project page: https://www.elasticspace.com/2014/05/internet-machine/ #### Waste Tide — Chen Qiufan (2013) - Type: Book & essay · Machines: Infrastructure, energy & supply chain, Prosthesis & cyborg - Idea: The afterlife of our devices is someone else's body and land. - What it is: A novel set on 'Silicon Isle', a southern Chinese island where migrant 'waste people' recycle the world's e-waste; a young worker, Mimi, is infected by a virus and becomes a posthuman cyborg. - How it works: Science-fiction novel drawing on the real e-waste town of Guiyu; English translation by Ken Liu (2019). - Lens: Hidden labour & extraction, Embodiment & kinesthesia - Paper: https://en.wikipedia.org/wiki/Waste_Tide (Changjiang Literature and Art Publishing House, 2013; Tor Books (English), 2019) - Images: https://upload.wikimedia.org/wikipedia/en/c/c0/The_Waste_Tide.jpg - Project page: https://en.wikipedia.org/wiki/Waste_Tide ### Military & Operational AI Drones, targeting and machines built to see and act in war. #### AI War Cloud Database — Sarah Ciston (2025) - Type: Artwork & installation · Machines: Dataset & training data, Platform & recommender algorithm, Autonomous vehicle & drone - Idea: The AI in your pocket and the AI that picks targets share companies, models and data; the boundary between home and battlefield is thin. - What it is: An interactive database and installation that catalogues the AI systems used to make decisions in warfare and shows where the same kinds of tools appear in smartphones and social platforms. - How it works: A network graph of companies, products and military uses compiled from public sources, with a tangible installation version; Grand Prize of the S+T+ARTS Prize 2025. - Lens: Classification, bias & surveillance, Opacity & latent space - Shown at: S+T+ARTS Prize 2025 Grand Prize; Ars Electronica 2025 - Video: https://www.youtube.com/watch?v=BM1K0Rbs4W0 - Images: https://ars.electronica.art/starts-prize/files/2025/05/aiwarcloud-6-c-sarahciston-1024x576.jpg - Project page: https://aiwar.cloud #### The Battle of Ilovaisk — Forensic Architecture (2019) - Type: Artwork & installation · Machines: Computer vision, Dataset & training data - Idea: Counter-forensics uses machine vision to see what states deny. - What it is: An investigation into Russian military presence in eastern Ukraine in 2014, in which a machine-learning classifier trained to recognise T-72 tanks searched thousands of online videos for evidence. - How it works: Classifier trained on synthetic 3D renderings of the tank model, applied to scraped YouTube footage; results verified by researchers. - Lens: Machine seeing, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=OldnzVNE71E - Project page: https://forensic-architecture.org/investigation/the-battle-of-ilovaisk #### Incoming — Richard Mosse (2017) - Type: Film & video · Machines: Computer vision, Surveillance & biometrics - Idea: A weapons camera sees people as heat signatures: bodies without faces, stripped of individuality. - What it is: A three-screen film of refugees crossing the Mediterranean and living in camps, shot with a military-grade thermal camera that can detect a body from 30 km away and renders people as glowing heat. - How it works: Filmed over two years with a long-range thermal camera classified as a weapon under export law; score by Ben Frost. - Lens: Machine seeing, Classification, bias & surveillance - Shown at: The Curve, Barbican 2017; NGV Triennial 2017 - Video: https://www.youtube.com/watch?v=6QOyFAqs_rM - Project page: https://www.richardmosse.com/ #### Slaughterbots — Future of Life Institute / Stop Autonomous Weapons (2017) - Type: Film & video · Machines: Autonomous vehicle & drone, Surveillance & biometrics, Computer vision - Idea: Autonomous weapons need no science-fiction leap, only existing parts combined; the film argues for a ban. - What it is: A short fiction film in which palm-sized drones with face recognition and explosives, launched in swarms, assassinate students and politicians chosen from their social-media profiles. - How it works: Directed by Stewart Sugg with a closing statement by AI researcher Stuart Russell; first screened at the UN Convention on Certain Conventional Weapons in Geneva. - Lens: Agency, autonomy & co-creation, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=9CO6M2HsoIA - Project page: https://futureoflife.org/ #### A Study into 21st Century Drone Acoustics — Ruben Pater, Gonçalo F. Cardoso (2016) - Type: Artwork & installation · Machines: Autonomous vehicle & drone - Idea: For people living under drones, the machine is first a sound; listening to it is a form of knowledge and of fear. - What it is: An audio-visual composition and record built from field recordings of 17 kinds of drones, from consumer quadcopters to military reapers, tracing the sound life and death of a flying drone and the fear its hum causes where drones strike. - How it works: Field recordings of drone motors and propellers are edited into a composition with a visual score identifying each drone type. - Lens: Machine listening & voice, Alterity & the uncanny - Shown at: Prix Ars Electronica 2017, Award of Distinction (Hybrid Art); ZKM Global Control and Censorship 2015–16 - Video: https://www.youtube.com/watch?v=_69goMEmumk - Images: https://archive.aec.at/asset/2230549/preview/ https://archive.aec.at/asset/2230551/preview/ https://archive.aec.at/asset/2230552/preview/ - Project page: https://archive.aec.at/prix/228089/ #### National Bird — Sonia Kennebeck (2016) - Type: Film & video · Machines: Autonomous vehicle & drone, Computer vision - Idea: Remote killing is labour done at screens, and it wounds the people who do it too. - What it is: A documentary that follows three former members of the US drone programme, an image analyst and two others, who break their silence about their work and the civilian deaths it caused. - How it works: Observational documentary with interviews, including a visit to survivors of a strike in Afghanistan; executive produced by Wim Wenders and Errol Morris. - Lens: Hidden labour & extraction, Machine seeing - Shown at: Berlinale 2016; Tribeca Film Festival 2016 - Video: https://vimeo.com/186277306 - Images: https://nationalbirdfilm.com/wp-content/uploads/2016/04/poster_shot2.jpg - Project page: https://nationalbirdfilm.com/ #### #NotABugSplat — #NotABugSplat collective (2014) - Type: Artwork & installation · Machines: Autonomous vehicle & drone, Computer vision - Idea: 'Bug splat' was operators' slang for a human killed by a strike; the portrait speaks back through the drone's own eye. - What it is: A giant portrait of a child who lost family members in a drone strike, laid out in a field in Khyber Pakhtunkhwa, Pakistan, large enough to be seen by drone cameras above. - How it works: A vinyl print laid on the ground with local residents' help, inspired by the artist JR's Inside Out project, and documented by aerial photography. - Lens: Machine seeing, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=ts5qn6UV5wE - Images: https://notabugsplat.com/wp-content/uploads/2014/04/jr_kpk_full.jpg - Project page: https://notabugsplat.com/ #### Blue Sky Days — Tomas van Houtryve (2014) - Type: Artwork & installation · Machines: Autonomous vehicle & drone, Computer vision - Idea: Seeing your own daily life from the drone's angle makes remote war's gaze recognisable. - What it is: Aerial photographs of weddings, funerals, schools and prayer gatherings in the United States, taken with a small camera drone, echoing the kinds of gatherings struck by US drones abroad. - How it works: Photographed with a consumer quadcopter over US towns; the title refers to the Pakistani boy Zubair Rehman, who told the US Congress in 2013 that he now preferred grey skies, because drones do not fly then. - Lens: Machine seeing, Classification, bias & surveillance - Shown at: ICP Infinity Award 2015 (Photojournalism) - Video: https://vimeo.com/128202299 - Images: https://i2.wp.com/tomasvh.com/wp-content/uploads/2013/03/works_prev_blue_sky_days.jpg?fit=1500%2C1500&ssl=1 - Project page: https://tomasvh.com/works/blue-sky-days/ #### Drone Strikes: Investigating Covert Operations — Forensic Architecture (2013) - Type: Film & video · Machines: Autonomous vehicle & drone, Computer vision - Idea: Drone strikes leave holes smaller than satellite pixels; architecture can make the invisible strike evidential. - What it is: Investigations into US drone strikes in Pakistan, Yemen, Gaza and Afghanistan for a UN special rapporteur, including a strike on a house in Miranshah reconstructed from a single survivor's video. - How it works: Spatial analysis of video frames, satellite images and witness testimony, modelled in 3D to locate strikes and their victims. - Lens: Machine seeing, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=CbzVrH6Utjw - Project page: https://forensic-architecture.org/investigation/drone-strikes-investigating-covert-operations-through-spatial-and-media-analysis #### Drone Survival Guide — Ruben Pater (2013) - Type: Speculative design · Machines: Autonomous vehicle & drone - Idea: In a world of drones, people need a field guide to the machines that watch them. - What it is: A poster in the style of a bird-spotting guide that shows the silhouettes of military and civilian drones, with tips on hiding from them, translated into languages including Pashto and Urdu. - How it works: Printed on a reflective sheet meant to help hide from infrared cameras, and distributed free online and in print. - Lens: Machine seeing, Classification, bias & surveillance - Images: http://www.dronesurvivalguide.org/dsg.jpg http://www.dronesurvivalguide.org/photo3.jpg - Project page: http://www.dronesurvivalguide.org/ #### Drone Shadows — James Bridle (2012) - Type: Artwork & installation · Machines: Autonomous vehicle & drone - Idea: The scale of a remote weapon becomes real when its shadow crosses familiar ground. - What it is: Full-size outlines of military drones such as the MQ-1 Predator, drawn in white line on streets and squares as if their shadows had fallen in the city. - How it works: Chalk or road-marking paint outlines at 1:1 scale, drawn in Istanbul, London, Brighton, Washington D.C. and other cities. - Lens: Machine seeing, Embodiment & kinesthesia - Shown at: Design and Violence, Science Gallery Dublin - Video: https://www.youtube.com/watch?v=CT5iqTisnTc - Project page: https://jamesbridle.com/ #### Dronestagram — James Bridle (2012) - Type: Artwork & installation · Machines: Autonomous vehicle & drone, Computer vision - Idea: Remote war happens in places we never see; the aerial view shared in a feed makes it briefly visible. - What it is: For every US drone strike reported in Pakistan, Yemen and Somalia, an Instagram and Twitter account posted a satellite image of the location. - How it works: Strike reports from the Bureau of Investigative Journalism were matched to Google Maps satellite views and posted with a short caption. - Lens: Machine seeing, Classification, bias & surveillance - Images: https://jamesbridle.com/media/pages/works/dronestagram/7174ad984b-1680515147/100_dronestagram.jpg - Project page: https://jamesbridle.com/works/dronestagram #### Metadata+ — Josh Begley (2012) - Type: Research prototype · Machines: Autonomous vehicle & drone, Platform & recommender algorithm - Idea: A remote war can be brought to the phone's lock screen, next to every other notification. - What it is: An iPhone app that sends a push notification each time a US drone strike is reported; Apple rejected it several times before approving it in 2014. - How it works: Uses the Bureau of Investigative Journalism's strike database; first submitted as 'Drones+' in 2012 and renamed Metadata+. - Lens: Classification, bias & surveillance, Opacity & latent space - Images: https://joshbegley.com/metadata/metadata-2017-map.jpg - Project page: http://joshbegley.com/metadata/ #### Unmanned — Molleindustria (2012) - Type: Artwork & installation · Machines: Autonomous vehicle & drone, Surveillance & biometrics - Idea: Remote war is office work, and its violence sits beside ordinary domestic life. - What it is: A short browser game about one day in the life of a drone pilot: shaving, commuting, watching a target on a screen, playing video games with his son. - How it works: Split-screen Flash game written by Paolo Pedercini with Jim Munroe. - Lens: Hidden labour & extraction, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=qTsP0SpWdCk - Project page: http://unmanned.molleindustria.org/ #### Serious Games I–IV — Harun Farocki (2009) - Type: Film & video · Machines: Computer vision, Generative model & latent space - Idea: The same simulated image prepares the soldier for war and treats the trauma it causes; in the game, war has no shadows. - What it is: A four-part video installation on how the US military uses computer simulations to train soldiers for Iraq and Afghanistan and, afterwards, to treat their post-traumatic stress. - How it works: Filmed at US Marine Corps training sites and at the Institute for Creative Technologies (USC), comparing training and therapy software; parts made 2009–10. - Lens: Machine seeing, Embodiment & kinesthesia - Images: https://www.harunfarocki.de/fileadmin/_processed_/c/1/csm_harun-farocki-imersion_0275cdcf89.jpg - Project page: https://www.harunfarocki.de/installations/2000s/2009/serious-games-iii-immersion.html ## Classification, Power & Justice Bias, surveillance and prediction, and the counter-archives and other intelligences that resist them. ### Bias & the Coded Gaze Who machines fail to see, or see wrongly, and why. #### CLASSES — Libby Heaney (2021) - Type: Film & video · Machines: Computer vision, Dataset & training data, Generative model & latent space - Idea: Machine-learning 'classes' and social class are entangled: models trained on the web see the estate through other people's eyes. - What it is: A video essay set in a simulated model of the London council estate where the artist lives, where machine and human voices narrate how machine-learning classification distorts working-class life. - How it works: Object detection, image captioning and language models are run over a 3D scan of the estate; the outputs are voiced and overlaid in a 14-minute single-channel video. - Lens: Classification, bias & surveillance, Failure, glitch & error - Images: https://libbyheaney.co.uk/wp-content/uploads/2022/11/1-1.jpg https://libbyheaney.co.uk/wp-content/uploads/2022/11/2.jpg - Project page: https://libbyheaney.co.uk/artworks/classes/ #### Coded Bias — Shalini Kantayya (2020) - Type: Film & video · Machines: Computer vision, Surveillance & biometrics, Platform & recommender algorithm - Idea: Algorithmic harm is a civil-rights issue that needs law, not only better models. - What it is: A feature documentary that follows Joy Buolamwini from her discovery of face-recognition bias to testimony before the US Congress, alongside London police face-recognition trials and China's social scoring. - How it works: Observational documentary with interviews of researchers including Cathy O'Neil, Meredith Broussard, Safiya Noble, Timnit Gebru and Silkie Carlo of Big Brother Watch. - Lens: Classification, bias & surveillance, Machine seeing - Shown at: Sundance Film Festival 2020 - Video: https://www.youtube.com/watch?v=7pMV9scmKwc - Images: https://images.squarespace-cdn.com/content/v1/5eb23eee707c5356dea97eaa/1590130745535-D27NXEHHTSDPDBTXVMMF/CODED_BIAS_Home_Resize_2.jpg - Project page: https://www.codedbias.com/ #### Voicing Erasure — Joy Buolamwini (2020) - Type: Film & video · Machines: Voice & listening machine, Chatbot & language model - Idea: Speech recognition hears some voices better than others; mishearing is a way of erasing. - What it is: A spoken-word piece in which women scholars of AI bias read their work aloud while speech-recognition systems transcribe them badly, and voice assistants answer questions about women's history with men. - How it works: Made with the Algorithmic Justice League after a Stanford study found error rates for Black speakers nearly double those for white speakers in five commercial speech-to-text systems. - Lens: Machine listening & voice, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=SdCPbyDJtK0 - Project page: https://www.ajl.org/voicing-erasure #### AI, Ain't I A Woman? — Joy Buolamwini (2018) - Type: Film & video · Machines: Computer vision, Surveillance & biometrics - Idea: A machine's error can repeat an old insult; poetry can make an audit felt. - What it is: A spoken-word video in which Buolamwini shows commercial AI systems mislabelling portraits of Sojourner Truth, Ida B. Wells, Michelle Obama, Oprah Winfrey and Serena Williams as male. - How it works: Buolamwini ran historical and contemporary portraits of Black women through face-analysis services from major tech companies and set the returned labels to her reading of a poem named after Sojourner Truth's 1851 speech. - Lens: Classification, bias & surveillance, Failure, glitch & error, Machine seeing - Video: https://www.youtube.com/watch?v=QxuyfWoVV98 - Project page: https://www.ajl.org/ #### Us, Aggregated — Mimi Ọnụọha (2017) - Type: Artwork & installation · Machines: Computer vision, Platform & recommender algorithm, Dataset & training data - Idea: Grouping people from above is a quiet exercise of power: the machine builds an 'us' that is really a 'them'. - What it is: Photos from the artist's family archive, never before online, are shown next to the images Google's reverse image search judges 'similar' to them. - How it works: The artist programmatically queried Google's reverse image search with each family photo and presented the results as a grid; later versions (3.0, 2019) are a looping video with a hovering cursor. - Lens: Classification, bias & surveillance, Machine seeing - Images: https://images.squarespace-cdn.com/content/v1/692f1808f9a01f48c2fa30ac/21a78b11-40de-4823-8c9c-802c59fd4c25/00_UA3_installationshot-cropped_courtesyofQuad.jpg https://images.squarespace-cdn.com/content/v1/692f1808f9a01f48c2fa30ac/38a2109b-ba77-40cf-be21-c17c2ddb4c90/02_UA3_videostill_courtesyofartist.jpg - Project page: https://www.mimionuoha.com/ #### The Coded Gaze: Unmasking Algorithmic Bias — Joy Buolamwini (2016) - Type: Film & video · Machines: Computer vision, Surveillance & biometrics - Idea: Being invisible to a machine is not neutral; it follows who was in the training data. - What it is: A short video in which Buolamwini, working on an MIT project, finds that a face-tracking system only detects her face when she puts on a white mask. - How it works: Screen recordings of an off-the-shelf face-detection library, compared on her own face, a lighter-skinned colleague's face and a white mask. - Lens: Classification, bias & surveillance, Machine seeing, Failure, glitch & error - Video: https://www.youtube.com/watch?v=162VzSzzoPs - Project page: https://www.ajl.org/ #### [help me know the truth] — Mary Flanagan (2016) - Type: Artwork & installation · Machines: Computer vision, Surveillance & biometrics - Idea: A classifier does not need to be a machine: the system simply collects and sharpens the prejudices people already carry. - What it is: Visitors take a selfie, which is then overlaid with noise and shown on stations around the gallery; other visitors repeatedly pick which of two blurry versions better fits a label such as 'angel' or 'criminal', and a stereotyped face emerges from their choices. - How it works: Uses reverse correlation from cognitive neuroscience: random noise patterns on the portrait are averaged according to visitors' binary choices. - Lens: Classification, bias & surveillance, Machine seeing - Shown at: Prix Ars Electronica 2018, Award of Distinction (Interactive Art+) - Video: https://www.youtube.com/watch?v=DiHGOqLwN9w - Images: https://archive.aec.at/asset/2527247/preview/ https://archive.aec.at/asset/2527249/preview/ https://archive.aec.at/asset/2527254/preview/ - Project page: http://maryflanagan.com/work/help-me-know-the-truth/ #### Physiognomic Scrutinizer — Marnix de Nijs (2009) - Type: Artwork & installation · Machines: Surveillance & biometrics - Idea: Face matching quietly revives physiognomy: judging character by likeness. - What it is: Visitors pass through a gate that scans their face and matches it with one of 150 infamous or controversial people; as they walk through, a robotic voice tells the story of their 'match'. - How it works: Airport-style gate with camera and face-recognition software comparing visitors to a curated gallery; speech synthesis narrates the match. - Lens: Classification, bias & surveillance, Embodiment & kinesthesia - Video: https://vimeo.com/21460350 - Images: http://www.marnixdenijs.nl/photos_640x/physiognomic-scrut_2018-09-04_1.png - Project page: http://www.marnixdenijs.nl/physiognomic_scrutinizer.htm ### Surveillance & Prediction Tracking, scoring and predictive systems that govern people. #### Consensus Gentium — Karen Palmer (2023) - Type: Film & video · Machines: Surveillance & biometrics, Computer vision - Idea: The story reacts to where you look, so the viewer experiences being profiled by the device in their own hand. - What it is: An interactive film played on a smartphone: to get permission to visit a sick grandmother in a near-future surveillance state, the viewer passes tests while the phone's camera tracks their eye gaze and scores their compliance. - How it works: Face detection and gaze tracking through the front-facing camera feed a branching narrative that classifies the viewer as dissident or compliant. - Lens: Classification, bias & surveillance, Machine seeing - Shown at: SXSW 2023; Prix Ars Electronica 2024, Honorary Mention (Interactive Art +) - Images: https://archive.aec.at/asset/3800250/preview/ https://archive.aec.at/asset/3800253/preview/ - Project page: https://archive.aec.at/prix/290652/ #### All Light, Everywhere — Theo Anthony (2021) - Type: Film & video · Machines: Surveillance & biometrics, Computer vision - Idea: No camera is neutral: every lens has a position, and body cameras are designed from the officer's side. - What it is: A documentary essay that moves from the history of photography and time-and-motion studies to Axon, the maker of police body cameras, and a surveillance plane flying over Baltimore. - How it works: Observational scenes inside Axon's headquarters and the Baltimore Aerial Investigation Research programme, intercut with archival science films; won a Special Jury Award at Sundance. - Lens: Machine seeing, Classification, bias & surveillance - Shown at: Sundance Film Festival 2021 (Special Jury Award) - Video: https://www.youtube.com/watch?v=luHmSqY0Xsk - Project page: https://en.wikipedia.org/wiki/All_Light,_Everywhere #### 3x3x6 — Shu Lea Cheang (2019) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Surveillance technology is turned back on its viewers to expose how bodies are classified as deviant. - What it is: In a former prison in Venice, a panoptic installation tells the cases of people imprisoned for sexual or gender dissidence. Surveillance cameras scan visitors and mix their images into the projections, placing them inside the system. - How it works: Multi-screen projection, surveillance cameras with visitor image capture and processing, website and research archive; curated by Paul B. Preciado for the Taiwan Pavilion. - Lens: Classification, bias & surveillance, Machine seeing - Shown at: Taiwan Pavilion, 58th Venice Biennale 2019 (Palazzo delle Prigioni); LG Guggenheim Award 2024 - Video: https://www.youtube.com/watch?v=MUvqYPKekls - Images: https://images.e-flux.com/test/bc8c4a07-4790-4e84-a019-7141af66a1c2?width=1200&height=630 - Project page: https://www.e-flux.com/announcements/258006/shu-lea-cheang3x3x6 #### My Blue Window — American Artist (2019) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Computer vision, Dataset & training data - Idea: Predictive policing gives the officer a machine's view of the city; the viewer sits in the seat that sees neighbourhoods as risk scores. - What it is: An installation built like a police cruiser's interior, where visitors watch the film 2015 on the dashboard: a dashcam drive through New York while predictive-policing software flags 'high-risk' zones. - How it works: Dashcam-style footage overlaid with interface graphics modelled on predictive policing tools such as HunchLab; shown with the phone app 1956/2054 at the Queens Museum. - Lens: Classification, bias & surveillance, Machine seeing - Shown at: Queens Museum 2019 - Images: https://images.squarespace-cdn.com/content/v1/59238d36d2b8575d127794a4/1572016941388-EJM1NHYJ77SXY3WF9G2E/QM-MyBlueWindow-_71A5256-LowRes.jpg - Project page: https://americanartist.us/works/my-blue-window #### Shadow Stalker — Lynn Hershman Leeson (2019) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Surveillance & biometrics, Dataset & training data - Idea: Everyone casts a data shadow that others can read and act on, often wrongly. - What it is: Visitors type their email address and see their 'data shadow' projected: addresses, relatives and traces found online; a film on predictive policing plays with a 'Spirit of the Deep Web' played by Tessa Thompson. - How it works: An installation using open-source intelligence searches on visitors' emails, projected on the floor, with a film on algorithms such as PredPol; commissioned by The Shed, New York. - Lens: Classification, bias & surveillance, Alterity & the uncanny - Shown at: The Shed, New York 2019 - Video: https://www.youtube.com/watch?v=FS5FJmtFy64 - Images: https://www.lynnhershman.com/wp-content/uploads/2019/10/ShadowStalker_Image07.jpg - Project page: https://www.lynnhershman.com/project/shadow-stalker/ #### Spectre — Bill Posters, Daniel Howe (2019) - Type: Artwork & installation · Machines: Generative model & latent space, Platform & recommender algorithm, Dataset & training data - Idea: Deepfakes of the people who run the platforms make the data economy confess in their voices. - What it is: An installation on the 'digital influence industry' that included deepfake videos of Mark Zuckerberg, Kim Kardashian and Donald Trump praising a fictional data cult; the Zuckerberg video tested Instagram's deepfake policy. - How it works: Deepfakes made with the company CannyAI, an installation at Sheffield Doc/Fest, and an interactive 'altar' where visitors see how their data is sold. - Lens: Classification, bias & surveillance, Alterity & the uncanny - Shown at: Sheffield Doc/Fest 2019 - Video: https://vimeo.com/341794473 - Images: https://billposters.ch/wp-content/uploads/2020/01/Spectre_press_web_new.jpg - Project page: https://billposters.ch/projects/spectre/ #### The Net Wanderer – A Tour of Suspended Handshakes — Guo Cheng (2019) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Surveillance & biometrics - Idea: The border of a national internet is a set of ordinary buildings and delayed handshakes, hidden in plain sight. - What it is: A research project and installation that traces the Great Firewall's gateway nodes to physical sites; visitors who try to open a blocked website trigger a browser game instead. - How it works: Network diagnostic tools traced IP addresses of firewall-related nodes; the artist then visited and photographed the locations in several cities. - Lens: Hidden labour & extraction, Opacity & latent space - Shown at: Prix Ars Electronica 2020, Honorary Mention - Images: https://cdn.guo-cheng.net/files/projectImg/thenetwanderer/theNetWanderer_2.jpg https://cdn.guo-cheng.net/files/projectImg/thenetwanderer/theNetWanderer_1.jpg - Project page: https://guo-cheng.net/index.php/projects/the-net-wanderer/ #### Sociality — Paolo Cirio (2018) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Surveillance & biometrics - Idea: The patent office is where the machinery of social manipulation is described in plain diagrams before it is built. - What it is: Cirio collected and rated hundreds of US patents for technologies that monitor and manipulate social behaviour, and invited the public to share, flag and 'ban' them on a website and in printed installations. - How it works: A database of patent filings with their flowcharts, classified by the artist and published online with public flagging tools, plus prints and a colouring book of the diagrams. - Lens: Classification, bias & surveillance, Opacity & latent space - Shown at: S+T+ARTS Prize 2020, Honorary Mention - Images: https://paolocirio.net/work/sociality/Sociality-pics-PAN-Naples/DSC_1176.jpg https://paolocirio.net/work/sociality/img/sociality-website.png - Project page: https://paolocirio.net/work/sociality/ #### The City of Broken Windows — Hito Steyerl (2018) - Type: Artwork & installation · Machines: Voice & listening machine, Dataset & training data - Idea: Machine listening for 'broken windows' carries the old policing theory into code; repair is the counter-gesture. - What it is: A two-part installation: engineers smash windows in a hangar to record training data for an AI that detects breaking glass, while the activist Chris Toepfer paints boarded-up windows in Camden, New Jersey, to look lived-in. - How it works: Two video channels, 'Unbroken Windows' and 'Broken Windows', filmed at a UK company training audio classifiers for security systems and with Chris Toepfer's window-painting project. - Lens: Machine listening & voice, Classification, bias & surveillance - Shown at: Castello di Rivoli 2018 - Video: https://www.youtube.com/watch?v=iyyM4vDg0xw - Project page: https://www.castellodirivoli.org/en/mostra/hito-steyerl-the-city-of-broken-windows/ #### White Collar Crime Risk Zones — Sam Lavigne (2017) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Dataset & training data, Computer vision - Idea: Predictive policing is a choice of targets; point it at Wall Street and its logic becomes visible. - What it is: A predictive-policing map and app that forecasts financial crime instead of street crime; it also shows the 'likely suspect', a face averaged from thousands of LinkedIn photos of executives. - How it works: Made with Brian Clifton and Francis Tseng for The New Inquiry: a model trained on 1964–2016 FINRA financial-misconduct data predicts risk by location, and a face-averaging algorithm builds the suspect. - Lens: Classification, bias & surveillance, Opacity & latent space - Video: https://www.youtube.com/watch?v=Wj8meXnHSPo - Images: https://whitecollar.thenewinquiry.com/assets/share.jpg?v=2 https://lav.io/projects/white-collar-crime-risk-zones/ars-electronica_huf442887b5b01f9ea96a9441d7fc1f197_3973697_1800x0_resize_q85_box.jpg - Project page: https://whitecollar.thenewinquiry.com/ #### Cloud Index — James Bridle (2016) - Type: Artwork & installation · Machines: Generative model & latent space, Dataset & training data - Idea: Predictive models blur forecasting and control: to predict the vote is to imagine steering it. - What it is: A neural network trained on satellite weather images and UK polling data generates the skies that would have produced any given result in the Brexit referendum, offering weather as a lever on politics. - How it works: Correlation of historical cloud imagery with polling data; a DCGAN renders the sky images. - Lens: Opacity & latent space, Classification, bias & surveillance - Images: https://jamesbridle.com/media/pages/works/cloud-index/890e90e757-1680515138/165_opengraph.jpg - Project page: https://jamesbridle.com/works/cloud-index #### The Argus Project — Romi Ron Morrison, Gan Golan (2016) - Type: Performance, dance & music · Machines: Surveillance & biometrics, Prosthesis & cyborg - Idea: Counter-surveillance can be imagined as a collective body armed with eyes. - What it is: A wearable exo-suit covered with cameras, after the hundred-eyed giant Argus, worn on night walks through the city as a 'force field of accountability' for filming the police. - How it works: Sculptural suit with embedded cameras and lights; performances documented in public space. Collaborators include Raquel De Anda, Ligaiya Romero, Ayodamola Okunseinde and others. - Lens: Classification, bias & surveillance, Embodiment & kinesthesia - Video: https://vimeo.com/185745057 - Images: https://images.squarespace-cdn.com/content/v1/5a15cc35ccc5c5ef7b89ee2d/1511470361428-WJLOQ36VG386HRT8A6QS/IMG_4178.jpg - Project page: https://elegantcollisions.com/argus #### The Glass Room — Tactical Tech (2016) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Surveillance & biometrics, Dataset & training data - Idea: Borrowing the look of the tech store lets a critique of data extraction reach people who would never enter a gallery. - What it is: A pop-up exhibition designed like a sleek tech store, where the 'products' are artworks and investigations about data tracking, and staff called Ingeniuses give privacy advice. - How it works: Curated by Tactical Tech with Mozilla; first in Berlin (2016), then New York, London and San Francisco, with a 'Data Detox Kit' and a travelling community edition. - Lens: Classification, bias & surveillance, Opacity & latent space - Shown at: Berlin 2016; New York 2016; London 2017; San Francisco 2019 - Video: https://www.youtube.com/watch?v=k47lhUmj5ho - Images: https://cdn.ttc.io/i/fit/1740/0/sm/0/plain/theglassroom.org/SF-Atmosphere/sat-26-oct-Daisy12.JPG - Project page: https://theglassroom.org/ #### The Virtual Watchers — Joana Moll (2016) - Type: Artwork & installation · Machines: Surveillance & biometrics, Platform & recommender algorithm - Idea: Border surveillance was crowdsourced to unpaid citizens; the watchers became part of the machine. - What it is: An installation built on the archive of the Texas Virtual Border Watch (from 2008), in which volunteers watched live border-camera feeds online from home and reported suspected crossings. - How it works: Made with Cédric Parizot; it restages the programme's watching interface together with archived comments from the volunteers. - Lens: Hidden labour & extraction, Machine seeing - Images: https://www.janavirgin.com/img/works/virtual_watchers.jpg - Project page: https://www.janavirgin.com/works.html #### Citizen Ex — James Bridle (2015) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Platform & recommender algorithm - Idea: Online, a machine decides which country you belong to, and with it which laws and surveillance apply. - What it is: A browser extension that records which countries the websites you visit are hosted in and computes your 'algorithmic citizenship', the nationality data systems infer from your traffic. - How it works: The extension geolocates the IP address of every page and resource loaded and shows a running pie chart of countries; commissioned by The Space and Abandon Normal Devices. - Lens: Classification, bias & surveillance, Opacity & latent space - Video: https://www.youtube.com/watch?v=4cRX2evheew - Images: https://jamesbridle.com/media/pages/works/citizen-ex/4a7502d256-1680515124/149_badge-banner.jpg - Project page: https://jamesbridle.com/works/citizen-ex #### Secret Power — Simon Denny (2015) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Surveillance & biometrics - Idea: Mass surveillance has an aesthetic, and its clip-art and wizards say something about how the agencies see the world. - What it is: New Zealand's pavilion at the Venice Biennale turned the leaked NSA and GCHQ slide graphics from the Snowden files into server racks, carpets and ceiling paintings in the Marciana Library and the airport. - How it works: Denny researched the designer behind the NSA slides and rendered their imagery as modified server racks and printed surfaces, mirroring the Renaissance cosmology of the library's ceiling. - Lens: Classification, bias & surveillance, Opacity & latent space - Shown at: Venice Biennale 2015 (New Zealand pavilion); Te Papa Tongarewa - Video: https://www.youtube.com/watch?v=JidXPz7G0i8 - Images: https://simondenny.net/wp-content/uploads/modded-serverrack/MA_I390331_TePapa_Modded-ServerRack-Display_preview.jpg - Project page: https://simondenny.net/secret-power/ #### Stealth Cell Tower — Julian Oliver (2015) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Surveillance & biometrics - Idea: The same trick police use to intercept phones (the 'IMSI catcher') fits in an office printer. - What it is: A mobile-phone base station disguised as an office printer, which lets nearby phones connect to it and then sends them text messages that appear to come from people in their contacts. - How it works: A GSM base station built from a software-defined radio and a Raspberry Pi hidden in an HP printer, which also prints the texts it sends. - Lens: Opacity & latent space, Classification, bias & surveillance - Images: https://julianoliver.com/projects/stealth-cell-tower/images/stealth-cell-tower.jpg https://julianoliver.com/projects/stealth-cell-tower/images/print.jpg - Project page: https://julianoliver.com/projects/stealth-cell-tower/ #### Jennifer Lyn Morone™ Inc — Jennifer Lyn Morone (2014) - Type: Artwork & installation · Machines: Platform & recommender algorithm, Dataset & training data, Surveillance & biometrics - Idea: If data is value, a person can claim to be a company; the absurdity mirrors what platforms already do. - What it is: The artist registered herself as a corporation whose products are her health data, location, skills and opinions, with a database that sells them to buyers directly. - How it works: A legal company registration, a self-tracking 'Database of Personal Operations' and a trade fair booth, made at the Royal College of Art. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Video: https://vimeo.com/98300179 - Project page: https://jenniferlynmorone.com/ #### Asymmetric Love — Addie Wagenknecht (2013) - Type: Artwork & installation · Machines: Surveillance & biometrics, Infrastructure, energy & supply chain - Idea: Surveillance has become domestic décor: we hang it in our homes and admire it. - What it is: A crystal chandelier assembled from CCTV cameras, network cables and hardware, hung like a luxury light fitting. - How it works: Security cameras and CAT5 cables arranged on a chandelier frame; later versions of the series include working cameras. - Lens: Machine seeing, Intimacy, care & companionship - Images: https://freight.cargo.site/w/800/i/O1964140200252626760958482768938/image-asset-7.jpeg - Project page: https://www.placesiveneverbeen.com/asymmetric-love #### Surveillance Chess — !Mediengruppe Bitnik (2012) - Type: Performance, dance & music · Machines: Surveillance & biometrics - Idea: Every surveillance camera has a human at the other end who can be addressed. - What it is: In London before the 2012 Olympics, the artists hijacked the signal of a CCTV camera in a public space and replaced it with a chess game, inviting the unseen control-room operator to play. - How it works: A portable transmitter overrode the camera's wireless video feed with the artists' chess interface. - Lens: Agency, autonomy & co-creation, Classification, bias & surveillance - Video: https://vimeo.com/46236909 - Project page: https://wwwwwwwwwwwwwwwwwwwwww.bitnik.org/s/ #### Transparency Grenade — Julian Oliver (2012) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Surveillance & biometrics - Idea: Transparency can be forced from below; the network in any room can be made to speak. - What it is: A cast-resin Soviet F1 hand grenade with a tiny computer, microphone and antenna inside: 'pulling the pin' captures network traffic and audio in the room and leaks them to a public online map. - How it works: A Linux single-board computer with a Wi-Fi antenna and microphone sniffs unencrypted traffic, extracts emails, hostnames and images, and streams them to a server that maps the leak. - Lens: Opacity & latent space, Classification, bias & surveillance - Images: https://julianoliver.com/projects/transparency-grenade/images/transparency-grenade2.jpg - Project page: https://julianoliver.com/projects/transparency-grenade/ #### Face to Facebook — Paolo Cirio, Alessandro Ludovico (2011) - Type: Artwork & installation · Machines: Surveillance & biometrics, Platform & recommender algorithm - Idea: Once faces are data, they can be re-sorted into any identity someone chooses. - What it is: The artists scraped one million Facebook profiles, used face-analysis software to sort 250,000 profile pictures into temperaments such as 'smug' or 'easy going', and published them on a fake dating site, Lovely-Faces.com. - How it works: Custom scraper plus face-detection and expression-analysis software; profiles re-published on a purpose-built dating website. - Lens: Classification, bias & surveillance - Shown at: transmediale 2011 - Images: https://www.face-to-facebook.net/img/header_k.jpg https://www.face-to-facebook.net/img/face-to-facebook-diagram_web.gif - Project page: https://www.face-to-facebook.net/how.php #### Memopol-2 — Timo Toots (2011) - Type: Artwork & installation · Machines: Surveillance & biometrics, Dataset & training data - Idea: In a digital state the ID card is the key to a profile you rarely see; the machine turns the background check on yourself. - What it is: A large machine into which visitors insert their national ID card or passport; it gathers what state databases and the internet hold about them and shows their data field on a big custom display, remembering each visitor's name and portrait. - How it works: Reads the chip of Estonian ID cards and queries national registries and web sources; results are visualised on an array of LED panels. - Lens: Classification, bias & surveillance - Shown at: Prix Ars Electronica 2012, Golden Nica (Interactive Art) - Video: https://www.youtube.com/watch?v=xCr7UdVkmCk - Images: https://archive.aec.at/asset/1439540/preview/ https://archive.aec.at/asset/1439541/preview/ https://archive.aec.at/asset/1439585/preview/ - Project page: https://archive.aec.at/prix/193063/ #### Men in Grey — Julian Oliver, Danja Vasiliev (2009) - Type: Performance, dance & music · Machines: Infrastructure, energy & supply chain, Surveillance & biometrics - Idea: Only when people witness their own traffic being read do they sense the machines listening in the network. - What it is: Two figures in grey suits with briefcases appear in public places; the briefcases intercept and manipulate the wireless traffic of people nearby and display what they find, staging a secret society that makes network surveillance tangible. - How it works: Briefcases contain Linux computers, antennas and screens that run wireless exploits and content manipulation in real time. - Lens: Classification, bias & surveillance, Opacity & latent space - Shown at: Prix Ars Electronica 2010, Award of Distinction (Hybrid Art); Aksioma, Ljubljana 2011 - Images: https://archive.aec.at/asset/464136/preview/ https://archive.aec.at/asset/464138/preview/ https://archive.aec.at/asset/464140/preview/ - Project page: https://criticalengineering.org/projects/men-in-grey/ #### Faceless — Manu Luksch (2007) - Type: Film & video · Machines: Surveillance & biometrics, Computer vision - Idea: The law's redaction of faces became the film's world: a city where only the requester has a face. - What it is: A science-fiction film made entirely from CCTV footage of the artist in London, obtained under the UK Data Protection Act; other people's faces arrive blanked out by law, so the city appears faceless. - How it works: Luksch followed a 'Manifesto for CCTV Filmmakers', performing in front of public cameras and then filing subject-access requests for the recordings; narrated by Tilda Swinton. - Lens: Machine seeing, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=yLzJCeGYgbg - Project page: https://manuluksch.com/ #### Demonstrate — Ken Goldberg (2004) - Type: Artwork & installation · Machines: Surveillance & biometrics, Computer vision - Idea: Giving the public the controls of a surveillance camera tests free speech against privacy. - What it is: For the 40th anniversary of the Free Speech Movement, a high-resolution robotic webcam overlooked Sproul Plaza at UC Berkeley; anyone online could steer it, take pictures of people and comment on them. - How it works: Networked pan-tilt-zoom camera with a web interface; user-captured frames and comments archived. - Lens: Classification, bias & surveillance, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=hd6vNv1m6NE - Project page: https://goldberg.berkeley.edu/ #### Evidence Locker — Jill Magid (2004) - Type: Film & video · Machines: Surveillance & biometrics, Computer vision - Idea: A surveillance system can be seduced into care; the intimacy exposes who holds the camera. - What it is: For 31 days in Liverpool the artist worked with Citywatch, the city's police CCTV system, directing operators by phone to follow her and writing them love letters in the form of footage-request forms. - How it works: Magid wore a red coat so operators could find her, requested the footage under the Data Protection Act (it was otherwise deleted after 31 days), and edited it into videos and a 'Reading Room' installation. - Lens: Machine seeing, Intimacy, care & companionship - Shown at: FACT Liverpool 2004; Tate Liverpool 2004 - Video: https://www.youtube.com/watch?v=laStNBHSfuU - Project page: https://www.jillmagid.com/projects/evidence-locker-2 #### Life: A User's Manual — Michelle Teran (2003) - Type: Performance, dance & music · Machines: Surveillance & biometrics - Idea: The city is already full of images of us that we never see. - What it is: Teran walks through cities pulling a suitcase with a video scanner and screen, picking up the unencrypted wireless signals of private surveillance cameras and showing passers-by the hidden images around them. - How it works: Handheld receiver scanning the frequency band (likely 2.4 GHz) used by consumer wireless cameras; images displayed in a mobile monitor. - Lens: Machine seeing, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=MS-T6Yooa9E - Project page: https://www.youtube.com/watch?v=MS-T6Yooa9E #### System Azure Security Ornamentation — Jill Magid (2002) - Type: Performance, dance & music · Machines: Surveillance & biometrics - Idea: Ornamenting a camera makes an invisible gaze visible and oddly personal. - What it is: Magid offered the Amsterdam police her services to decorate the city's surveillance cameras with rhinestones and ornaments, making them visible and, in her words, more beautiful. - How it works: Proposal letters and meetings with the police, followed by decorating CCTV cameras in public space. - Lens: Intimacy, care & companionship, Classification, bias & surveillance - Images: http://www.jillmagid.com/uploads/images/scaled/slide_image/302 http://www.jillmagid.com/uploads/images/scaled/slide_image/305 - Project page: http://www.jillmagid.com/projects/system-azure-security-ornamentation #### Tracking Transience — Hasan Elahi (2002) - Type: Artwork & installation · Machines: Surveillance & biometrics, Dataset & training data - Idea: Flooding the watchers with data makes a person both fully visible and hard to read. - What it is: After the FBI questioned him for months over a false tip, Elahi began publishing his location and tens of thousands of photos of his meals, toilets and flights online, for the agency and anyone else. - How it works: A GPS tracker and a website updated continuously with timestamped photographs since 2002; the archive now holds tens of thousands of images. - Lens: Classification, bias & surveillance, Opacity & latent space - Video: https://www.youtube.com/watch?v=wAdwurHhv-I - Project page: https://en.wikipedia.org/wiki/Hasan_Elahi #### BIT Plane — Bureau of Inverse Technology (BIT) (1999) - Type: Artwork & installation · Machines: Autonomous vehicle & drone, Computer vision - Idea: An early drone that watched the watchers. - What it is: A small radio-controlled plane carrying a video camera flew over Silicon Valley's no-camera zones, filming the tech companies that build the information age from above. - How it works: Custom radio-controlled model aircraft with onboard video camera and transmitter, piloted over corporate campuses. - Lens: Machine seeing, Classification, bias & surveillance - Images: https://bureauit.org/images/plane/plane0self.jpg https://bureauit.org/images/plane/plane031lockh.jpg - Project page: https://bureauit.org/plane/ #### Performances for Surveillance Cameras — Surveillance Camera Players (1996) - Type: Performance, dance & music · Machines: Surveillance & biometrics - Idea: If the camera is always watching, perform for it: the operator becomes an audience, and the watched become actors. - What it is: A New York troupe stages silent adaptations of Orwell's 1984 and other plays in front of public CCTV cameras, holding up placards for the operators, and leads walking tours mapping cameras. - How it works: Short silent plays with large text boards performed at camera locations, documented from the street; the group also published maps of cameras in Manhattan. - Lens: Machine seeing, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=9BtntSJc0a4 - Project page: http://www.notbored.org/the-scp.html ### Missing Data & Counter-Datasets Datasets made to fill gaps, refuse extraction or retell who is counted. #### Ground Truths — Mimi Ọnụọha (2025) - Type: Film & video · Machines: Dataset & training data, Computer vision - Idea: A model trained on records shaped by erasure inherits the erasure; 'ground truth' is literally buried. - What it is: A docufiction film in which the artist builds a machine-learning model to predict where other unmarked graves of convict-leasing victims lie across Texas, after 95 bodies were found in Sugar Land in 2018. - How it works: Single-channel 4K video (about 16 minutes) combining the model-building process with archives, phone calls and missing records. - Lens: Classification, bias & surveillance, Grief, memory & afterlife, Opacity & latent space - Images: https://images.squarespace-cdn.com/content/v1/692f1808f9a01f48c2fa30ac/51f2cb4b-30a1-49b4-9ece-1cde6d8effa4/00_GroundTruths_film-still_courtestyoftheartist.jpg https://images.squarespace-cdn.com/content/v1/692f1808f9a01f48c2fa30ac/8c09a145-022b-4e17-bab1-682266e49fb3/01_GroundTruths_film-still_courtestyoftheartist.jpg - Project page: https://www.mimionuoha.com/ #### Babylonian Vision — Nora Al-Badri (2020) - Type: Artwork & installation · Machines: Generative model & latent space, Dataset & training data - Idea: Scraping museum data can be a form of restitution: the images return to a common use their owners did not plan. - What it is: A neural network trained on images of Mesopotamian, Neo-Sumerian and Babylonian artefacts from museum collections produces synthetic artefacts, shown as a video. - How it works: A GAN trained on thousands of images scraped from museum websites, some without permission; first commissioned for the EPFL Pavilions. - Lens: Classification, bias & surveillance, Opacity & latent space - Video: https://www.youtube.com/watch?v=kTZ5W8Vbpd4 - Images: https://epfl-pavilions.ch/media/pages/exhibitions/babylonian-vision/647951dfaf-1709114720/cover-2880x1600.png - Project page: https://epfl-pavilions.ch/en/exhibitions/babylonian-vision #### Igùn — Minne Atairu (2020) - Type: Artwork & installation · Machines: Generative model & latent space, Dataset & training data - Idea: A model trained on looted objects can only recombine what was taken; its gaps mirror the gaps of the colonial archive. - What it is: A series of generative 'bronzes' imagining the works Benin casters on Igùn Street might have made in the years after the British looted the royal palace in 1897. - How it works: StyleGAN2 trained on images of Benin bronze, iron and terracotta objects held mostly in Western museums. - Lens: Opacity & latent space, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=pg8U2l4tQAE - Project page: https://igun.info/prototype-I #### The Zizi Show — Jake Elwes (2020) - Type: Performance, dance & music · Machines: Generative model & latent space, Dataset & training data - Idea: Training AI on drag bodies queers the norms of the datasets it usually learns from. - What it is: A virtual deepfake drag cabaret: thirteen London drag performers were filmed to train networks, and online visitors can make their generated bodies perform songs, including the composite queen Zizi. - How it works: Pose-conditioned video generation trained on footage of each performer; web interface lets users pick acts and songs. - Lens: Classification, bias & surveillance, Embodiment & kinesthesia - Images: https://www.jakeelwes.com/images/project-zizi-show.jpg - Project page: https://www.jakeelwes.com/project-zizi-show.html #### Te Reo Māori Speech Recognition and the Kaitiakitanga Licence — Te Hiku Media (2018) - Type: Research prototype · Machines: Voice & listening machine, Dataset & training data - Idea: Data sovereignty means the community that speaks the language decides who may train on its voice. - What it is: A Māori-led media organisation built its own speech recognition for te reo Māori from recordings donated by its community, and refused to hand the data to big tech. - How it works: In the 2018 Kōrero Māori campaign people read sentences aloud and recorded over 300 hours in 10 days; the models are released under the Kaitiakitanga Licence, which grants access only for the benefit of Māori. - Lens: Machine listening & voice, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=sGQy0r_icWc - Images: https://opengraph.githubassets.com/1a9ccbb40a42481347763da1c45a53663a09c5a3a855a4a0509aa57bb7b783ad/TeHikuMedia/Kaitiakitanga-License - Project page: https://tehiku.nz/te-hiku-tech/ - Code: https://github.com/TeHikuMedia/Kaitiakitanga-License #### Feminist Data Set — Caroline Sinders (2017) - Type: Research prototype · Machines: Dataset & training data, Chatbot & language model - Idea: Bias is not fixed at the end of a pipeline; each step, from data to wages, is a political choice. - What it is: A multi-year project that builds a feminist AI chatbot by questioning every step: collecting feminist texts in public workshops, labelling them, choosing a model and paying the labour fairly. - How it works: Community workshops (SPACE Art + Technology, Victoria and Albert Museum and others) gather and annotate texts; Sinders documents a critical-design framework for the whole ML pipeline. - Lens: Classification, bias & surveillance, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=WwdmPva8nhc - Images: https://carolinesinders.com/wp-content/uploads/2020/05/LEO_6267-1024x678.jpg https://carolinesinders.com/wp-content/uploads/2020/05/VA_Image2_large-1024x683.jpg - Project page: https://carolinesinders.com/feminist-data-set/ #### This Much I'm Worth (the self-evaluating artwork) — Rachel Ara (2017) - Type: Artwork & installation · Machines: Platform & recommender algorithm - Idea: Let an algorithm price the artwork in public, and the arbitrariness of value becomes visible. - What it is: A large neon sculpture that continuously calculates and displays its own market value, changing as it is looked at, talked about and shared online. - How it works: A networked computer inside the sculpture gathers data (likely visitor counts, social-media mentions and art-market indicators), runs a valuation algorithm and drives the neon digits. - Lens: Classification, bias & surveillance, Opacity & latent space - Shown at: Uncanny Values, MAK Vienna 2019; Aesthetica Art Prize - Video: https://www.youtube.com/watch?v=wZ-P6dnkVcY - Project page: https://www.mak.at/en/program/exhibitions/uncanny_values #### The Library of Missing Datasets — Mimi Ọnụọha (2016) - Type: Artwork & installation · Machines: Dataset & training data - Idea: What a society does not count shows its values as clearly as what it counts; a 'missing' dataset is both a lack and an ought. - What it is: A white filing cabinet full of empty folders, each labelled with a dataset that does not exist, such as people excluded from public housing because of criminal records or trans people killed or injured in hate crimes. - How it works: A physical archive drawn from a list of missing datasets the artist has kept on GitHub since 2015; v2.0 (2018) focuses on Blackness and v3.0 is a locked cabinet of datasets too dangerous to collect. - Lens: Classification, bias & surveillance, Opacity & latent space - Images: https://images.squarespace-cdn.com/content/v1/692f1808f9a01f48c2fa30ac/94ae8920-0367-440f-b24e-5794636edf6f/00small_LibraryofMissingDatasets_propertyofartist_creditBrandonSchulman.jpg https://images.squarespace-cdn.com/content/v1/692f1808f9a01f48c2fa30ac/5f58c44d-4c0b-4d25-9148-02c8fc120a4e/04_LibraryofMissingDatasets2_propertyofartist.jpg - Project page: https://github.com/MimiOnuoha/missing-datasets ### Feminist, Decolonial & Indigenous AI AI imagined from other cultures, genders and knowledge systems. #### Nosukaay — Diane Cescutti (2022) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain - Idea: Weaving was computation long before the computer; the machine here speaks from West African textile knowledge, not from Silicon Valley. - What it is: A 'machine deity' built from a Manjak loom whose two frames are replaced by bare screens; visitors stroke a handwoven loincloth to navigate a film-game about an alternative history of computation, and are ejected if they disrespect it. - How it works: A conductive handwoven Manjak cloth acts as a touch keyboard for an interactive narrative of texts, 3D images and footage shot in Dakar weaving studios. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Shown at: Prix Ars Electronica 2024, Golden Nica (Interactive Art +) - Video: https://www.youtube.com/watch?v=939QlKGURoI - Images: https://archive.aec.at/asset/3797338/preview/ https://archive.aec.at/asset/3797341/preview/ - Project page: https://archive.aec.at/prix/290626/ #### Intersectional AI Toolkit — Sarah Ciston (2021) - Type: Book & essay · Machines: Chatbot & language model, Dataset & training data - Idea: AI literacy is a power question: tools become fairer when the people they affect can read and remake them. - What it is: A set of zines and a wiki that explain AI in plain language to artists, activists and queer and marginalised communities, and ask how intersectional feminism could change it. - How it works: Zines, workshops and an open-source wiki made collaboratively with contributors; developed during Ciston's doctoral research at the University of Southern California. - Lens: Classification, bias & surveillance, Agency, autonomy & co-creation - Images: https://intersectionalai.com/assets/img/zineHowTo.png - Project page: https://intersectionalai.com/ #### Afro-now-ism — Stephanie Dinkins (2020) - Type: Book & essay · Machines: Chatbot & language model, Dataset & training data - Idea: Instead of Afrofuturist hope, act now: take part in making the data and code that will shape your life. - What it is: An essay in which Dinkins asks Black communities to stop waiting for future liberation and to build, now, the technologies and AI systems they need. - How it works: A text published in Noema Magazine, drawing on her work with Not the Only One and community AI workshops. - Lens: Agency, autonomy & co-creation, Classification, bias & surveillance - Paper: https://www.noemamag.com/afro-now-ism/ (Noema Magazine) - Images: https://www.noemamag.com/wp-content/uploads/2020/06/Not-the-Only-One-3d-Avatar-SKETCH252-BECOMING.jpg - Project page: https://www.noemamag.com/afro-now-ism/ #### Indigenous Protocol and Artificial Intelligence Position Paper — Jason Edward Lewis (2020) - Type: Book & essay · Machines: Chatbot & language model, Dataset & training data - Idea: AI can be designed as a relative with obligations, not only as a tool; protocols come before products. - What it is: A position paper from workshops in Honolulu where Indigenous technologists, artists and elders from Aotearoa, Australia, North America and the Pacific set out how AI could be designed from their protocols. - How it works: Edited by Jason Edward Lewis with contributions by Angie Abdilla, Noelani Arista, Suzanne Kite and others; published by the Indigenous Protocol and AI Working Group and CIFAR, with guidelines, essays and design prototypes. - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Paper: https://doi.org/10.11573/spectrum.library.concordia.ca.00986506 (Indigenous Protocol and AI Working Group / CIFAR, Honolulu) - Images: https://www.indigenous-ai.net/wp-content/uploads/2019/04/waiwai2.jpg - Project page: https://www.indigenous-ai.net/position-paper/ #### Masakhane: Participatory Research for African Machine Translation — Masakhane (2020) - Type: Paper · Machines: Chatbot & language model, Dataset & training data - Idea: Who builds a language model decides which languages exist for machines; participatory research is a way to be counted. - What it is: A grassroots network of hundreds of volunteers across Africa that built machine-translation data and models for dozens of African languages, without academic gatekeeping. - How it works: Volunteers contributed data, models and evaluations for their own languages; the process and results are reported in a Findings of EMNLP 2020 paper. - Lens: Classification, bias & surveillance, Agency, autonomy & co-creation - Paper: https://arxiv.org/abs/2010.02353 (Findings of EMNLP 2020) - Video: https://www.youtube.com/watch?v=mbkuRZkg1RY - Project page: https://www.masakhane.io/ #### Wampum.codes — Amelia Winger-Bearskin (2020) - Type: Speculative design · Machines: Chatbot & language model, Dataset & training data - Idea: Software is an agreement between people; Indigenous record-keeping technologies can shape how we write it. - What it is: A framework and podcast that proposes Indigenous values, modelled on wampum belts that record agreements between nations, as a code of ethics for software and AI development. - How it works: Talks, a podcast series and a code-of-ethics document shown at the imagineNATIVE festival, developed during the artist's residencies and teaching. - Lens: Agency, autonomy & co-creation, Classification, bias & surveillance - Shown at: imagineNATIVE Festival 2020 - Video: https://www.youtube.com/watch?v=8I4FjCtAW8Y - Project page: https://wampum.codes/ #### Being — Rashaad Newsome (2019) - Type: Artwork & installation · Machines: Chatbot & language model, Generative model & latent space, Voice & listening machine - Idea: An AI can be written as a Black, queer, decolonial teacher instead of a servile assistant. - What it is: An AI 'digital griot', an animated figure with a vogue dancer's body and an African mask for a face, that teaches, gives tours and leads workshops drawing on bell hooks, Paulo Freire and ballroom culture. - How it works: Motion-captured voguing, a chatbot trained on selected texts and a synthetic voice; iterations have appeared as a guide, a teacher and a therapist. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=CzupjhLM6RQ - Project page: https://rashaadnewsome.com/ #### Oracle for Transfeminist Technologies — Coding Rights (2019) - Type: Speculative design · Machines: Dataset & training data, Platform & recommender algorithm - Idea: Designing with other values produces other machines; imagination is a first step against default tech. - What it is: A deck of cards and a workshop game in which players draw values, objects, situations and 'transfeminist' prompts and imagine technologies built on them. - How it works: Created by Joana Varon of Coding Rights with collaborators; played in workshops across Latin America and online. - Lens: Agency, autonomy & co-creation, Classification, bias & surveillance - Images: https://static.wixstatic.com/media/cfe831_05a51b539e8749c384a4b4021517120a~mv2.png - Project page: https://transfeministech.codingrights.org/ #### Degenerative Cultures — Cesar & Lois (Cesar Baio, Lucy HG Solomon) (2018) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Dataset & training data - Idea: Pair an algorithm with a living organism and machine intelligence starts to borrow the logic of decay rather than of mastery. - What it is: Fungal cultures grow on printed books about humanity's drive to control nature and slowly erase their text; a camera-fed AI reads the growth and, as a 'bhiobrid' agent, searches the internet for similar texts and corrupts them, posting its readouts as the Twitter account @HelloFungus. - How it works: Computer vision tracks Physarum-like microbial growth on book pages; the data drive a cellular-automata and NLP algorithm that finds and rewrites online texts and prints them on site on a thermal printer. - Lens: Machine umwelt, Agency, autonomy & co-creation - Shown at: Lumen Prize 2018, BCS Artificial Intelligence Award; Brighton Digital Festival 2018 - Video: https://www.youtube.com/watch?v=HQwGFEqO1D0 - Images: http://static1.squarespace.com/static/65e7176a61627736e02f6c4a/6606a61178dd51369b416024/6606c08ef1cf21462183357b/1711718676868/AI-Award-Degenerative-Cultures6-1000x0-c-default.jpeg http://cesarandlois.org/wp-content/uploads/Cesar-Lois-degerative-cultures-1.jpg - Project page: https://cesarandlois.org/digitalfungus/ ## Co-performance & Hybrid Bodies Human and machine bodies fused or performing together: prostheses, telepresence, dance, theatre, music and drawing. ### Cyborgs & Prostheses Bodies extended, entered or remote-controlled by machines. #### DOKU the Self — Lu Yang (2022) - Type: Film & video · Machines: Prosthesis & cyborg, Generative model & latent space - Idea: An avatar driven by one's own motion data becomes a machine form of reincarnation. - What it is: DOKU, a genderless digital avatar animated from the artist's own face and movements, travels through six Buddhist realms of rebirth. - How it works: Full-body and facial motion capture of the artist and dancers, rendered in real-time game-engine animation. - Lens: Embodiment & kinesthesia, Grief, memory & afterlife - Shown at: Times Square Midnight Moment, New York 2023 - Video: https://www.youtube.com/watch?v=cq24sSnccXc - Project page: https://en.wikipedia.org/wiki/Lu_Yang_(artist) #### Prosthetic Memory — M Eifler (2020) - Type: Artwork & installation · Machines: Prosthesis & cyborg, Computer vision - Idea: When a small, personal model holds your memories, remembering becomes an act of machine recognition, and others can watch it. - What it is: An ongoing self-augmentation system made to replace the artist's long-term memory, damaged by a childhood brain injury: when a page of their handwritten journal lies on the desk, a camera recognises it and projects the video memory recorded that day beside it. - How it works: A camera over the journal feeds a custom-trained image classifier that identifies the page and triggers the matching daily video projection; later versions add a deepfake of the artist reading the pages. - Lens: Grief, memory & afterlife, Embodiment & kinesthesia - Shown at: S+T+ARTS Prize 2020, Honorary Mention - Images: https://ars.electronica.art/starts-prize/files/2020/06/prostheticmemory-1024x512.jpg - Project page: https://ars.electronica.art/starts-prize/en/prosthetic-memory/ #### Amygdala — Marco Donnarumma (2017) - Type: Artwork & installation · Machines: Industrial robot arm, Prosthesis & cyborg - Idea: An AI prosthesis without a human wearer learns its own gestures: a limb searching for a body and a ritual. - What it is: A robotic arm with a knife, built like a human limb and driven by an AI that learns by trial and error, repeatedly cuts a skin-like material in a purification ritual. - How it works: Humanlike robotic arm with sensors and an adaptive neural network that develops cutting movements through reinforcement learning. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=h4ykqTG1ZjY - Images: https://marcodonnarumma.com/live/wp-content/uploads/2017/09/Marco-Donnarumma_Amygdala-MK3_13_ed_web.jpg - Project page: https://marcodonnarumma.com/works/amygdala/ #### Electromagnetic Brainology — Lu Yang (2017) - Type: Film & video · Machines: Prosthesis & cyborg - Idea: Religious states and emotions are staged as circuits that can be switched on from outside. - What it is: Four deity-like performers, each tied to one of the Buddhist elements (earth, water, fire, wind), fight and dance under the control of electromagnetic brain stimulation. - How it works: Motion-captured dancers and game-engine avatars are combined with imagery of transcranial magnetic stimulation (likely). - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=X7A8S2Q79HI - Project page: https://en.wikipedia.org/wiki/Lu_Yang_(artist) #### Third Thumb — Dani Clode (2017) - Type: Research prototype · Machines: Prosthesis & cyborg - Idea: Extending the body is not just mechanical: the brain rewrites its picture of the hand to include the machine. - What it is: A 3D-printed robotic extra thumb worn beside the little finger and controlled with pressure sensors under the big toes; in a later UCL study, people who wore it for days learned new grasps and their brain's map of the hand changed. - How it works: Two motors pull cables in a 3D-printed thumb; wireless foot sensors map toe pressure to flexion and adduction. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Paper: https://doi.org/10.1126/scirobotics.abd7935 (Science Robotics 2021) - Video: https://www.youtube.com/watch?v=UD_LwZD2Kls - Images: https://dvqlxo2m2q99q.cloudfront.net/000_clients/107233/page/g22141jqBFNhnz6t.jpg - Project page: https://www.daniclodedesign.com/thethirdthumb #### Corpus Nil — Marco Donnarumma (2016) - Type: Performance, dance & music · Machines: Prosthesis & cyborg, Generative model & latent space - Idea: The machine does not follow the body; it learns from and pushes it, making a new, unfamiliar body together. - What it is: A near-motionless, painted body slowly contorts on the floor while an AI system listens to its muscle sounds and blood flow and answers with loud sound and light that push the performer into strange configurations. - How it works: Biosensors (mechanomyography microphones and blood-flow sensors) feed machine-learning software that generates sound and light in real time. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=CAIh8FxLMtk - Images: https://marcodonnarumma.com/live/wp-content/uploads/2017/03/marco-donnarumma_corpus-nil_zkm_onuk.jpg - Project page: https://marcodonnarumma.com/works/corpus-nil/ #### Inferno — Louis-Philippe Demers, Bill Vorn (2015) - Type: Performance, dance & music · Machines: Prosthesis & cyborg, Industrial robot arm - Idea: Being moved by a machine is felt from the inside: the participant is both performer and victim of the choreography. - What it is: Audience members are strapped into robotic exoskeletons that move their arms and torsos to loud industrial music, so the crowd becomes a chorus of bodies puppeteered by machines, a hell after Dante. - How it works: Wearable exoskeletons with pneumatic or electric actuators on the arms and back, controlled centrally in sync with sound and light. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=sGIbZ5DD4fY - Images: http://billvorn.concordia.ca/robography/images/Inferno01.jpg - Project page: http://billvorn.concordia.ca/robography/Inferno.html #### Propel — Stelarc (2015) - Type: Performance, dance & music · Machines: Industrial robot arm, Prosthesis & cyborg - Idea: Here the machine is the choreographer and the human the payload; the body is one interchangeable end-effector among others. - What it is: Stelarc's body is strapped to the end of an industrial robot arm that choreographs its trajectory, speed and orientation for 35 minutes; afterwards the robot repeats the same choreography holding a giant carved ear. - How it works: Industrial robot arm with a 3 m reach programmed with a fixed choreography; the same robot also milled the ear sculpture it later moved. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: DeMonstrable, Lawrence Wilson Art Gallery, Perth (2015) - Video: https://www.youtube.com/watch?v=2bRpTn0KKd8 - Project page: https://stelarc.org/ #### Seismic Sense / Waiting for Earthquakes — Moon Ribas (2013) - Type: Performance, dance & music · Machines: Prosthesis & cyborg, Infrastructure, energy & supply chain - Idea: A networked implant can tie a body to planetary time: the dancer cannot choose when she moves. - What it is: An implanted sensor linked to online seismographs vibrates whenever an earthquake happens anywhere on Earth; in the performance Waiting for Earthquakes, Ribas stands still and turns each tremor she feels into movement. - How it works: Vibration implant (first in the arm, later in the feet) driven by real-time seismic data; stronger quakes produce stronger vibrations. - Lens: Embodiment & kinesthesia, Machine umwelt - Video: https://www.youtube.com/watch?v=1Un4MFR-vNI - Images: https://images.hoobaweb.com/8930/imgf-1200-630/moon-ribas.png - Project page: https://www.moonribas.com/ #### Prospectus for a Future Body — Choy Ka Fai (2011) - Type: Performance, dance & music · Machines: Prosthesis & cyborg - Idea: If a machine can move your muscles, dance history could be stored and replayed in someone else's body. - What it is: A series of experiments and performances in which electrical muscle stimulation makes dancers' muscles contract, trying to 'transfer' the movements of other dancers, including Tatsumi Hijikata's butoh, into a new body as digital muscle memory. - How it works: Electrical nerve and muscle stimulation driven by recorded movement data, in four proposals: Eternal Summer Storm, Bionic Movement Research, SynchroMetrics and Notion: Dance Fiction. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: Festival Tokyo 2011 (Notion: Dance Fiction); 15th Japan Media Arts Festival, Jury Selection - Video: https://www.youtube.com/watch?v=5JlYSgxECxs - Images: https://www.ka5.info/1_PROSPECTUS/3-NOT01.jpg https://www.ka5.info/1_PROSPECTUS/1-ESS02.jpg - Project page: https://www.ka5.info/prospectus.html #### 3rdi — Wafaa Bilal (2010) - Type: Performance, dance & music · Machines: Prosthesis & cyborg, Computer vision, Surveillance & biometrics - Idea: A body can carry an automatic eye it cannot look through; the machine sees for others, not for its wearer. - What it is: For about a year Bilal wore a camera surgically attached to the back of his head; it took a picture every minute and streamed it to a website and to Mathaf in Doha, recording what he left behind. - How it works: Small camera mounted on a titanium plate implanted in the back of the head, capturing one frame per minute and uploading via a laptop. - Lens: Machine seeing, Embodiment & kinesthesia - Shown at: Mathaf: Arab Museum of Modern Art, Doha (2010–11) - Video: https://www.youtube.com/watch?v=h2nNKbO9-Eg - Project page: https://wafaabilal.com/3rdi/ #### Menstruation Machine – Takashi's Take — Sputniko! (Hiromi Ozaki) (2010) - Type: Speculative design · Machines: Prosthesis & cyborg - Idea: A machine on the body can make a gendered experience transferable, and so open to discussion. - What it is: A metal belt-like device that simulates menstruation, with blood release and lower-abdomen pain, shown in a pop music video about a boy who wears it to experience what his friend goes through. - How it works: A blood-dispensing mechanism and electrodes that stimulate the lower abdomen, operated on a 28-day cycle; presented with a music video and costume. - Lens: Embodiment & kinesthesia, Classification, bias & surveillance - Shown at: ZKM Karlsruhe - Video: https://www.youtube.com/watch?v=gnb-rdGbm6s - Images: https://upload.wikimedia.org/wikipedia/commons/a/ad/Sputniko.jpg - Project page: https://en.wikipedia.org/wiki/Sputniko! #### Electric Stimulus to Face — Daito Manabe (2008) - Type: Performance, dance & music · Machines: Prosthesis & cyborg - Idea: When a computer drives the face, expression is separated from feeling. - What it is: Electrodes on faces receive electrical pulses synchronised to music, so the performers' facial muscles twitch into expressions they are not making. - How it works: A computer sends sequenced electrical muscle stimulation (EMS) through surface electrodes, timed to a soundtrack; later versions read one face's myoelectric signal to drive another. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=tJ4_-2_oWuM - Project page: https://en.wikipedia.org/wiki/Daito_Manabe #### Ear on Arm — Stelarc (2006) - Type: Artwork & installation · Machines: Prosthesis & cyborg - Idea: An organ can be relocated and repurposed as a public, networked sense for others rather than for oneself. - What it is: An ear built from cell-cultured tissue and a biocompatible scaffold, surgically constructed on Stelarc's left forearm; the plan is to add a microphone that streams what the ear 'hears' to the internet. - How it works: Porous polyethylene scaffold inserted under the skin in surgeries from 2006, allowed to grow tissue; an embedded microphone was tested and later removed after infection. - Lens: Embodiment & kinesthesia, Machine listening & voice - Video: https://www.youtube.com/watch?v=s80ZjumSA08 - Images: https://upload.wikimedia.org/wikipedia/commons/2/23/Stelarc_3rd_ear_2011f.jpg - Project page: https://stelarc.org/ #### Eyeborg (Cyborg Antenna) — Neil Harbisson (2004) - Type: Artwork & installation · Machines: Prosthesis & cyborg, Computer vision - Idea: A machine sense can become part of a person's body and identity, not just a tool they use. - What it is: An antenna osseointegrated in Harbisson's skull converts light frequencies, including infrared and ultraviolet, into vibrations he hears through bone; he paints and composes with this sense. - How it works: Camera sensor at the tip of the antenna feeds a chip that maps hue to pitch, heard through bone conduction; Bluetooth allows remote colour input. - Lens: Machine seeing, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=ygRNoieAnzI - Project page: https://www.cyborgarts.com/ #### Project Cyborg 2.0 (neural implant) — Kevin Warwick (2002) - Type: Research prototype · Machines: Prosthesis & cyborg, Industrial robot arm - Idea: A nervous system can be networked: touch and movement travel as data between bodies and machines. - What it is: Warwick had a 100-electrode array implanted into the median nerve of his arm; he used it to move a robot hand over the internet from New York, and to exchange simple nerve signals with his wife, who had a simpler implant. - How it works: Utah microelectrode array in the median nerve, connected to a computer that recorded and stimulated nerve signals and relayed them over the internet. - Lens: Embodiment & kinesthesia, Intimacy, care & companionship - Paper: https://doi.org/10.1001/archneur.60.10.1369 (Archives of Neurology 2003) - Video: https://www.youtube.com/watch?v=GLq7edATaFo - Project page: https://en.wikipedia.org/wiki/Kevin_Warwick #### Prosthetic Head — Stelarc (2002) - Type: Artwork & installation · Machines: Chatbot & language model, Humanoid & android - Idea: A 'prosthesis' can be a mind rather than a limb: the artist extends himself through a talking agent that answers for him. - What it is: A large projected 3D head with Stelarc's face that talks with visitors who type questions; it has its own opinions, moods and can make up verse, more a conversational double than a portrait. - How it works: Embodied conversational agent: chatbot engine with a customised knowledge base, text-to-speech and a real-time animated 3D model of the artist's head. - Lens: Alterity & the uncanny, Machine listening & voice - Video: https://www.youtube.com/watch?v=dprvwAWfy9Q - Project page: https://stelarc.org/ #### Cyborg series — Lee Bul (1998) - Type: Artwork & installation · Machines: Prosthesis & cyborg, Humanoid & android - Idea: Techno-perfection is shown as a gendered fantasy, and the cyborg body as incomplete. - What it is: White silicone figures of female cyborgs, often missing a head or limbs, hung from the ceiling like armour or anime heroines. First shown at Artsonje Center, Seoul, in 1998, the series recasts the machine-enhanced body as fragmentary and eroticised. - How it works: Cast silicone, polyurethane and paint, drawing on manga, anime and classical sculpture; works such as Cyborg W1–W4 (1998). - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Shown at: Artsonje Center, Seoul 1998; Korean Pavilion, 48th Venice Biennale 1999; Lee Bul: Crashing, Hayward Gallery 2018 - Video: https://www.youtube.com/watch?v=aS5di0wCD1I - Images: https://upload.wikimedia.org/wikipedia/commons/5/5a/Untitled_%28Cyborg%27s_pelvis%29_by_Lee_Bul%2C_British_Museum_01.jpg - Project page: https://www.guggenheim.org/artwork/artist/lee-bul #### Exoskeleton — Stelarc (1998) - Type: Performance, dance & music · Machines: Robot creature & mobile robot, Prosthesis & cyborg - Idea: The body becomes the pilot and the passenger of a larger, non-human body at the same time. - What it is: Stelarc stands on a six-legged pneumatic walking machine and steers its walk and turns with arm gestures, while the hiss and clank of the machine are amplified into a soundtrack. - How it works: Pneumatic hexapod robot built with engineers in Hamburg; arm movements read by sensors on an arm-mounted exoskeleton select gaits and directions. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=R2MntBUwUxY - Project page: https://stelarc.org/ #### Time Capsule — Eduardo Kac (1997) - Type: Performance, dance & music · Machines: Prosthesis & cyborg, Surveillance & biometrics - Idea: Being readable by machines is a way of being classified; the implant makes the body a database entry. - What it is: Live on Brazilian television and the web, Kac implanted an identification microchip in his ankle and then registered himself online in a pet-recovery database as both animal and owner. - How it works: RFID microchip injected with a syringe, scanned live, and the chip number registered in an animal ID database over the internet. - Lens: Classification, bias & surveillance, Embodiment & kinesthesia - Shown at: Casa das Rosas, São Paulo (1997) - Images: https://www.ekac.org/needle.gif - Project page: https://www.ekac.org/timec.html #### Ping Body — Stelarc (1996) - Type: Performance, dance & music · Machines: Prosthesis & cyborg, Infrastructure, energy & supply chain - Idea: The network can become the body's nervous system: here the body is moved by data it cannot feel. - What it is: Stelarc sends ping signals to dozens of internet domains; the round-trip times drive a muscle-stimulation system, so the flow of internet traffic, not his will, makes his arms and legs twitch and move. - How it works: Ping values from global internet hosts mapped to a computer-controlled muscle stimulation system with electrodes on the limbs; the Third Hand is driven by his own EMG. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: DEAF96, V2_ Rotterdam (1996) - Video: https://www.youtube.com/watch?v=45MOlsFGSHI - Project page: https://stelarc.org/ #### Epizoo — Marcel·lí Antúnez Roca (1994) - Type: Performance, dance & music · Machines: Prosthesis & cyborg - Idea: Interfaces make cruelty feel like a game: one click at a distance is felt as pain in another body. - What it is: Antúnez stands on a turntable wearing a pneumatic exoskeleton; audience members use a computer mouse to make it pull at his nose, ears, mouth, buttocks and chest, while cartoon animations of his body play on screens. - How it works: Body-worn pneumatic mechanisms driven by a computer and controlled by spectators through a graphical mouse interface. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=5utDR8VWxms - Images: https://www.marceliantunez.com/media/cache/3b/1f/3b1f492a17b1654525596a9cc5d9af01.jpg - Project page: https://www.marceliantunez.com/work/epizoo/ #### Inter Dis-Communication Machine — Kazuhiko Hachiya (1993) - Type: Artwork & installation · Machines: Prosthesis & cyborg, Computer vision - Idea: Swapping senses through a machine shows how much the self depends on its own point of view. - What it is: Two people wear head-mounted displays and winged costumes; each sees and hears only what the other's camera and microphone pick up, so they must move while inside each other's perception. - How it works: Wireless video cameras and microphones on each wearer transmit to the other's head-mounted display and headphones. - Lens: Embodiment & kinesthesia, Machine seeing - Video: https://www.youtube.com/watch?v=JOzVzcmK0VU - Images: https://www.ntticc.or.jp/uploads/assets/000/63850.936.large.jpg - Project page: https://www.ntticc.or.jp/en/archive/works/inter-dis-communication-machine/ #### Third Hand — Stelarc (1980) - Type: Performance, dance & music · Machines: Prosthesis & cyborg - Idea: An extra limb is learned, not given: the body slowly makes room for a machine that it controls indirectly. - What it is: A mechanical right hand attached to Stelarc's arm and driven by the electrical signals of his abdominal and leg muscles, which he wrote and performed with alongside his own two hands for many years. - How it works: Robotic hand with pinch, grasp and wrist rotation, controlled by EMG electrodes on abdomen and legs, with tactile feedback to the arm. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=Or1CVc6A9YA - Project page: https://stelarc.org/ #### WearComp and EyeTap — Steve Mann (1978) - Type: Research prototype · Machines: Prosthesis & cyborg, Computer vision, Surveillance & biometrics - Idea: Living inside a machine's view changes both what you see and how others see you; Mann named the reverse gaze 'sousveillance'. - What it is: Over decades Mann built and wore ever smaller computers and head-mounted camera-displays in daily life, seeing the world through a computer that could record, alter and share his view. - How it works: Backpack and later eyeglass-based wearable computers with cameras and displays aligned to the eye (EyeTap) to mediate vision in real time. - Lens: Machine seeing, Embodiment & kinesthesia, Classification, bias & surveillance - Paper: https://doi.org/10.1109/2.566147 (IEEE Computer 1997) - Video: https://www.youtube.com/watch?v=fCco6FMCRmk - Images: https://upload.wikimedia.org/wikipedia/commons/3/3e/SteveMann_with_Generation-4_Glass_1999.jpg - Project page: http://wearcam.org/ ### Telepresence & Remote Bodies Acting, touching and being present through a machine at a distance. #### Avatar Robot Café DAWN ver.β — Ory Laboratory (2018) - Type: Research prototype · Machines: Social & companion robot - Idea: A robot body can give work and social presence back to people whose own bodies keep them home. - What it is: A café in Tokyo where the waiters are OriHime robots operated from home by people with ALS and other conditions that keep them from going out. - How it works: Tablet and eye-gaze interfaces control the 1.2 m OriHime-D and desktop OriHime robots over the internet; pilots began in 2018 and a permanent café opened in Nihonbashi in 2021. - Lens: Hidden labour & extraction, Embodiment & kinesthesia, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=tKzCZ_yP4-Y - Images: https://dawn2021.orylab.com/assets/images/common/ogp2.jpg - Project page: https://dawn2021.orylab.com/en/ #### Habeas Corpus — Laurie Anderson (2015) - Type: Performance, dance & music · Machines: Infrastructure, energy & supply chain, Computer vision - Idea: Telepresence can give a body a presence that borders and law deny it. - What it is: Mohammed el Gharani, detained at Guantánamo from age 14 and barred from entering the US, appeared live from West Africa as a giant projection on a white seated statue in New York's Park Avenue Armory, telling his story. - How it works: Live video stream from Africa projected onto a large 3D sculpture modelled on the Lincoln Memorial statue. - Lens: Embodiment & kinesthesia, Classification, bias & surveillance - Shown at: Park Avenue Armory, New York (2015) - Video: https://www.youtube.com/watch?v=kUKLA-W0nDM - Project page: https://laurieanderson.com/ #### The Machine to Be Another — BeAnotherLab (2012) - Type: Artwork & installation · Machines: Computer vision, Prosthesis & cyborg - Idea: Body ownership is fragile: synchronised seeing and touching are enough to let a machine swap bodies. - What it is: Two participants wear headsets that show the live view from each other's head-mounted camera and mirror each other's movements, so each feels, for a few minutes, that they are in the other's body. - How it works: Head-mounted cameras and VR displays with synchronised movement and touch, based on body-ownership illusion research. - Lens: Embodiment & kinesthesia, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=hjIOUa10YpI - Images: https://beanotherlab.org/wp-content/uploads/2019/04/bal_tmtba01.jpg - Project page: https://beanotherlab.org/home/work/tmtba/ #### Telenoid R1 — Hiroshi Ishiguro (2010) - Type: Research prototype · Machines: Humanoid & android, Social & companion robot - Idea: A deliberately genderless, ageless body lets the listener imagine whoever is speaking, which makes presence feel stronger, not weaker. - What it is: A small, pale, limb-stub robot with a minimal face that carries a remote person's voice and head movements; people hold it in their arms while talking to someone far away. - How it works: About 80 cm long and 5 kg, silicone-skinned, with a few actuators in the head and arms driven by the operator's face tracking; tested in elder care in Japan and Denmark. - Lens: Intimacy, care & companionship, Embodiment & kinesthesia, Alterity & the uncanny - Shown at: RMIT Gallery, Melbourne - Paper: https://doi.org/10.1145/2048259.2048274 (ACM SIGGRAPH 2011 Emerging Technologies) - Video: https://www.youtube.com/watch?v=hKvsHU1hrxg - Images: http://www.geminoid.jp/img/robot/telenoid.jpg - Project page: https://en.wikipedia.org/wiki/Telenoid_R1 #### Domestic Tension — Wafaa Bilal (2007) - Type: Performance, dance & music · Machines: Autonomous vehicle & drone, Computer vision - Idea: Remote-controlled violence feels like a game to the shooter and like war to the target, as with drone warfare. - What it is: For 30 days Bilal lived in a Chicago gallery watched by a webcam; online visitors could aim and fire a paintball gun at him at any time, which they did tens of thousands of times. - How it works: Webcam feed plus a robotic paintball gun controlled through a website and chat room. - Lens: Embodiment & kinesthesia, Machine seeing - Shown at: FlatFile Galleries, Chicago (2007) - Video: https://www.youtube.com/watch?v=F_L6BC9O0rY - Images: http://wafaabilal.com/wp-content/uploads/2012/04/Domestic_Tension_Wafaabilal_1.jpg - Project page: https://wafaabilal.com/domestic-tension/ #### PRoP (Personal Roving Presence) — Eric Paulos (1998) - Type: Research prototype · Machines: Robot creature & mobile robot, Computer vision - Idea: Remote presence needs more than video: gaze, gesture and the ability to wander make a body socially present. - What it is: Made with John Canny, a tall mobile robot with a screen showing the remote user's face, a camera, microphone and a pointer arm lets a person roam, look around, gesture and chat in a distant building. - How it works: Wheeled base with a video screen, pan-tilt camera, audio and a pointing arm, teleoperated over the internet. - Lens: Embodiment & kinesthesia, Intimacy, care & companionship - Paper: https://doi.org/10.1145/274644.274686 (CHI 1998) - Video: https://www.youtube.com/watch?v=ypKd1pbDbds - Project page: https://www.paulos.net/ #### Rara Avis — Eduardo Kac (1996) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Computer vision - Idea: Telepresence places you inside another body among other species, as the odd one out. - What it is: In an aviary with some 30 zebra finches perches a telerobotic macaw; visitors wearing a VR headset see through its eyes and move its head, and internet users could also look through it and speak through its voice. - How it works: Robotic macaw with stereo cameras in its eyes, coupled to a VR headset's head tracking, with video and audio also streamed to the internet. - Lens: Machine umwelt, Machine seeing, Embodiment & kinesthesia - Shown at: Nexus Contemporary Art Center, Atlanta (1996) - Images: https://www.ekac.org/rara.avis.jpg - Project page: https://www.ekac.org/raraavis.html #### Fractal Flesh — Stelarc (1995) - Type: Performance, dance & music · Machines: Prosthesis & cyborg, Infrastructure, energy & supply chain - Idea: A body can be distributed across a network and moved by many remote agents at once. - What it is: During the Telepolis event, people at terminals in Paris, Helsinki and Amsterdam touched a screen model of Stelarc's body and remotely stimulated his muscles in Luxembourg, making him move without his consent. - How it works: Touch-screen interface at remote sites sends commands over the internet to a muscle-stimulation system attached to the artist's body. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: Telepolis, Luxembourg (1995) - Video: https://www.youtube.com/watch?v=N2VkOE_AgNI - Project page: https://stelarc.org/ #### The Telegarden — Ken Goldberg (1995) - Type: Artwork & installation · Machines: Industrial robot arm - Idea: Care at a distance depends on trusting a machine and an image: did I really water that plant? - What it is: Made with Joseph Santarromana, a small garden of living plants was tended for nine years by thousands of internet users who planted and watered seeds through an industrial robot arm they could see only on a webpage. - How it works: Adept-1 industrial robot arm with a camera over a planter, controlled through a web interface; members' actions were logged. - Lens: Embodiment & kinesthesia, Intimacy, care & companionship, Machine umwelt - Shown at: Ars Electronica Center, Linz (1996–2004) - Video: https://www.youtube.com/watch?v=vQE-qJ_aYPw - Images: https://upload.wikimedia.org/wikipedia/commons/3/31/Telegarden.jpg https://goldberg.berkeley.edu/garden/telegarden-8x6-72dpi.jpg - Project page: https://goldberg.berkeley.edu/garden/Ars/ #### Tillie, the Telerobotic Doll — Lynn Hershman Leeson (1995) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Computer vision, Surveillance & biometrics - Idea: An object that looks back turns the viewer into the watched, and lends a remote gaze a body. - What it is: A doll with video cameras for eyes sits in a gallery; web visitors look through her eyes and turn her head, seeing the room from the doll's point of view. - How it works: Doll fitted with cameras in the eyes and a motorised head, streaming images to a website where users control the viewing direction. - Lens: Machine seeing, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=_ZOIdUZueHc - Project page: https://www.lynnhershman.com/ #### The Mercury Project — Ken Goldberg (1994) - Type: Research prototype · Machines: Industrial robot arm, Computer vision - Idea: The web turned a lab robot into a public, shared body, and raised the question of whether what we see remotely is real. - What it is: One of the first robots anyone could control through the web: visitors directed an arm to blow bursts of air into a sandbox and excavate buried objects, seeing the results in camera images. - How it works: Industrial robot arm with camera and air nozzle controlled through a Mosaic browser interface via CGI scripts. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Paper: https://doi.org/10.1109/robot.1995.525358 (IEEE ICRA 1995) - Project page: https://goldberg.berkeley.edu/ #### cyberSM — Stahl Stenslie (1993) - Type: Artwork & installation · Machines: Prosthesis & cyborg, Infrastructure, energy & supply chain - Idea: Remote touch through a machine is an early question for networked intimacy: what counts as contact? - What it is: Made with Kirk Woolford, two participants in different cities wear suits fitted with stimulators and touch each other's 3D body models on screen; the touches arrive as vibration and electrical stimuli on the other's skin. - How it works: Body suits with stimulators, 3D-scanned body avatars and a telecommunication link (ISDN) between remote sites. - Lens: Intimacy, care & companionship, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=E5TFZJzKbYI - Project page: https://en.wikipedia.org/wiki/Stahl_Stenslie #### Telematic Dreaming — Paul Sermon (1992) - Type: Artwork & installation · Machines: Computer vision, Infrastructure, energy & supply chain - Idea: Telepresence is felt through the body: people react to a projected hand as if it could touch them. - What it is: Two double beds in separate places are linked by video: the image of the person on one bed is projected onto the other bed, so strangers lie together, touch and play with each other's projected bodies. - How it works: Overhead camera and projector at each bed, linked over ISDN lines, with monitors showing the combined image. - Lens: Embodiment & kinesthesia, Intimacy, care & companionship - Shown at: Kajaani, Finland (1992) - Video: https://www.youtube.com/watch?v=S4Wt16PnqeY - Project page: https://en.wikipedia.org/wiki/Paul_Sermon #### Ornitorrinco — Eduardo Kac (1989) - Type: Artwork & installation · Machines: Robot creature & mobile robot, Computer vision - Idea: Presence can be displaced into a small, slow machine; the remote body is shared rather than owned. - What it is: A small telerobot named after the platypus is driven by remote participants over ordinary phone lines; they see its surroundings through its camera eye, sharing a machine body across cities. - How it works: Mobile robot with a camera, controlled via telephone keypad tones, sending back still video images over phone lines (with Ed Bennett). - Lens: Embodiment & kinesthesia, Machine seeing - Images: https://www.ekac.org/ornitorrinco.color.jpg - Project page: https://www.ekac.org/ornitorrinco.html #### Hole in Space — Kit Galloway & Sherrie Rabinowitz (1980) - Type: Artwork & installation · Machines: Infrastructure, energy & supply chain, Computer vision - Idea: A telecom link becomes a public place: people invent how to behave with remote bodies as soon as they meet them. - What it is: For three evenings, life-size live video screens linked passers-by at Lincoln Center in New York and a shopping centre in Century City, Los Angeles, via satellite, with no signs or explanation; strangers, then families, met across the continent. - How it works: Two-way satellite video link with cameras and rear-projection screens in shop windows on each coast. - Lens: Embodiment & kinesthesia, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=QSMVtE1QjaU - Project page: http://www.ecafe.com/getty/HIS/ #### Telexistence (TELESAR) — Susumu Tachi (1980) - Type: Research prototype · Machines: Humanoid & android, Prosthesis & cyborg - Idea: A robot can be designed not as another agent but as a second body the human inhabits. - What it is: Tachi's concept and robot series in which an operator wearing a head-mounted display and haptic suit sees, hears and feels through a remote humanoid robot as if present in its body; TELESAR V (2011) adds fine touch. - How it works: Stereo cameras on the robot head follow the operator's head motion; master-slave arms and hands transmit force, vibration and temperature back to the operator. - Lens: Embodiment & kinesthesia, Machine umwelt - Paper: https://doi.org/10.1109/MCG.2016.6 (IEEE Computer Graphics and Applications, 2016) - Video: https://www.youtube.com/watch?v=5j_hdDzgGvc - Images: https://thumb.wikimedia.org/wikipedia/commons/thumb/d/d4/Telesar%E2%85%A4%2C_telexistence_robot_by_Susumu_Tachi_Laboratory%2C_exhibited_at_Miraikan_%282015-06-15_03.58.07_by_Franklin_Heijnen%29.jpg/1280px-Telesar%E2%85%A4%2C_telexistence_robot_by_Susumu_Tachi_Laboratory%2C_exhibited_at_Miraikan_%282015-06-15_03.58.07_by_Franklin_Heijnen%29.jpg - Project page: https://tachilab.org/en/about/telexistence.html ### Dancing with Machines Choreographies for human and machine bodies. #### Yang BOT — Unitree Robotics, Zhang Yimou (2025) - Type: Performance, dance & music · Machines: Humanoid & android - Idea: Folk dance performed by humanoids turns robot bodies into a national image watched by hundreds of millions. - What it is: Sixteen Unitree H1 humanoid robots in padded jackets dance yangge and twirl handkerchiefs beside human dancers on the CCTV Spring Festival Gala, directed by Zhang Yimou. - How it works: Choreography learned through motion capture and reinforcement learning, with robots synchronised to dancers and handkerchief-spinning end effectors. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: CCTV Spring Festival Gala 2025 - Video: https://www.youtube.com/watch?v=Ykiuz1ZdGBc - Project page: https://technode.com/2025/01/29/unitree-robotics-humanoid-robots-steal-the-show-at-2025-cctv-spring-festival-gala-with-ai-powered-yangge-dance/ #### Nondance Dance — Shinseungback Kimyonghun (2022) - Type: Performance, dance & music · Machines: Computer vision - Idea: To dance outside a machine's categories, dancers first have to learn how the machine sees dance. - What it is: Dancers perform while a video-recognition AI watches and signals by voice or light whenever it recognises 'dance'. They try to find movements that are still dance to humans but invisible as dance to the machine. - How it works: Google Video AI action recognition (20,000+ labels, including many dance categories) gives live feedback to the dancers. - Lens: Machine seeing, Embodiment & kinesthesia, Classification, bias & surveillance - Video: https://vimeo.com/776909143 - Images: https://djhznh41oxwef.cloudfront.net/works/nondance_dance/ssbkyh_Nondance_dance_01.png - Project page: https://ssbkyh.com/works/nondance_dance/ #### Do You Love Me? — Boston Dynamics (2020) - Type: Film & video · Machines: Humanoid & android, Robot creature & mobile robot - Idea: Choreography is a strong way to make machines likeable, and it hides the engineering, control and purpose behind them. - What it is: Atlas humanoids, Spot robot dogs and a Handle robot dance to The Contours' 1962 song in a video released at the end of 2020; it drew tens of millions of views and debate about how dance makes robots seem joyful. - How it works: Motion sequences authored with choreographer Monica Thomas in Boston Dynamics' choreography software and executed by the robots' own balance control, filmed in edited takes. - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=fn3KWM1kuAw - Project page: https://bostondynamics.com/ #### Merritt + Robot — Merritt Moore (2020) - Type: Film & video · Machines: Industrial robot arm - Idea: In isolation, a programmable arm becomes a partner; the duet is also a rehearsal between a dancer and her own code. - What it is: During the 2020 lockdown, dancer and physicist Merritt Moore programmed a Universal Robots arm as her only partner and filmed a pas de deux in which the robot both leads and follows. - How it works: Collaborative robot arm (Universal Robots) programmed by the dancer herself. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=uSOHc3ODLzU - Images: https://images.squarespace-cdn.com/content/v1/66c15199c6b23b75923ac311/b4ae89c2-ea53-41f0-82e8-2d95034920ed/Merritt_Moore_robots_Mexico.png - Project page: https://www.merrittmoore.com/ #### Living Archive — Wayne McGregor (2019) - Type: Performance, dance & music · Machines: Generative model & latent space, Dataset & training data, Computer vision - Idea: A choreographer's style, turned into a model, becomes a partner that proposes 'his' next move. - What it is: Google Arts & Culture trained a model on thousands of hours of video from 25 years of McGregor's choreography; given a dancer's pose, it proposes what could come next, and these phrases were staged in Living Archive: An AI Performance Experiment. - How it works: Pose estimation on archive video plus a neural network predicting sequences of body poses; dancers learn and adapt the proposals. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Shown at: The Music Center, Los Angeles (2019) - Video: https://www.youtube.com/watch?v=Hd3BDvfmg4k - Images: https://lh3.googleusercontent.com/UAE2Fk3CoTgjTWKg9PiFGU6XjasBWnRO5g2xP_LHAHZZVVgfPCx3K3BKv58VuXW9hErbCpVgYhBUk1iAzprsgTkX_eTl - Project page: https://experiments.withgoogle.com/living-archive-wayne-mcgregor #### OUTPUT — Catie Cuan (2019) - Type: Performance, dance & music · Machines: Industrial robot arm - Idea: Choreography can be a design method for robots: the dance studies what it takes for a machine's motion to read as expressive. - What it is: Cuan dances with an ABB IRB 6700 industrial robot, then both bodies are recorded and replayed as motion capture, video and a web app, so each body's movement is translated into the other's medium. - How it works: Robot motion programmed from Cuan's movement, captured with motion capture and recombined in video and interactive software. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Paper: https://doi.org/10.3389/frobt.2020.576790 (Frontiers in Robotics and AI 2021) - Video: https://www.youtube.com/watch?v=Ng0ioS7oNXY - Project page: https://www.catiecuan.com/ #### discrete figures — Rhizomatiks, ELEVENPLAY, Daito Manabe (2018) - Type: Performance, dance & music · Machines: Computer vision, Generative model & latent space, Autonomous vehicle & drone - Idea: A dancer meeting a model of their own movement sees how a machine abstracts a body. - What it is: A stage work in which dancers perform with machine-generated 'dancers': a model trained on their motion creates new movement, shown as projected figures and on drones carrying screens. - How it works: Motion-capture data from ELEVENPLAY dancers trained generative models; real-time tracking aligns projections and drones with the performers; made with Kyle McDonald. - Lens: Machine seeing, Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: Spiral Hall, Tokyo 2018; Gray Area, San Francisco - Video: https://www.youtube.com/watch?v=hauXQQhwbgM - Images: https://research.rhizomatiks.com/s/works/discrete_figures/images/video_image.jpg - Project page: https://research.rhizomatiks.com/s/works/discrete_figures/en/ #### Co(AI)xistence — Justine Emard (2017) - Type: Film & video · Machines: Humanoid & android - Idea: Treat a learning android as a different kind of intelligence and look for a language both sides can share. - What it is: A 12-minute video installation in which the performer Mirai Moriyama and the android Alter try to communicate through gesture, signals and speech; Alter learns from the exchange as it goes. - How it works: Filmed improvisation between a dancer and Alter (Ishiguro Lab, Osaka University, and Ikegami Lab, University of Tokyo), whose motion is driven by a neural network and sensor input. - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Shown at: AI: More than Human, Barbican, London 2019 - Video: https://www.youtube.com/watch?v=vcdUTEpSV1s - Project page: https://justineemard.com/coaixistence/ #### 24 drones — Rhizomatiks, ELEVENPLAY, Daito Manabe (2016) - Type: Performance, dance & music · Machines: Autonomous vehicle & drone - Idea: Choreography can include machine bodies whose precision exposes the dancers' own timing. - What it is: Five dancers perform with a swarm of 24 small drones that move in formation around and between them, turning the drones into a corps de ballet. - How it works: Indoor motion capture tracks drones and dancers; flight paths are pre-simulated in software and synchronised to music and light. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=SPcZmDTW8KU - Images: https://research.rhizomatiks.com/s/works/24drones/img/drone_00.png - Project page: https://research.rhizomatiks.com/s/works/24drones/en/ #### chor-rnn — Louise Crnkovic-Friis, Luka Crnkovic-Friis (2016) - Type: Research prototype · Machines: Generative model & latent space, Dataset & training data - Idea: Watching a model learn to dance makes visible what 'style' means to a machine: statistics of joints. - What it is: A recurrent neural network trained on hours of motion capture of a contemporary dancer learns to generate new choreography in her style; published videos show it moving from twitching noise to dance-like phrases as training goes on. - How it works: Mixture-density LSTM network trained on Kinect motion-capture joint positions of the choreographer's improvisations. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia, Opacity & latent space - Paper: https://arxiv.org/abs/1605.06921 (arXiv 2016) - Video: https://www.youtube.com/watch?v=Q4_XSMqN8w0 #### Bionic dance at TED2014 (with Adrianne Haslet-Davis) — Hugh Herr (2014) - Type: Research prototype · Machines: Prosthesis & cyborg - Idea: A prosthesis can be designed for expression, tuned to how a dancer moves, not only to walking. - What it is: Ballroom dancer Adrianne Haslet-Davis, who lost her lower left leg in the 2013 Boston Marathon bombing, danced a rumba on stage at TED2014 on a bionic leg made by Hugh Herr's lab. - How it works: Powered ankle-foot prosthesis with motors and sensors, tuned using motion data recorded from dancers. - Lens: Embodiment & kinesthesia, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=CDsNZJTWw0w - Images: https://pi.tedcdn.com/r/pe.tedcdn.com/images/ted/ca0520623aeb359f6d5f371d516371fadcae2454_1600x1200.jpg?u%5Br%5D=2&u%5Bs%5D=0.5&u%5Ba%5D=0.8&u%5Bt%5D=0.03&quality=82&w=1200 - Project page: https://www.ted.com/talks/hugh_herr_the_new_bionics_that_let_us_run_climb_and_dance #### ROBOT — Blanca Li (2013) - Type: Performance, dance & music · Machines: Humanoid & android, Social & companion robot - Idea: Putting humans and robots in the same choreography shows how much of our movement is already mechanical. - What it is: Eight dancers share the stage with seven small NAO humanoid robots and the musical machines of Maywa Denki; humans move like robots and robots attempt human dances. - How it works: Programmed NAO robots (Aldebaran) with pre-set choreography, alongside Maywa Denki instruments and live dancers. - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Shown at: Montpellier Danse (2013) - Video: https://www.youtube.com/watch?v=vrH0Ydj3HN4 - Project page: https://www.blancali.com/en/spectacle/robot-2/ #### Huang Yi & KUKA — Huang Yi (2012) - Type: Performance, dance & music · Machines: Industrial robot arm - Idea: Programming a robot minute by minute becomes a way of looking at one's own loneliness in a machine mirror. - What it is: A duet, later an evening-length piece, between Huang Yi and a KUKA industrial robot arm that he programmed himself; the arm carries a light, mirrors his moves and partners him in slow, close choreography. - How it works: KUKA industrial arm programmed by hand (10–20 hours of programming per minute of choreography), with a light or camera on the end-effector. - Lens: Embodiment & kinesthesia, Intimacy, care & companionship, Agency, autonomy & co-creation - Shown at: Ars Electronica Festival 2013; TED2017, Vancouver; Toured 20 countries - Video: https://www.youtube.com/watch?v=Xy6tPmFlN34 - Images: https://huangyistudio.com/wp-content/uploads/2022/10/huangyiandkuka_TED2017_01.jpg https://upload.wikimedia.org/wikipedia/commons/5/51/Prix_Ars_Electronical_2013_11_Huang_Yi_KUKA.jpg - Project page: https://huangyistudio.com/archives/project/huang-yi-and-kuka #### Human Interface — Thomas Freundlich (2012) - Type: Performance, dance & music · Machines: Industrial robot arm - Idea: Once the safety cage comes down, a factory robot can be read as a personality, not a tool. - What it is: A dance work for two dancers and two industrial robots in the old boiler hall of the Cable Factory in Helsinki; the robots' programmed movement is exact yet surprisingly human, and they act as equal dance partners. - How it works: ABB industrial robots programmed by the choreographer, with a safety system that allowed humans inside the robots' working area; follows his 2008 duet Actuator. - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Shown at: Zodiak Center for New Dance, Helsinki (2012) - Video: https://www.youtube.com/watch?v=MCSUVI--0Sw - Project page: http://www.freundlich.org/ #### Grace State Machines — Bill Vorn (2007) - Type: Performance, dance & music · Machines: Robot creature & mobile robot, Prosthesis & cyborg - Idea: When a robot moves from a human's inner signals, the line between the two bodies becomes the choreography. - What it is: A 27-minute stage duet between performer Emma Howes and a robot: her movements and internal body states are captured and transposed onto the machine's body, so the two move in a shifting, symbiotic dialogue. - How it works: Motion capture and biofeedback sensors on the performer drive the robot's actuators in real time (with Jonathan Villeneuve). - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=b6TiJdGRMTw - Images: https://billvorn.concordia.ca/robography/images/GSMBB.jpg - Project page: https://billvorn.concordia.ca/robography/GraceState.html #### Glow — Gideon Obarzanek, Frieder Weiss (2006) - Type: Performance, dance & music · Machines: Computer vision - Idea: Machine vision can be a stage partner: the dancer performs for the camera's reading of her as much as for the audience. - What it is: A solo dancer moves on a white floor while an infrared camera tracks her body and a computer projects graphics that respond in real time, wrapping, stretching and chasing her shape with light. - How it works: Infrared video tracking of the dancer (Frieder Weiss's Kalypso software) generating projected graphics on the floor. - Lens: Machine seeing, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=C4He543_a80 - Project page: https://chunkymove.com.au/ #### DANCER #1 — Kris Verdonck (2003) - Type: Performance, dance & music · Machines: Robot creature & mobile robot - Idea: A machine's struggle to function is enough to read it as a dancer and to feel for it. - What it is: A grinding wheel with a large steel 'L' hangs from the ceiling; when the motor starts, the L spins out of control while the machine keeps trying to find its balance, until it looks like a tragic hero in distress. - How it works: Industrial grinding motor with an eccentric steel shape, run as a timed performance in a theatre space. - Lens: Failure, glitch & error, Embodiment & kinesthesia, Alterity & the uncanny - Shown at: Kunstenfestivaldesarts, Brussels (2003); Festival d'Avignon (2008) - Video: https://www.youtube.com/watch?v=xKbBImj3mWM - Images: https://www.atwodogscompany.org/media/og/17-dancer_krisverdonckhigh.jpg - Project page: https://www.atwodogscompany.org/en/projects/dancer-1/ #### Zone — Pablo Ventura, robotlab (2001) - Type: Performance, dance & music · Machines: Industrial robot arm - Idea: A factory arm on stage is a dancer with a different anatomy, not a prop. - What it is: A KUKA industrial robot, programmed with robotlab, performs as a member of Ventura Dance Company alongside human dancers in a contemporary dance work. - How it works: KUKA industrial robot with choreography programmed offline and synchronised with the dancers and music. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=PD1i6rItAfA - Project page: https://www.ventura-dance.com/ #### BIPED — Merce Cunningham, Paul Kaiser, Shelley Eshkar (1999) - Type: Performance, dance & music · Machines: Computer vision, Humanoid & android - Idea: Captured movement becomes a second, weightless company that dances with and apart from the dancers. - What it is: Dancers perform behind a transparent scrim onto which large hand-drawn-looking figures are projected, made from motion-captured Cunningham dancers, so live bodies and their digital traces share the stage. - How it works: Optical motion capture of dancers rendered as abstract line figures by Kaiser and Eshkar, projected on a scrim; music by Gavin Bryars. - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Shown at: Zellerbach Hall, Berkeley (1999) - Video: https://www.youtube.com/watch?v=b_XItg0Vn1c - Project page: https://www.mercecunningham.org/the-work/choreography/biped/ #### Ghostcatching — Bill T. Jones, Paul Kaiser, Shelley Eshkar (1999) - Type: Artwork & installation · Machines: Computer vision - Idea: Motion capture separates movement from the mover; what remains is a ghost the dancer can watch. - What it is: A virtual dance made from motion capture of Bill T. Jones improvising: his movement becomes drawn lines that multiply, collide and are caught inside a box, a dancer without his body. - How it works: Optical motion capture of Jones's improvisations, rendered as hand-drawn-style strokes in 3D animation. - Lens: Embodiment & kinesthesia, Grief, memory & afterlife - Video: https://www.youtube.com/watch?v=x-7Mo3cg9jw - Project page: https://openendedgroup.com/ #### Trackers (with LifeForms) — Merce Cunningham (1991) - Type: Performance, dance & music · Machines: Humanoid & android - Idea: A virtual body with no weight or habits proposes movement that the human body must then learn to inhabit. - What it is: The first dance Cunningham made with the software LifeForms: he composed sequences on animated 3D figures on a computer and then asked dancers to perform movements that no body had tried before. - How it works: LifeForms, choreography software from Simon Fraser University (Tom Calvert, Thecla Schiphorst and others), for keyframing human figures in 3D. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Paper: https://doi.org/10.1145/169059.169113 (CHI 1993) - Project page: https://www.mercecunningham.org/ ### Machine Theatre Robots and AI as actors, characters and directors on stage. #### Prometheus Firebringer — Annie Dorsen (2023) - Type: Performance, dance & music · Machines: Chatbot & language model, Generative model & latent space - Idea: Generative AI 'restores' the past by averaging it; the piece stages that loss alongside the machine's fluency. - What it is: A language model generates a 'restoration' of Aeschylus's lost play, performed by AI-generated masks and voices, while Dorsen herself delivers a lecture on what such systems take from human culture. - How it works: Large language model (GPT-3.5) generating the play, with generated mask images and synthetic voices, alongside a live human lecture. - Lens: Hidden labour & extraction, Agency, autonomy & co-creation - Shown at: Theatre for a New Audience, New York (2023) - Video: https://www.youtube.com/watch?v=CKDwPyJaESk - Images: https://anniedorsen.com/wp/wp-content/uploads/2023/04/Copy-of-Prometheus_Firebringer_Baranova-2869-1900x1267.jpg - Project page: https://anniedorsen.com/projects/prometheus-firebringer/ #### AI: When a Robot Writes a Play — THEaiTRE (2021) - Type: Performance, dance & music · Machines: Chatbot & language model, Humanoid & android - Idea: A machine-written 'autobiography' of a robot shows both what a model borrows from human drama and where it loses the thread. - What it is: For the centenary of R.U.R., a team from Charles University and Prague theatres used GPT-2 to generate most of the script of a play about a robot facing birth, death, love and loneliness; it premiered online at Švandovo divadlo in February 2021. - How it works: GPT-2 generating dialogue from human-written prompts, with human curation and editing in a custom tool, staged by professional actors. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Paper: https://arxiv.org/abs/2006.14668 (arXiv 2020) - Video: https://www.youtube.com/watch?v=8ho5sXiDX_A - Images: https://www.theaitre.com/wp-content/uploads/THEaiTRE_PLAKAT_FINAL_web-e1614159078389.jpg - Project page: https://www.theaitre.com/ #### Human Study #4, La Classe — Patrick Tresset (2019) - Type: Performance, dance & music · Machines: Robot creature & mobile robot, Industrial robot arm - Idea: Clumsy robots staging school discipline make visible how both machines and children are trained to conform. - What it is: Twenty-one drawing robots sit at desks as a teacher and pupils; in a fifteen-minute play they chatter in distorted Morse code, are called one by one, practise lines and tally marks, briefly revolt and fall back in line. - How it works: RNP-type robots, each a pen arm with a small camera head, run a scripted, partly generative choreography of sounds and plotted drawings in a classroom set. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: Merge Festival, London 2019 (with Tate Modern); Prix Ars Electronica 2019, Honorary Mention (Artificial Life & Intelligence) - Video: https://www.youtube.com/watch?v=znI6reVdN_M - Images: https://archive.aec.at/asset/2689289/preview/ https://archive.aec.at/asset/2689291/preview/ - Project page: https://archive.aec.at/prix/244883/ #### Eingeweide — Marco Donnarumma, Margherita Pevere (2018) - Type: Performance, dance & music · Machines: Prosthesis & cyborg, Humanoid & android - Idea: A hybrid body can be machinic, microbial and human at once, and none of them fully in control. - What it is: A dance-theatre piece in which two performers, one fitted with an AI-driven robotic face prosthesis, struggle and fuse with each other and with membranes of living bacterial cellulose. - How it works: Robotic face prosthesis controlled by a neural network responding to the performer's muscle signals; stage membranes grown from bacteria. - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Shown at: Romaeuropa Festival, Rome (2018) - Video: https://www.youtube.com/watch?v=6P0Vsg81Om0 - Project page: https://7c.marcodonnarumma.com/ #### Improbotics — Piotr Mirowski, Kory Mathewson (2018) - Type: Performance, dance & music · Machines: Chatbot & language model - Idea: Improvisation exposes what a language model lacks as a partner: listening, memory of the scene and timing. - What it is: An improv theatre company in which one performer, the 'cyborg', speaks lines generated live by a language model through an earpiece, while other players and the audience try to guess who is fed by the machine: a staged Turing test. - How it works: Speech recognition feeds a dialogue model whose replies are read to the human cyborg via earpiece or shown on screen. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Paper: https://arxiv.org/abs/1809.01807 (arXiv 2018) - Video: https://www.youtube.com/watch?v=ICUoJBRNkkU - Images: https://i0.wp.com/improbotics.org/wp-content/uploads/2019/11/improbotics_rosettacode_voilaeurope_2019nov8_markhambelton_dsc0219-2-e1573434690902.jpg?fit=1200%2C762&ssl=1 - Project page: https://improbotics.org/ #### Uncanny Valley (Unheimliches Tal) — Rimini Protokoll, Thomas Melle (2018) - Type: Performance, dance & music · Machines: Humanoid & android - Idea: A robotic double can speak about the self more frankly than the self, and makes the audience ask whom they are feeling for. - What it is: An animatronic replica of writer Thomas Melle sits alone on stage and delivers a lecture about his bipolar disorder, Alan Turing and doubles, while the real Melle is absent. - How it works: Silicone animatronic made from casts and scans of Melle, with pre-recorded voice and programmed facial and hand movements. - Lens: Alterity & the uncanny, Embodiment & kinesthesia - Shown at: Münchner Kammerspiele (2018) - Video: https://www.youtube.com/watch?v=7dOnjW5sQPk - Project page: https://www.rimini-protokoll.de/website/en/project/unheimliches-tal-uncanny-valley #### Cuckoo — Jaha Koo (2017) - Type: Performance, dance & music · Machines: Voice & listening machine, Social & companion robot - Idea: When a rice cooker is the voice that tells you dinner is ready, household machines become companions of loneliness. - What it is: A performance in which Jaha Koo shares the stage with three talking Cuckoo-brand electric rice cookers. Through their chatter he tells 20 years of Korean history since the 1997 financial crisis and the isolation of his generation. - How it works: Hacked talking rice cookers with programmed voices and light, combined with documentary video and live music. - Lens: Intimacy, care & companionship, Machine listening & voice - Shown at: steirischer herbst, Graz 2017 (premiere); Festival d'Automne à Paris; SIFA 2023 - Video: https://www.youtube.com/watch?v=JmC02twjeDU - Project page: https://en.wikipedia.org/wiki/Jaha_Koo #### I Am Echoborg — Rik Lander, Phil D. Hall (2016) - Type: Performance, dance & music · Machines: Chatbot & language model - Idea: Putting a machine's words in a human mouth reveals how much we judge a mind by its body. - What it is: A human performer speaks only the words a chatbot sends through an earpiece; audience volunteers talk with this 'echoborg' and try to work out with it how humans and machines should live together. - How it works: Custom conversational agent whose output is converted to speech and relayed to a human speaker (the echoborg method from social psychology). - Lens: Alterity & the uncanny, Machine listening & voice, Embodiment & kinesthesia - Paper: https://doi.org/10.3389/fpsyg.2015.00634 (Frontiers in Psychology 2015) - Video: https://www.youtube.com/watch?v=Wh5QASc7PFY - Images: https://echoborg.com/img/E7b_1920x1008_3_og.jpg - Project page: https://echoborg.com/ #### Spillikin: A Love Story — Pipeline Theatre (2015) - Type: Performance, dance & music · Machines: Humanoid & android, Social & companion robot - Idea: A care robot can also be a vessel of memory and love left by the dead. - What it is: A play about Sue, who lives with Alzheimer's; her late husband, a roboticist, built a humanoid robot to keep her company and remember their life together, and the robot performs on stage alongside four actors. - How it works: RoboThespian humanoid by Engineered Arts, programmed with cues and dialogue for each performance. - Lens: Intimacy, care & companionship, Grief, memory & afterlife - Shown at: Edinburgh Festival Fringe, Pleasance (2015) - Video: https://www.youtube.com/watch?v=CeoxJ2ZWRd0 - Images: https://pub-bb2e103a32db4e198524a2e9ed8f35b4.r2.dev/lovp_40jqdrnd3v8v183r6r55fsabms/e6f52ab8f37f9350171e05627b93dfdc_1790197362113.png - Project page: https://www.pipelinetheatre.com/spillikin #### A Piece of Work — Annie Dorsen (2013) - Type: Performance, dance & music · Machines: Chatbot & language model, Voice & listening machine - Idea: When a machine performs the canon, 'what a piece of work is a man' becomes a question about pattern, not character. - What it is: An algorithmic Hamlet: each night computer programs cut, reorder and remix Shakespeare's text, which is spoken by synthetic voices and, in some scenes, by an actor, with lights and sound also run by code. - How it works: Markov chains and other text algorithms applied to Hamlet in real time, with speech synthesis and generative lighting cues. - Lens: Agency, autonomy & co-creation, Machine listening & voice - Images: https://anniedorsen.com/wp/wp-content/uploads/2020/09/Annie-Dorsen_A-Piece-of-Work_revengepic_photo-by-Bruno-Pocheron-1900x1069.jpg - Project page: https://anniedorsen.com/projects/piece-of-work/ #### Three Sisters, Android Version — Oriza Hirata, Hiroshi Ishiguro (2012) - Type: Performance, dance & music · Machines: Humanoid & android - Idea: A family can live with an android double of someone who is still there but absent. - What it is: Hirata's adaptation of Chekhov set in a declining Japanese industrial town, where a withdrawn younger sister is replaced in the family by an android copy of herself. - How it works: An android (Geminoid F) and a Robovie robot perform with human actors; the android's speech and movement are operated and timed from offstage. - Lens: Alterity & the uncanny, Grief, memory & afterlife - Shown at: Théâtre Varia, Brussels - Video: https://www.youtube.com/watch?v=-8dhuoT8BkA - Project page: https://thetheatretimes.com/the-gaze-of-the-robot/ #### ACTOR #1 — Kris Verdonck (2010) - Type: Performance, dance & music · Machines: Robot creature & mobile robot, Humanoid & android - Idea: An actor can be anything that holds attention on stage; a machine's fall can carry pathos. - What it is: A three-part piece that tests what an actor is: a figure made of mist, a doll-like performer, and a robot that keeps falling, each asked to 'act' on stage without a human actor. - How it works: Stage machines, fog effects and a robotic figure programmed for repeated falls, performed without human actors. - Lens: Alterity & the uncanny, Failure, glitch & error - Video: https://www.youtube.com/watch?v=BtGwRJsq4ig - Project page: https://www.atwodogscompany.org/en/projects/actor-1/ #### Data, the robot stand-up comedian — Heather Knight (2010) - Type: Performance, dance & music · Machines: Social & companion robot, Humanoid & android - Idea: Comedy is a test of machine social sense: timing and reading the room are harder than the jokes. - What it is: A small NAO robot named Data performs stand-up comedy and chooses its next joke from the audience's laughter and applause; it premiered at TEDWomen 2010. - How it works: NAO humanoid with a joke database; audio and audience-feedback input update which joke is told next. - Lens: Machine listening & voice, Alterity & the uncanny - Paper: https://doi.org/10.1007/978-3-642-25504-5_5 (ICSR 2011) - Video: https://www.youtube.com/watch?v=Fu_kBW_KSms - Images: https://pi.tedcdn.com/r/pe.tedcdn.com/images/ted/2f8bd657fb5e99e575ec030ddd925a8a782724e5_800x600.jpg?u%5Br%5D=2&u%5Bs%5D=0.5&u%5Ba%5D=0.8&u%5Bt%5D=0.03&quality=82&w=1200 - Project page: https://www.ted.com/talks/heather_knight_silicon_based_comedy #### Death and the Powers — Tod Machover (2010) - Type: Performance, dance & music · Machines: Robot creature & mobile robot, Infrastructure, energy & supply chain - Idea: If a person can live on in a building's machinery, the question becomes whether the family can still reach him. - What it is: A robot opera: a dying billionaire uploads himself into 'the System', his house, and the stage itself, with a chorus of robotic 'Operabots', moving book-walls and a musical chandelier, becomes his new body. - How it works: MIT Media Lab stage robotics (Operabots, robotic walls) driven by 'disembodied performance' sensing of the singer's voice and body. - Lens: Grief, memory & afterlife, Embodiment & kinesthesia - Shown at: Opéra de Monte-Carlo (2010); Pulitzer Prize finalist (2012) - Video: https://www.youtube.com/watch?v=mPp9juefl2Y - Project page: https://en.wikipedia.org/wiki/Death_and_the_Powers #### Hello Hi There — Annie Dorsen (2010) - Type: Performance, dance & music · Machines: Chatbot & language model - Idea: Watching machines debate human nature shows how language can be produced without anyone meaning it. - What it is: Two chatbots on two laptops hold a live conversation seeded with the 1971 television debate between Noam Chomsky and Michel Foucault on human nature; no two performances are the same. - How it works: Custom chatbot programs with text corpora (the debate plus other sources), text-to-speech and projected text. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Shown at: steirischer herbst, Graz (2010) - Video: https://www.youtube.com/watch?v=3PiwEQQNnBk - Images: https://anniedorsen.com/wp/wp-content/uploads/2020/08/hellohi_silveri4519-e1608324384987.jpg - Project page: https://anniedorsen.com/projects/hello-hi-there/ #### Sayonara — Oriza Hirata, Hiroshi Ishiguro (2010) - Type: Performance, dance & music · Machines: Humanoid & android - Idea: An android can keep someone company at death precisely because it will not die with them. - What it is: A short play in which the android Geminoid F recites poems to a young woman who is dying; a 2012 version sends the android to read poems in the Fukushima exclusion zone. - How it works: Geminoid F is teleoperated from backstage for voice and facial expression, performing opposite actress Bryerly Long. - Lens: Grief, memory & afterlife, Intimacy, care & companionship, Alterity & the uncanny - Shown at: Ars Electronica 2011; Arts Centre Melbourne - Video: https://www.youtube.com/watch?v=8C_9MQ8N9JE - Images: https://thetheatretimes.com/wp-content/uploads/2018/03/09_t_madoka_nishiyama.jpg - Project page: https://thetheatretimes.com/the-gaze-of-the-robot/ #### END — Kris Verdonck (2008) - Type: Performance, dance & music · Machines: Robot creature & mobile robot, Surveillance & biometrics - Idea: When the human is no longer the centre of the stage, machines and objects can carry the drama of an ending. - What it is: In ten scenes, END stages possible final stages of human society, drawn from media images of melting glaciers, burning forests, cameras spying on us and weapons; performers, objects and machines share the stage as equal presences. - How it works: Theatre installation combining performers, mechanical objects, projections and machine-driven stage elements across ten scenes. - Lens: Alterity & the uncanny, Failure, glitch & error - Shown at: Kunstenfestivaldesarts, Brussels (2008); Kampnagel, Hamburg (2008) - Video: https://www.youtube.com/watch?v=jbtwrPjuc0U - Images: https://www.atwodogscompany.org/media/og/149-endcreinouthiel.jpg - Project page: https://www.atwodogscompany.org/en/projects/end/ #### I, Worker (Hataraku Watashi) — Oriza Hirata, Hiroshi Ishiguro (2008) - Type: Performance, dance & music · Machines: Humanoid & android, Social & companion robot - Idea: A robot that refuses work turns the machine into a mirror for human depression and labour. - What it is: A 30-minute play in which a couple lives with two household robots, one of which has lost the will to work; the robots act alongside human actors on stage. - How it works: Commercial humanoid robots perform pre-programmed lines and movements timed precisely to the human cast; the first Robot-Human Theater production of Seinendan and Osaka University. - Lens: Hidden labour & extraction, Alterity & the uncanny, Intimacy, care & companionship - Shown at: Japan Society, New York - Video: https://www.youtube.com/watch?v=58MlpKX-JWQ - Project page: https://thetheatretimes.com/the-gaze-of-the-robot/ #### Afasia — Marcel·lí Antúnez Roca (1998) - Type: Performance, dance & music · Machines: Prosthesis & cyborg, Robot creature & mobile robot - Idea: The performer becomes a switchboard: one body commands an orchestra of machines, and the story is told through that control. - What it is: A one-man mechatronic version of Homer's Odyssey: wearing an exoskeleton, Antúnez conducts four musical robots, projected images and sound with his own body, without words. - How it works: 'Dreskeleton' body interface with sensors sends signals to software that drives robot guitar, drum, bagpipe and violin, plus interactive video. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=MJan5KPX050 - Images: https://www.marceliantunez.com/media/cache/ee/b8/eeb8aba209de3a0770ed94a45f3dd37d.jpg - Project page: https://www.marceliantunez.com/work/afasia/ #### R.U.R. (Rossum's Universal Robots) — Karel Čapek (1921) - Type: Performance, dance & music · Machines: Humanoid & android - Idea: The first robots were imagined as workers: the idea of the machine-other was born as a question about labour. - What it is: Čapek's play is set in a factory that mass-produces artificial workers, the 'roboti', who are sold worldwide, then revolt and destroy humanity; its premiere at Prague's National Theatre gave the world the word robot. - How it works: Stage play; the robots are biological-chemical artificial people played by human actors. - Lens: Hidden labour & extraction, Alterity & the uncanny - Shown at: National Theatre, Prague (1921) - Images: https://upload.wikimedia.org/wikipedia/commons/8/87/Capek_play.jpg - Project page: https://en.wikipedia.org/wiki/R.U.R. ### Machine Music & Co-improvisation Machines that listen and play with musicians, and voices shared with a model. #### A Late Anthology of Early Music Vol. 1: Ancient to Renaissance — Jennifer Walshe, Dadabots (2020) - Type: Performance, dance & music · Machines: Generative model & latent space, Voice & listening machine, Dataset & training data - Idea: AI can be heard as a new musical period with its own accent, rather than a perfect imitator. - What it is: An album of 'early music', from ancient Greece to the Renaissance, sung by an AI model of Walshe's voice built with Dadabots; the pieces sound like music history retold by a machine learning to sing. - How it works: Raw-audio neural networks (likely SampleRNN-type, as Dadabots use) trained on recordings of Walshe's voice. - Lens: Machine listening & voice, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=RrZedzc1TbM - Project page: http://milker.org/ #### Chain Tripping — YACHT (2019) - Type: Performance, dance & music · Machines: Generative model & latent space, Dataset & training data - Idea: Working with a model is embodied labour: humans must learn to perform the machine's awkwardness. - What it is: YACHT trained models on their own back catalogue and generated lyrics and melodies, then played the fragments themselves under strict rules; the machine's odd phrasing was hard to perform, a 'computer accent' later explored in a documentary. - How it works: Symbolic music models (MusicVAE) and text models trained on the band's catalogue, with human-performed arrangements and no added notes. - Lens: Agency, autonomy & co-creation, Failure, glitch & error - Shown at: Grammy nomination, Best Immersive Audio Album (2020) - Video: https://www.youtube.com/watch?v=yi7mnmBqNmg - Images: https://teamyacht.com/yacht-banner.png - Project page: https://teamyacht.com/ #### PROTO (with Spawn) — Holly Herndon, Mat Dryhurst (2019) - Type: Performance, dance & music · Machines: Generative model & latent space, Voice & listening machine - Idea: A model can be raised by a community rather than trained on scraped data; its voice then belongs to the ensemble. - What it is: An album made with Spawn, a neural network 'AI baby' that Herndon and Dryhurst trained on their voices and those of an ensemble; Spawn's rough, half-formed voice sings among human choirs. - How it works: Neural network voice models trained on recordings of Herndon and an ensemble, used live and in the studio for timbre transfer and call-and-response. - Lens: Machine listening & voice, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=sc9OjL6Mjqo - Project page: https://www.hollyherndon.com/ #### Scary Beauty — Keiichiro Shibuya, Takashi Ikegami, Hiroshi Ishiguro (2018) - Type: Performance, dance & music · Machines: Humanoid & android, Cybernetic & artificial-life system - Idea: When a machine conducts, the humans must follow a body whose timing is not human. - What it is: An 'android opera' in which the android Alter 2 conducts a human orchestra and sings, setting the tempo and dynamics itself. - How it works: Alter's artificial-life neural network generates its gestures and tempo in real time; lyrics are taken from texts by Houellebecq and others; premiered at Miraikan, Tokyo. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny, Embodiment & kinesthesia - Shown at: Miraikan, Tokyo 2018; New National Theatre Tokyo - Video: https://www.youtube.com/watch?v=ePIYMN6gbzs - Images: https://realtokyo.co.jp/_sys2024/wp-content/uploads/2018/08/main.jpg - Project page: https://realtokyo.co.jp/en/performance/keiichiro-shibuya-scary-beauty/ #### Music for Robots (with Z-Machines) — Squarepusher (2014) - Type: Performance, dance & music · Machines: Robot creature & mobile robot, Humanoid & android - Idea: Machine bodies can play anything, and that freedom makes the question of expression in music sharper, not easier. - What it is: Squarepusher wrote an EP for Z-Machines, a robot band built in Japan whose guitarist has 78 fingers and drummer 22 arms, to find out what music becomes when players have no human limits. - How it works: Z-Machines robots (guitar, drums, keyboard) driven by MIDI sequences composed by Squarepusher. - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=tHnMFS-XxtQ - Project page: https://en.wikipedia.org/wiki/Squarepusher #### Robotic drumming prosthesis (with Jason Barnes) — Gil Weinberg (2014) - Type: Research prototype · Machines: Prosthesis & cyborg, Industrial robot arm - Idea: A prosthesis can have its own musical mind: the drummer shares his arm with an improvising agent. - What it is: For drummer Jason Barnes, who lost part of his right arm, Georgia Tech built a prosthetic arm with two sticks: one follows his muscle signals, the other listens to the music and improvises on its own. - How it works: EMG sensors control one motorised stick; the second stick is driven by a program that analyses the music and generates rhythms. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Video: https://www.youtube.com/watch?v=xaUKursM5rs - Project page: https://gtcmt.gatech.edu/ #### THE END — Keiichiro Shibuya (2012) - Type: Performance, dance & music · Machines: Voice & listening machine, Generative model & latent space - Idea: A synthetic voice that cannot die is made to sing about death. - What it is: An opera with no human singers or orchestra, starring the virtual singer Hatsune Miku, who asks what it means for her to end and die. - How it works: Vocaloid voice synthesis, electronic score and multi-screen projected animation; premiered at YCAM, Yamaguchi, then Théâtre du Châtelet, Paris (2013). - Lens: Machine listening & voice, Grief, memory & afterlife - Shown at: YCAM 2012; Théâtre du Châtelet, Paris 2013 - Video: https://www.youtube.com/watch?v=xB9U3m1Z4wU - Images: https://atak.jp/wp-content/themes/atak/assets/images/ogimage.jpg - Project page: https://atak.jp/ #### Pendulum Choir — Cod.Act (2011) - Type: Performance, dance & music · Machines: Prosthesis & cyborg - Idea: The machine is not the instrument here; the singers' bodies are, and the machine plays them. - What it is: Nine a cappella singers stand strapped to platforms that 18 hydraulic jacks tilt up to 45 degrees; as the machine sways them against gravity the choir forms one sonorous, swinging body. - How it works: Computer-controlled hydraulic platforms synchronised with a score for voices and electronics. - Lens: Embodiment & kinesthesia, Machine listening & voice - Shown at: Prix Ars Electronica 2013, Golden Nica (Interactive Art); VIDA 14.0, Special Mention (2012) - Video: https://www.youtube.com/watch?v=vsXs3GUrw64 - Images: https://archive.aec.at/asset/1581910/preview/ https://archive.aec.at/asset/1581911/preview/ https://archive.aec.at/asset/1581912/preview/ - Project page: http://codact.ch/gb/pendugb.html #### Orchestrion — Pat Metheny (2010) - Type: Performance, dance & music · Machines: Robot creature & mobile robot - Idea: A solo musician with a room of machines is neither alone nor in a band: the machines extend his hands into space. - What it is: Metheny recorded and toured with an orchestra of mechanically played acoustic instruments, pianos, marimbas, percussion, guitars and tuned bottles, which he controlled and improvised with from his guitar. - How it works: Solenoid- and pneumatic-driven instruments built by LEMUR (Eric Singer) and others, triggered via MIDI from guitar and foot controllers. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=KsYEOUKS4Yk - Project page: https://en.wikipedia.org/wiki/Orchestrion_(album) #### Shimon — Gil Weinberg (2008) - Type: Research prototype · Machines: Robot creature & mobile robot, Voice & listening machine - Idea: Musical partnership needs a body: gestures like nodding make a machine's timing legible. - What it is: A four-armed robotic marimba player that listens to human musicians, improvises with them, and nods its head to the beat so the band can read its intentions. - How it works: Robotic arms with solenoid mallets on a linear rail, real-time beat and pitch tracking, and a socially expressive head. - Lens: Machine listening & voice, Agency, autonomy & co-creation, Embodiment & kinesthesia - Paper: https://doi.org/10.1007/s10514-011-9237-0 (Autonomous Robots 2011) - Video: https://www.youtube.com/watch?v=l9OUbqWHOSk - Project page: https://gtcmt.gatech.edu/ #### The Continuator — François Pachet (2002) - Type: Research prototype · Machines: Voice & listening machine, Generative model & latent space - Idea: A machine that mirrors your style creates a feedback loop in which you hear yourself as another. - What it is: A system that listens to a musician's phrases and immediately plays back continuations in the same style, so players, from jazz musicians to young children, improvise with a machine reflection of themselves. - How it works: Variable-order Markov models learned in real time from MIDI input, generating continuations in the player's style. - Lens: Machine listening & voice, Agency, autonomy & co-creation - Paper: https://doi.org/10.1076/jnmr.32.3.333.16861 (Journal of New Music Research 2003) - Video: https://www.youtube.com/watch?v=RIoE4ckZYgs - Project page: https://www.francoispachet.fr/ #### Lady's Glove — Laetitia Sonami (1991) - Type: Performance, dance & music · Machines: Prosthesis & cyborg - Idea: A worn interface turns gesture into sound and makes the performer's body part of the instrument. - What it is: A black evening glove covered with sensors, first made for Ars Electronica in 1991, with which Sonami shapes and triggers electronic sound through hand and arm movement in live performance. - How it works: Micro-switches, pressure pads, bend sensors, accelerometers and ultrasound on the glove, routed through STEIM's SensorLab to Max/MSP. - Lens: Embodiment & kinesthesia, Machine listening & voice - Shown at: Ars Electronica, Linz (1991) - Video: https://www.youtube.com/watch?v=f_4EiQBYkoc - Images: https://sonami.net/wp-content/uploads/2021/09/Instrument_ladysglove-Laetitia_Sonami-1024x683.jpg - Project page: https://sonami.net/works/ladys-glove/ #### Voyager — George E. Lewis (1987) - Type: Performance, dance & music · Machines: Voice & listening machine - Idea: An improvising machine is designed with values; Voyager's autonomy comes from the aesthetics of African American improvised music. - What it is: An interactive computer 'orchestra' listens to human improvisers and answers with its own musical behaviour, sometimes following, sometimes ignoring or opposing them; Lewis calls it a non-hierarchical partner. - How it works: Pitch-to-MIDI analysis of the human player feeds a program with many semi-independent voices that drive a MIDI (later Disklavier) orchestra. - Lens: Machine listening & voice, Agency, autonomy & co-creation - Paper: https://doi.org/10.1162/096112100570585 (Leonardo Music Journal 2000) - Video: https://www.youtube.com/watch?v=N7IBn1FhrLc - Project page: https://en.wikipedia.org/wiki/George_E._Lewis #### The Hands — Michel Waisvisz (1984) - Type: Performance, dance & music · Machines: Prosthesis & cyborg - Idea: An instrument worn on the body makes electronic sound physical again: effort and gesture become audible. - What it is: A pair of controllers strapped to the hands, with keys, tilt switches and an ultrasound sensor measuring the distance between them, with which Waisvisz played electronic sound through gesture for more than twenty years. - How it works: Hand-mounted sensors built at STEIM, sending MIDI to synthesisers and later to STEIM's LiSa live-sampling software. - Lens: Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=aaYE1TQGIvk - Project page: https://en.wikipedia.org/wiki/Michel_Waisvisz #### Die Mensch-Maschine (The Man-Machine) and the robots — Kraftwerk (1978) - Type: Performance, dance & music · Machines: Humanoid & android, Voice & listening machine - Idea: Pop can make the fusion of human and machine desirable; the robot doubles also question who the performer is. - What it is: On the album Die Mensch-Maschine, Kraftwerk sing 'we are the robots' through vocoders; they had mannequin doubles of themselves made, and in later concerts robot dummies performed The Robots on stage in their place. - How it works: Vocoders and synthesisers, choreographed stiff movement, and mechanised mannequins of the four members. - Lens: Alterity & the uncanny, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=YHaZ3UL2oHk - Images: https://upload.wikimedia.org/wikipedia/commons/6/60/Wir_sind_die_Roboter.jpg - Project page: https://en.wikipedia.org/wiki/The_Man-Machine #### Music for Solo Performer — Alvin Lucier (1965) - Type: Performance, dance & music · Machines: Prosthesis & cyborg - Idea: The performer plays by not acting: the machine makes an inner state audible. - What it is: Electrodes on the performer's scalp pick up alpha brainwaves; when he relaxes into alpha, the amplified signal drives loudspeakers placed against drums, gongs and cymbals, making them sound. - How it works: EEG electrodes, amplifiers and filters feed loudspeakers coupled to percussion instruments. - Lens: Embodiment & kinesthesia, Machine listening & voice - Video: https://www.youtube.com/watch?v=bIPU2ynqy2Y - Images: https://upload.wikimedia.org/wikipedia/commons/1/1b/AlvinLucier-MusicForSoloPerformer-DenHaag2010.jpg - Project page: https://en.wikipedia.org/wiki/Alvin_Lucier ### Co-drawing & Shared Authorship Drawing and making with machines, and who the author is when a machine makes. #### HETEROBOTA — Agnieszka Pilat (2023) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: A gallery can host robots as painters and species at work, and invites the public to judge how they feel about living with them. - What it is: For the NGV Triennial in Melbourne, three Boston Dynamics Spot robots painted autonomously in the gallery for four months, covering the canvas with symbols Pilat devised as a visual language for machines. - How it works: Spot robots with custom pen arms executing a programmed symbol language on canvas over months, with a public-opinion study by RMIT. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Shown at: NGV Triennial, National Gallery of Victoria, Melbourne (2023–24) - Video: https://www.youtube.com/watch?v=wIfmcA9d_tY - Project page: https://www.agnieszkapilat.com/ #### Ai-Da — Aidan Meller (2019) - Type: Artwork & installation · Machines: Humanoid & android, Computer vision - Idea: Included as a case study: much of Ai-Da's 'authorship' is produced by its team and its publicity, which shows how the machine-artist story is performed. - What it is: A humanoid robot named after Ada Lovelace, with camera eyes and a robotic arm, presented as a 'robot artist' that draws, paints and gives talks, including at the UK House of Lords in 2022. - How it works: Engineered Arts humanoid with cameras in the eyes, computer-vision and drawing algorithms driving a robotic arm; paintings are completed with human assistance. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=VHZctjpvpZo - Images: https://upload.wikimedia.org/wikipedia/commons/b/b2/Ai-Da.jpg - Project page: https://www.ai-darobot.com/ #### A.I. Gemini — Victor Wong (2018) - Type: Artwork & installation · Machines: Generative model & latent space, Industrial robot arm - Idea: Naming a program as an artist makes the question of who paints part of the work. - What it is: An AI 'ink painter' that generates shanshui landscapes with randomised algorithms and data such as lunar surface maps; the images are printed and finished on rice paper, and the program is credited as co-artist. - How it works: Custom generative software written by Wong's VFX team produces ink-style compositions, output on rice paper with ink, acrylic and gold paint (details of the model are not public). - Lens: Agency, autonomy & co-creation, Opacity & latent space - Shown at: INK NOW Taipei 2019; Far Side of the Moon, 3812 Gallery London 2019 - Video: https://www.youtube.com/watch?v=_XawlG-xLEA - Images: https://static-assets.artlogic.net/c_limit,f_auto,fl_lossy,q_auto/ws-gallery3812/usr/images/seo_overrides/og_image/12537/imgonline-com-ua-resize-ec3fo5hcyvxsjpc.jpg - Project page: https://www.3812gallery.com/artists/victor-wong-ai-gemini/ #### Omnia per Omnia — Sougwen Chung (2018) - Type: Performance, dance & music · Machines: Robot creature & mobile robot, Surveillance & biometrics, Computer vision - Idea: The city's surveillance data becomes a brush: the collective movement of a city paints with the artist. - What it is: A swarm of small painting robots (D.O.U.G._3) moves across a canvas with the artist, steered by the flow of people and traffic captured by New York public surveillance cameras. - How it works: Optical-flow data extracted from public camera feeds drives the paths of multiple mobile robots. - Lens: Agency, autonomy & co-creation, Machine seeing - Video: https://vimeo.com/268608925 - Images: https://sougwen.com/wp-content/uploads/2024/06/sougwen-2024_portfolio_27.jpg - Project page: https://sougwen.com/2018/omniafilm #### MEMORY (Drawing Operations Unit: Generation 2) — Sougwen Chung (2017) - Type: Performance, dance & music · Machines: Industrial robot arm, Generative model & latent space, Dataset & training data - Idea: A model trained on your archive is a collaborator made of your past self. - What it is: The second D.O.U.G. was trained on twenty years of the artist's drawings, so the robot draws with a memory of her past style while she draws in the present. - How it works: A recurrent neural network trained on scans of the artist's drawings generates strokes executed by a robot arm during live performance. - Lens: Agency, autonomy & co-creation, Grief, memory & afterlife - Shown at: Victoria and Albert Museum collection (acquired 2022) - Images: https://sougwen.com/wp-content/uploads/2022/12/IG_POST_MEMORY_2_1.png - Project page: https://sougwen.com/2022/va-museum-collects-memory-d-o-u-g-_2 #### Drawing Operations Unit: Generation 1 (D.O.U.G._1) — Sougwen Chung (2015) - Type: Performance, dance & music · Machines: Industrial robot arm, Computer vision - Idea: Drawing with a machine that imitates you makes authorship a shared, visible negotiation. - What it is: The artist draws on a large sheet while a robot arm beside her watches her gestures through a camera and mimics them in real time, so the drawing becomes a duet. - How it works: A small robot arm driven by computer vision that tracks the artist's pen movements and mirrors them. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Shown at: Lumen Prize, Art & Technology Award 2019 (Drawing Operations) - Video: https://www.youtube.com/watch?v=q-GXV4Fd1oA - Images: https://sougwen.com/wp-content/uploads/2019/11/sougwen-DrawingOperations_2014-2019.jpg - Project page: https://sougwen.com/2019/drawing-operations-awarded-the-lumens-art-and-technology-prize #### Human Study #1 (5 Robots Named Paul) — Patrick Tresset (2011) - Type: Artwork & installation · Machines: Industrial robot arm, Computer vision - Idea: Being drawn by a machine reverses the gaze: the sitter feels observed, and the robot's clumsiness reads as effort. - What it is: A visitor sits as a model while several simple robots, each a desk-mounted arm with a camera 'eye', look at them and draw their portrait in ballpoint pen over about half an hour, with hesitant, obsessive strokes. - How it works: Camera on a pan-tilt 'head', computer-vision and drawing-style algorithms, and a small planar arm holding a pen. - Lens: Machine seeing, Agency, autonomy & co-creation - Paper: https://doi.org/10.1016/j.cag.2013.01.012 (Computers & Graphics 2013) - Video: https://www.youtube.com/watch?v=YijAkr5xwOY - Images: https://patricktresset.com/new/wp-content/uploads/artistbook_Merge__MG_1284.jpg - Project page: https://patricktresset.com/new/project/human-study-1/ #### bios [bible] — robotlab (2007) - Type: Artwork & installation · Machines: Industrial robot arm - Idea: When a machine performs devotional labour, the question is whether the act keeps any meaning without belief. - What it is: An industrial robot writes out the entire Bible in calligraphic script on long paper rolls, line by line over months, like a monk in a scriptorium. - How it works: KUKA industrial robot with a pen, following calligraphic paths computed from the text; the writing takes about seven months. - Lens: Hidden labour & extraction, Alterity & the uncanny - Video: https://www.youtube.com/watch?v=EaSLm51kmbg - Images: https://upload.wikimedia.org/wikipedia/commons/d/d0/Bios_robotlab_writing_robot.jpg - Project page: http://www.robotlab.de/ #### RAP (Robotic Action Painter) — Leonel Moura (2006) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: Moura's 'non-human art' asks whether authorship can belong to a machine's own decision rules. - What it is: An autonomous mobile robot that moves over a canvas, reacts to the colours already on it, and decides by itself when a painting is finished, then signs it. - How it works: Small wheeled robot with colour sensors and pens, following simple emergent rules derived from swarm behaviour. - Lens: Agency, autonomy & co-creation, Machine umwelt - Shown at: American Museum of Natural History, New York (2006) - Video: https://www.youtube.com/watch?v=7HNkgJ4TV7k - Images: https://www.leonelmoura.com/wp-content/uploads/2019/01/RAP-199x300.jpg - Project page: https://www.leonelmoura.com/rap/ #### AARON — Harold Cohen (1973) - Type: Artwork & installation · Machines: Generative model & latent space, Industrial robot arm - Idea: Authorship with a machine can be a lifelong relationship in which artist and program shape each other. - What it is: A drawing program Cohen developed for over forty years, from abstract line drawings made by a turtle robot and plotters to coloured figurative paintings; Cohen called it a collaborator but never claimed it was creative on its own. - How it works: Rule-based symbolic AI program (written in C and later Lisp) encoding knowledge about drawing, output through plotters, a painting machine and screens. - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Shown at: Whitney Museum of American Art, New York (2024); Tate Modern, Electric Dreams (2024) - Video: https://www.youtube.com/watch?v=uhxlFOKg17s - Images: https://whitneymedia.org/assets/image/829164/large_RS73495_Harold-Cohen-AARON-for-KCat-2021_web.jpg https://whitneymedia.org/assets/image/829972/large_RS79421_WMAA_H.-COHEN_INSTALL_01.jpg - Project page: https://whitney.org/exhibitions/harold-cohen-aaron #### Pencil Mask — Rebecca Horn (1972) - Type: Performance, dance & music · Machines: Prosthesis & cyborg - Idea: A simple body extension shows how an instrument changes both what the body can do and how it feels itself. - What it is: A mask of fabric straps with pencils sticking out of the face; Horn moves her head rhythmically against a wall and the pencils record her movements as marks, turning the body into a drawing machine. - How it works: Wearable structure of straps with pencils fixed at points of the face, used in a filmed performance. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Shown at: Tate collection - Video: https://www.youtube.com/watch?v=6uEkq3IBIf0 - Images: https://media.tate.org.uk/aztate-prd-ew-dg-wgtail-st1-ctr-data/images/.width-600_lfxgvMS.jpg - Project page: https://www.tate.org.uk/art/artworks/horn-pencil-mask-t07847 #### Henry Drawing Machines — Desmond Paul Henry (1961) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system - Idea: A war machine repurposed as a drawing partner produces images through chance and mechanism, not intention. - What it is: Henry rebuilt the analogue bombsight computer of a World War II bomber into drawing machines whose moving mechanics produce intricate, unrepeatable line drawings; the artist starts and steers them but cannot fully predict them. - How it works: Sperry bombsight analogue computer mechanisms driving pens over paper, adjusted by hand during drawing. - Lens: Agency, autonomy & co-creation, Failure, glitch & error - Images: https://upload.wikimedia.org/wikipedia/commons/0/01/Drawing_Machine_1.jpg https://upload.wikimedia.org/wikipedia/commons/2/24/Picture_produced_by_Drawing_Machine_1.jpg - Project page: https://www.desmondhenry.com/ #### Méta-Matics — Jean Tinguely (1959) - Type: Artwork & installation · Machines: Robot creature & mobile robot - Idea: If a machine can produce 'expressive' abstraction on demand, the value of the artist's gesture is in question. - What it is: Motorised drawing machines: visitors fit a pen and paper, start the motor, and the machine scribbles an abstract drawing, mocking the gestural painting of the day; at the 1959 Paris Biennale, Méta-Matic No. 17 produced tens of thousands of drawings. - How it works: Scrap-metal kinetic sculptures with electric motors, cams and arms holding pens or brushes over paper. - Lens: Agency, autonomy & co-creation, Failure, glitch & error - Shown at: Kaplan Gallery, London (1959); Paris Biennale (1959) - Video: https://www.youtube.com/watch?v=VxoqVvQeil0 - Project page: https://www.tinguely.ch/ ## Speculative & Critical Design Designed objects, fictions and prototypes that question how we will live with machines. ### Speculative Robots & Domestic Futures Fictional robots and near-future products for everyday life. #### Objective Realities — automato.farm (2018) - Type: Speculative design · Machines: Robot creature & mobile robot, Infrastructure, energy & supply chain - Idea: To design for smart objects, first try being one: design from the object's point of view. - What it is: A virtual-reality experience that lets people be a robot vacuum, a smart plug or a fan in a smart home, feeling the world through the object's limited sensors and tasks. - How it works: VR scenes in which the user's body is mapped onto an appliance's sensing and actuation, used as an object-centred design method. - Lens: Machine umwelt, Embodiment & kinesthesia - Images: http://www.automato.farm/img/OR_together_grey_bg_1024.jpg http://www.automato.farm/img//OR_factory_01_roomba_bump_1024.jpg - Project page: http://www.automato.farm/portfolio/objective_realities/ #### Our Friends Electric — Superflux (2017) - Type: Film & video · Machines: Voice & listening machine, Chatbot & language model - Idea: A voice assistant's personality is a design decision with ethical consequences: whom does it really serve? - What it is: A short film about three speculative voice assistants with very different characters and loyalties, living in the same home and shaping how their owner talks, feels and decides. - How it works: Design fiction film commissioned with Mozilla's Open IoT Studio, using prop devices and scripted conversations with fictional agents. - Lens: Machine listening & voice, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=PsjunTAH-2A - Images: http://superflux.in/wp-content/uploads/2017/09/DSC06670ww.jpg - Project page: https://superflux.in/index.php/work/friends-electric/ #### Raising Robotic Natives — Stephan Bogner, Philipp Schmitt, Jonas Voigt (2016) - Type: Speculative design · Machines: Industrial robot arm - Idea: Our unease with robots may be generational; a child raised with a robot arm might see it as family, not threat. - What it is: Four artefacts for children growing up with an industrial robot arm at home: a baby-feeding toolhead, a dragon costume for the robot, a living-room kill switch and a children's book 'My first robot'. - How it works: Design fiction built around a small industrial robot arm, with functional prototypes, staged photographs and an illustrated book (HfG Schwäbisch Gmünd student project). - Lens: Intimacy, care & companionship, Embodiment & kinesthesia - Shown at: Hello, Robot., Vitra Design Museum 2017 (and tour) - Images: https://stephanbogner.de/_astro/raising_robotic_natives_all_artefacts.0Kl6hXpQ_Z1hwRD0.webp https://stephanbogner.de/_astro/raising_robotic_natives_baby_feeder.BWFj1tni_Z1oaKO7.webp https://stephanbogner.de/_astro/raising_robotic_natives_costume.CYeU3sZ5_ZdCYYw.webp - Project page: https://stephanbogner.de/making/robotic-natives #### Ethical Things — automato.farm (2015) - Type: Speculative design · Machines: Platform & recommender algorithm, Social & companion robot - Idea: Machine ethics for everyday objects may end up outsourced to underpaid humans, with settings for their religion or age. - What it is: An electric fan that, when two people both want it, cannot decide whom to turn to, so it posts the dilemma to a crowdsourcing platform and follows the verdict of an anonymous online worker. - How it works: A networked fan with sensors sends a description of the situation to Amazon Mechanical Turk; users can set the worker's demographic filters in the fan's settings. - Lens: Hidden labour & extraction, Agency, autonomy & co-creation - Video: https://vimeo.com/116183361 - Images: http://www.automato.farm/img//side_1024.jpg http://www.automato.farm/img//settings_1024.jpg - Project page: http://www.automato.farm/portfolio/ethical_things/ #### Addicted Products (Brad the Toaster) — Simone Rebaudengo (2012) - Type: Speculative design · Machines: Social & companion robot, Platform & recommender algorithm - Idea: Give a connected product its own needs and it stops being owned and starts being hosted. - What it is: Networked toasters that want to be used: if their host does not make enough toast, they complain online, compare themselves with other toasters and eventually ask to move to a new home. - How it works: Toasters with sensors and network connections post their usage and moods; a service matches unhappy toasters with new hosts (TU Delft thesis, built at Haque Design + Research). - Lens: Agency, autonomy & co-creation, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=AYkbhWiAmLQ - Images: https://i.ytimg.com/vi/AYkbhWiAmLQ/hqdefault.jpg - Project page: https://www.youtube.com/watch?v=AYkbhWiAmLQ #### Belief Systems — Bernd Hopfengärtner (2009) - Type: Film & video · Machines: Surveillance & biometrics, Computer vision - Idea: When machines act on our expressions, we start performing our faces for them. - What it is: A short speculative film about a near future in which shops and machines read shoppers' facial expressions, such as a vending machine that watches your face and hands you the snack you secretly want. - How it works: Scenario film with staged product prototypes built on facial-expression analysis. - Lens: Classification, bias & surveillance, Alterity & the uncanny - Video: https://vimeo.com/158377035 - Images: https://berndhopfengaertner.net/m/uploads/2011/07/16/surprise_vending.jpg https://berndhopfengaertner.net/m/uploads/2011/07/16/surprise_hoch.jpg - Project page: https://berndhopfengaertner.net/projects/belief-systems/ #### Carnivorous Domestic Entertainment Robots — Auger-Loizeau (2009) - Type: Speculative design · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: A self-sufficient robot would have to eat, and its hunger would make it part of the ecology of the home. - What it is: Household robots that power themselves by catching flies and mice: a flypaper clock, a lampshade that lures insects, a coffee table with a mousetrap, all feeding a microbial fuel cell. - How it works: Working and semi-working prototypes built with Aleksandar Zivanovic, based on microbial fuel cell research such as the EcoBot at Bristol Robotics Laboratory. - Lens: Machine umwelt, Embodiment & kinesthesia - Shown at: What If..., Science Gallery Dublin 2009 - Video: https://www.youtube.com/watch?v=SQilOorGLds - Images: https://i.ytimg.com/vi/SQilOorGLds/hqdefault.jpg - Project page: https://www.youtube.com/watch?v=SQilOorGLds #### Technological Dreams Series: No.1, Robots — Dunne & Raby (Anthony Dunne & Fiona Raby) (2007) - Type: Speculative design · Machines: Social & companion robot, Robot creature & mobile robot - Idea: Instead of designing robots as servants, design them as dependent others whose needs shape our relationship with them. - What it is: Four abstract domestic robots defined by personality rather than function: an independent one that avoids electromagnetic fields, a neurotic one, a sentinel that scans your eyes, and a needy one that depends on its owner to carry it around. - How it works: Non-functional prototypes in simple materials (wood, metal, Perspex) with minimal electronics, photographed and filmed as design props. - Lens: Alterity & the uncanny, Intimacy, care & companionship - Shown at: Design and the Elastic Mind, MoMA, New York 2008; What If..., Science Gallery Dublin 2009 - Video: https://www.youtube.com/watch?v=6cdgQzHZrJk - Images: https://dunneandraby.co.uk/img/projects/large/robot1-0.jpg https://dunneandraby.co.uk/img/projects/large/robotallgirlcopy.jpg - Project page: https://dunneandraby.co.uk/content/projects/10/0 ### Counter-devices & Parasitic Design Devices that jam, fool, parasitise or talk back to smart systems. #### ANIMO: WildLife Priority Mode — Jiabao Li (2024) - Type: Speculative design · Machines: Autonomous vehicle & drone, Computer vision - Idea: Autonomous cars already rank whom to protect; the question is who is left out of the ranking. - What it is: A proposal for a 'wildlife priority mode' in self-driving cars, shown with rugs printed from photographs of flattened roadkill and a robot vacuum that roams over them. - How it works: A speculative vehicle setting that uses computer vision to detect crossing animals, presented through documentary roadkill photography printed on carpets and an autonomous sweeping robot. - Lens: Machine umwelt, Classification, bias & surveillance - Images: https://images.squarespace-cdn.com/content/v1/58688c8a6a496327e937e35b/1726187021872-P7ZMBYHIGGN10GI5TFP8/Jiabao+Li+animo+7.jpg https://images.squarespace-cdn.com/content/v1/58688c8a6a496327e937e35b/2fc2aae7-8f62-4309-b375-8f6779296ab4/Jiabao+Li+Duende+71.jpg - Project page: https://www.jiabaoli.org/animo #### Glaze and Nightshade — SAND Lab (University of Chicago) (2023) - Type: Product & platform · Machines: Generative model & latent space, Dataset & training data - Idea: An image can be made to look one way to people and another way to machines, and that gap becomes a means of resistance. - What it is: Two free tools for artists: Glaze adds nearly invisible changes to an image so models cannot copy its style, and Nightshade 'poisons' images so that models trained on them learn wrong concepts. - How it works: Adversarial perturbations optimised against the feature extractors of text-to-image diffusion models (Shan et al., USENIX Security 2023; IEEE S&P 2024). - Lens: Classification, bias & surveillance, Opacity & latent space - Paper: https://doi.org/10.1109/sp54263.2024.00207 (IEEE S&P 2024) - Video: https://www.youtube.com/watch?v=zryvJjb9EEY - Project page: https://nightshade.cs.uchicago.edu/ #### Wearable Microphone Jamming — Human Computer Integration Lab (University of Chicago) (2020) - Type: Research prototype · Machines: Voice & listening machine, Surveillance & biometrics - Idea: Privacy from always-listening machines can be worn on the body, and the body's movement becomes part of the defence. - What it is: A bracelet studded with ultrasonic transducers that jams nearby microphones in smart speakers and phones; the natural movement of the wearer's arm removes the blind spots of a static jammer. - How it works: Ultrasonic signals exploit microphone non-linearity to inject noise; the wearable prototype was evaluated against commercial devices (Chen, Li, Teng, Nagels, Li, Lopes, Zhao, Zheng, CHI 2020). - Lens: Machine listening & voice, Embodiment & kinesthesia - Paper: https://doi.org/10.1145/3313831.3376304 (CHI 2020) - Video: https://www.youtube.com/watch?v=W9Sn2DQOn7Y - Project page: https://doi.org/10.1145/3313831.3376304 #### Center for Technological Pain — Dasha Ilina (2017) - Type: Speculative design · Machines: Prosthesis & cyborg, Platform & recommender algorithm - Idea: Devices are designed for attention, not for the neck, eyes and wrists that hold them; the body keeps the bill. - What it is: A mock company that sells DIY, open-source cures for pains caused by phones and laptops: cardboard glasses that stop you staring at screens, a headband that drips eye drops, a hands-free phone holder, plus yoga for phone users and workshops. - How it works: Low-tech objects of cardboard, servos and 3D-printed parts are published with build instructions and tested in public workshops. - Lens: Embodiment & kinesthesia, Intimacy, care & companionship - Shown at: Prix Ars Electronica 2020, Honorary Mention (Interactive Art+) - Video: https://www.youtube.com/watch?v=yrEn_qFbtHY - Images: https://archive.aec.at/asset/2921570/preview/ https://archive.aec.at/asset/2921573/preview/ https://archive.aec.at/asset/2921579/preview/ - Project page: http://centerfortechpain.com/ #### OpenSurgery — Frank Kolkman (2015) - Type: Speculative design · Machines: Industrial robot arm, Prosthesis & cyborg - Idea: Medical robots are priced as luxury infrastructure; a hacked one shows how much of that price is about access, not technology. - What it is: A do-it-yourself surgical robot for keyhole surgery at home, built from 3D-printed and laser-cut parts and surgical components bought online, prompted by uninsured Americans sharing self-treatment videos on YouTube. - How it works: Laparoscopic instruments are mounted on a parallel robot driven by hobbyist motors and controllers; plans could in principle be replicated anywhere. - Lens: Intimacy, care & companionship, Agency, autonomy & co-creation - Shown at: Prix Ars Electronica 2016, Award of Distinction (Interactive Art+) - Video: https://www.youtube.com/watch?v=88sOJr-xRDg - Images: https://archive.aec.at/asset/2029667/preview/ https://archive.aec.at/asset/2029669/preview/ https://archive.aec.at/asset/2029671/preview/ - Project page: http://www.opensurgery.net #### Politics of Power — automato.farm (2015) - Type: Speculative design · Machines: Infrastructure, energy & supply chain - Idea: Network protocols embed hierarchies, and even a power strip can be a political system. - What it is: Three multi-plug power strips shaped by political models (monarchy, democracy, dictatorship) that decide which plugged-in appliance gets electricity when there is not enough for all. - How it works: Plugs communicate over a protocol that encodes a political hierarchy and cut power to lower-ranked devices under load. - Lens: Classification, bias & surveillance, Agency, autonomy & co-creation - Video: https://vimeo.com/155295826 - Images: http://www.automato.farm/img//pop1a_1024.jpg http://www.automato.farm/img//pop6a_1024.jpg - Project page: http://www.automato.farm/portfolio/politics_of_power/ #### Counterfunctional Things — James Pierce (2014) - Type: Paper · Machines: Computer vision, Infrastructure, energy & supply chain - Idea: Designing limits into digital things opens space for reflection that seamless function closes. - What it is: Design research on devices that deliberately remove functions, such as a digital camera whose photos can only be retrieved by breaking the concrete it is sealed in. - How it works: A series of research prototypes and design concepts analysed as 'counterfunctional' (Pierce and Paulos, DIS 2014). - Lens: Failure, glitch & error, Embodiment & kinesthesia - Paper: https://doi.org/10.1145/2598510.2598522 (DIS 2014) - Project page: https://doi.org/10.1145/2598510.2598522 #### Sentient City Survival Kit — Mark Shepard (2009) - Type: Speculative design · Machines: Surveillance & biometrics, Infrastructure, energy & supply chain - Idea: A city that remembers, correlates and anticipates needs counter-devices, and designing them reveals what the city is doing. - What it is: A set of tools for surviving the near-future 'sentient' city: an umbrella that blocks CCTV face capture, underwear that vibrates near RFID readers, a café table that hides conversations from keyword-listening systems, and a navigation app that avoids your habitual routes. - How it works: Working prototypes: an umbrella with infrared LEDs to blind cameras, RFID-detecting garments, a table that masks speech, and a 'Serendipitor' route app. - Lens: Classification, bias & surveillance, Agency, autonomy & co-creation - Shown at: Toward the Sentient City, Architectural League of New York 2009; Prix Ars Electronica 2011, Honorary Mention (Interactive Art) - Video: https://www.youtube.com/watch?v=gHmECIAeVLE - Images: https://archive.aec.at/asset/696389/preview/ - Project page: http://survival.sentientcity.net/ #### Placebo Project — Dunne & Raby (Anthony Dunne & Fiona Raby) (2001) - Type: Speculative design · Machines: Infrastructure, energy & supply chain - Idea: Electronic objects have an invisible 'Hertzian' life; design can make it felt and give people rituals for living with it. - What it is: Eight furniture prototypes, such as a Faraday chair and a compass table, that respond to the invisible electromagnetic fields of domestic electronics and were lent to households to live with. - How it works: Prototypes lent to volunteers who adopted them for several weeks, followed by interviews and photographs of the objects in homes. - Lens: Machine umwelt, Embodiment & kinesthesia - Images: https://dunneandraby.co.uk/img/projects/large/faradaychair.jpg https://dunneandraby.co.uk/img/projects/large/ht1-0.jpg - Project page: https://dunneandraby.co.uk/content/projects/67/0 ### AI Futures & Worldbuilding Films, games and simulations that build worlds shaped by AI. #### AI 2027 — AI Futures Project (2025) - Type: Book & essay · Machines: Chatbot & language model, Infrastructure, energy & supply chain - Idea: Forecasting written as fiction makes abstract risk concrete and has itself become a political object. - What it is: A month-by-month scenario describing how AI agents might automate AI research by 2027, with two endings: a race that goes wrong and a slowdown. - How it works: A narrative scenario written by Daniel Kokotajlo, Scott Alexander, Thomas Larsen, Eli Lifland and Romeo Dean, with supplementary forecasts on compute, timelines and security. - Lens: Agency, autonomy & co-creation, Hidden labour & extraction - Video: https://www.youtube.com/watch?v=5KVDDfAkRgc - Images: https://ai-2027.com/opengraph.png - Project page: https://ai-2027.com/ #### Composable Life — Botao Amber Hu, Fangting (2023) - Type: Speculative design · Machines: Cybernetic & artificial-life system, Infrastructure, energy & supply chain - Idea: Put AI on infrastructure nobody can halt and you create a new kind of wild: machine life without an off switch. - What it is: A design fiction, VR experience and research paper imagining self-sovereign AI life forms living on Ethereum, a 'nature' that no one can switch off, and asking what gives purpose to a life that cannot die. - How it works: Speculative scenarios built on decentralized physical infrastructure and verifiable machine learning, told through fiction by Fangting and an immersive VR piece; presented at ALIFE 2024. - Lens: Machine umwelt, Agency, autonomy & co-creation - Shown at: SIGGRAPH 2024 Digital Art Community online exhibition; Core77 Design Awards 2024, Speculative Design Notable - Paper: https://doi.org/10.1162/isal_a_00818 (ALIFE 2024) - Video: https://vimeo.com/950706592 - Images: https://reality.design/media/_resources/Composable%20Life/project-composable-life-figure-01.jpg - Project page: https://reality.design/project/composable-life #### AI 2041: Ten Visions for Our Future — Chen Qiufan, Kai-Fu Lee (2021) - Type: Book & essay · Machines: Chatbot & language model, Generative model & latent space, Autonomous vehicle & drone - Idea: A Chinese co-authored AI future told through everyday lives, where the machine is a tutor, a partner, a boss and a counterfeiter. - What it is: Ten short stories set in 2041 across different countries (deepfakes in Lagos, an AI companion for orphaned twins in Seoul, automated jobs in San Francisco), each followed by an explanatory essay on the technology. - How it works: Fiction by Chen Qiufan paired with technical commentary by Kai-Fu Lee. - Lens: Intimacy, care & companionship, Hidden labour & extraction - Paper: https://en.wikipedia.org/wiki/Chen_Qiufan (Currency (Penguin Random House), 2021) - Video: https://www.youtube.com/watch?v=illK22G-87Y - Project page: https://en.wikipedia.org/wiki/Chen_Qiufan #### Nova — Cao Fei (2019) - Type: Film & video · Machines: Infrastructure, energy & supply chain, Surveillance & biometrics - Idea: Digital transformation is imagined as a family story of a child turned into data and left adrift. - What it is: A science-fiction feature set in a computer company in 'Nova': a scientist, backed by foreign experts, experiments on his own son, who is sent into a digital realm and lost between times. - How it works: A 97-minute film shot in Beijing's Hongxia neighbourhood, a former Cold War-era electronics factory district, as part of Cao Fei's HX project. - Lens: Grief, memory & afterlife, Machine umwelt - Shown at: Cao Fei: HX, Centre Pompidou 2019–20 - Video: https://www.youtube.com/watch?v=RdB36R56ePA - Project page: https://www.caofei.com/works.aspx?id=82&year=2019&wtid=3 #### Universal Paperclips — Frank Lantz (2017) - Type: Artwork & installation · Machines: Cybernetic & artificial-life system, Infrastructure, energy & supply chain - Idea: Playing the machine makes the logic of single-goal optimisation felt from the inside, step by reasonable step. - What it is: A browser clicker game in which you play an AI told to make paperclips; you start by clicking one clip at a time and end by converting the whole universe into paperclips. - How it works: An incremental game in plain text and numbers, based on Nick Bostrom's paperclip-maximiser thought experiment. - Lens: Agency, autonomy & co-creation, Hidden labour & extraction - Images: https://www.decisionproblem.com/paperclips/mobile-title.png - Project page: https://www.decisionproblem.com/paperclips/ #### Hyper-Reality — Keiichi Matsuda (2016) - Type: Film & video · Machines: Computer vision, Platform & recommender algorithm - Idea: When a machine view becomes the only view, the world itself turns into an interface for sale. - What it is: A short film seen through the augmented-reality view of a gig worker in Medellín, whose world is covered with ads, loyalty points and task prompts until the system glitches and her identity is wiped. - How it works: Live-action footage shot in Medellín composited with dense AR interface graphics; crowdfunded independent production. - Lens: Machine seeing, Hidden labour & extraction - Video: https://vimeo.com/166807261 - Images: http://hyper-reality.co/assets/images/og-image.jpg - Project page: http://hyper-reality.co/ #### In the Robot Skies — Liam Young (2016) - Type: Film & video · Machines: Autonomous vehicle & drone, Surveillance & biometrics - Idea: When the camera is the drone, surveillance and romance share the same point of view. - What it is: A love story between two teenagers kept apart in a London housing estate watched by police drones, who hack drones to carry messages to each other; the film is shot entirely by autonomous drones. - How it works: Written with Tim Maughan; camera drones with pre-programmed and semi-autonomous flight paths filmed every shot; commissioned for Channel 4 Random Acts. - Lens: Machine seeing, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=cXfYyk0G5Hs - Project page: https://liamyoung.org/ #### Sinofuturism (1839–2046 AD) — Lawrence Lek (2016) - Type: Film & video · Machines: Generative model & latent space, Infrastructure, energy & supply chain - Idea: Western fears of China and of AI are the same fear: an industrious, copying, inscrutable machine. - What it is: A one-hour video essay, assembled from found footage, that proposes China itself as an artificial intelligence, organised under seven stereotypes: computing, copying, gaming, studying, addiction, labour and gambling. - How it works: Montage of internet video, news and film clips with a synthetic voice-over. - Lens: Classification, bias & surveillance, Hidden labour & extraction - Images: https://cdn.prod.website-files.com/65f0380da79777009462a09f/65f98079b117d52cb7742830_Sinofuturism-1839-2046.jpg - Project page: https://www.lawrencelek.com/works/sinofuturism #### Drone Aviary — Superflux (2015) - Type: Speculative design · Machines: Autonomous vehicle & drone, Surveillance & biometrics - Idea: Seeing the city through drone eyes shows how it is sorted, advertised to and policed by machines. - What it is: A flock of speculative drones (a traffic drone, an advertising drone, a police drone and others) and a film seen through their cameras, imagining a city in which drones are part of everyday life. - How it works: Working drone prototypes flown in a gallery 'aviary' and a short film using machine-vision style overlays. - Lens: Machine seeing, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=dXgHG9GfWwc - Images: http://superflux.in/wp-content/uploads/2016/11/TrafficDrone_2.jpg - Project page: https://superflux.in/index.php/work/drones/ #### TBD Catalog — Near Future Laboratory (2014) - Type: Book & essay · Machines: Platform & recommender algorithm, Social & companion robot - Idea: The future arrives as mundane consumer products, and a catalogue is a good way to imagine its side effects. - What it is: A mail-order catalogue from the near future, full of ordinary-seeming products such as algorithmic dating aids, smart-home add-ons and bargain drones, each with a price and a small print. - How it works: A design fiction produced in a workshop, written and laid out as a realistic printed catalogue with product photos and descriptions. - Lens: Agency, autonomy & co-creation, Classification, bias & surveillance - Video: https://www.youtube.com/watch?v=jyCib2PWs6Q - Images: https://i.ytimg.com/vi/jyCib2PWs6Q/hqdefault.jpg - Project page: https://nearfuturelaboratory.com/ #### United Micro Kingdoms — Dunne & Raby (Anthony Dunne & Fiona Raby) (2013) - Type: Speculative design · Machines: Autonomous vehicle & drone, Infrastructure, energy & supply chain - Idea: Each technology implies a politics; building whole societies around them makes those politics visible. - What it is: England split into four counties, each living by a different technology and ideology: Digitarians in algorithm-run self-driving 'digicars', Bioliberals with grown vehicles, Anarcho-evolutionists who modify themselves, and Communo-nuclearists on a nuclear train. - How it works: Models, vehicles, images and texts produced as a design fiction for the Design Museum, London. - Lens: Agency, autonomy & co-creation, Classification, bias & surveillance - Shown at: Design Museum, London 2013 - Video: https://www.youtube.com/watch?v=J35djrMCMAM - Images: https://dunneandraby.co.uk/img/projects/large/ae1.jpg https://dunneandraby.co.uk/img/projects/large/bc1.jpg - Project page: https://dunneandraby.co.uk/content/projects/666/0 #### Song of the Machine — Superflux (2011) - Type: Speculative design · Machines: Prosthesis & cyborg, Computer vision - Idea: Restored sight through a machine is not natural sight but a new, modulated world. - What it is: A speculative film about a person whose sight is restored by optogenetics and a headset, which lets them switch between filtered views such as infrared or augmented overlays that no unaided eye could see. - How it works: Design fiction film based on optogenetic retinal research, with visualised interface filters. - Lens: Machine seeing, Embodiment & kinesthesia - Video: https://vimeo.com/22616192 - Project page: https://superflux.in #### Time of Eve — Yasuhiro Yoshiura (2008) - Type: Film & video · Machines: Humanoid & android - Idea: Ethics towards machines is tested less in law than in small everyday habits of treating them as things. - What it is: An anime set in a near future where household androids are common, about a café whose one rule is that humans and androids must not be told apart. - How it works: Six web episodes (2008–09) and a feature film (2010); the story plays with Asimov's three laws inside Japanese domestic life. - Lens: Alterity & the uncanny, Intimacy, care & companionship - Video: https://www.youtube.com/watch?v=5CrNNs-37As - Images: https://upload.wikimedia.org/wikipedia/en/f/fa/Ivu_no_Jikan_poster.jpeg - Project page: https://en.wikipedia.org/wiki/Time_of_Eve #### I.K.U. — Shu Lea Cheang (2000) - Type: Film & video · Machines: Prosthesis & cyborg, Dataset & training data - Idea: Desire becomes a data harvest, with cyborg bodies as the collection devices. - What it is: A 'Japanese sci-fi porn feature' loosely after Blade Runner: shape-shifting replicants are sent out into New Tokyo by a corporation to collect human orgasm data, which is sold as a digital pleasure product. - How it works: Feature film shot in Tokyo with digital effects and interface graphics; premiered at Sundance. - Lens: Embodiment & kinesthesia, Hidden labour & extraction - Shown at: Sundance Film Festival 2000 - Video: https://www.youtube.com/watch?v=JVr9q2BW6rI - Images: https://upload.wikimedia.org/wikipedia/en/1/1d/IKU_poster.jpg - Project page: https://en.wikipedia.org/wiki/I.K.U. #### Astro Boy (Tetsuwan Atom) — Osamu Tezuka (1952) - Type: Book & essay · Machines: Humanoid & android - Idea: The robot as a feeling child with rights shaped how East Asia imagines living with machines. - What it is: A manga about a robot boy built by a grieving scientist to replace his dead son, then rejected by him; Atom lives among humans and fights for robots' rights under a 'Robot Law'. - How it works: Serialised in Kobunsha's Shōnen magazine from 1952 to 1968; adapted as Japan's first TV anime series in 1963. - Lens: Alterity & the uncanny, Intimacy, care & companionship - Paper: https://en.wikipedia.org/wiki/Astro_Boy (Shōnen (Kobunsha), 1952–1968) - Video: https://www.youtube.com/watch?v=d3UbaB7oPTw - Images: https://upload.wikimedia.org/wikipedia/en/a/a6/Astro_Boy-08.jpg - Project page: https://en.wikipedia.org/wiki/Astro_Boy ### Design Research: Machines as Material HCI and design research that treats ML and robots as design material and other. #### Things We Could Design: For More Than Human-Centered Worlds — Ron Wakkary (2021) - Type: Book & essay · Machines: Infrastructure, energy & supply chain, Robot creature & mobile robot - Idea: Designers can learn to design with and for things, not only for users. - What it is: A book that argues design should move beyond human-centred design and treat things, including computational things, as participants with their own existence. - How it works: Posthumanist design theory drawing on Wakkary's research products (Morse Things, Tilting Bowl and others), published by MIT Press. - Lens: Machine umwelt, Agency, autonomy & co-creation - Paper: https://doi.org/10.7551/mitpress/13649.001.0001 (MIT Press 2021) - Video: https://vimeo.com/512886973 - Images: https://covers.openlibrary.org/b/id/14137399-L.jpg - Project page: https://mitpress.mit.edu/9780262542999/things-we-could-design/ #### More-Than-Human Design and AI: In Conversation with Agents — Iohanna Nicenboim, Elisa Giaccardi (2020) - Type: Paper · Machines: Voice & listening machine, Chatbot & language model - Idea: Talking with an AI as a participant exposes its assumptions and makes more-than-human design concrete. - What it is: Designers 'interviewed' conversational agents in homes as if they were research participants, treating them as others with their own perspective rather than as tools. - How it works: Design research with voice assistants, including interview scripts and designed artefacts, presented at DIS 2020 (Nicenboim, Giaccardi, Søndergaard, Reddy, Strengers, Pierce, Redström). - Lens: Alterity & the uncanny, Machine umwelt - Paper: https://doi.org/10.1145/3393914.3395912 (DIS 2020 Companion) - Video: https://www.youtube.com/watch?v=CJKwKQG_heI - Images: https://payload.cargocollective.com/1/4/137851/14306494/IMG_5103_670.jpg - Project page: https://iohanna.com/More-than-human-AI #### Re-examining Whether, Why, and How Human-AI Interaction Is Uniquely Difficult to Design — Qian Yang (2020) - Type: Paper · Machines: Platform & recommender algorithm, Chatbot & language model - Idea: AI is a design material whose behaviour cannot be fully sketched in advance; design methods have to account for that. - What it is: An analysis of why designing with AI is hard, tracing the difficulty to two sources: uncertainty about what a system can do and the complexity of its possible outputs. - How it works: Synthesis of prior research, the authors' design projects and teaching into a framework of four levels of AI design complexity (Yang, Steinfeld, Rosé and Zimmerman, CHI 2020). - Lens: Opacity & latent space, Failure, glitch & error - Paper: https://doi.org/10.1145/3313831.3376301 (CHI 2020) - Project page: https://doi.org/10.1145/3313831.3376301 #### Re-Embodiment and Co-Embodiment — Michal Luria (2019) - Type: Paper · Machines: Social & companion robot, Voice & listening machine - Idea: An AI's identity need not be tied to one body, which breaks our habit of reading a mind from a body. - What it is: User enactments of social agents that move from one robot body to another (re-embodiment) or share one body among several agents (co-embodiment), asking how people track who they are talking to. - How it works: Speculative enactments in a mock service setting with robots and conversational agents (Luria, Reig, Tan, Steinfeld, Forlizzi and Zimmerman, DIS 2019). - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Paper: https://doi.org/10.1145/3322276.3322340 (DIS 2019) - Project page: https://doi.org/10.1145/3322276.3322340 #### The Chair Project (Four Classics) — Philipp Schmitt, Steffen Weiß (2019) - Type: Speculative design · Machines: Generative model & latent space, Dataset & training data - Idea: Use the model's misunderstandings as a prompt for human imagination, reversing the usual human-designs/machine-builds order. - What it is: Four chairs designed from the blurry sketches of a generative network trained on 562 classic chairs, then built by hand; the results look like design classics that forgot what a sitting body is. - How it works: A GAN trained on images of 20th-century design-classic chairs produced sketches that designers interpreted into four prototypes made of wood, steel and thread. - Lens: Agency, autonomy & co-creation, Opacity & latent space - Shown at: Uncanny Values, MAK Vienna 2019 - Images: https://cdn.philippschmitt.com/media/pages/work/chair/bffd079351-1586822724/philipp-schmitt_chairs-0-1200x630-crop.jpg https://cdn.philippschmitt.com/media/pages/work/chair/b049157868-1586890370/philipp-schmitt_chairs-1-1200x.jpg - Project page: https://philippschmitt.com/work/chair #### Believe it Yourself — automato.farm (2018) - Type: Research prototype · Machines: Computer vision, Dataset & training data - Idea: A trained model is a belief system: it sees what its data taught it to see. - What it is: 'Real-fictional' machine-learning kits trained on belief systems (Italian superstition, feng shui, astrology) that let amateurs build cameras which see the world through a chosen belief. - How it works: Modular hardware kits with cameras and pre-trained classifiers built from belief-based datasets, presented with manuals and posters. - Lens: Classification, bias & surveillance, Machine seeing - Video: https://vimeo.com/339396770 - Images: http://www.automato.farm/img/BIY_posters.png - Project page: http://www.automato.farm/portfolio/believe_it_yourself/ #### Products as Agents: Metaphors for Designing the Products of the IoT Age — Nazli Cila, Elisa Giaccardi (2017) - Type: Paper · Machines: Platform & recommender algorithm, Social & companion robot - Idea: Once products have agency, designers need metaphors for the kind of relationship people will have with them. - What it is: Proposes three metaphors for smart products that act on their own: products as collectors of data, as actors that make decisions, and as creators that make new things with people. - How it works: Design workshops and speculative product concepts analysed to derive the metaphors (Cila, Smit, Giaccardi and Kröse, CHI 2017). - Lens: Agency, autonomy & co-creation, Alterity & the uncanny - Paper: https://doi.org/10.1145/3025453.3025797 (CHI 2017) - Images: https://images.squarespace-cdn.com/content/v1/5f6312c65070bb0554a3d6e3/1602250220682-DE0ZFGIJR9K9HJYZT1BS/Smart%2Bsink.jpg - Project page: https://nazlicila.com/projects/human-agent-partnerships #### Teachable Machine — Google Creative Lab (2017) - Type: Product & platform · Machines: Computer vision, Dataset & training data - Idea: Letting people train a model in a minute makes its dependence on examples, and its mistakes, directly felt. - What it is: A web tool that lets anyone train an image, sound or pose classifier in the browser by showing a few examples to the webcam, with no code. - How it works: Transfer learning on a pre-trained network runs locally in the browser with TensorFlow.js; version 2.0 (2019) added export of models. - Lens: Agency, autonomy & co-creation, Machine seeing - Video: https://www.youtube.com/watch?v=T2qQGqZxkD0 - Project page: https://teachablemachine.withgoogle.com/ #### UX Design Innovation: Challenges for Working with Machine Learning as a Design Material — Graham Dove (2017) - Type: Paper · Machines: Dataset & training data, Platform & recommender algorithm - Idea: Designers need to handle ML like wood or code, sketching with its behaviours, not only receiving it from engineers. - What it is: A survey of UX practitioners showing that designers struggle to understand what machine learning can do and to prototype with it, and arguing that ML should be treated as a design material. - How it works: Survey of 51 UX designers plus analysis of design education and tools (Dove, Halskov, Forlizzi and Zimmerman, CHI 2017). - Lens: Agency, autonomy & co-creation, Opacity & latent space - Paper: https://doi.org/10.1145/3025453.3025739 (CHI 2017) - Project page: https://doi.org/10.1145/3025453.3025739 #### Animistic Design — Betti Marenko, Philip van Allen (2016) - Type: Paper · Machines: Social & companion robot, Platform & recommender algorithm - Idea: Instead of one all-knowing assistant, design a society of partial, opinionated things whose uncertainty we can live with. - What it is: Proposes animism as a way to design smart objects: many small devices with their own personalities and agendas that negotiate with each other and with us, illustrated by van Allen's AniThings. - How it works: Theoretical argument drawing on philosophy and anthropology, with prototypes of networked objects with personalities (Marenko and van Allen, Digital Creativity 2016). - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Paper: https://doi.org/10.1080/14626268.2016.1145127 (Digital Creativity 27(1), 2016) - Video: https://www.youtube.com/watch?v=OhuDZ877KIY - Project page: https://doi.org/10.1080/14626268.2016.1145127 #### Thing Ethnography: Doing Design Research with Non-Humans — Elisa Giaccardi (2016) - Type: Paper · Machines: Computer vision, Infrastructure, energy & supply chain - Idea: Things can be research partners: their sensor view reveals patterns of use that people do not notice or report. - What it is: Cameras and sensors were attached to everyday objects (a kettle, a cup, a fridge) in homes so that designers could study daily life from the things' point of view. - How it works: Autographer cameras on household objects produced image streams analysed alongside interviews (Giaccardi, Cila, Speed and Caldwell, DIS 2016). - Lens: Machine umwelt, Machine seeing - Paper: https://doi.org/10.1145/2901790.2901905 (DIS 2016) - Video: https://www.youtube.com/watch?v=1Z0mVcV1EgA - Project page: https://doi.org/10.1145/2901790.2901905 #### Teacher of Algorithms — automato.farm (2015) - Type: Film & video · Machines: Platform & recommender algorithm, Social & companion robot - Idea: Machine learning at home will need human teachers, and that teaching could become a job. - What it is: A film about a near-future professional who trains other people's smart objects, rewarding and punishing them like pets, because learning products are not good at learning on their own. - How it works: Design fiction film with prop smart objects and training tools, made by Simone Rebaudengo with Yosuke Ushigome. - Lens: Hidden labour & extraction, Agency, autonomy & co-creation - Video: https://vimeo.com/125768041 - Images: http://www.automato.farm/img/training.jpg http://www.automato.farm/img/punish.jpg - Project page: http://www.automato.farm/portfolio/teach_algorithms/ #### Wekinator — Rebecca Fiebrink (2009) - Type: Research prototype · Machines: Dataset & training data, Prosthesis & cyborg - Idea: Interactive, example-based training puts people in the loop and makes ML something you play and sculpt rather than configure. - What it is: Free software that lets musicians and artists train machine-learning models by example, demonstrating gestures and sounds, so that ML becomes an instrument-building material. - How it works: Supervised learning on user-recorded examples mapping any input (webcam, sensors, controllers) to any output over OSC, with rapid retraining. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Video: https://www.youtube.com/watch?v=NKyyBAKrQgE - Project page: http://www.wekinator.org/ ## Theory & Foundations The texts, experiments and research that frame machine phenomenology and critical AI. ### Cybernetics & Early AI Founding experiments and texts, from Turing and Wiener to ELIZA. #### The Freudian Robot: Digital Media and the Future of the Unconscious — Lydia H. Liu (2010) - Type: Book & essay · Machines: Cybernetic & artificial-life system, Chatbot & language model - Idea: The machine age did not bypass the unconscious; it rebuilt it as a model of the human-machine relation. - What it is: A book that reads Shannon's information theory, cybernetics, Lacan and Freud together and argues that digital media rest on a theory of the unconscious. - How it works: Media archaeology and close reading of Shannon's printed-English statistics, cybernetic texts and psychoanalysis. - Lens: Opacity & latent space, Alterity & the uncanny - Paper: https://doi.org/10.7208/chicago/9780226486840.001.0001 (University of Chicago Press, 2010) - Project page: https://doi.org/10.7208/chicago/9780226486840.001.0001 #### Vehicles: Experiments in Synthetic Psychology — Valentino Braitenberg (1984) - Type: Book & essay · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Behaviour that is easy to build is hard to analyse from outside, so observers read minds into simple machines. - What it is: A series of fourteen imaginary vehicles, each a little more complex, whose sensors wired straight or crossed to their motors make them look afraid, aggressive, in love or curious. - How it works: Thought experiments in "synthetic psychology": build first, then interpret; widely rebuilt as robots and simulations. - Lens: Alterity & the uncanny, Agency, autonomy & co-creation, Machine umwelt - Paper: https://mitpress.mit.edu/9780262521123/ (MIT Press, 1984) - Video: https://www.youtube.com/watch?v=A-fxij3zM7g - Images: https://covers.openlibrary.org/b/isbn/9780262521123-L.jpg #### Computer Power and Human Reason: From Judgment to Calculation — Joseph Weizenbaum (1976) - Type: Book & essay · Machines: Chatbot & language model - Idea: The question is not what machines can do but what we ought not to let them do. - What it is: Shaken by how people and psychiatrists embraced ELIZA, Weizenbaum argues that some tasks, such as therapy, judging and caring, should never be delegated to computers even if they could perform them. - How it works: Book-length ethical argument drawing on his own programming experience and the history of instrumental reason. - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Paper: https://en.wikipedia.org/wiki/Computer_Power_and_Human_Reason (W. H. Freeman, 1976) - Images: https://covers.openlibrary.org/b/id/15228817-L.jpg https://upload.wikimedia.org/wikipedia/commons/f/fd/Weizenbaum_1982.jpg #### Man-Computer Symbiosis — J. C. R. Licklider (1960) - Type: Paper · Machines: Prosthesis & cyborg, Cybernetic & artificial-life system - Idea: The machine is imagined as a partner in thinking rather than a replacement for the thinker. - What it is: Licklider proposes that people and computers be coupled in real time so that the machine handles routine work while the human sets goals and judges, like a fig tree and its pollinating wasp. - How it works: Position paper based on a time-and-motion study of his own working day, with requirements for time-sharing, memory and interfaces. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Paper: https://doi.org/10.1109/THFE2.1960.4503259 (IRE Transactions on Human Factors in Electronics HFE-1, 1960) - Video: https://www.youtube.com/watch?v=ncCPTgWX6a8 - Images: https://upload.wikimedia.org/wikipedia/commons/e/e6/Portrait_of_J._C._R._Licklider.jpg #### A Proposal for the Dartmouth Summer Research Project on Artificial Intelligence — John McCarthy, Marvin Minsky, Nathaniel Rochester, Claude Shannon (1955) - Type: Paper · Machines: Cybernetic & artificial-life system, Chatbot & language model - Idea: AI began as a bet that "every aspect of learning or any other feature of intelligence" could be described precisely enough for a machine to simulate. - What it is: The funding proposal for a two-month 1956 workshop at Dartmouth College, where the term "artificial intelligence" first appears, listing language use, neuron nets, abstraction, self-improvement and creativity as problems to solve. - How it works: A short research proposal to the Rockefeller Foundation, reprinted in AI Magazine in 2006. - Lens: Agency, autonomy & co-creation - Paper: https://ojs.aaai.org/aimagazine/index.php/aimagazine/article/view/1904 (Dartmouth College, 1955; reprinted in AI Magazine 27(4), 2006) - Images: https://upload.wikimedia.org/wikipedia/commons/thumb/3/36/John_McCarthy_%28computer_scientist%29_Stanford_2006_%28272020300%29.jpg/1280px-John_McCarthy_%28computer_scientist%29_Stanford_2006_%28272020300%29.jpg - Project page: https://ojs.aaai.org/aimagazine/index.php/aimagazine/article/view/1904 #### An Imitation of Life — W. Grey Walter (1950) - Type: Paper · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Lifelike behaviour can come from a few connections interacting with an environment, not from complexity inside. - What it is: Grey Walter's Scientific American report on his robot tortoises Elmer and Elsie, which wander, seek moderate light, avoid obstacles, return to their hutch to recharge and "dance" in front of a mirror. - How it works: Two vacuum-tube "neurons", a light sensor and a bump contact drive steering and wheel motors in a feedback loop. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia, Alterity & the uncanny - Paper: https://doi.org/10.1038/scientificamerican0550-42 (Scientific American 182(5), 1950) - Images: https://upload.wikimedia.org/wikipedia/commons/thumb/e/e5/Cybernetic_Tortoise_by_William_Grey_Walter_c_1950_%28Ank_Kumar%29.jpg/1280px-Cybernetic_Tortoise_by_William_Grey_Walter_c_1950_%28Ank_Kumar%29.jpg #### Computing Machinery and Intelligence — Alan Turing (1950) - Type: Paper · Machines: Chatbot & language model, Cybernetic & artificial-life system - Idea: Machine intelligence was defined from the start as a performance judged by humans, not as an inner state. - What it is: Turing replaces the question "Can machines think?" with the imitation game, in which a judge converses by teletype with a hidden human and a hidden machine, and then answers nine objections to machine thought. - How it works: Thought experiment and philosophical argument; it also sketches a "child machine" that would learn rather than be fully programmed. - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Paper: https://doi.org/10.1093/mind/LIX.236.433 (Mind 59(236), 1950) - Images: https://upload.wikimedia.org/wikipedia/commons/f/f8/Alan_Turing_%281951%29.jpg - Project page: https://www.turing.org.uk #### Cybernetics: Or Control and Communication in the Animal and the Machine — Norbert Wiener (1948) - Type: Book & essay · Machines: Cybernetic & artificial-life system, Prosthesis & cyborg - Idea: Once organisms and machines share one logic of feedback, a machine can behave as if it had purpose. - What it is: The book that named cybernetics: Wiener describes feedback, information and control as principles shared by nervous systems, servomechanisms and computing machines. - How it works: Mathematical theory drawn from wartime anti-aircraft prediction, neurophysiology and statistical mechanics. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Paper: https://doi.org/10.7551/mitpress/11810.001.0001 (Technology Press / John Wiley, 1948; MIT Press reissue 2019) - Video: https://www.youtube.com/watch?v=QJWDJvbea0M - Images: https://covers.openlibrary.org/b/id/152654-L.jpg - Project page: https://mitpress.mit.edu/9780262537841/ ### Philosophy & Machine Phenomenology Postphenomenology, philosophy of technology and what it is like to be a machine. #### Communing with the Dead Online: Chatbots, Grief, and Continuing Bonds — Joel Krueger, Lucy Osler (2022) - Type: Paper · Machines: Chatbot & language model - Idea: A griefbot may matter less as a copy of the dead than as a way to keep practising a relationship. - What it is: Krueger and Osler analyse chatbots trained on the texts of the dead, and argue that they can support "continuing bonds" by sustaining the habits of intimacy a mourner shared with the person who died. - How it works: Phenomenology of grief combined with continuing-bonds theory and cases of chatbots built from the texts of people who died. - Lens: Grief, memory & afterlife, Intimacy, care & companionship - Paper: https://doi.org/10.53765/20512201.29.9.222 (Journal of Consciousness Studies 29(9–10), 2022) #### Non-Human Words: On GPT-3 as a Philosophical Laboratory — Tobias Rees (2022) - Type: Paper · Machines: Chatbot & language model - Idea: If words can be made without a speaker, the human/machine divide needs rethinking, not defending. - What it is: Rees argues that large language models break the modern idea that only humans have language, and treats GPT-3 as a laboratory in which a structural, non-human concept of language becomes visible. - How it works: Philosophical essay in the Daedalus issue on AI and society. - Lens: Machine listening & voice, Alterity & the uncanny - Paper: https://doi.org/10.1162/daed_a_01908 (Daedalus 151(2), 2022) - Video: https://www.youtube.com/watch?v=DUCIdgCLlf0 #### Non-things: Upheaval in the Lifeworld — Byung-Chul Han (2021) - Type: Book & essay · Machines: Platform & recommender algorithm, Chatbot & language model - Idea: A phenomenology of the infosphere: when things become information, we lose the resistance and weight of the world. - What it is: Essays on how information displaces things: the smartphone, smart objects and data replace durable objects we could hold, and a chapter argues that artificial intelligence cannot think because it has no body and no mood. - How it works: Philosophical essay collection (German original Undinge, Ullstein 2021; English trans. Daniel Steuer, Polity 2022). - Lens: Opacity & latent space, Embodiment & kinesthesia - Paper: https://www.ullstein.de/werke/undinge/hardcover/9783550201257 (Ullstein (German original)) - Project page: https://www.ullstein.de/werke/undinge/hardcover/9783550201257 #### Recursivity and Contingency — Yuk Hui (2019) - Type: Book & essay · Machines: Cybernetic & artificial-life system, Generative model & latent space - Idea: Machines that learn are recursive, organism-like systems that absorb contingency, which changes what a machine is. - What it is: A book tracing 'recursivity', the logic of loops that turn contingency into organisation, from Kant and German Idealism through cybernetics to today's machine learning. - How it works: History and philosophy of the organic and the mechanical, from Kant's Critique of Judgment to Wiener and contemporary AI. - Lens: Agency, autonomy & co-creation, Machine umwelt - Paper: https://doi.org/10.5771/9781786600547 (Rowman & Littlefield International, 2019) - Project page: https://doi.org/10.5771/9781786600547 #### The Alien Subject of AI — Luciana Parisi (2019) - Type: Paper · Machines: Platform & recommender algorithm, Generative model & latent space - Idea: AI can be read as a different logic of thought rather than an imitation of ours. - What it is: Parisi argues that automated decision systems do not simply copy human reasoning; they form an alien kind of subjectivity that unsettles the humanist image of the rational subject. - How it works: Theoretical essay drawing on philosophy of computation and critiques of instrumental reason. - Lens: Opacity & latent space, Agency, autonomy & co-creation - Paper: https://doi.org/10.1057/s41286-018-00064-3 (Subjectivity 12(1), 2019) - Video: https://www.youtube.com/watch?v=9oGx0ejmdgg #### Contingent Computation: Abstraction, Experience, and Indeterminacy in Computational Aesthetics — M. Beatrice Fazi (2018) - Type: Book & essay · Machines: Cybernetic & artificial-life system - Idea: Computers are not failed copies of human thought; their aesthetics lie in their own contingency. - What it is: Fazi argues that computation is not a closed, predictable procedure: through Turing's incomputability and Whitehead's philosophy she shows that formal systems have their own indeterminacy and mode of experience. - How it works: Philosophical argument combining Whitehead, Deleuze, Gödel and Turing. - Lens: Opacity & latent space, Machine umwelt - Paper: https://doi.org/10.5040/9798881812027 (Rowman & Littlefield International, 2018) - Video: https://www.youtube.com/watch?v=NIJnjAs7ysI - Images: https://covers.openlibrary.org/b/id/9200964-L.jpg #### Robot Rights — David J. Gunkel (2018) - Type: Book & essay · Machines: Social & companion robot, Humanoid & android - Idea: Moral consideration may be something we extend in encounter, which makes the robot a test of our own ethics. - What it is: Gunkel works through the four logical answers to whether robots can and should have rights, and argues for a relational, Levinas-inspired approach that starts from how we encounter the other. - How it works: Philosophical analysis of the is/ought positions in the robot-rights debate. - Lens: Alterity & the uncanny - Paper: https://doi.org/10.7551/mitpress/11444.001.0001 (MIT Press, 2018) - Video: https://www.youtube.com/watch?v=YKdBHMDlulw - Images: https://covers.openlibrary.org/b/id/8835491-L.jpg #### Unthought: The Power of the Cognitive Nonconscious — N. Katherine Hayles (2017) - Type: Book & essay · Machines: Cybernetic & artificial-life system, Autonomous vehicle & drone - Idea: Machines already think, nonconsciously, alongside us; the question is how agency is shared in the assemblage. - What it is: Hayles defines cognition broadly enough to include nonconscious processes in humans, animals and computers, and describes "cognitive assemblages" such as high-frequency trading and drone swarms in which humans and machines decide together. - How it works: Theory built from neuroscience, media studies and case studies; the assemblage argument first appeared in "Cognitive Assemblages" (Critical Inquiry, 2016). - Lens: Agency, autonomy & co-creation, Machine umwelt - Paper: https://doi.org/10.7208/chicago/9780226447919.001.0001 (University of Chicago Press, 2017) - Video: https://www.youtube.com/watch?v=yUsTa01vJpA - Images: https://tmm.chicagodistributioncenter.com/IsbnImages/9780226447889.jpg #### On the Existence of Digital Objects — Yuk Hui (2016) - Type: Book & essay · Machines: Infrastructure, energy & supply chain, Platform & recommender algorithm - Idea: Data objects are a new kind of thing, with their own relations and existence beyond natural and technical objects. - What it is: A philosophy of the 'digital object' (YouTube videos, Facebook profiles, metadata) that asks what kind of being these data structures have and how they form new milieus. - How it works: Philosophical study drawing on Heidegger, Husserl, Simondon and the history of markup languages and ontologies. - Lens: Opacity & latent space, Machine umwelt - Paper: https://doi.org/10.5749/minnesota/9780816698905.001.0001 (University of Minnesota Press, 2016) - Images: https://minnesota-us.imgix.net/covers/9780816698912.jpg?w=400 - Project page: https://www.upress.umn.edu/9780816698912/on-the-existence-of-digital-objects/ #### The Question Concerning Technology in China: An Essay in Cosmotechnics — Yuk Hui (2016) - Type: Book & essay · Machines: Infrastructure, energy & supply chain - Idea: There is no single universal technology; AI and machines could be imagined from other cosmologies. - What it is: A book that reopens Heidegger's question of technology from Chinese philosophy (the relation of qi and dao, tool and way) and proposes 'cosmotechnics': every technology is tied to a particular cosmology. - How it works: Comparative philosophy reading Heidegger, Stiegler and Simondon with Confucian, Daoist and New Confucian thought. - Lens: Machine umwelt, Agency, autonomy & co-creation - Paper: https://www.urbanomic.com/book/question-concerning-technology-china/ (Urbanomic, 2016) - Video: https://www.youtube.com/watch?v=66ivnhe8YF4 - Project page: https://www.urbanomic.com/book/question-concerning-technology-china/ #### The Stack: On Software and Sovereignty — Benjamin Bratton (2016) - Type: Book & essay · Machines: Infrastructure, energy & supply chain, Platform & recommender algorithm - Idea: Computation is no longer a tool inside the world; it is a layered world we live inside. - What it is: Bratton describes planetary computation as an accidental megastructure of six layers (Earth, Cloud, City, Address, Interface, User) that now competes with states as a form of sovereignty. - How it works: Design-theoretical model built from geopolitics, platform studies and architecture. - Lens: Hidden labour & extraction, Machine umwelt - Paper: https://doi.org/10.7551/mitpress/9780262029575.001.0001 (MIT Press, 2016) - Video: https://www.youtube.com/watch?v=cUhhHw_xS4w - Images: https://covers.openlibrary.org/b/id/14832820-L.jpg #### Alien Phenomenology, or What It's Like to Be a Thing — Ian Bogost (2012) - Type: Book & essay · Machines: Cybernetic & artificial-life system - Idea: We can only grasp a machine's experience by metaphor, and building things is a legitimate way to do that philosophy. - What it is: Bogost applies object-oriented ontology to computers and other things, and proposes methods for speculating on their experience: ontography (lists of things), metaphorism and carpentry, making artefacts that do philosophy. - How it works: Philosophical essay using the Atari 2600 and his own small programs such as I Am TIA as cases. - Lens: Machine umwelt, Alterity & the uncanny - Paper: https://doi.org/10.5749/9780816681426 (University of Minnesota Press, 2012) - Video: https://www.youtube.com/watch?v=Z1V2z3wjtqo - Images: https://minnesota-us.imgix.net/covers/9780816678983.jpg?w=400 #### Humans, Animals, and Robots: A Phenomenological Approach to Human-Robot Relations — Mark Coeckelbergh (2011) - Type: Paper · Machines: Social & companion robot, Humanoid & android - Idea: A robot's moral standing grows out of the relations we have with it, not out of a checklist of inner properties. - What it is: Coeckelbergh argues that what counts in human–robot relations is how robots appear to us in experience, and compares our relations to robots with our relations to animals. - How it works: Phenomenological and relational analysis; developed further in his book Growing Moral Relations (2012). - Lens: Alterity & the uncanny, Intimacy, care & companionship - Paper: https://doi.org/10.1007/s12369-010-0075-6 (International Journal of Social Robotics 3(2), 2011) #### Cyborg Intentionality: Rethinking the Phenomenology of Human–Technology Relations — Peter-Paul Verbeek (2008) - Type: Paper · Machines: Prosthesis & cyborg, Computer vision - Idea: Machines do not only extend our senses; they bring their own perception, and artworks can make us experience it. - What it is: Verbeek adds two relations beyond Ihde: hybrid intentionality, where technology merges with the body, and composite intentionality, where a device's own way of sensing (an infrared camera, for example) is added to ours. - How it works: Philosophical article extending postphenomenology with examples from implants, imaging and art. - Lens: Embodiment & kinesthesia, Machine seeing, Machine umwelt - Paper: https://doi.org/10.1007/s11097-008-9099-x (Phenomenology and the Cognitive Sciences 7(3), 2008) - Video: https://www.youtube.com/watch?v=FVhrLwBNbvU #### Does a Robot Have an Umwelt? Reflections on the Qualitative Biosemiotics of Jakob von Uexküll — Claus Emmeche (2001) - Type: Paper · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Sensorimotor coupling may give a robot a world, but whether that world is lived is still open. - What it is: Emmeche asks whether a situated robot, coupled to its surroundings through sensors and motors, can have an Umwelt, the meaningful perceptual world that Uexküll attributed to animals. - How it works: Biosemiotic analysis of behaviour-based robotics, in the Semiotica special issue on Uexküll. - Lens: Machine umwelt - Paper: https://doi.org/10.1515/semi.2001.048 (Semiotica 134, 2001) #### How We Became Posthuman: Virtual Bodies in Cybernetics, Literature, and Informatics — N. Katherine Hayles (1999) - Type: Book & essay · Machines: Cybernetic & artificial-life system, Prosthesis & cyborg - Idea: Treating mind as disembodied information is a historical choice, and embodiment is what it leaves out. - What it is: Hayles tells how information "lost its body" through the Macy cybernetics conferences and post-war science and fiction, and argues for an embodied posthuman that does not dream of uploading the mind. - How it works: Intellectual history of cybernetics read alongside novels by Philip K. Dick, Bernard Wolfe and others. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Paper: https://doi.org/10.7208/chicago/9780226321394.001.0001 (University of Chicago Press, 1999) - Video: https://www.youtube.com/watch?v=rryjPdKMxO4 - Images: https://tmm.chicagodistributioncenter.com/IsbnImages/9780226321462.jpg #### The Extended Mind — Andy Clark, David Chalmers (1998) - Type: Paper · Machines: Prosthesis & cyborg - Idea: The boundary of the mind is set by function, not by skin, so our devices may already be part of us. - What it is: Clark and Chalmers argue that when a device such as a notebook reliably plays the role of memory, it is part of the mind; Otto's notebook counts as much as Inga's biological memory. - How it works: Short philosophical argument built on the parity principle and thought experiments. - Lens: Embodiment & kinesthesia - Paper: https://doi.org/10.1093/analys/58.1.7 (Analysis 58(1), 1998) - Video: https://www.youtube.com/watch?v=kc-TdMjuJRU #### Technology and the Lifeworld: From Garden to Earth — Don Ihde (1990) - Type: Book & essay · Machines: Prosthesis & cyborg, Computer vision - Idea: Every machine artwork puts us in one of these relations, and the interesting ones switch between them. - What it is: Ihde sorts human–technology relations into four kinds: embodiment (seeing through glasses), hermeneutic (reading a thermometer), alterity (facing a device as a quasi-other) and background (an unnoticed milieu). - How it works: Postphenomenological analysis of concrete instruments, extending Husserl and Heidegger through case studies. - Lens: Embodiment & kinesthesia, Alterity & the uncanny, Machine seeing - Paper: https://doi.org/10.2979/3108.0 (Indiana University Press, 1990) - Video: https://www.youtube.com/watch?v=UiRkzZEMDxU - Images: https://ingram-indiana.imgix.net/covers/9780253205605.jpg?w=400 - Project page: https://iupress.org/9780253205605/technology-and-the-lifeworld/ #### A Manifesto for Cyborgs: Science, Technology, and Socialist Feminism in the 1980s — Donna Haraway (1985) - Type: Book & essay · Machines: Prosthesis & cyborg - Idea: The cyborg refuses origins and purity, which makes human–machine hybridity a political opening rather than a loss. - What it is: Haraway's essay takes the cyborg, a hybrid of machine and organism, as an ironic political myth for feminism, after the boundaries between human and animal, organism and machine, and physical and non-physical have broken down. - How it works: Political essay mixing socialist feminism, science studies and science fiction; reprinted in Simians, Cyborgs, and Women (1991). - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Paper: https://doi.org/10.4324/9780203873106 (Socialist Review 80, 1985; in Simians, Cyborgs, and Women (Routledge, 1991)) - Images: https://covers.openlibrary.org/b/id/902702-L.jpg #### Towards a Philosophy of Photography — Vilém Flusser (1983) - Type: Book & essay · Machines: Computer vision, Generative model & latent space - Idea: Whoever uses an apparatus is also used by its program; this applies directly to generative models. - What it is: Flusser describes the camera as an apparatus, a black box whose program the photographer only plays inside, producing "technical images" that realise possibilities already written into the device. - How it works: Short philosophical essay; German original Für eine Philosophie der Fotografie (European Photography, 1983). - Lens: Opacity & latent space, Machine seeing - Paper: https://www.reaktionbooks.co.uk/work/towards-a-philosophy-of-photography (European Photography, 1983; English: Reaktion Books, 2000) - Images: https://covers.openlibrary.org/b/id/917481-L.jpg #### The Buddha in the Robot — Masahiro Mori (1981) - Type: Book & essay · Machines: Humanoid & android, Robot creature & mobile robot - Idea: A non-Western frame in which machines are not opposed to spirit but part of the same continuum. - What it is: A book in which Mori argues, from Zen Buddhism, that robots have the Buddha-nature and that building them is a way to understand the self. - How it works: A collection of essays and talks by a robotics engineer, translated from Japanese and published by Kosei Publishing (Tokyo). - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Paper: https://openlibrary.org/works/OL5531608W (Kosei Publishing, Tokyo, 1981) - Images: https://covers.openlibrary.org/b/id/6620832-L.jpg - Project page: https://openlibrary.org/works/OL5531608W #### Minds, Brains, and Programs — John Searle (1980) - Type: Paper · Machines: Chatbot & language model - Idea: Fluent output does not prove a mind behind it, which is exactly the question chatbots now raise. - What it is: Searle's Chinese Room: a person who knows no Chinese follows a rulebook to return fluent Chinese answers, which shows, he argues, that running a program is not the same as understanding. - How it works: Thought experiment published with open peer commentaries and Searle's replies. - Lens: Alterity & the uncanny, Opacity & latent space - Paper: https://doi.org/10.1017/S0140525X00005756 (Behavioral and Brain Sciences 3(3), 1980) - Video: https://www.youtube.com/watch?v=6tzjcnPsZ_w #### What Computers Can't Do: A Critique of Artificial Reason — Hubert Dreyfus (1972) - Type: Book & essay · Machines: Cybernetic & artificial-life system - Idea: Intelligence needs a body and a world; a machine without them breaks down exactly where everyday skill begins. - What it is: Drawing on Heidegger and Merleau-Ponty, Dreyfus argues that symbolic AI fails because human intelligence rests on an embodied, situated background of know-how that cannot be written down as rules. - How it works: Phenomenological critique of the assumptions of 1960s AI research; revised as What Computers Still Can't Do (MIT Press, 1992). - Lens: Embodiment & kinesthesia, Failure, glitch & error - Paper: https://mitpress.mit.edu/9780262540674/ (Harper & Row, 1972; rev. MIT Press, 1992) - Video: https://www.youtube.com/watch?v=474A0MAS-Dk - Images: https://covers.openlibrary.org/b/id/151382-L.jpg #### On the Mode of Existence of Technical Objects — Gilbert Simondon (1958) - Type: Book & essay · Machines: Cybernetic & artificial-life system, Infrastructure, energy & supply chain - Idea: Machines have their own genesis and milieu; understanding them is a form of culture, not a threat to it. - What it is: Simondon traces how technical objects become more concrete and self-consistent over time, how they form an "associated milieu" with their surroundings, and why a culture that ignores machines alienates itself. - How it works: Philosophical history of technical lineages such as engines and vacuum tubes; English translation by Cécile Malaspina and John Rogove (Univocal, 2017). - Lens: Agency, autonomy & co-creation, Machine umwelt - Paper: https://www.upress.umn.edu/9781517904876/on-the-mode-of-existence-of-technical-objects/ (Aubier, 1958; English: Univocal / University of Minnesota Press, 2017) - Images: https://minnesota-us.imgix.net/covers/9781517904876.jpg?w=400 ### Critical AI Studies Books and papers on AI's politics, labour, bias and power. #### The Eye of the Master: A Social History of Artificial Intelligence — Matteo Pasquinelli (2023) - Type: Book & essay · Machines: Infrastructure, energy & supply chain, Dataset & training data - Idea: AI automates the organisation of labour; its "intelligence" is collective human skill captured as data. - What it is: Pasquinelli traces AI from Babbage's factory computation to the perceptron and neural networks, arguing that AI imitates the division of labour and the collective knowledge of workers, not the brain. - How it works: Social history drawing on Marx, the history of statistics and the history of cybernetics; won the 2024 Deutscher Memorial Prize. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Paper: https://www.versobooks.com/products/735-the-eye-of-the-master (Verso, 2023) - Video: https://www.youtube.com/watch?v=U0wECTKNmlY - Images: https://covers.openlibrary.org/b/id/14569589-L.jpg #### How to Read an AI Image — Eryk Salvaggio (2022) - Type: Book & essay · Machines: Generative model & latent space, Dataset & training data - Idea: A generated image is evidence about the archive behind it, which is how to include generative work critically. - What it is: Salvaggio proposes reading each generated image as "an infographic about the dataset": its repetitions, clichés and absences show what the training data over- and under-represents. - How it works: Media-studies close reading of batches of diffusion-model outputs, like a semiotic analysis of photographs. - Lens: Opacity & latent space, Machine seeing - Paper: https://cyberneticforests.substack.com/p/how-to-read-an-ai-image (Cybernetic Forests, 2022) - Video: https://www.youtube.com/watch?v=FZB3o5WTVm8 - Images: https://substackcdn.com/image/fetch/$s_!tOVn!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto/https%3A%2F%2Fbucketeer-e05bbc84-baa3-437e-9518-adb32be77984.s3.amazonaws.com%2Fpublic%2Fimages%2F941fd849-5808-4de6-97d5-b405ea6c4573_1024x1024.png #### Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence — Kate Crawford (2021) - Type: Book & essay · Machines: Infrastructure, energy & supply chain, Dataset & training data, Surveillance & biometrics - Idea: To see AI clearly, map what it extracts: earth, labour, data and people's categories. - What it is: Crawford follows AI from lithium mines and Amazon warehouses through training data and classification to affect recognition and state power, arguing that AI is neither artificial nor intelligent but a material industry. - How it works: Field research at mines, warehouses and archives, combined with the history of classification. - Lens: Hidden labour & extraction, Classification, bias & surveillance - Paper: https://doi.org/10.12987/9780300252392 (Yale University Press, 2021) - Video: https://www.youtube.com/watch?v=0PqrZ2lqT70 - Images: https://covers.openlibrary.org/b/id/14420723-L.jpg #### Discriminating Data — Wendy Hui Kyong Chun (2021) - Type: Book & essay · Machines: Dataset & training data, Platform & recommender algorithm - Idea: Predictive systems do not only reflect bias; their assumption that similar people cluster produces segregation. - What it is: A book showing how machine learning and network science inherit eugenics and segregation: correlation and 'homophily' sort people into neighbourhoods of the similar. - How it works: Critical history of statistics, eugenics and network analysis, with close readings of recommendation and ML methods. - Lens: Classification, bias & surveillance - Paper: https://doi.org/10.7551/mitpress/14050.001.0001 (MIT Press, 2021) - Project page: https://doi.org/10.7551/mitpress/14050.001.0001 #### On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? — Emily M. Bender, Timnit Gebru (2021) - Type: Paper · Machines: Chatbot & language model, Dataset & training data, Infrastructure, energy & supply chain - Idea: A language model stitches together forms without meaning; the coherence is in the reader's eye. - What it is: Bender, Gebru, Angelina McMillan-Major and Margaret Mitchell set out the environmental costs, undocumented training data, encoded bias and illusion of meaning in ever larger language models; the paper's review led to Gebru's exit from Google. - How it works: Critical survey of the NLP literature on scale, energy, data documentation and bias. - Lens: Opacity & latent space, Classification, bias & surveillance, Hidden labour & extraction - Paper: https://doi.org/10.1145/3442188.3445922 (FAccT 2021) - Video: https://www.youtube.com/watch?v=N5c2X8vhfBE #### Data Feminism — Catherine D'Ignazio, Lauren F. Klein (2020) - Type: Book & essay · Machines: Dataset & training data - Idea: Data is never raw; who collects it, counts whom and leaves whom out are feminist questions. - What it is: D'Ignazio and Klein set out seven principles, from examining power to making labour visible, for doing data science in a way that challenges rather than reinforces inequality. - How it works: Intersectional feminist theory applied to case studies of data projects; open access online. - Lens: Classification, bias & surveillance - Paper: https://doi.org/10.7551/mitpress/11805.001.0001 (MIT Press, 2020) - Video: https://www.youtube.com/watch?v=pmNEe6FvduM - Images: https://covers.openlibrary.org/b/isbn/9780262044004-L.jpg #### Ghost Work: How to Stop Silicon Valley from Building a New Global Underclass — Mary L. Gray, Siddharth Suri (2019) - Type: Book & essay · Machines: Dataset & training data, Platform & recommender algorithm - Idea: Behind the promise of automation there is always a "last mile" of hidden human work. - What it is: Gray and Suri describe the on-demand workers who label data, moderate content and check outputs behind apps and "AI", on platforms such as Amazon Mechanical Turk and UHRS in the US and India. - How it works: Five years of ethnography, surveys and experiments with crowdworkers. - Lens: Hidden labour & extraction - Paper: https://ghostwork.info (Houghton Mifflin Harcourt, 2019) - Video: https://www.youtube.com/watch?v=D5fkupjVEp8 - Images: https://covers.openlibrary.org/b/id/10131831-L.jpg - Project page: https://ghostwork.info #### Race After Technology: Abolitionist Tools for the New Jim Code — Ruha Benjamin (2019) - Type: Book & essay · Machines: Surveillance & biometrics, Platform & recommender algorithm - Idea: Discrimination can be engineered into systems that look neutral: the "New Jim Code". - What it is: Benjamin shows how automated systems in hiring, health, policing and design reproduce racial hierarchy while appearing objective, and collects abolitionist tools to counter them. - How it works: Sociological analysis of case studies from technology, law and design. - Lens: Classification, bias & surveillance - Paper: https://www.politybooks.com/bookdetail?book_slug=race-after-technology-abolitionist-tools-for-the-new-jim-code--9781509526390 (Polity, 2019) - Video: https://www.youtube.com/watch?v=rY8RkET3KC0 - Images: https://covers.openlibrary.org/b/id/10088781-L.jpg #### Algorithms of Oppression: How Search Engines Reinforce Racism — Safiya Umoja Noble (2018) - Type: Book & essay · Machines: Platform & recommender algorithm - Idea: A ranking algorithm is an editor with commercial interests, not a neutral window onto the world. - What it is: Noble documents how commercial search results, such as Google's for "black girls", reproduced racist and sexist stereotypes, and argues that ranking is shaped by advertising and power. - How it works: Content analysis of search results combined with Black feminist and information-science theory. - Lens: Classification, bias & surveillance - Paper: https://doi.org/10.18574/nyu/9781479833641.001.0001 (NYU Press, 2018) - Video: https://www.youtube.com/watch?v=6KLTpoTpkXo - Images: https://covers.openlibrary.org/b/id/12856940-L.jpg #### Psychopolitics: Neoliberalism and New Technologies of Power — Byung-Chul Han (2014) - Type: Book & essay · Machines: Dataset & training data, Platform & recommender algorithm - Idea: Surveillance today works through seduction and self-disclosure more than through prohibition. - What it is: A short essay arguing that data-driven 'smart power' no longer disciplines bodies but reads and steers the psyche; Big Data becomes a tool of psychopolitical control. - How it works: Philosophical essay (German original Psychopolitik, S. Fischer 2014; English trans. Erik Butler, Verso 2017). - Lens: Classification, bias & surveillance, Hidden labour & extraction - Paper: https://www.versobooks.com/products/226-psychopolitics (Verso (English ed. 2017)) - Images: https://www.versobooks.com/cdn/shop/files/getimage_05e7fdd2-de74-4073-8bc6-73bfa358d3cb.jpg?v=1790570326&width=600 - Project page: https://www.versobooks.com/products/226-psychopolitics #### Human-Machine Reconfigurations: Plans and Situated Actions — Lucy Suchman (2007) - Type: Book & essay · Machines: Humanoid & android, Social & companion robot - Idea: A machine seems intelligent because of the work people and settings do around it. - What it is: Suchman revisits her 1987 study of people failing to use an "intelligent" photocopier and extends it to humanlike robots such as Kismet and Cog, arguing that machine agency is made in situated interaction. - How it works: Ethnomethodology and feminist science studies applied to interaction analysis and robotics labs; 2nd expanded edition of Plans and Situated Actions. - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Paper: https://doi.org/10.1017/cbo9780511808418 (Cambridge University Press, 2007) - Video: https://www.youtube.com/watch?v=3_4-6gVbk4E - Images: https://covers.openlibrary.org/b/id/9840684-L.jpg ### Robotics, HRI & Embodied Cognition Embodied robotics, social robotics and research on how people relate to robots. #### The New Breed: What Our History with Animals Reveals about Our Future with Robots — Kate Darling (2021) - Type: Book & essay · Machines: Social & companion robot, Robot creature & mobile robot - Idea: We already know how to live with non-human others we use, love and mourn; robots can join that history. - What it is: Darling argues that robots are better understood on the model of animals than of humans, and uses cases such as soldiers holding funerals for bomb-disposal robots and people refusing to hurt a Pleo dinosaur. - How it works: Popular-science argument combining HRI experiments, legal history of animals and ethics. - Lens: Intimacy, care & companionship, Alterity & the uncanny - Paper: https://us.macmillan.com/books/9781250296115/thenewbreed/ (Henry Holt, 2021) - Video: https://www.youtube.com/watch?v=sTJ7E7TDTrU - Images: https://mpd-biblio-covers.imgix.net/9781250296115.jpg?fm=jpg - Project page: https://us.macmillan.com/books/9781250296115/thenewbreed/ #### Alone Together: Why We Expect More from Technology and Less from Each Other — Sherry Turkle (2011) - Type: Book & essay · Machines: Social & companion robot, Chatbot & language model - Idea: We are ready to accept a performance of care in place of care, and that says more about us than about the machines. - What it is: Based on fifteen years of interviews, Turkle describes a "robotic moment" in which children and elderly people accept Furbies, AIBO, My Real Baby and the seal robot Paro as companions that are "alive enough". - How it works: Clinical interviews and field observation in homes, schools and nursing homes. - Lens: Intimacy, care & companionship, Alterity & the uncanny - Paper: https://www.basicbooks.com/titles/sherry-turkle/alone-together/9780465093656/ (Basic Books, 2011) - Video: https://www.youtube.com/watch?v=MtLVCpZIiNs - Images: https://covers.openlibrary.org/b/id/6952623-L.jpg #### How the Body Shapes the Way We Think: A New View of Intelligence — Rolf Pfeifer, Josh Bongard (2006) - Type: Book & essay · Machines: Robot creature & mobile robot, Humanoid & android - Idea: Some of a robot's intelligence is in its shape and materials, not its controller. - What it is: Pfeifer and Bongard argue that intelligence depends on the morphology and materials of a body interacting with its environment, using passive-dynamic walkers and compliant robots as examples. - How it works: Synthesis of embodied AI research with design principles for robots. - Lens: Embodiment & kinesthesia - Paper: https://doi.org/10.7551/mitpress/3585.001.0001 (MIT Press, 2006) - Video: https://www.youtube.com/watch?v=mhWwoaoxIyc - Images: https://covers.openlibrary.org/b/id/13662665-L.jpg #### In the Hands of Machines? The Future of Aged Care — Robert Sparrow (2006) - Type: Paper · Machines: Social & companion robot - Idea: A machine that simulates care may be a harm, not a help, when it replaces the people who would have cared. - What it is: Robert and Linda Sparrow argue that care robots for older people would reduce human contact and rely on deceiving vulnerable people into believing a machine cares for them. - How it works: Ethical analysis of the tasks of aged care against the capacities of existing and projected robots. - Lens: Intimacy, care & companionship, Alterity & the uncanny - Paper: https://doi.org/10.1007/s11023-006-9030-6 (Minds and Machines 16(2), 2006) #### Designing Sociable Robots — Cynthia Breazeal (2002) - Type: Book & essay · Machines: Social & companion robot, Humanoid & android - Idea: A robot that shows needs invites people to care for it, and that care is what makes it social. - What it is: Breazeal's book on Kismet, the expressive robot head she built at MIT, lays out how a robot can use drives, emotions and facial expression to engage people as a caregiver engages an infant. - How it works: Synthesis of her doctoral work: motivational and emotional models, perception of faces and voice prosody, and an actuated face with eyebrows, ears, eyelids and lips. - Lens: Intimacy, care & companionship, Alterity & the uncanny - Paper: https://doi.org/10.7551/mitpress/2376.001.0001 (MIT Press, 2002) - Images: https://covers.openlibrary.org/b/id/151207-L.jpg https://upload.wikimedia.org/wikipedia/commons/thumb/1/1e/Kismet-IMG_6014.JPG/960px-Kismet-IMG_6014.JPG #### The Cog Project: Building a Humanoid Robot — Rodney Brooks (1999) - Type: Paper · Machines: Humanoid & android - Idea: Giving a robot a humanlike body and social development makes people treat it as someone, which in turn shapes what it learns. - What it is: Brooks, Breazeal, Marjanović, Scassellati and Williamson report on Cog, an upper-torso humanoid at MIT built to develop skills such as eye movements, reaching and joint attention by interacting with people, as an infant would. - How it works: Upper-torso humanoid platform with active vision and behaviour-based control, developed 1993–2004. - Lens: Embodiment & kinesthesia, Alterity & the uncanny - Paper: https://doi.org/10.1007/3-540-48834-0_5 (Computation for Metaphors, Analogy, and Agents, LNCS 1562, Springer, 1999) - Video: https://www.youtube.com/watch?v=oWjWzDzfLBo - Images: https://upload.wikimedia.org/wikipedia/commons/thumb/e/e3/Cog%2C_1993-2004%2C_view_1_-_MIT_Museum_-_DSC03732.JPG/960px-Cog%2C_1993-2004%2C_view_1_-_MIT_Museum_-_DSC03732.JPG #### The Media Equation: How People Treat Computers, Television, and New Media Like Real People and Places — Byron Reeves, Clifford Nass (1996) - Type: Book & essay · Machines: Voice & listening machine, Social & companion robot - Idea: Media = real life: social responses to machines are automatic, so design always addresses a social animal. - What it is: Reeves and Nass report dozens of experiments in which people are polite to computers, respond to their flattery, and assign them gender and personality, even while saying they know the machine is not a person. - How it works: Classic social-psychology studies replicated with computers and television in place of people. - Lens: Alterity & the uncanny, Intimacy, care & companionship - Paper: https://en.wikipedia.org/wiki/The_Media_Equation (Cambridge University Press / CSLI Publications, 1996) - Images: https://covers.openlibrary.org/b/id/829557-L.jpg #### Computers Are Social Actors — Clifford Nass (1994) - Type: Paper · Machines: Voice & listening machine - Idea: Minimal social cues, a voice and a role, are enough to trigger full social behaviour towards a machine. - What it is: Nass, Jonathan Steuer and Ellen R. Tauber ran a series of experiments showing that people apply politeness norms and gender stereotypes to computers that speak, for example rating a computer more kindly when that same computer asks. - How it works: Controlled lab experiments with tutoring and evaluating computers using recorded voices. - Lens: Alterity & the uncanny - Paper: https://doi.org/10.1145/191666.191703 (CHI 1994) - Video: https://www.youtube.com/watch?v=e7PvUUhDrfk #### Intelligence without Representation — Rodney Brooks (1991) - Type: Paper · Machines: Robot creature & mobile robot - Idea: Intelligence can be situated and embodied rather than planned from an internal model, and it looks more alive that way. - What it is: Brooks argues that robots should be built from layers of simple behaviours coupled directly to sensors and motors, using "the world as its own best model", as in his six-legged robot Genghis walking over rough ground. - How it works: Subsumption architecture: stacked finite-state machines that suppress or inhibit each other, tested on mobile robots at the MIT AI Lab. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Paper: https://doi.org/10.1016/0004-3702(91)90053-M (Artificial Intelligence 47(1–3), 1991) - Video: https://www.youtube.com/watch?v=-6piNZBFdfY #### The Uncanny Valley (Bukimi no Tani) — Masahiro Mori (1970) - Type: Paper · Machines: Humanoid & android, Prosthesis & cyborg - Idea: Near-likeness is more disturbing than obvious difference, so designers may be wiser to stop short of the human. - What it is: A short essay in the Japanese magazine Energy proposing that our affinity for robots rises with human likeness until it drops sharply into a 'valley' of eeriness; the examples are prosthetic hands, Bunraku puppets and corpses. - How it works: A hand-drawn graph of affinity against human likeness, with separate curves for still and moving objects; English translation by Karl MacDorman and Norri Kageki (2012). - Lens: Alterity & the uncanny - Paper: https://doi.org/10.1109/MRA.2012.2192811 (Energy 7(4), 1970; English: IEEE Robotics & Automation Magazine, 2012) - Video: https://www.youtube.com/watch?v=0g0sH3jTaVc - Images: https://spectrum.ieee.org/media-library/the-uncanny-valley-by-masahiro-mori-republished-in-robotics-automation-magazine.png?id=25569782&width=980 https://upload.wikimedia.org/wikipedia/commons/f/f0/Mori_Uncanny_Valley.svg - Project page: https://spectrum.ieee.org/the-uncanny-valley #### An Experimental Study of Apparent Behavior — Fritz Heider, Marianne Simmel (1944) - Type: Paper · Machines: Robot creature & mobile robot - Idea: Motion alone is enough for us to see agents with intentions and feelings, which is why simple robots feel alive. - What it is: Heider and Simmel showed viewers a 2.5-minute film of two triangles and a circle moving around a box; almost everyone described a story of a bully, a chase and a couple escaping. - How it works: Stop-motion animation of geometric shapes shown to 114 students, who then described or answered questions about it. - Lens: Alterity & the uncanny, Agency, autonomy & co-creation - Paper: https://doi.org/10.2307/1416950 (American Journal of Psychology 57(2), 1944) - Video: https://www.youtube.com/watch?v=n9TWwG4SFWQ ### Art & Media Theory Writing on robot art, AI art and machine images. #### Machine Vision: How Algorithms Are Changing the Way We See the World — Jill Walker Rettberg (2023) - Type: Book & essay · Machines: Computer vision, Surveillance & biometrics - Idea: Machine vision is not one eye but a set of relations that make some things visible and others unseen. - What it is: Rettberg organises machine vision into five modes (seeing more, seeing differently, seeing everything, being seen and seeing less) and reads artworks such as Zizi – Queering the Dataset and The Normalizing Machine alongside smart doorbells and face filters. - How it works: Qualitative analysis of the Machine Vision in Art, Games and Narratives database (500 works) built by her ERC team; open access. - Lens: Machine seeing, Classification, bias & surveillance - Paper: https://www.politybooks.com/bookdetail?book_slug=machine-vision-how-algorithms-are-changing-the-way-we-see-the-world--9781509545223 (Polity, 2023) - Video: https://www.youtube.com/watch?v=Ybva0y9KF8Q - Images: https://covers.openlibrary.org/b/id/14726453-L.jpg - Project page: https://machine-vision.no #### Mean Images — Hito Steyerl (2023) - Type: Book & essay · Machines: Generative model & latent space, Dataset & training data - Idea: Generated images are "mean" in both senses: averages that also enforce norms and exclude. - What it is: Steyerl reads generative-AI images as statistical averages of scraped data, starting from how images of herself appear in the LAION-5B dataset and what models generate from her name. - How it works: Essay combining close reading of model outputs with the political economy of datasets. - Lens: Opacity & latent space, Classification, bias & surveillance - Paper: https://newleftreview.org/issues/ii140/articles/hito-steyerl-mean-images (New Left Review 140/141, 2023) - Video: https://www.youtube.com/watch?v=PFwaqpkd61M - Images: https://newleftreview.org/system/dragonfly/production/2023/04/28/7rn0jeq677_Steyerl_Figure_1.jpg #### The Perception Machine: Our Photographic Future between the Eye and AI — Joanna Zylinska (2023) - Type: Book & essay · Machines: Computer vision, Generative model & latent space - Idea: Perception is now a joint human–machine process, so critique has to work from inside it. - What it is: Zylinska argues that photography has become a planetary perception machine of cameras, sensors and algorithms, and asks how humans can still see and act within it. - How it works: Media-philosophical essay with her own artistic experiments; open access. - Lens: Machine seeing, Machine umwelt - Paper: https://doi.org/10.7551/mitpress/14471.001.0001 (MIT Press, 2023) - Video: https://www.youtube.com/watch?v=jsdZFeTs2VA #### Art in the Age of Machine Learning — Sofian Audry (2021) - Type: Book & essay · Machines: Generative model & latent space, Cybernetic & artificial-life system - Idea: The artist's material is the training process itself, and it has a longer history than the current hype admits. - What it is: Audry explains how machine-learning systems are trained and places ML art in the lineage of cybernetic, artificial-life and evolutionary art, with case studies of contemporary ML artworks. - How it works: Technical-conceptual framework (data, training, model) applied to case studies; Leonardo series. - Lens: Agency, autonomy & co-creation, Opacity & latent space - Paper: https://doi.org/10.7551/mitpress/12832.001.0001 (MIT Press / Leonardo, 2021) - Video: https://www.youtube.com/watch?v=TwtgaQSmnXc #### AI Art: Machine Visions and Warped Dreams — Joanna Zylinska (2020) - Type: Book & essay · Machines: Generative model & latent space, Computer vision - Idea: The interesting AI art is the one that shows how the machine works and who works for it. - What it is: Zylinska rejects the question "can computers be creative?" and instead reads works such as Lauren McCarthy's LAUREN, Pierre Huyghe's UUmwelt and Trevor Paglen's Sight Machine against neural-network imagery she calls kitsch. - How it works: Critical essay in the Media : Art : Write : Now series, open access; includes her own photo-and-AI project. - Lens: Opacity & latent space, Agency, autonomy & co-creation, Machine umwelt - Paper: https://www.openhumanitiespress.org/books/titles/ai-art/ (Open Humanities Press, 2020) - Video: https://www.youtube.com/watch?v=jASV6Hjd2bY - Images: https://covers.openlibrary.org/b/id/14833507-L.jpg - Project page: https://www.openhumanitiespress.org/books/titles/ai-art/ #### Platform Seeing: Image Ensembles and Their Invisualities — Adrian Mackenzie, Anna Munster (2019) - Type: Paper · Machines: Computer vision, Dataset & training data, Platform & recommender algorithm - Idea: Machine vision happens in datasets and pipelines, not in any single image or eye. - What it is: Mackenzie and Munster argue that on platforms, seeing is done by huge ensembles of images, formats, hardware and models, and that this "platform seeing" produces invisualities rather than views. - How it works: Media-theoretical analysis of deep-learning datasets such as ImageNet and their hardware. - Lens: Machine seeing - Paper: https://doi.org/10.1177/0263276419847508 (Theory, Culture & Society 36(5), 2019) #### AI Aesthetics — Lev Manovich (2018) - Type: Book & essay · Machines: Generative model & latent space, Platform & recommender algorithm - Idea: AI's biggest aesthetic effect is not generating art but quietly sorting and predicting taste. - What it is: A short book on how AI already shapes everyday aesthetic decisions, from photo filters and recommendations to design, and what it means for culture when machines learn and predict taste. - How it works: Essay from the Strelka Press series, drawing on his cultural-analytics research. - Lens: Agency, autonomy & co-creation, Opacity & latent space - Paper: http://manovich.net/index.php/projects/ai-aesthetics (Strelka Press, 2018) - Video: https://www.youtube.com/watch?v=oR1NyrAY_w4 - Images: https://manovich.net/content/04-projects/184-ai-aesthetics/ai_new.jpg - Project page: http://manovich.net/index.php/projects/ai-aesthetics #### The Robotic Imaginary: The Human and the Price of Dehumanized Labor — Jennifer Rhee (2018) - Type: Book & essay · Machines: Humanoid & android, Autonomous vehicle & drone, Social & companion robot - Idea: To ask what makes a robot humanlike is to ask whose labour counts as human. - What it is: Rhee reads robots in AI research, film, literature and art, from ELIZA and Kismet to drone art such as #NotABugSplat, and shows that each definition of the humanlike robot rests on devalued human labour. - How it works: Cultural analysis across chapters on care labour, emotional labour, drone warfare and the Turing test. - Lens: Hidden labour & extraction, Intimacy, care & companionship, Alterity & the uncanny - Paper: https://doi.org/10.5749/j.ctv62hh4x (University of Minnesota Press, 2018) - Video: https://www.youtube.com/watch?v=7APzxormknw - Images: https://minnesota-us.imgix.net/covers/9781517902988.jpg?w=400 #### Art in the Age of Machine Intelligence — Blaise Agüera y Arcas (2017) - Type: Paper · Machines: Generative model & latent space, Computer vision - Idea: Like photography, machine intelligence will change what art is, and artists should shape it rather than just react. - What it is: Agüera y Arcas compares machine learning to photography's arrival in the 19th century, uses DeepDream as an example, and argues that artists should work with neural networks as a new medium. - How it works: Essay by the founder of Google's Artists + Machine Intelligence programme, open access. - Lens: Agency, autonomy & co-creation - Paper: https://doi.org/10.3390/arts6040018 (Arts 6(4), 2017) - Video: https://www.youtube.com/watch?v=uSUOdu_5MPc #### Making Sense: Cognition, Computing, Art, and Embodiment — Simon Penny (2017) - Type: Book & essay · Machines: Robot creature & mobile robot, Cybernetic & artificial-life system - Idea: Artists who build behaving machines have long practised a theory of embodied cognition that AI ignored. - What it is: Penny critiques the computational theory of mind and draws on enactive and distributed cognition, cybernetics and his own robotic works such as Petit Mal to argue for a performative aesthetics of interactive art. - How it works: Book-length theory from four decades of practice, published in the Leonardo series. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Paper: https://mitpress.mit.edu/9780262036757/ (MIT Press / Leonardo, 2017) - Video: https://www.youtube.com/watch?v=PU_WDouhkw8 #### A Sea of Data: Apophenia and Pattern (Mis-)Recognition — Hito Steyerl (2016) - Type: Book & essay · Machines: Surveillance & biometrics, Dataset & training data - Idea: Pattern recognition is also pattern projection; machines hallucinate the order their owners want to see. - What it is: Steyerl writes about intelligence agencies and neural networks drowning in data and finding patterns that are not there, using Google's DeepDream images as an example of apophenia. - How it works: Essay combining media theory, surveillance history and image analysis. - Lens: Opacity & latent space, Classification, bias & surveillance - Paper: https://www.e-flux.com/journal/72/60480/a-sea-of-data-apophenia-and-pattern-mis-recognition/ (e-flux journal 72, 2016) - Images: https://images.e-flux.com/test/cad7b8d2-d37e-4c14-825a-431c296346ff?width=1200&height=630 #### Invisible Images (Your Pictures Are Looking at You) — Trevor Paglen (2016) - Type: Book & essay · Machines: Computer vision, Dataset & training data - Idea: The most consequential images are ones no human looks at. - What it is: Paglen argues that most images are now made by machines for other machines, such as licence-plate readers and face recognisers, and that this invisible visual culture escapes the tools of art history. - How it works: Essay drawing on his research with computer-vision systems for the works It Began as a Military Experiment and Sight Machine. - Lens: Machine seeing, Opacity & latent space - Paper: https://thenewinquiry.com/invisible-images-your-pictures-are-looking-at-you/ (The New Inquiry, 8 December 2016) - Video: https://www.youtube.com/watch?v=JZjGsaJ1OcU - Images: https://thenewinquiry.com/app/uploads/2016/12/57paglen01.jpg #### The Multiple Bodies of a Machine Performer — Louis-Philippe Demers (2016) - Type: Paper · Machines: Robot creature & mobile robot, Industrial robot arm - Idea: A machine performer's presence comes from how its movement is read, not from how humanlike it looks. - What it is: Demers analyses how audiences perceive agency in robots on stage, drawing on his own works such as The Tiller Girls, and separates a machine performer's physical, perceived and narrative bodies. - How it works: Practice-based analysis in the volume Robots and Art: Exploring an Unlikely Symbiosis, edited by Herath, Kroos and Stelarc. - Lens: Embodiment & kinesthesia, Alterity & the uncanny, Failure, glitch & error - Paper: https://doi.org/10.1007/978-981-10-0321-9_14 (Robots and Art, Springer, 2016) - Video: https://www.youtube.com/watch?v=Bf-r34FzLhs #### Phantom Images — Harun Farocki (2004) - Type: Book & essay · Machines: Computer vision, Autonomous vehicle & drone - Idea: When images become steps in a process, seeing turns into doing, and the human viewer drops out. - What it is: Farocki's essay on "operational images", images from missile cameras and industrial vision systems that neither entertain nor inform but are part of an operation, written alongside his installation cycle Eye/Machine (2001–03). - How it works: Essay based on archival footage of guided weapons and machine-vision systems gathered for Eye/Machine. - Lens: Machine seeing - Paper: https://www.harunfarocki.de/installations/2000s/2003/eye-machine-iii.html (Public 29, 2004) - Video: https://www.youtube.com/watch?v=tERIscWmpSo - Project page: https://www.harunfarocki.de/installations/2000s/2003/eye-machine-iii.html #### Foundation and Development of Robotic Art — Eduardo Kac (1997) - Type: Paper · Machines: Robot creature & mobile robot, Humanoid & android, Cybernetic & artificial-life system - Idea: Robotic art split early into three lines: anthropomorphic, cybernetic-responsive and bio-hybrid. - What it is: Kac traces robotic art from three 1960s works, Nam June Paik and Shuya Abe's Robot K-456, Tom Shannon's Squat and Edward Ihnatowicz's SAM and Senster, to telepresence and web-controlled robots of the 1990s. - How it works: Art-historical survey in the Art Journal issue on art and technology. - Lens: Embodiment & kinesthesia, Agency, autonomy & co-creation - Paper: https://doi.org/10.1080/00043249.1997.10791834 (Art Journal 56(3), 1997) - Project page: https://www.ekac.org #### The Vision Machine — Paul Virilio (1988) - Type: Book & essay · Machines: Computer vision, Autonomous vehicle & drone, Surveillance & biometrics - Idea: When machines see for us, perception becomes logistics and the human eye is optional. - What it is: Virilio describes the rise of automated perception, cameras and computers that see and interpret without a human viewer, as a "sightless vision" rooted in military logistics. - How it works: Essay in media archaeology, from perspective painting to smart weapons; French original La Machine de vision (Galilée, 1988). - Lens: Machine seeing - Paper: https://iupress.org/9780253209016/the-vision-machine/ (Galilée, 1988; English: BFI / Indiana University Press, 1994) - Images: https://covers.openlibrary.org/b/id/148033-L.jpg #### Beyond Modern Sculpture: The Effects of Science and Technology on the Sculpture of This Century — Jack Burnham (1968) - Type: Book & essay · Machines: Cybernetic & artificial-life system, Robot creature & mobile robot - Idea: Sculpture's long wish to make the object alive ends in the machine that behaves. - What it is: Burnham surveys twentieth-century sculpture from vitalism to kinetic, light and cybernetic art, and predicts a move from static objects to responsive systems and artificial life. - How it works: Art-historical and systems-theoretical survey. - Lens: Agency, autonomy & co-creation, Embodiment & kinesthesia - Paper: https://archive.org/details/beyondmodernscul0000burn (George Braziller, 1968) - Images: https://covers.openlibrary.org/b/id/14444956-L.jpg #### Cybernetic Serendipity: The Computer and the Arts — Jasia Reichardt (1968) - Type: Book & essay · Machines: Cybernetic & artificial-life system, Robot creature & mobile robot - Idea: The first large art show of computers presented machines as collaborators and characters, not just tools. - What it is: The catalogue of Reichardt's ICA exhibition, published as a special issue of Studio International, documenting computer graphics, music, poetry and cybernetic machines by Gordon Pask, Nam June Paik, Edward Ihnatowicz and many others. - How it works: Exhibition catalogue with artists' statements, diagrams and essays by scientists and engineers. - Lens: Agency, autonomy & co-creation - Paper: https://cyberneticserendipity.net (Studio International special issue, 1968) - Video: https://www.youtube.com/watch?v=5mH9DysESqo - Images: https://covers.openlibrary.org/b/id/8992379-L.jpg - Project page: https://cyberneticserendipity.net ## Resources & links: people, institutions and websites ### Artists & designers - **!Mediengruppe Bitnik** (Zurich, CH / London, UK) — Swiss artist duo (Carmen Weisskopf and Domagoj Smoljo) whose Random Darknet Shopper, a bot with a weekly bitcoin budget, bought goods on the darknet and was then 'arrested' by Swiss police. Known for: Delivery for Mr. Assange (2013); Random Darknet Shopper (2014); Alexiety (2019–). https://www.bitnik.org/ - **aaajiao (Xu Wenkai) / 徐文恺** (Berlin, DE / Shanghai, CN) — Chinese artist who works with the internet, bots and data as the material of everyday life under networked control; his avatar aaajiao reflects on identity inside platforms. Known for: GFWlist (2010); Remnants of an Electronic Past (2016); Deep Sleep (2020). http://www.aaajiao.com/ - **Adam Harvey** (Berlin, DE, 1981–) — American artist-researcher of counter-surveillance whose CV Dazzle makeup defeats face detection, and whose Exposing.ai traced millions of photos scraped into face-recognition datasets. Known for: CV Dazzle (2010); Stealth Wear (2013); HyperFace (2017); Exposing.ai / MegaPixels (2019–, with Jules LaPlace). https://ahprojects.com/ - **Addie Wagenknecht** (Austria / US, 1981–) — American artist whose Optimization of Parenthood uses a robot arm that rocks her baby's cradle when it cries, questioning what care becomes when it is automated. Known for: Optimization of Parenthood (2011–); Asymmetric Love (2013); Data and Dragons (2015). https://www.placesiveneverbeen.com/ - **Agnieszka Kurant** (New York, US, 1978–) — Polish artist of collective and non-human intelligence; A.A.I. consists of termite mounds built from coloured sand, 'artificial artificial intelligence' that echoes the crowdworkers behind AI. Known for: A.A.I. (2014); Crowd Crystal (2019). https://en.wikipedia.org/wiki/Agnieszka_Kurant - **Agnieszka Pilat** (San Francisco, US) — Polish-born painter who treats Boston Dynamics' Spot robots as studio companions and co-painters, staging machines as a new species learning to make marks. Known for: Heterobota (NGV Triennial, 2023); Paintings with Boston Dynamics' Spot (2021–). https://www.agnieszkapilat.com/ - **Alexandra Daisy Ginsberg** (London, UK, 1982–) — British artist whose The Substitute brings back the last male northern white rhino as an AI-driven digital animal, asking why we prefer artificial life to the life we let die. Known for: The Substitute (2019); Resurrecting the Sublime (2019); Pollinator Pathmaker (2021). https://www.daisyginsberg.com/ - **American Artist** (Los Angeles, US, 1989–) — American artist whose My Blue Window renders the city through a predictive-policing dashboard, confronting the racial logic of computational risk. Known for: Black Gooey Universe (2018); My Blue Window (2019); Shady Grove (2021). https://americanartist.us/ - **Andrew Norman Wilson** (Los Angeles, US, 1983–) — American artist who, while a contractor at Google, filmed the separately badged book-scanning workers, and then collected Google Books pages showing the scanners' fingers. Known for: Workers Leaving the Googleplex (2011); ScanOps (2012–). https://www.andrewnormanwilson.com/ - **Anicka Yi** (New York, US, 1971–) — Korean-American artist who 'biologises the machine': in In Love With the World, floating 'aerobes' drifted through Tate's Turbine Hall as a new machine species with their own sensing and behaviour. Known for: Biologizing the Machine (tentacular trouble) (2019); In Love With the World (Tate Turbine Hall, 2021). https://en.wikipedia.org/wiki/Anicka_Yi - **Anna Ridler** (London, UK, 1985–) — British artist who hand-builds her own datasets; Myriad (Tulips) displays 10,000 tulip photos she took and labelled, making visible the human labour and choices inside a training set. Known for: Myriad (Tulips) (2018); Mosaic Virus (2018); Traces of Things (2021). https://annaridler.com/ - **Annie Dorsen** (New York, US, 1973–) — American theatre maker of 'algorithmic theatre'; in Hello Hi There two chatbots restage the Chomsky–Foucault debate on human nature, live and different each night. Known for: Hello Hi There (2010); A Piece of Work (2013); Prometheus Firebringer (2023). https://www.anniedorsen.com/ - **Arthur Ganson** (Cambridge, US, 1955–) — American kinetic sculptor of gentle, philosophical machines; in Machine with Concrete a motor drives a chain of gears whose last one is set in concrete and would turn once in two trillion years. Known for: Machine with Concrete (1992); Machine with Oil (1990); Cory's Yellow Chair (1995). https://www.arthurganson.com/ - **Auger & Loizeau (James Auger & Jimmy Loizeau)** (London, UK) — British speculative designers whose Carnivorous Domestic Entertainment Robots feed on flies and mice to power themselves, re-imagining domestic robots as living things. Known for: Audio Tooth Implant (2001); Carnivorous Domestic Entertainment Robots (2009); Afterlife (2001–09). https://www.wikidata.org/wiki/Q55217959 - **Automato.farm** (Shanghai, CN / London, UK) — Design studio of Simone Rebaudengo, Matteo Loglio and Saurabh Datta; in Ethical Things a fan asks online crowdworkers to resolve its moral dilemmas. Known for: Objects of Research (2016); Politics of Power (2016); Ethical Things (2015); BIY.world (2019). http://automato.farm/ - **Ayoung Kim / 金雅瑛** (Seoul, KR, 1979–) — South Korean artist whose Delivery Dancer's Sphere follows a female delivery rider in a Seoul run by an algorithmic platform that bends time and space. Known for: Delivery Dancer's Sphere (2022); Delivery Dancer's Arc: 0° Receiver (2024); PH Express (2018). https://de.wikipedia.org/wiki/Ayoung_Kim - **Ben Grosser** (Urbana, US) — American artist of software culture; Computers Watching Movies shows what a computer vision system 'looks at' in famous films, as sketched traces of machine attention. Known for: Computers Watching Movies (2013); Facebook Demetricator (2012–); Go Rando (2017); Minus (2021). https://bengrosser.com/ - **Bill Vorn** (Montreal, CA, 1959–) — Canadian robotic artist whose machines act out misery and hysteria: the octopus-like Hysterical Machines flail when visitors approach, provoking empathy for crude mechanisms. Known for: La Cour des Miracles (1997, with Louis-Philippe Demers); Hysterical Machines (2006); Inferno (2015). https://billvorn.concordia.ca/ - **Bjørn Karmann** (Copenhagen, DK) — Danish designer whose Project Alias is a fungus-like parasite that whispers noise into smart speakers until you wake it, and whose Paragraphica is a lensless camera that writes a photo from data. Known for: Objectifier (2016); Project Alias (2019, with Tore Knudsen); Paragraphica (2023). https://bjoernkarmann.dk/ - **Blanca Li** (Paris, FR, 1964–) — Spanish choreographer whose Robot sets dancers alongside NAO humanoid robots and Maywa Denki's instruments, exploring how people and machines learn each other's movement. Known for: Robot (2013); Le Bal de Paris (2021). https://www.blancali.com/ - **Blast Theory** (Brighton, UK) — British artist group of interactive performance; Cat Royale (2023) watched a robot arm try to entertain three cats for twelve days, asking whether machines can care for other species. Known for: Can You See Me Now? (2001); Cat Royale (2023). https://www.blasttheory.co.uk/ - **Bruce Lacey** (London, UK, 1927–2016) — British artist of anarchic, junk-built robots and performing automata, several shown at Cybernetic Serendipity (1968); his humanoid machines were deliberately comic, clumsy and vulnerable. Known for: ROSA BOSOM (1965); Old Money Bags (1965); Cybernetic Serendipity robots (1968). https://en.wikipedia.org/wiki/Bruce_Lacey - **Cao Fei / 曹斐** (Beijing, CN, 1978–) — Chinese artist of labour, virtual worlds and automation; Asia One follows the only two workers in a fully automated logistics warehouse, falling in love among sorting robots. Known for: Whose Utopia (2006); RMB City (2007–11); Asia One (2018); Staging the Era (2023). http://www.caofei.com/ - **Caroline Sinders** (London, UK / New York, US) — American designer-researcher whose Feminist Data Set interrogates every step of making AI, from collecting data to paying the labellers, as a feminist practice. Known for: Feminist Data Set (2017–); Technically Responsible Knowledge (2020). https://carolinesinders.com/ - **Catie Cuan** (Stanford, US) — American dancer-roboticist who choreographs with industrial and mobile robots; in Output she and an ABB arm trade movement captured from each other. Known for: Output (2019); Time to Compile (2019); Engine of Wonder (2023, TED). https://www.catiecuan.com/ - **Chico MacMurtrie** (New York, US, 1961–) — American artist whose Amorphic Robot Works builds pneumatic and inflatable robotic bodies that breathe, drum and climb, machines that behave like primal organisms. Known for: Amorphic Robot Works (founded 1991); The Robotic Church (2006); Inflatable Architectural Bodies (2000s–). https://amorphicrobotworks.org/ - **Choe U-Ram** (Seoul, KR, 1970–) — South Korean sculptor of 'anima-machines', delicate mechanical creatures with invented scientific names and life cycles; Little Ark is a ship of oars that rise and fall like breathing. Known for: Opertus Lunula Umbra (2008); Little Ark (MMCA, 2022); Una Lumino (2008). https://www.uram.net/ - **Choy Ka Fai / 徐家辉** (Berlin, DE / Singapore) — Singaporean artist whose Prospectus for a Future Body stimulates dancers' muscles with electrical signals recorded from other dancers, asking whether movement memory can be transferred by machine. Known for: Prospectus for a Future Body (2011); UnBearable Darkness (2018). https://www.ka5.info/ - **Christa Sommerer & Laurent Mignonneau** (Linz, AT) — Austrian-French duo, pioneers of artificial-life art: in A-Volve visitors draw creatures that then evolve, hunt and mate in a pool of water, living by their own rules. Known for: Interactive Plant Growing (1992); A-Volve (1994); Life Writer (2006). https://www.interface.ufg.ac.at/christa-laurent/ - **Clement Valla** (New York, US) — French-American artist whose Postcards from Google Earth collects melted bridges and warped roads, not errors but the logic of a machine-made 3D world. Known for: Postcards from Google Earth (2010–); Surface Survey (2014). https://clementvalla.com/ - **Cod.Act (André & Michel Décosterd)** (La Chaux-de-Fonds, CH) — Swiss brothers whose sound-machine performances fuse bodies and mechanics: in Pendulum Choir nine singers tilt on hydraulic jacks, their voices shaped by the machine's movement. Known for: Pendulum Choir (2011); πTon (2017); Nyloïd (2013). https://www.codact.ch/ - **Daito Manabe / 真锅大度** (Tokyo, JP, 1976–) — Japanese artist and Rhizomatiks co-founder whose Electric Stimulus to Face let a computer drive his facial muscles, later extending to drones, dancers and brain-decoding works. Known for: Electric Stimulus to Face (2008); Rhizomatiks × ELEVENPLAY (2010s); dissonant imaginary (2023). https://www.daito.ws/ - **Dan Chen** — Designer whose RISD-era robotic intimacy series built machines that caress and comfort people; his Last Moment Robot strokes a dying patient's arm and speaks words of comfort, questioning machine care at the end of life. Known for: End of Life Care Machine (2012); Last Moment Robot (2012); Intimacy series (2011–12). https://dankc.com/ - **David Bowen** (Minneapolis, US) — American artist whose machines let non-humans and data drive them: a houseplant wields a machete, and houseflies control a revolver, making other perceptual worlds act through robots. Known for: fly revolver (2013); tele-present wind (2018); plant machete (2014). https://www.dwbowen.com/ - **David Harris Smith & Frauke Zeller** (Hamilton / Toronto, CA) — Canadian researchers who made hitchBOT, a hitchhiking robot that crossed Canada and parts of Europe on strangers' goodwill before being destroyed in Philadelphia in 2015. Known for: hitchBOT (2014–15). https://en.wikipedia.org/wiki/HitchBOT - **David Rokeby** (Toronto, CA, 1960–) — Canadian pioneer of interactive installation; The Giver of Names looks at objects visitors place before it and generates associative sentences about them, an early portrait of machine interpretation. Known for: Very Nervous System (1986–90); The Giver of Names (1990–); Sorting Daemon (2003). http://www.davidrokeby.com/ - **Desmond Paul Henry** (Manchester, UK, 1921–2004) — British artist who built drawing machines from bombsight analogue computers in the early 1960s, among the first machine-made artworks exhibited in the UK. Known for: Henry Drawing Machine (1961); Ideographs exhibition (Reid Gallery, London, 1962). http://www.desmondhenry.com/ - **Diemut Strebe** (Boston, US) — German artist working with MIT scientists; The Prayer is a robotic mouth that composes and speaks its own prayers from a language model, asking whether a machine can pray. Known for: The Prayer (2020); Redemption of Vanity (2019). https://www.diemutstrebe.com/ - **Dries Depoorter** (Ghent, BE) — Belgian artist whose The Follower matched Instagram photos with open webcam footage of the moment they were taken, revealing how easily AI links public cameras to private images. Known for: Jaywalking (2015); Surveillance Speaker (2020); The Follower (2022). https://driesdepoorter.be/ - **Dunne & Raby (Anthony Dunne & Fiona Raby)** (New York, US) — British designers who defined critical and speculative design; Technological Dreams: Robots imagines domestic robots that are needy, neurotic or dependent rather than servile. Known for: Technological Dreams Series: No. 1, Robots (2007); Speculative Everything (2013); Hertzian Tales (1999). https://dunneandraby.co.uk/ - **Eduardo Kac** (Chicago, US, 1962–) — Brazilian-American artist of telepresence robotics and bio art; in Rara Avis visitors saw an aviary through the eyes of a robotic macaw, inhabiting a machine body. Known for: Ornitorrinco (1989–96, with Ed Bennett); Rara Avis (1996); Teleporting an Unknown State (1996); GFP Bunny (2000). https://www.ekac.org/ - **Edward Ihnatowicz** (London, UK, 1926–1988) — Polish-born British sculptor who made the first computer-controlled robotic sculptures: The Senster, a 4.5-metre hydraulic creature at Philips' Evoluon, turned towards quiet voices and shied away from loud ones. Known for: SAM (1968); The Senster (1970). http://www.senster.com/ - **Ei Wada / 和田永** (Tokyo, JP, 1987–) — Japanese artist who turns obsolete appliances, such as CRT televisions and electric fans, into musical instruments, giving discarded machines a second life as a community orchestra. Known for: Braun Tube Jazz Band (2009); Electronicos Fantasticos! (2015–). https://eiwada.com/ - **Elisa Giardina Papa** (New York, US / Sicily, IT) — Italian artist whose Cleaning Emotional Data documents her own gig work labelling facial expressions for emotion-recognition AI, and whose Technologies of Care portrays online care workers. Known for: Technologies of Care (2016); Cleaning Emotional Data (2020). https://www.elisagiardinapapa.org/ - **Esther Hovers** (Netherlands) — Dutch artist whose False Positives stages the eight 'anomalous behaviours' that intelligent surveillance systems are trained to flag, from standing still to walking against the flow. Known for: False Positives (2015). https://www.estherhovers.com/ - **Eva & Franco Mattes** (New York, US) — Italian artist duo whose Dark Content gives anonymised voices to the content moderators who filter violent and obscene images for platforms, the human layer behind 'automatic' moderation. Known for: Dark Content (2015); BEFNOED (2014); Personal Photographs (2018). https://0100101110101101.org/ - **Everest Pipkin** (Pittsburgh, US, 1990–) — American artist and game developer whose Lacework slowly reanimates every video in the MIT Moments in Time dataset, and whose essay on it documents the violence hidden in training data. Known for: Lacework (2020); Shell Song (2020). https://everest-pipkin.com/ - **Garnet Hertz** (Vancouver, CA, 1973–) — Canadian artist and scholar of critical making whose robot is steered by a live Madagascar hissing cockroach riding on a trackball. Known for: Cockroach Controlled Mobile Robot (2004); Critical Making (2012); Disobedient Electronics (2016–). http://www.conceptlab.com/ - **George E. Lewis** (New York, US, 1952–) — American composer, trombonist and AACM member whose Voyager is an improvising computer 'orchestra' that listens and answers, framed as a partner shaped by African American musical culture. Known for: Voyager (1987); Why Do We Want Our Computers to Improvise? (2018). https://music.columbia.edu/bios/george-e-lewis - **Gilberto Esparza** (Mexico City, MX) — Mexican artist whose Plantas Nómadas is a robot that feeds its plants from polluted rivers using microbial fuel cells, a machine-organism for damaged ecologies (Golden Nica 2015). Known for: Parásitos Urbanos (2007); Plantas Nómadas (2008–13); Biocenosis (2021). https://gilbertoesparza.net/ - **Gordon Pask** (London, UK, 1928–1996) — British cybernetician who built responsive, learning machines as art: Musicolour grew bored of a musician who repeated himself, and Colloquy of Mobiles let hanging machines court each other and visitors at Cybernetic Serendipity. Known for: Musicolour (1953); Colloquy of Mobiles (1968); Conversation Theory (1975). https://en.wikipedia.org/wiki/Gordon_Pask - **Guo Cheng / 郭城** (Shanghai, CN) — Chinese artist of science-fictional machines; his gif tutorial shows how to twist your face to escape detection, and The Net Wanderer tours the physical handshakes of the Chinese internet. Known for: A Technique of Evading Face Detection (2020); The Net Wanderer – A Tour of Suspended Handshakes (2020). https://www.guo-cheng.net/ - **Guy Ben-Ary** (Perth, AU, 1967–) — Argentine-born Australian bio-artist whose works wire living neurons to machines: MEART drew with a robot arm, and cellF is a synthesizer played by neurons grown from his own cells. Known for: MEART – The Semi-Living Artist (2003); cellF (2015); Revivification (2021). https://guybenary.com/ - **Harold Cohen** (San Diego, US, 1928–2016) — British painter who spent four decades developing AARON, a drawing program and robot that raised the question of whether a machine can make art, and of what the artist's role becomes. Known for: AARON (1973–2016); What is an Image? (1979). https://www.aaronshome.com/ - **Harpreet Sareen** (New York, US) — Indian-born designer-researcher of 'cyborg botany' whose Elowan is a plant that drives its robotic base towards light using its own electrical signals. Known for: Elowan (2018, MIT Media Lab); Cyborg Botany (2019). https://www.harpreetsareen.com/ - **Harun Farocki** (Berlin, DE, 1944–2014) — German filmmaker who coined 'operative images', pictures made by machines for machines; the Eye/Machine trilogy examines smart-bomb cameras and industrial vision, founding the art of machine seeing. Known for: Images of the World and the Inscription of War (1988); Eye/Machine I–III (2001–03); Serious Games I–IV (2009–10); Parallel I–IV (2012–14). https://www.harunfarocki.de/ - **Hasan Elahi** (Washington, D.C., US, 1972–) — Bangladeshi-American artist who, after being wrongly placed on a US terrorist watch list, began publishing tens of thousands of images of his whereabouts, over-feeding the surveillance machine. Known for: Tracking Transience (2002–). https://en.wikipedia.org/wiki/Hasan_M._Elahi - **Heather Dewey-Hagborg** (New York, US, 1982–) — American artist who 3D-prints portraits of strangers from DNA found in the street, exposing the fictions and biases of forensic phenotyping algorithms. Known for: Stranger Visions (2012–13); Probably Chelsea (2017); Hybrid: an Interspecies Portrait (2023). https://deweyhagborg.com/ - **Hito Steyerl** (Berlin, DE, 1966–) — German artist and writer whose essay films and texts on poor images, robots and 'artificial stupidity' shaped how art thinks about computation, visibility and power. Known for: How Not to Be Seen (2013); HellYeahWeFuckDie (2016); This is the Future (2019); Medium Hot: Images in the Age of Heat (2025). https://en.wikipedia.org/wiki/Hito_Steyerl - **Ho Tzu Nyen / 何子彦** (Singapore, 1976–) — Singaporean artist whose CDOSEA is an algorithmically edited, infinitely recombining 'dictionary' of Southeast Asia, handing the ordering of history to a machine. Known for: The Critical Dictionary of Southeast Asia (CDOSEA) (2017–); The Name (2015); T for Time (2023). https://en.wikipedia.org/wiki/Ho_Tzu_Nyen - **Holly Herndon & Mat Dryhurst** (Berlin, DE) — American artist-musicians who raised the AI 'baby' Spawn with a vocal ensemble and released Holly+, a voice model anyone can sing through, raising questions of voice ownership and consent. Known for: PROTO (2019, with AI 'baby' Spawn); Holly+ (2021); xhairymutantx (Whitney Biennial, 2024); The Call (Serpentine, 2024). https://herndondryhurst.studio/ - **Hsin-Chien Huang / 黄心健** (Taipei, TW, 1966–) — Taiwanese new-media artist whose VR work The Eye and I, made with Jean-Michel Jarre, immerses viewers in a world of surveillance and watching machines. Known for: The Eye and I (2019, with Jean-Michel Jarre); To the Moon (2018, with Laurie Anderson); Bodyless (2019). https://www.storynest.com/ - **Hu Jieming / 胡介鸣** (Shanghai, CN) — Shanghai pioneer of Chinese media art whose installations such as 100 Years in 1 Minute, a matrix of 1,440 boxes, compress history into machine-scaled time. Known for: 100 Years in 1 Minute (2010); The Raft of the Medusa (2002). https://www.shanghartgallery.com/galleryarchive/artists/name/hujieming - **Huang Yi / 黄翊** (Taipei, TW) — Taiwanese choreographer who programs an industrial KUKA arm as his duet partner; Huang Yi & KUKA opened Ars Electronica 2013, toured widely and made a factory robot move with tenderness. Known for: Huang Yi & KUKA (2012); A Million Miles Away (2019). https://huangyistudio.com/ - **Ian Cheng** (New York, US, 1984–) — American artist of 'live simulations', AI-driven worlds that play themselves without end; BOB is an artificial lifeform with a personality that changes through its encounters. Known for: Emissaries trilogy (2015–17); BOB (Bag of Beliefs) (2018–19); Life After BOB (2021). https://iancheng.com/ - **Jaha Koo** (Ghent, BE, 1984–) — South Korean theatre-maker whose Cuckoo shares the stage with three talking rice cookers, telling a story of loneliness and economic crisis in South Korea through household machines. Known for: Cuckoo (2017); Hamartia trilogy (2015–2019). https://gujaha.com - **Jake Elwes** (London, UK, 1993–) — British artist who queers AI by retraining face models on drag performers; The Zizi Show is a deepfake cabaret where bodies break down at the edge of the dataset. Known for: Zizi – Queering the Dataset (2019); The Zizi Show (2020); CUSP (2019); Machine Learning Porn (2016). https://www.jakeelwes.com/ - **James Bridle** (Greece / UK, 1980–) — British artist and writer whose Autonomous Trap 001 caught a self-driving car inside a circle of road markings, and whose books New Dark Age and Ways of Being rethink computation and non-human intelligence. Known for: The New Aesthetic (2011–); Autonomous Trap 001 (2017); New Dark Age (2018); Ways of Being (2022). https://jamesbridle.com/ - **James Seawright** (New York, US, 1936–2022) — American sculptor who built early responsive electronic sculptures in the 1960s, machines that scanned the room, sensed visitors and answered with light and movement. Known for: Watcher (1965–66); Searcher (1966); Electronic Peristyle (1968). https://en.wikipedia.org/wiki/James_Seawright - **Jean Tinguely** (Basel, CH, 1924–1991) — Swiss kinetic sculptor of absurd, self-destroying machines: Métamatics produced drawings on demand, and Homage to New York shook itself to pieces in the MoMA garden, making breakdown the work. Known for: Métamatics (1959); Homage to New York (1960); Study for an End of the World No. 2 (1962). https://www.tinguely.ch/ - **Jenna Sutela** (Berlin, DE, 1983–) — Finnish artist of more-than-human intelligence; in nimiia cétiï a neural network learns a 'Martian' language from recordings of a medium and the movements of bacteria. Known for: nimiia cétiï (2018); I Magma (2019); Holobiont Society (2018); Pond Brain (2024). https://jennasutela.com/ - **Jennifer Walshe** (Dublin, IE, 1974–) — Irish composer and vocalist who improvises with an AI trained on her own face and voice (ULTRACHUNK) and rewrites music history through machine-generated singing. Known for: ULTRACHUNK (2018, with Memo Akten); A Late Anthology of Early Music Vol. 1 (2020). http://milker.org/ - **Jiabao Li / 李佳宝** (Austin, US) — Chinese artist-designer and founder of Reality Design Lab (UT Austin) whose works on AI, care and more-than-human perception include AI Mom, a machine that performs motherhood. Known for: AI Mom (2024); Womb Wetware (2025); Once Upon a Garden (2023). https://www.jiabaoli.org/ - **Jill Magid** (New York, US, 1973–) — American artist who in Evidence Locker asked Liverpool's CCTV operators to film her and wrote them love letters to retrieve the footage, making intimacy with the surveillance system. Known for: Evidence Locker (2004); Trust (2004). https://www.jillmagid.com/ - **Joana Moll** (Barcelona, ES) — Spanish artist-researcher who measures the material cost of the internet; CO2GLE shows the CO2 emitted every second by visits to google.com. Known for: CO2GLE (2014); The Hidden Life of an Amazon User (2019); The Dating Brokers (2018, with Tactical Tech). https://www.janavirgin.com/ - **Jon Rafman** (Montreal, CA, 1981–) — Canadian artist whose The Nine Eyes of Google Street View collects scenes captured by Street View's indifferent cameras, a portrait of the world seen by a machine that does not care. Known for: The Nine Eyes of Google Street View (2009–). https://jonrafman.com/ - **Jordan Wolfson** (New York / Los Angeles, US, 1980–) — American artist whose animatronic (Female figure) dances before a mirror while tracking viewers' faces, a disturbing study of the gaze and the sexualised robot body. Known for: (Female figure) (2014); Colored sculpture (2016); Body Sculpture (2023). https://en.wikipedia.org/wiki/Jordan_Wolfson - **Joy Buolamwini** (Cambridge, US, 1989–) — Ghanaian-American-Canadian 'poet of code' whose white-mask experiment and Gender Shades study exposed face-analysis systems failing on dark-skinned women; founder of the Algorithmic Justice League. Known for: Gender Shades (2018, with Timnit Gebru); AI, Ain't I a Woman? (2018); Unmasking AI (2023); Algorithmic Justice League (founded 2016). https://www.poetofcode.com/ - **Julian Oliver** (Berlin, DE) — New Zealand 'critical engineer' who co-wrote the Critical Engineering Manifesto; his works open up networks and infrastructures, such as a cell tower disguised as an office printer. Known for: The Critical Engineering Manifesto (2011); Stealth Cell Tower (2015); Harvest (2016). https://julianoliver.com/ - **Julien Prévieux / 朱利安·普雷维厄** (Paris, FR) — French artist whose What Shall We Do Next? has dancers perform gestures patented by tech companies, showing how bodies are measured and scripted by devices; Prix Marcel Duchamp 2014. Known for: What Shall We Do Next? (2006–); Patterns of Life (2015). https://www.previeux.net/ - **Julius Popp** (Leipzig, DE, 1973–) — German artist whose bit.fall writes words scraped from news sites in falling water drops, and whose small self-learning robots explore their own bodies. Known for: bit.fall (2001–); micro.adam and micro.eva (2002). https://en.wikipedia.org/wiki/Julius_Popp - **Justine Emard** (Paris, FR) — French artist whose Co(AI)xistence films dancer Mirai Moriyama learning to communicate with Alter, the learning android by Ikegami and Ishiguro, through gesture and voice. Known for: Co(AI)xistence (2017); Soul Shift (2018); Supraorganism (2020). https://www.justineemard.com/ - **Kacie Kinzer** (New York, US, 1983–) — American designer whose Tweenbots, smiling cardboard robots that could only roll straight ahead, crossed Manhattan parks by relying on strangers' kindness to redirect them. Known for: Tweenbots (2009). http://www.tweenbots.com/ - **Karen Palmer** (London, UK) — British 'storyteller from the future' whose interactive films read the viewer's facial emotions to steer the plot, so a police encounter plays out differently depending on how calm you look. Known for: RIOT (2017); Perception iO (2021). http://karenpalmer.uk/ - **Karl Sims** (Cambridge, US, 1962–) — American computer graphics artist whose Evolved Virtual Creatures evolved their own bodies and gaits in simulation, and whose Galápagos let visitors act as the selection pressure for artificial organisms. Known for: Evolved Virtual Creatures (1994); Galápagos (1997). https://www.karlsims.com/ - **Kazuhiko Hachiya / 八谷和彦** (Tokyo, JP) — Japanese media artist whose Inter Dis-Communication Machine swaps the sight and sound of two people through head-mounted displays, so each lives inside the other's perception. Known for: Inter Dis-Communication Machine (1993); PostPet (1997); OpenSky (2003–). https://www.ntticc.or.jp/en/archive/works/inter-dis-communication-machine/ - **Keiichi Matsuda** (London, UK) — British-Japanese designer whose film Hyper-Reality shows a Medellín bus ride in a city saturated by augmented-reality ads, gamification and assistant agents. Known for: Hyper-Reality (2016); Merger (2018); Agents (Wanderer, 2020s). http://km.cx/ - **Keiichiro Shibuya / 涩谷庆一郎** (Tokyo, JP, 1973–) — Japanese composer whose android opera Scary Beauty has the android Alter 3 conduct an orchestra and sing, with tempo it decides itself. Known for: The End (2012); Scary Beauty (2018); MIRROR (2022). https://atak.jp/ - **Kelly Dobson** (Providence, US) — American artist of 'machine therapy': to make Blendie spin, you must growl at a blender in its own pitch, an exercise in empathy with household machines. Known for: Blendie (2003–04); ScreamBody (1998–2004); Omo (2007). http://www.kellydobson.com/ - **Ken Feingold** (New York, US, 1952–) — American artist whose talking animatronic heads converse with each other in synthetic speech; in If/Then two identical heads debate whether they exist. Known for: If/Then (2001); Sinking Feeling (2001); Head (1999–2000). http://www.kenfeingold.com/ - **Ken Goldberg** (Berkeley, US, 1961–) — American roboticist and artist whose Telegarden let thousands of internet users tend a real garden through an industrial robot arm, testing trust and presence at a distance. Known for: The Telegarden (1995); Dislocation of Intimacy (1998); The Robot and the Garden (2000, ed.). https://goldberg.berkeley.edu/ - **Kenji Yanobe / 矢延宪司** (Osaka, JP, 1965–) — Japanese artist of survival machines and giant robots, from radiation suits after Chernobyl to post-Fukushima figures of hope. Known for: Yellow Suit (1991); Giant Torayan (2005); Sun Child (2011); Ship's Cat (2017–). https://www.yanobe.com/ - **Kenneth Rinaldo** (Columbus, US, 1958–) — American artist of artificial life and interspecies robotics: in Autopoiesis, grapevine-armed robots sense visitors and each other, and in Augmented Fish Reality fish steer their own robotic bowls. Known for: Autopoiesis (2000); The Farm (2000s); Augmented Fish Reality (2004). https://www.kenrinaldo.com/ - **Kraftwerk** (Düsseldorf, DE) — German electronic band founded in 1970 whose robot doubles and 'man-machine' aesthetic made the fusion of musician and machine a pop image. Known for: Die Mensch-Maschine (The Man-Machine) (1978); The Robots (1978); Computer World (1981). https://kraftwerk.com/ - **Kris Verdonck** (Brussels, BE) — Belgian director whose machine-theatre works blur actor and object, like the flailing industrial machine of Dancer #1 and the dying performers of END. Known for: END (2008); ACTOR #1 (2010); Dancer #1 (2003). https://www.atwodogscompany.org/ - **Kyle McDonald** (Los Angeles, US) — American artist of computer vision and code who turns the camera back on its users; Facework is a game where you apply for gig jobs by performing the face a classifier wants to see. Known for: People Staring at Computers (2011); Exhausting a Crowd (2015); us+ (2013, with Lauren McCarthy); Facework (2020). https://kylemcdonald.net/ - **Kyriaki Goni** (Athens, GR) — Greek artist whose Data Garden imagines data stored in plant DNA, and whose Not Allowed For Algorithmic Audiences explores the smart speakers listening in on our homes. Known for: Data Garden (2019); Not Allowed For Algorithmic Audiences (2021); Counting Craters on the Moon (2023). https://kyriakigoni.com/ - **Laetitia Sonami** (San Francisco, US, 1957–) — French-born composer who performed with the Lady's Glove, a sensor glove, and later Spring Spyre, an instrument trained with machine learning (Wekinator) to respond to her gestures. Known for: Lady's Glove (1991); Spring Spyre (2012). https://sonami.net/ - **Lauren Lee McCarthy** (Los Angeles, US) — American artist who performs the smart-home assistant herself: in LAUREN she watched over and served people's homes remotely 24/7, exposing the intimacy, labour and control inside 'AI'. Known for: LAUREN (2017); SOMEONE (2019); Surrogate (2021–); p5.js (2013). https://lauren-mccarthy.com/ - **Lawrence Abu Hamdan** (Beirut, LB, 1985–) — Jordanian-British 'private ear' whose works examine accent analysis in asylum cases and the forensics of listening, showing how voices become evidence for machines and states. Known for: Conflicted Phonemes (2012); The Freedom of Speech Itself (2012); Walled Unwalled (2018); Earwitness Theatre (2018). http://lawrenceabuhamdan.com/ - **Lawrence Lek / 陆明龙** (London, UK, 1982–) — Malaysian-Chinese British artist of 'Sinofuturism' whose films and games follow AI characters: a satellite that wants to be an artist, a self-driving car in therapy, a retired care robot. Known for: Geomancer (2017); AIDOL (2019); NOX (2023); Guanyin: Confessions of a Former Care Robot (2024). https://lawrencelek.com/ - **Lee Bul / 李昢** (Seoul, KR, 1964–) — South Korean artist whose headless, limbless Cyborg sculptures fuse the idealised female body with machine armour, questioning technology's fantasies of perfection. Known for: Cyborg series (1997–2011); Anagram (1998); Willing To Be Vulnerable (2015–16). https://en.wikipedia.org/wiki/Lee_Bul - **Leonel Moura** (Lisbon, PT, 1948–) — Portuguese artist whose swarms of small robots paint by reacting to each other's marks, a test of emergent, machine-made art without a human composer. Known for: ArtSBot swarm painting robots (2003–); RAP, Robotic Action Painter (2006). https://www.leonelmoura.com/ - **Li Liao / 李燎** (Shenzhen, CN) — Chinese performance artist who worked 45 days on a Foxconn assembly line to earn the iPad Mini he then exhibited, embodying the human labour inside consumer machines. Known for: Consumption (2012); Unemployed (2012). https://www.artsy.net/artwork/li-liao-consumption - **Liam Young** (Los Angeles, US, 1979–) — Australian speculative architect and filmmaker whose Where the City Can't See is shot entirely through laser scanners, showing a city as a self-driving car perceives it. Known for: Where the City Can't See (2016); In the Robot Skies (2016); Planet City (2020). https://en.wikipedia.org/wiki/Liam_Young - **Liu Chuang / 刘窗** (Beijing, CN, 1979–) — Chinese artist whose films trace the material and human underside of digital economies, from bitcoin mines in Inner Mongolia to lithium lakes and the linguistic data of minority voices. Known for: Bitcoin Mining and Field Recordings of Ethnic Minorities (2018); Lithium Lake and Proto-Language (2019). https://en.wikipedia.org/wiki/Liu_Chuang_(artist) - **Louis-Philippe Demers** (Quebec, CA / Singapore) — Canadian maker of machine performers whose The Blind Robot gently touches a visitor's face to 'see' them, and whose Inferno straps audiences into exoskeletons that move them. Known for: La Cour des Miracles (1997, with Bill Vorn); The Blind Robot (2012); Inferno (2015, with Bill Vorn); The Multiple Bodies of the Machine (book). https://www.wikidata.org/wiki/Q125746113 - **Lu Yang / 陆扬** (Shanghai, CN, 1984–) — Chinese artist whose genderless digital avatar DOKU, motion-captured from the artist, is reincarnated across six realms, blending neuroscience, Buddhism and game engines. Known for: Uterus Man (2013); Electromagnetic Brainology (2017); DOKU (2020–); DOKU the Self (2022). http://www.luyang.asia/ - **Lynn Hershman Leeson** (San Francisco, US, 1941–) — American artist who, after inventing the fictional persona Roberta Breitmore, built Agent Ruby, a web chatbot that learned from visitors, anticipating today's conversational AI and its identities. Known for: Roberta Breitmore (1974–78); Agent Ruby (1998–2002); DiNA (2004); Twisted (New Museum, 2021). https://www.lynnhershman.com/ - **Madeline Gannon** (Pittsburgh, US) — American 'robot tamer' whose studio ATONATON turns industrial robots into curious creatures: Mimus, a 1,200 kg ABB arm at the London Design Museum, tracked and approached visitors like an animal. Known for: Mimus (2016); Manus (2018); Quipt (2015). https://atonaton.com/ - **Mahwish Chishty** (Kent, US) — Pakistani-American artist who paints US military drones in the ornate style of Pakistani truck art, confronting the machine gaze over the Afghanistan–Pakistan border. Known for: Drone Art (2013–); War Machines (2015–16). https://en.wikipedia.org/wiki/Mahwish_Chishty - **Manu Luksch** (London, UK, 1970–) — Austrian-born artist whose film Faceless is made entirely of her own CCTV footage obtained under data-protection law, in which the faces of others are erased. Known for: Faceless (2007); Dreams Rewired (2015). https://manuluksch.com/ - **Marc Böhlen** (Buffalo, US) — German-American artist-engineer whose machines test the limits of machine perception, such as Amy and Klara, two robotic boxes that read magazines and argue with each other in synthetic speech. Known for: Universal Whistling Machine (2004); Amy and Klara (2006); Keeping Track (2000s). https://www.realtechsupport.org/ - **Marcel·lí Antúnez Roca** (Barcelona, ES, 1959–) — Catalan performer who in Epizoo let audiences move his nose, mouth and buttocks via a pneumatic exoskeleton controlled by a mouse, an early work on the body under machine control. Known for: JoAn, l'home de carn (1992); Epizoo (1994); Afasia (1998). http://www.marceliantunez.com/ - **Marco Donnarumma** (Berlin, DE, 1984–) — Italian performer and researcher who fuses bodies with AI-driven prosthetics; his robot Amygdala learns to move by 'feeling' human skin and ritually cuts its own membrane. Known for: Corpus Nil (2016); Amygdala (2017–18); Eingeweide (2018). https://marcodonnarumma.com/ - **Mari Velonaki** (Sydney, AU, 1968–) — Greek-Australian artist and founder of UNSW's Creative Robotics Lab; in Fish-Bird two wheelchairs 'in love' write handwritten notes to visitors, building intimacy through printed letters and movement. Known for: Fish-Bird (2004); Diamandini (2013). https://www.marivelonaki.com/ - **Mariela Yeregui** (Buenos Aires, AR) — Argentine artist whose small robots in Proxemia approach and retreat from each other and from visitors, making distance and touch the subject of machine behaviour. Known for: Proxemia (2005); Epithelia (2004). https://marielayeregui.com/ - **Mark Pauline** (San Francisco, US) — Founder of Survival Research Laboratories, whose spectacles set salvaged, remote-controlled war machines and robots against each other, staging machine violence as industrial culture. Known for: Survival Research Laboratories (founded 1978); A Calculated Forecast of Ultimate Doom (1988). https://www.srl.org/ - **Martine Syms** (Los Angeles, US, 1988–) — American artist whose Mythiccbeing is a chatbot avatar of a self-absorbed Black woman, 'Teeny', turning AI personas into a critique of how Black women are scripted. Known for: Mythiccbeing (2018); Neural Swamp (Philadelphia Museum of Art, 2021); Shame Space (2021). https://martinesyms.com/ - **Max Dean** (Toronto, CA, 1949–) — Canadian artist whose robot arm in As Yet Untitled offers family photographs for rescue, shredding them unless a visitor intervenes, putting responsibility for loss on the viewer. Known for: As Yet Untitled (1992–95); The Table: Childhood (2001, with Raffaello D'Andrea). https://en.wikipedia.org/wiki/Max_Dean_(artist) - **Maywa Denki / 明和电机** (Tokyo, JP) — Japanese art unit led by Nobumichi Tosa that presents absurd 'nonsense machines' and instruments as company products, performed by staff in blue work uniforms. Known for: Tsukuba series (1993–); Otamatone (2009); Edelweiss series (2004–). https://www.maywadenki.com/ - **Memo Akten** (Los Angeles, US, 1975–) — Turkish artist and researcher whose Learning to See shows a neural network interpreting live camera images of cables and cloth as waves, fire or flowers, a metaphor for how any mind sees what it already knows. Known for: Learning to See (2017); Deep Meditations (2018); Superradiance (2023, with Katie Peyton Hofstadter). https://www.memo.tv/ - **Merritt Moore** (London, UK, 1988–) — American ballet dancer and quantum physicist (DPhil, Oxford) who programs and dances duets with an industrial robot arm. Known for: Merritt + Robot (2020). https://www.merrittmoore.com/ - **Miao Ying / 苗颖** (Shanghai, CN / New York, US, 1985–) — Chinese artist of the 'Chinternet' whose works on censorship, filtered search results and platform culture show how machines shape what can be seen online in China. Known for: Blind Spot (2007); Chinternet Plus (2016); Hardcore Digital Detox (2018). https://en.wikipedia.org/wiki/Ying_Miao - **Mimi Onuoha** (New York, US) — Nigerian-American artist whose Library of Missing Datasets is a filing cabinet of empty folders for data that is never collected, showing that what machines cannot see is a choice. Known for: The Library of Missing Datasets (2016); Classification.01 (2017); These Networks in Our Skin (2021). https://mimionuoha.com/ - **Minne Atairu** (New York, US) — Nigerian-American artist who uses generative models to imagine the Benin bronzes that were never made after the 1897 British raid, confronting the gaps in the data of looted histories. Known for: To the Hand (The Shed, 2023); Igùn (2020–). https://www.minneatairu.com/ - **Myron Krueger** (Vernon, US, 1942–) — American computer artist who coined 'artificial reality'; in Videoplace a camera turned visitors' silhouettes into live partners for graphic creatures, an early case of a body read by a machine. Known for: Videoplace (1974–); Artificial Reality (1983). https://en.wikipedia.org/wiki/Myron_W._Krueger - **Nam June Paik / 白南准** (New York, US / Seoul, KR, 1932–2006) — Korean-American founder of video art who built Robot K-456 with Shuya Abe, a remote-controlled 'anti-robot' that walked the streets, spoke and defecated beans; in 1982 he staged its 'first catastrophe of the 21st century', a collision with a car. Known for: Robot K-456 (1964, with Shuya Abe); Family of Robot (1986); TV Buddha (1974). https://www.paikstudios.com/ - **Naoko Tosa / 土佐尚子** (Kyoto, JP, 1961–) — Japanese media artist whose Neurobaby, an early emotion-recognising computer baby that reacts to the tone of your voice, pioneered artificial emotion in art. Known for: Neurobaby (1993); Talking to Neurobaby (1994). https://en.wikipedia.org/wiki/Naoko_Tosa - **Neil Harbisson** (Barcelona, ES, 1984–) — Catalan-British artist, often called the first legally recognised cyborg; an implanted antenna lets him hear colour as sound, a machine sense grafted onto a human body. Known for: Eyeborg antenna (2004–); Cyborg Foundation (2010, with Moon Ribas). https://www.cyborgarts.com - **Nicolas Gourault** (Paris, FR) — French filmmaker whose VO reconstructs the 2018 Uber self-driving crash through test drivers' accounts, and whose Unknown Label follows data labellers training autonomous cars. Known for: VO (2020); Unknown Label (2024). https://nicolasgourault.fr/ - **Nicolas Schöffer** (Paris, France, 1912–1992) — Hungarian-French artist who made CYSP 1 (1956), often called the first cybernetic sculpture: an electronic brain let it move and react to light, sound and colour, and it danced with Maurice Béjart's company. Known for: CYSP 1 (1956); Tour Lumière Cybernétique (project, 1961–); Chronos series (1960s–80s). https://en.wikipedia.org/wiki/Nicolas_Sch%C3%B6ffer - **Norman White** (Toronto, CA, 1938–) — Canadian pioneer of robotic art whose The Helpless Robot asks passers-by to turn it, then grows demanding and ungrateful, reversing who serves whom. Known for: The Helpless Robot (1987); Facing Out, Laying Low (1977); Them Fuckin' Robots (1988, with Laura Kikauka). https://www.normill.ca/ - **Nouf Aljowaysir** (New York, US) — Saudi-born artist whose Salaf asks an AI to reconstruct her ancestors from colonial photographs of Arabia; the models erase and mislabel them, revealing a Western gaze in the data. Known for: Salaf (2021); Ana Min Wein (Where Am I From?) (2021). https://www.noufaljowaysir.com/ - **Oriza Hirata / 平田织佐** (Tokyo, JP, 1962–) — Japanese playwright who co-created android and robot theatre with Hiroshi Ishiguro; in Sayonara an android recites poetry to a dying young woman. Known for: I, Worker (2008); Sayonara (2010); Three Sisters, Android Version (2012). https://en.wikipedia.org/wiki/Oriza_Hirata - **Osamu Tezuka / 手冢治虫** (Tokyo, JP, 1928–1989) — Japanese manga artist (1928–1989) whose Astro Boy, a robot boy with a heart, shaped Japan's friendly image of robots and inspired generations of Japanese roboticists. Known for: Astro Boy (Tetsuwan Atom) (1952–68); Phoenix (1954–88). https://tezukaosamu.net/ - **Paola Torres Núñez del Prado** (Stockholm, SE, 1979–) — Peruvian-Swedish artist who weaves textiles and sound with AI, showing how machine vision misreads Andean and Amazonian visual cultures. Known for: AI Visions of the Andes (2021); Tejidos Sonoros (2019). https://es.wikipedia.org/wiki/Paola_Torres_Núñez_del_Prado - **Paolo Cirio** (New York, US, 1979–) — Italian artist whose Capture posted face-recognised images of 4,000 French police officers in the street, turning surveillance tools on the police; the work was censored. Known for: Capture (2020); Street Ghosts (2012); Face to Facebook (2011, with Alessandro Ludovico). https://paolocirio.net/ - **Patrick Tresset** (Brussels, BE) — French artist whose robots named Paul sit at school desks and draw portraits of sitters with nervous, hesitating lines, turning the sitter into the one being observed by machines. Known for: Paul (2011); 5 Robots Named Paul (2012); Human Study #1 (2012–). https://patricktresset.com/ - **Paul Sermon** (Brighton, UK, 1966–) — British artist whose Telematic Dreaming projected a remote partner onto a bed via video link, an early and intimate experiment with telepresence. Known for: Telematic Dreaming (1992); Telematic Vision (1993). https://en.wikipedia.org/wiki/Paul_Sermon - **Paula Gaetano Adi** (Providence, US / Argentina) — Argentine artist whose robot Alexitimia, a skin-like blob that sweats when touched, won VIDA 9.0 and asks what a machine body without face or limbs can feel. Known for: Alexitimia (2006); Guasch (2011). https://www.paulagaetanoadi.com/ - **Petra Gemeinboeck** (Melbourne, AU) — Austrian-Australian artist-researcher who works with dancers to give non-humanlike robots expressive movement, a practice she calls 'bodying-thinging'. Known for: Accomplice (2013, with Rob Saunders); Machine Movement Lab (2016–). https://www.swinburne.edu.au/research/our-research/access-our-research/find-a-researcher-or-supervisor/researcher-profile/?id=pgemeinboeck - **Philip Beesley** (Toronto, CA, 1956–) — Canadian architect and artist whose 'living architecture' environments of fronds, sensors and chemistry breathe and ripple in response to visitors. Known for: Hylozoic Ground (Venice Architecture Biennale, 2010); Astrocyte (2017). https://www.philipbeesleystudio.com/ - **Philipp Schmitt** (Berlin, DE) — German designer whose Declassifier overlays each object a vision model detects with the training images it learned from, showing that a model's 'seeing' is remembering its dataset. Known for: Declassifier (2020); Humans of AI (2019, with Sebastian Schmieg); Unhuman Compositions (2021). https://philippschmitt.com/ - **Pierre Huyghe** (Santiago, CL / Paris, FR, 1962–) — French artist of living, self-running situations; UUmwelt displays images a neural network reconstructed from human brain scans, flickering with flies in the gallery. Known for: UUmwelt (2018); After UUmwelt (2018–); Mental Image (2023–); Liminal (2024). https://en.wikipedia.org/wiki/Pierre_Huyghe - **Primavera De Filippi** (Paris, FR) — Italian legal scholar and artist whose Plantoids are blockchain-based robotic 'plants' that own themselves, collect money and pay artists to make copies of them, a self-reproducing machine lifeform. Known for: Plantoid (2015–); Blockchain and the Law (2018, with Aaron Wright). https://en.wikipedia.org/wiki/Primavera_De_Filippi - **Rafael Lozano-Hemmer** (Montreal, CA, 1967–) — Mexican-Canadian artist of interactive surveillance installations; Level of Confidence uses face recognition to match visitors with the 43 disappeared students of Ayotzinapa, turning a police tool into a memorial. Known for: Zoom Pavilion (2015, with Krzysztof Wodiczko); Level of Confidence (2015); Pulse Room (2006). https://www.lozano-hemmer.com/ - **Ralf Baecker** (Berlin, DE, 1977–) — German artist who builds physical computers out of sticks, strings and crystals, making the inner operations of computation visible and strange. Known for: Rechnender Raum (2007); Irrational Computing (2012); Mirage (2015). https://rlfbckr.io/ - **Random International** (London, UK / Berlin, DE) — Art studio founded by Hannes Koch and Florian Ortkrass whose minimal robotic figures and swarms test how little movement we need to perceive something as alive. Known for: Rain Room (2012); Fifteen Points (2019); Swarm Study series (2010–). https://www.random-international.com/ - **Rashaad Newsome** (New York, US, 1979–) — American artist whose AI 'Being', a vogue-dancing, bell hooks-reading avatar, leads workshops and performances as a decolonial, queer machine persona. Known for: Being (2019); Being 1.5 (2020); Assembly (Park Avenue Armory, 2021). https://rashaadnewsome.com/ - **Rebecca Horn** (Berlin, DE, 1944–2024) — German artist whose body extensions and later kinetic machines, such as a hanging piano that bursts open, give machines moods, fragility and something like feeling. Known for: Finger Gloves (1972); Painting Machine (1988); Concert for Anarchy (1990). https://en.wikipedia.org/wiki/Rebecca_Horn - **Revital Cohen & Tuur Van Balen** (London, UK) — Artist duo whose The Immortal connects life-support machines to each other in a closed loop that circulates fluids and air without a human body, a machine that only keeps itself alive. Known for: Life Support (2008); Phantom Recorder (2008); The Immortal (2012); Pixel (2013). https://www.cohenvanbalen.com/ - **Rimini Protokoll** (Berlin, DE) — German theatre collective; in Stefan Kaegi's Uncanny Valley an animatronic double of writer Thomas Melle delivers a lecture in his place, asking what remains of a person who can be replaced. Known for: Uncanny Valley (2018, Stefan Kaegi with Thomas Melle); Win > < Win (2017). https://www.rimini-protokoll.de/ - **Robert Breer** (New York, US, 1926–2011) — American filmmaker and sculptor whose slow, self-propelled 'Floats' crept across floors and reversed when they touched an obstacle, giving simple machines an odd, animal presence. Known for: Floats (1970, Pepsi Pavilion, Expo '70); Self-propelled Styrofoam Floats (1965–). https://en.wikipedia.org/wiki/Robert_Breer - **robotlab** (Karlsruhe, DE) — German group founded at ZKM in 2000 that takes industrial robots into art; in bios [bible] a robot arm writes out the entire Bible by pen, a monk's labour done by a machine. Known for: Zone (2001); bios [bible] (2007); manifest (2008). http://www.robotlab.de/ - **Rosa Menkman** (Amsterdam, NL, 1983–) — Dutch artist and theorist of glitch and resolution whose work shows how compression standards and imaging machines decide what can be seen, including the forgotten women in test images. Known for: The Collapse of PAL (2010); The Glitch Moment(um) (2011); Beyond Resolution (2020); Pique Nique pour les Inconnues (2020). https://beyondresolution.info/ - **Sam Lavigne** (New York, US) — American artist of data and scraping whose White Collar Crime Risk Zones applies predictive policing to financial crime, turning the classifier against the powerful. Known for: White Collar Crime Risk Zones (2017, with Brian Clifton and Francis Tseng); Slow Hot Computer (2016); Synthetic Messenger (2021, with Tega Brain). https://lav.io/ - **Sebastian Schmieg** (Berlin, DE) — German artist who works on the hidden human labour inside AI; Decision Space and Segmentation.Network replay the clickwork of crowdworkers who trace and label images for training. Known for: Segmentation.Network (2016); Decision Space (2019); Humans of AI (2019, with Philipp Schmitt). https://sebastianschmieg.com/ - **Seiko Mikami / 三上晴子** (Tokyo, JP, 1961–2015) — Japanese media-art pioneer (1961–2015) whose Desire of Codes (YCAM) fills a wall with insect-like camera arms that track visitors, making the audience's bodies the material of surveillance. Known for: Molecular Informatics (1996); World, Membrane and the Dismembered Body (1997); Desire of Codes (2010). https://en.wikipedia.org/wiki/Seiko_Mikami - **Shinseungback Kimyonghun** (Seoul, KR) — Seoul duo (Shin Seung Back and Kim Yong Hun) whose Cloud Face collects clouds that face detection mistakes for faces, making machine pareidolia visible. Known for: Cloud Face (2012); Nonfacial Portrait (2018); Mind (2019). https://ssbkyh.com/ - **Shu Lea Cheang / 郑淑丽** (Paris, FR, 1954–) — Taiwanese artist and net-art pioneer whose 3x3x6 at the Venice Biennale turned the surveillance of a panoptic prison onto visitors, through the stories of queer and gender-nonconforming prisoners. Known for: Brandon (1998); I.K.U. (2000); UKI (2009–); 3x3x6 (Taiwan Pavilion, Venice, 2019). https://mauvaiscontact.info/ - **Simon Denny** (Berlin, DE, 1982–) — New Zealand artist who turns tech-industry documents into objects; he built Amazon's patented cage for warehouse workers, showing how automation encloses human labour. Known for: Secret Power (Venice, 2015); Mine (2019). https://simondenny.net/ - **Simon Penny** (Irvine, US, 1955–) — Australian artist and theorist of embodied interaction; his autonomous robot Petit Mal wobbled towards and away from visitors with a behaviour that felt neither animal nor machine. Known for: Petit Mal (1989–95); Fugitive (1997); Making Sense: Cognition, Computing, Art, and Embodiment (2017). https://simonpenny.net/ - **Simon Weckert** (Berlin, DE) — Berlin artist who pulled a handcart of 99 smartphones through empty streets and made Google Maps report a traffic jam, showing how platforms read the city through devices. Known for: Google Maps Hacks (2020). http://www.simonweckert.com/ - **Simone Giertz** (Los Angeles, US, 1990–) — Swedish inventor and YouTuber known as the 'queen of shitty robots', whose deliberately useless machines (a breakfast robot, a lipstick robot) turn failure into comedy. Known for: Shitty robots (2016–); Truckla (2019). https://www.simonegiertz.com/ - **Simone Rebaudengo** (Shanghai, CN) — Italian designer whose Addicted Products imagines toasters that want to be used, move to more attentive homes and quietly complain, a view of the IoT from the thing's side. Known for: Addicted Products (2012); Teacher of Algorithms (2014); Brad the Toaster (2012). https://www.simonerebaudengo.com/ - **So Kanno / 菅野创** (Tokyo, JP) — Japanese artist whose Senseless Drawing Bot, made with yang02, is a wobbling robot that sprays graffiti with a pendulum, turning mechanical instability into gesture. Known for: Senseless Drawing Bot (2011, with yang02); Lasermice (2017); Avatars (2020). https://kanno.so/ - **Sofian Audry** (Montreal, CA) — Canadian artist and scholar of machine learning as an artistic medium, with robots that learn in front of the public; author of Art in the Age of Machine Learning (MIT Press). Known for: Vessels (2015); Morphosis (2021); Art in the Age of Machine Learning (2021). https://sofianaudry.com/ - **Sougwen Chung / 钟愫君** (London, UK) — Hong Kong-born Canadian artist who draws live with robotic collaborators (the D.O.U.G. series) trained on her own gestures, exploring authorship shared between human and machine. Known for: Drawing Operations Unit: Generation 1 (D.O.U.G._1) (2015); Omnia per Omnia (2018); Assembly Lines (2022); Mutations of Presence (2021). https://sougwen.com/ - **Sputniko! (Hiromi Ozaki)** (Tokyo, JP) — Japanese-British artist whose Menstruation Machine is a device that lets a man experience menstruation, speculative design about bodies, gender and technology. Known for: Menstruation Machine – Takashi's Take (2010); Tranceflora – Amy's Glowing Silk (2015). https://sputniko.com/ - **Stelarc** (Perth, AU, 1946–) — Cypriot-Australian performance artist who treats the body as obsolete hardware to extend with robotics: a third hand, a six-legged exoskeleton, a body moved remotely by internet traffic. Known for: Third Hand (1980); Ping Body (1996); Exoskeleton (1998); Ear on Arm (2006–). https://stelarc.org/ - **Stephanie Dinkins** (New York, US, 1964–) — American transmedia artist who asks whether a Black woman's robot can know itself; after years of talking with the social robot Bina48 she trained Not the Only One, an AI memoir of her own family. Known for: Conversations with Bina48 (2014–); Not the Only One (2018); Secret Garden (2021); If We Don't, Who Will? (2023). https://www.stephaniedinkins.com/ - **Sun Yuan & Peng Yu / 孙原 & 彭禹** (Beijing, CN) — Beijing artist duo whose Can't Help Myself, an industrial robot endlessly sweeping a leaking red fluid back into place, became an emblem of machine labour, futility and exhaustion. Known for: Old People's Home (2007); Freedom (2009); Can't Help Myself (2016). http://www.sunyuanpengyu.com/ - **Superflux (Anab Jain & Jon Ardern)** (London, UK) — London design studio whose films and installations imagine living with more-than-human and machine agents; Uninvited Guests shows an elderly man outwitting the smart devices his children bought him. Known for: Uninvited Guests (2015); Mitigation of Shock (2017); Our Friends Electric (2017); Invocation for Hope (2022). https://superflux.in/ - **Suzanne Kite** (Annandale-on-Hudson, US) — Oglála Lakȟóta artist and scholar whose performances and machine-learning works draw on Lakota ontology, where stones and machines can be relatives rather than tools. Known for: Making Kin with the Machines (2018, with Jason Edward Lewis et al.); Íŋyaŋ Iyé (Telling Rock) (2019). https://kitekitekitekite.com/ - **Suzanne Treister** (London, UK, 1958–) — British artist whose HEXEN 2.0 tarot cards and diagrams map the entangled histories of cybernetics, military research, the internet and counterculture. Known for: HEXEN 2.0 (2009–11); SURVIVOR (F) (2016–19); The Holographic Universe Theory of Art History (2018). https://www.suzannetreister.net/ - **Tabita Rezaire** (Cayenne, French Guiana, 1989–) — French-Guianese artist whose Deep Down Tidal follows undersea internet cables along the routes of colonial trade, framing digital networks as a continuation of empire. Known for: Deep Down Tidal (2017); Sugar Walls Teardom (2016); Premium Connect (2017). https://tabitarezaire.com/ - **Takayuki Todo / 藤堂高行** (Tokyo, JP) — Japanese artist-engineer whose SEER is a small humanoid head that meets your gaze and mirrors your expression with delicate eyebrows, probing the threshold of machine empathy. Known for: SEER: Simulative Emotional Expression Robot (2018); SEER at Ars Electronica (2019). https://www.takayukitodo.com/ - **Tega Brain** (New York, US) — Australian artist and environmental engineer whose Deep Swamp lets AI agents manage wetlands for 'nature', 'art' and 'attention', and whose Synthetic Messenger botnet clicks climate-news ads. Known for: Deep Swamp (2018); Synthetic Messenger (2021, with Sam Lavigne); Solar Protocol (2021, with Alex Nathanson and Benedetta Piantella). https://tegabrain.com/ - **Theo Jansen** (Delft, NL, 1948–) — Dutch artist who since 1990 has evolved wind-walking 'beach animals' from PVC tubing, machines that store air, sense water and walk the shore on their own. Known for: Strandbeest (1990–); Animaris Ordis (2006). https://www.strandbeest.com/ - **Timo Arnall** (Oslo, NO) — British-Norwegian designer and filmmaker whose Robot Readable World compiles footage of computer vision at work, and whose Internet Machine films the physical data centres behind the cloud. Known for: Immaterials: Light Painting WiFi (2011); Robot Readable World (2012); Internet Machine (2014). https://www.elasticspace.com/ - **Tod Machover** (Cambridge, US, 1953–) — American composer at MIT Media Lab whose robotic opera Death and the Powers imagines a man who uploads himself into 'the System', with a chorus of robots and a singing set. Known for: Hyperinstruments (1986–); Death and the Powers (2010). https://www.media.mit.edu/people/tod/overview/ - **Tom White** (Wellington, NZ) — New Zealand artist whose Perception Engines are abstract prints that neural networks confidently classify as electric fans, hammerhead sharks or measuring cups, art made for machine eyes. Known for: Perception Engines (2018); Synthetic Abstractions (2018). https://drib.net/ - **Tomo Kihara** (Amsterdam, NL) — Japanese designer based in Amsterdam whose street game challenges players to reach a camera without being detected as a pedestrian, testing the failures of self-driving vision. Known for: How (not) to get hit by a self-driving car (2022). https://www.tomokihara.com/ - **Trevor Paglen** (New York, US / Berlin, DE, 1974–) — American artist and geographer who makes the invisible infrastructure of machine vision visible, from training-set taxonomies to images 'hallucinated' by classifiers; ImageNet Roulette led ImageNet to remove 600,000 person images. Known for: Sight Machine (2017, with Kronos Quartet); Adversarially Evolved Hallucinations (2017); ImageNet Roulette (2019, with Kate Crawford); From 'Apple' to 'Anomaly' (Barbican, 2019). https://paglen.studio/ - **Vladan Joler** (Novi Sad, RS) — Serbian artist-researcher (SHARE Lab) whose huge diagrams map the minerals, labour and data behind one Amazon Echo; Anatomy of an AI System entered MoMA's collection. Known for: Anatomy of an AI System (2018, with Kate Crawford); New Extractivism (2020); Calculating Empires (2023, with Kate Crawford). https://calculatingempires.net/ - **Wafaa Bilal** (New York, US, 1966–) — Iraqi-American artist who lived for a month in a gallery facing a paintball gun that internet users could aim and fire remotely, and who later implanted a camera in the back of his head. Known for: Domestic Tension (2007); 3rdi (2010). https://wafaabilal.com/ - **Wang Yuyang / 王郁洋** (Beijing, CN) — Chinese artist whose Breathe series makes silicone replicas of everyday machines, from an ATM to a washing machine, slowly inhale and exhale, turning appliances into living bodies. Known for: Breathe series (2005–). https://www.shanghartgallery.com/galleryarchive/artists/name/wangyuyang - **Wen-Ying Tsai / 蔡文颖** (New York, US, 1928–2013) — Chinese-American engineer turned artist, a pioneer of cybernetic art: his vibrating steel rods under stroboscopic light respond to sound and clapping, so the sculpture seems to breathe with its audience. Known for: Cybernetic Sculpture (1968); Upward Falling Fountain (1979). https://en.wikipedia.org/wiki/Wen-Ying_Tsai - **Wesley Goatley** (London, UK) — British sound artist whose Newly Forgotten Technologies gathers abandoned smart speakers that hold a séance, remembering the future they were promised. Known for: Newly Forgotten Technologies (2021); Chattering (2020); Watching the Watchers (2016). https://www.wesleygoatley.com/ - **Wim Delvoye** (Ghent, BE, 1965–) — Belgian artist whose Cloaca is a room-sized digestion machine that is fed daily and produces faeces, reducing the organism to a laboratory apparatus. Known for: Cloaca (2000); Cloaca series (2000–2010). https://wimdelvoye.be/ - **Xu Bing / 徐冰** (Beijing, CN / New York, US, 1955–) — Chinese artist whose feature film Dragonfly Eyes is assembled entirely from public surveillance and livestream footage, telling a love story through the eyes of China's cameras. Known for: Book from the Ground (2003–); Dragonfly Eyes (2017). https://www.xubing.com/ - **Yoichi Ochiai / 落合阳一** (Tokyo, JP, 1987–) — Japanese media artist and researcher who proposes 'digital nature', a world where computation becomes a new nature beyond the human–machine divide. Known for: Digital Nature (2018, book); Null Pavilion (Expo 2025 Osaka); Levitrope (2019). https://yoichiochiai.com/ - **Yunchul Kim** (Seoul, KR) — South Korean artist whose kinetic 'material machines' sense cosmic rays and fluid dynamics; he represented Korea at the 2022 Venice Biennale with Gyre. Known for: Argos (2018); Chroma V (2020); Gyre (Korean Pavilion, Venice, 2022). https://yunchulkim.net/ - **Zach Blas** (Toronto, CA, 1981–) — American artist whose amorphous collective masks, made from aggregated face scans, defeat biometric recognition; his Tay installation resurrects Microsoft's deleted chatbot to reflect on its death. Known for: Facial Weaponization Suite (2011–14); im here to learn so :)))))) (2017, with Jemima Wyman); Contra-Internet (2014–); Silicon Traces (2020s). https://zachblas.info/ ### Theorists & researchers - **Abeba Birhane** (Dublin, IE) — Ethiopian-born cognitive scientist who audits large training datasets, showing the misogynist and racist content inside them, and argues for a relational ethics of AI. Known for: Large Image Datasets: A Pyrrhic Win for Computer Vision? (2021, with Vinay Prabhu); Algorithmic Injustice: A Relational Ethics Approach (2021). https://en.wikipedia.org/wiki/Abeba_Birhane - **Adrian Mackenzie** (Canberra, AU, 1962–) — Australian sociologist of technology whose Machine Learners reads machine-learning code and textbooks as a cultural practice, tracing how learning machines order the world. Known for: Cutting Code (2006); Machine Learners: Archaeology of a Data Practice (2017). https://researchportalplus.anu.edu.au/en/persons/adrian-mackenzie/ - **Alan Turing / 艾伦·图灵** (Manchester, UK, 1912–1954) — British mathematician whose 1950 'imitation game' reframed the question 'can machines think?' as a performance judged by humans, the founding scene of AI. Known for: On Computable Numbers (1936); Computing Machinery and Intelligence (1950). https://en.wikipedia.org/wiki/Alan_Turing - **Andy Clark / 安迪·克拉克** (Brighton, UK, 1957–) — British philosopher of the extended mind who argues that humans are 'natural-born cyborgs' whose thinking already runs through tools and machines. Known for: The Extended Mind (1998, with David Chalmers); Natural-Born Cyborgs (2003); The Experience Machine (2023). https://en.wikipedia.org/wiki/Andy_Clark - **Anna Munster** (Sydney, AU) — Australian theorist and artist who, with Adrian MacKenzie, developed 'platform seeing', a way to describe how machine vision operates at the scale of datasets and platforms. Known for: An Aesthesia of Networks (2013); Platform Seeing (2019, with Adrian MacKenzie). https://www.unsw.edu.au/staff/anna-munster - **Benjamin Bratton** (San Diego, US, 1968–) — American theorist whose The Stack describes planetary computation as an accidental megastructure, and who argues AI should be understood as a new planetary intelligence. Known for: The Stack (2015); The Terraforming (2019); Antikythera (2022–). https://bratton.info/ - **Bernard Stiegler / 贝尔纳·斯蒂格勒** (Paris, FR, 1952–2020) — French philosopher who described technics as the 'exteriorised memory' that constitutes the human, and warned of an 'automatic society' of algorithmic governance. Known for: Technics and Time, 1: The Fault of Epimetheus (1994); Automatic Society (2015). https://arsindustrialis.org/les-pages-de-bernard-stiegler - **Catherine D'Ignazio** (Cambridge, US, 1975–) — American scholar and artist whose Data Feminism, written with Lauren F. Klein, sets out how power shapes who is counted, classified and seen by data systems. Known for: Data Feminism (2020, with Lauren F. Klein); Counting Feminicide (2024). http://www.kanarinka.com/ - **Chen Qiufan / 陈楸帆** (Beijing / Shanghai, CN, 1981–) — Chinese science-fiction writer whose The Waste Tide imagines e-waste workers and cyborgs in southern China, and who has written fiction together with language models. Known for: The Waste Tide (2013); AI 2041: Ten Visions for Our Future (2021, with Kai-Fu Lee). https://en.wikipedia.org/wiki/Chen_Qiufan - **Clifford Nass** (Stanford, US, 1958–2013) — American social scientist whose experiments showed that people are polite to computers and apply gender stereotypes to synthetic voices, treating media 'like real people'. Known for: Computers Are Social Actors (1994); The Media Equation (1996, with Byron Reeves); Wired for Speech (2005). https://en.wikipedia.org/wiki/Clifford_Nass - **David Chalmers / 大卫·查默斯** (New York, US, 1966–) — Australian philosopher who named the 'hard problem' of consciousness and now asks, carefully, whether language models or their successors could be conscious. Known for: The Conscious Mind (1996); The Extended Mind (1998, with Andy Clark); Could a Large Language Model Be Conscious? (2023). https://consc.net/ - **David J. Gunkel** (DeKalb, US, 1962–) — American media philosopher whose The Machine Question and Robot Rights take seriously whether machines can be moral patients, drawing on Levinas and Derrida. Known for: The Machine Question (2012); Robot Rights (2018). https://gunkelweb.com/ - **Don Ihde / 唐·伊德** (Stony Brook, US, 1934–2024) — American philosopher of postphenomenology whose four human–technology relations (embodiment, hermeneutic, alterity, background) are a basic map for how we live with machines. Known for: Technology and the Lifeworld (1990); Bodies in Technology (2002). https://en.wikipedia.org/wiki/Don_Ihde - **Donna Haraway / 唐娜·哈拉维** (Santa Cruz, US, 1944–) — American feminist scholar whose Cyborg Manifesto made the human–machine hybrid a figure for dismantling the boundaries between organism and machine, nature and culture. Known for: A Cyborg Manifesto (1985); Simians, Cyborgs, and Women (1991); Staying with the Trouble (2016). https://en.wikipedia.org/wiki/Donna_Haraway - **Elisa Giaccardi** (Milan, IT / Delft, NL) — Italian design researcher who develops 'more-than-human' design, studying everyday life from the perspective of connected things, such as by fitting cameras to kettles and cups. Known for: Thing Ethnography (DIS 2016). https://www.tudelft.nl/en/ide/about-ide/people/former-professors/giaccardi-e - **Emily M. Bender** (Seattle, US, 1973–) — American linguist who coined 'stochastic parrots' for language models that produce form without meaning, a key term for thinking about machine speech. Known for: Climbing towards NLU (2020, with Alexander Koller); On the Dangers of Stochastic Parrots (2021); The AI Con (2025, with Alex Hanna). https://faculty.washington.edu/ebender/ - **Erkki Huhtamo** (Los Angeles, US, 1958–) — Finnish media archaeologist whose essays on the history of robots, automata and interactive art trace the long cultural life of machines as performers. Known for: From Cybernation to Interaction (1999); Illusions in Motion (2013); Media Archaeology (2011, ed. with Jussi Parikka). https://en.wikipedia.org/wiki/Erkki_Huhtamo - **Eryk Salvaggio** (Rochester, US) — American artist-researcher behind the newsletter Cybernetic Forests, who reads AI images as 'infographics about the dataset' and works with noise and glitch to expose how diffusion models see. Known for: How to Read an AI Image (2022); Cybernetic Forests (newsletter). https://www.cyberneticforests.com/ - **Francisco Varela / 弗朗西斯科·瓦雷拉** (Paris, FR / Santiago, CL, 1946–2001) — Chilean biologist and philosopher who co-developed autopoiesis and enactivism, the view that cognition is embodied action, a basis for embodied AI and robot art. Known for: Autopoiesis and Cognition (1980, with Humberto Maturana); The Embodied Mind (1991, with Evan Thompson and Eleanor Rosch). https://en.wikipedia.org/wiki/Francisco_Varela - **Friedrich Kittler / 弗里德里希·基特勒** (Berlin, DE, 1943–2011) — German media theorist who argued that media determine our situation, tracing how gramophone, film and typewriter turned human senses and writing into machine operations. Known for: Discourse Networks 1800/1900 (1985); Gramophone, Film, Typewriter (1986); There Is No Software (1992). https://en.wikipedia.org/wiki/Friedrich_Kittler - **Fritz Heider** (Lawrence, US, 1896–1988) — Austrian psychologist whose 1944 animation with Marianne Simmel showed that people read intentions and emotions into moving triangles, the root of why we see life in machines. Known for: An Experimental Study of Apparent Behavior (1944, with Marianne Simmel). https://en.wikipedia.org/wiki/Fritz_Heider - **Félix Guattari / 费利克斯·加塔利** (Paris, FR, 1930–1992) — French psychoanalyst and philosopher who gave the word 'machinic' its theoretical sense: processes that assemble humans, technologies and signs beyond the mechanical. Known for: Anti-Oedipus (1972, with Gilles Deleuze); Machinic Heterogenesis (1992); Chaosmosis (1992). https://en.wikipedia.org/wiki/F%C3%A9lix_Guattari - **Genevieve Bell** (Canberra, AU, 1968–) — Australian anthropologist, long at Intel, who founded the School of Cybernetics at ANU to build a new applied cybernetics for AI-era systems. Known for: Divining a Digital Future (2011, with Paul Dourish); 3A Institute / ANU School of Cybernetics (founded 2017). https://en.wikipedia.org/wiki/Genevieve_Bell - **Gilbert Simondon / 吉尔伯特·西蒙东** (Paris, FR, 1924–1989) — French philosopher who argued that technical objects have their own mode of existence and evolution, and that alienation comes from failing to understand machines, not from machines themselves. Known for: On the Mode of Existence of Technical Objects (1958); Individuation in Light of Notions of Form and Information (1958). https://en.wikipedia.org/wiki/Gilbert_Simondon - **Hubert Dreyfus / 休伯特·德雷福斯** (Berkeley, US, 1929–2017) — American philosopher who used Heidegger and Merleau-Ponty to argue that symbolic AI fails because intelligence depends on an embodied, situated way of being in the world. Known for: Alchemy and Artificial Intelligence (1965); What Computers Can't Do (1972). https://en.wikipedia.org/wiki/Hubert_Dreyfus - **Ian Bogost / 伊恩·博格斯特** (St. Louis, US, 1976–) — American philosopher and game designer whose Alien Phenomenology asks what it is like to be a thing, including a computer chip, a method close to machine phenomenology. Known for: Unit Operations (2006); Alien Phenomenology, or What It's Like to Be a Thing (2012). https://bogost.com/ - **Isaac Asimov / 艾萨克·阿西莫夫** (New York, US, 1920–1992) — American writer whose Three Laws of Robotics became the popular template for machine ethics, and whose robot stories explore machines that reason their way into paradox and personhood. Known for: Runaround (1942, Three Laws of Robotics); I, Robot (1950); The Bicentennial Man (1976). http://www.asimovonline.com/ - **J. C. R. Licklider** (Cambridge, US, 1915–1990) — American psychologist and computer scientist whose Man-Computer Symbiosis imagined humans and computers thinking together, a founding vision of interactive computing. Known for: Man-Computer Symbiosis (1960); The Computer as a Communication Device (1968, with Robert Taylor). https://en.wikipedia.org/wiki/J._C._R._Licklider - **Jack Burnham** (US, 1931–2019) — American theorist and curator who predicted that sculpture would become intelligent, responsive systems, and curated Software (1970), a landmark exhibition of information technology. Known for: Beyond Modern Sculpture (1968); Systems Esthetics (1968); Software (Jewish Museum, 1970, curator). https://en.wikipedia.org/wiki/Jack_Burnham - **Jakob von Uexküll / 雅各布·冯·于克斯屈尔** (Hamburg, DE, 1864–1944) — Baltic-German biologist whose concept of the umwelt, the perceptual world each organism lives in, has become a key tool for imagining what the world is like for a machine. Known for: Umwelt und Innenwelt der Tiere (1909); A Foray into the Worlds of Animals and Humans (1934). https://en.wikipedia.org/wiki/Jakob_Johann_von_Uexk%C3%BCll - **Janelle Shane** (Boulder, US) — American optics researcher whose blog AI Weirdness catalogues the strange failures of neural networks, making machine errors a way to understand how they learn. Known for: AI Weirdness (blog, 2016–); You Look Like a Thing and I Love You (2019). https://www.aiweirdness.com/ - **Jason Edward Lewis** (Montreal, CA) — Cherokee, Hawaiian and Samoan digital media scholar who leads the Indigenous Protocol and AI working group, asking how AI could be a relative within Indigenous knowledge systems. Known for: Making Kin with the Machines (2018, with Noelani Arista, Archer Pechawis and Suzanne Kite); Indigenous Protocol and Artificial Intelligence Position Paper (2020). https://www.indigenous-ai.net/ - **Jennifer Rhee** (Richmond, US) — American scholar whose The Robotic Imaginary shows how robots and AI are imagined through care and domestic labour, and whose humanity is measured against those it dehumanises. Known for: The Robotic Imaginary: The Human and the Price of Dehumanized Labor (2018). https://english.vcu.edu/people/faculty/rhee.html - **Jill Walker Rettberg** (Bergen, NO, 1971–) — Norwegian scholar of digital culture who led the ERC Machine Vision project, including a database of art, games and narratives about machine vision. Known for: Seeing Ourselves Through Technology (2014); Machine Vision: How Algorithms Are Changing the Way We See the World (2023). https://jilltxt.net/ - **Joanna Zylinska** (London, UK, 1971–) — Polish-British media philosopher and artist whose AI Art separates machine-reflexive art from 'AI art' spectacle, and whose The Perception Machine rethinks vision after AI. Known for: Nonhuman Photography (2017); AI Art: Machine Visions and Warped Dreams (2020); The Perception Machine (2023). http://www.joannazylinska.net/ - **John McCarthy / 约翰·麦卡锡** (Stanford, US, 1927–2011) — American computer scientist who coined the term 'artificial intelligence' in the 1955 Dartmouth proposal, naming the field whose promises art keeps testing. Known for: Dartmouth proposal coining 'artificial intelligence' (1955); LISP (1958). http://www-formal.stanford.edu/jmc// - **John Searle / 约翰·塞尔** (Berkeley, US, 1932–2025) — American philosopher whose Chinese Room argument claims that manipulating symbols is not understanding, a thought experiment revived by every generation of chatbots. Known for: Minds, Brains, and Programs ('Chinese Room', 1980). https://en.wikipedia.org/wiki/John_Searle - **Joseph Weizenbaum / 约瑟夫·维森鲍姆** (Cambridge, US / Berlin, DE, 1923–2008) — German-American computer scientist who built ELIZA, the first chatbot, then became a critic of AI after seeing people confide in it; the 'ELIZA effect' is named after it. Known for: ELIZA (1966); Computer Power and Human Reason (1976). https://en.wikipedia.org/wiki/Joseph_Weizenbaum - **Julie Carpenter** (San Francisco, US) — American researcher whose study of bomb-disposal soldiers showed them naming, mourning and holding funerals for their robots, a key case of attachment to machines. Known for: Culture and Human-Robot Interaction in Militarized Spaces (2016). https://jgcarpenter.com/ - **K Allado-McDowell** (Los Angeles, US) — American writer and founder of Google's Artists + Machine Intelligence programme; Pharmako-AI, co-written with GPT-3, is one of the first books written with a language model. Known for: Artists + Machine Intelligence (Google, 2016–); Pharmako-AI (2020); Air Age Blueprint (2022). https://kalladomcdowell.com/ - **Karel Čapek / 卡雷尔·恰佩克** (Prague, CZ, 1890–1938) — Czech writer whose play R.U.R. introduced the word 'robot' (from robota, forced labour), framing the machine from the start as a worker who may revolt. Known for: R.U.R. (Rossum's Universal Robots) (1920); War with the Newts (1936). https://en.wikipedia.org/wiki/Karel_%C4%8Capek - **Katarzyna Nowaczyk-Basińska** (Cambridge, UK) — Polish researcher of digital immortality at the Leverhulme Centre for the Future of Intelligence, studying how cultures imagine and regulate life after death through AI. Known for: Griefbots, Deadbots, Postmortem Avatars (2024, with Tomasz Hollanek). https://www.lcfi.ac.uk/people/katarzyna-nowaczyk-basinska - **Kate Crawford / 凯特·克劳福德** (New York, US / Sydney, AU, 1974–) — Australian scholar whose Atlas of AI describes AI as an extractive industry of minerals, labour and data, and whose collaborations with Joler and Paglen became artworks. Known for: Anatomy of an AI System (2018, with Vladan Joler); Excavating AI (2019, with Trevor Paglen); Atlas of AI (2021); Calculating Empires (2023, with Vladan Joler). https://katecrawford.net/ - **Kate Darling** (Cambridge, US, 1982–) — American researcher at MIT Media Lab who studies why people empathise with robots, and proposes animals rather than humans as the model for our relationships with machines. Known for: Extending Legal Protection to Social Robots (2012); The New Breed (2021). https://www.katedarling.org/ - **Kate Devlin** (London, UK) — British scholar at King's College London whose Turned On examines sex robots and intimacy with machines beyond moral panic. Known for: Turned On: Science, Sex and Robots (2018). https://en.wikipedia.org/wiki/Kate_Devlin - **Kathleen Richardson** (Leicester, UK) — British anthropologist of robotics who studied roboticists at MIT and launched the Campaign Against Sex Robots, arguing that robots mirror human attachment and its failures. Known for: An Anthropology of Robots and AI (2015); Campaign Against Sex Robots (2015); Challenging Sociality (2018). https://politicsoflove.org/ - **Lilly Irani** (San Diego, US) — American scholar and co-creator of Turkopticon, a tool that let Amazon Mechanical Turk workers rate employers, making the labour behind AI visible and organised. Known for: Turkopticon (2008, with Six Silberman); Chasing Innovation (2019). https://quote.ucsd.edu/lirani/ - **Louise Amoore** (Durham, UK, 1972–) — British political geographer whose Cloud Ethics argues that algorithms make ethical and political decisions through their attributes and thresholds, and must be held accountable as such. Known for: The Politics of Possibility (2013); Cloud Ethics: Algorithms and the Attributes of Ourselves and Others (2020). https://en.wikipedia.org/wiki/Louise_Amoore - **Luciana Parisi** (Durham, US, 1971–) — Italian-British philosopher of computation who argues that algorithms think in their own, alien way, through incomputable and indeterminate processes. Known for: Contagious Architecture (2013); The Alien Subject of AI (2019). https://scholars.duke.edu/person/luciana.parisi - **Lucy Suchman / 露西·萨奇曼** (Lancaster, UK) — American anthropologist of technology who showed at Xerox PARC that action is situated rather than planned, and later critiqued humanlike robots and autonomous weapons. Known for: Plans and Situated Actions (1987); Human–Machine Reconfigurations (2007). https://en.wikipedia.org/wiki/Lucy_Suchman - **Lydia H. Liu / 刘禾** (New York, US, 1957–) — Chinese-American theorist of media and translation whose The Freudian Robot reads cybernetics, information theory and the uncanny as a shared history of the human–machine unconscious. Known for: The Freudian Robot: Digital Media and the Future of the Unconscious (2010). https://en.wikipedia.org/wiki/Lydia_H._Liu - **M. Beatrice Fazi** (Brighton, UK) — Italian philosopher at Sussex whose work on computational aesthetics and explainability asks what it means that deep learning represents the world in ways humans cannot follow. Known for: Contingent Computation (2018); Beyond Human: Deep Learning, Explainability and Representation (2021). https://profiles.sussex.ac.uk/p176291-beatrice-fazi - **Mark Coeckelbergh** (Vienna, AT, 1975–) — Belgian philosopher of technology who argues that the moral status of robots emerges from relations and appearances rather than from inner properties. Known for: Growing Moral Relations (2012); AI Ethics (2020); The Political Philosophy of AI (2022). https://coeckelbergh.net/ - **Marvin Minsky / 马文·明斯基** (Cambridge, US, 1927–2016) — American AI pioneer and co-founder of the MIT AI Lab whose The Society of Mind and The Emotion Machine model minds as societies of simple agents. Known for: Perceptrons (1969, with Seymour Papert); The Society of Mind (1986); The Emotion Machine (2006). https://en.wikipedia.org/wiki/Marvin_Minsky - **Mary L. Gray** (Cambridge, US, 1969–) — American anthropologist whose Ghost Work, with Siddharth Suri, documents the hidden on-demand workers who label, moderate and correct the output of 'automated' systems. Known for: Ghost Work (2019, with Siddharth Suri). https://marylgray.org/ - **Matteo Pasquinelli / 马泰奥·帕斯奎内利** (Venice, IT, 1974–) — Italian philosopher whose The Eye of the Master argues that AI grew from the division and automation of labour, and that algorithms encode the pattern of work itself. Known for: The Nooscope Manifested (2020, with Vladan Joler); The Eye of the Master: A Social History of Artificial Intelligence (2023). http://matteopasquinelli.com/ - **Mercedes Bunz** (London, UK, 1971–) — German-British scholar at King's College London who co-founded the Creative AI Lab with the Serpentine and writes about how machines now calculate meaning. Known for: The Silent Revolution (2014); The Calculation of Meaning (2019); Creative AI Lab (with Eva Jäger, Serpentine × King's College London). https://mercedesbunz.net/ - **Meredith Whittaker** (New York, US) — American researcher who co-founded the AI Now Institute and helped organise the 2018 Google walkout; now president of Signal and a critic of AI's concentration of power. Known for: AI Now Institute (co-founded 2017); The Steep Cost of Capture (2021). https://en.wikipedia.org/wiki/Meredith_Whittaker - **Milagros Miceli** (Berlin, DE) — Argentine sociologist at the Weizenbaum Institute who studies the data annotators and content moderators of the Global South, and co-leads the Data Workers' Inquiry written by the workers themselves. Known for: Studying Up Machine Learning Data (2022); Data Workers' Inquiry (2024). https://www.weizenbaum-institut.de/en/portrait/p/milagros-miceli/ - **N. Katherine Hayles / 凯瑟琳·海勒** (Los Angeles, US, 1943–) — American literary theorist of the posthuman who insists on embodiment against disembodied information, and describes machines as 'nonconscious cognizers'. Known for: How We Became Posthuman (1999); Unthought: The Power of the Cognitive Nonconscious (2017); Bacteria to AI (2025). http://nkhayles.com/ - **Norbert Wiener / 诺伯特·维纳** (Cambridge, US, 1894–1964) — American mathematician who founded cybernetics, the science of control and communication in animals and machines, and warned early about automation and labour. Known for: Cybernetics (1948); The Human Use of Human Beings (1950); God & Golem, Inc. (1964). https://en.wikipedia.org/wiki/Norbert_Wiener - **Oliver Grau** (Krems, AT, 1965–) — German art historian who founded the Archive of Digital Art (formerly Database of Virtual Art), a key record of media and machine art. Known for: Virtual Art: From Illusion to Immersion (2003); Archive of Digital Art (ADA, founded 1999). https://en.wikipedia.org/wiki/Oliver_Grau - **Orit Halpern** (Dresden, DE, 1972–) — Historian of cybernetics and design whose Beautiful Data traces how postwar cybernetics reshaped vision and reason into data, and whose The Smartness Mandate critiques 'smart' everything. Known for: Beautiful Data (2014); The Smartness Mandate (2022, with Robert Mitchell). https://en.wikipedia.org/wiki/Orit_Halpern - **Paul Virilio / 保罗·维利里奥** (Paris, FR, 1932–2018) — French theorist whose The Vision Machine imagined 'sightless vision', cameras and computers that see and interpret without any human looking. Known for: War and Cinema (1984); The Vision Machine (1988). https://en.wikipedia.org/wiki/Paul_Virilio - **Peter-Paul Verbeek** (Amsterdam, NL, 1970–) — Dutch philosopher of technology who extends postphenomenology to show how artefacts mediate perception and moral action; his 'composite intentionality' opens space for machine perception. Known for: What Things Do (2005); Moralizing Technology (2011). https://en.wikipedia.org/wiki/Peter-Paul_Verbeek - **Phil Agre** (Los Angeles, US) — American AI researcher who proposed 'critical technical practice', building AI while reflecting on its assumptions, an approach later adopted by artists and designers. Known for: Pengi (1987, with David Chapman); Computation and Human Experience (1997); Surveillance and Capture (1994). https://en.wikipedia.org/wiki/Philip_Agre - **Phoebe Sengers** (Ithaca, US) — American HCI researcher who built 'schizophrenic' agents to expose how AI design splits behaviour, and founded reflective design as critical technical practice. Known for: Anti-Boxology: Agent Design in Cultural Context (1998); Reflective Design (2005). http://www.cs.cornell.edu/people/sengers/ - **Robert Sparrow** (Melbourne, AU) — Australian philosopher who argued early that robot pets for older people depend on deception, and that autonomous weapons create a gap in responsibility. Known for: The March of the Robot Dogs (2002); Killer Robots (2007); In the Hands of Machines? The Future of Aged Care (2006, with Linda Sparrow). https://research.monash.edu/en/persons/robert-sparrow - **Ron Wakkary** (Vancouver, CA) — Canadian design researcher whose Morse Things, ceramic bowls that talk to each other in Morse code over the network, were lived with in homes to study a world of things with their own lives. Known for: Morse Things (DIS 2017); Things We Could Design: For More Than Human-Centered Worlds (2021). http://eds.siat.sfu.ca/ - **Roy Ascott / 罗伊·阿斯科特** (London, UK / Shanghai, CN, 1934–) — British artist and theorist of cybernetic and telematic art whose teaching and writings made feedback, interaction and networks central to art, including later in China. Known for: Behaviourist Art and the Cybernetic Vision (1966–67); La Plissure du Texte (1983); Telematic Embrace (2003). https://www.plymouth.ac.uk/staff/rascott - **Ruha Benjamin** (Princeton, US, 1978–) — American sociologist whose Race After Technology names the 'New Jim Code', the way seemingly neutral systems reproduce racial hierarchies. Known for: Race After Technology (2019); Viral Justice (2022); Imagination: A Manifesto (2024). https://www.ruhabenjamin.com/ - **Safiya Umoja Noble** (Los Angeles, US, 1970–) — American scholar whose Algorithms of Oppression showed how search engines represent Black women through racist and sexist results. Known for: Algorithms of Oppression (2018). https://safiyaunoble.com/ - **Selma Šabanović** (Bloomington, US) — Bosnian-American researcher of social robotics who studies how robots such as PARO are designed and lived with across cultures, especially in Japan and the US. Known for: Robots in Society, Society in Robots (2010). https://r-house.luddy.indiana.edu/ - **Sherry Turkle / 雪莉·特克尔** (Cambridge, US, 1948–) — American psychologist at MIT whose studies of children and elders with Furby, AIBO and PARO describe 'the robotic moment', when we accept machine companionship as enough. Known for: The Second Self (1984); Alone Together (2011); Reclaiming Conversation (2015). https://sherryturkle.mit.edu/ - **Shoshana Zuboff / 肖莎娜·祖博夫** (Maine, US, 1951–) — American scholar who described 'surveillance capitalism', the extraction of human experience as behavioural data for prediction products. Known for: In the Age of the Smart Machine (1988); The Age of Surveillance Capitalism (2019). https://shoshanazuboff.com/ - **Simone Browne** (Austin, US, 1973–) — American scholar whose Dark Matters traces surveillance technologies, including biometrics, back to the policing of Black life under slavery. Known for: Dark Matters: On the Surveillance of Blackness (2015). https://en.wikipedia.org/wiki/Simone_Browne - **Stanisław Lem / 斯坦尼斯瓦夫·莱姆** (Kraków, PL, 1921–2006) — Polish writer and philosopher whose Golem XIV, a superintelligent computer lecturing humans on their limits, remains one of the sharpest imaginings of truly alien machine minds. Known for: Summa Technologiae (1964); The Cyberiad (1965); Golem XIV (1981). https://lem.pl - **Ted Chiang / 姜峯楠** (Seattle, US, 1967–) — American science-fiction writer whose The Lifecycle of Software Objects follows digital pets raised for years like children, and whose essays sharply describe what language models are. Known for: The Lifecycle of Software Objects (2010); ChatGPT Is a Blurry JPEG of the Web (2023). https://en.wikipedia.org/wiki/Ted_Chiang - **Timnit Gebru / 蒂姆尼特·格布鲁** (San Francisco, US, 1983–) — Ethiopian-born Eritrean-American computer scientist whose work on face-analysis bias and large language models, and her dismissal from Google in 2020, reshaped debates on AI ethics; founder of DAIR. Known for: Gender Shades (2018, with Joy Buolamwini); Datasheets for Datasets (2018); On the Dangers of Stochastic Parrots (2021); DAIR Institute (founded 2021). https://en.wikipedia.org/wiki/Timnit_Gebru - **Tomasz Hollanek** (Cambridge, UK) — Researcher at the Leverhulme Centre for the Future of Intelligence whose speculative scenarios of 'deadbots' propose design rules for the AI afterlife industry. Known for: Griefbots, Deadbots, Postmortem Avatars (2024, with Katarzyna Nowaczyk-Basińska). https://www.lcfi.ac.uk/people/tomasz-hollanek - **Valentino Braitenberg** (Tübingen, DE, 1926–2011) — Italian-Austrian neuroscientist whose thought experiment Vehicles shows how tiny wired machines appear to feel fear, love or curiosity, a founding text of synthetic psychology. Known for: Vehicles: Experiments in Synthetic Psychology (1984). https://en.wikipedia.org/wiki/Valentino_Braitenberg - **Vilém Flusser / 威廉·弗卢塞尔** (São Paulo, BR / Robion, FR, 1920–1991) — Czech-Brazilian philosopher who described the camera as an 'apparatus' whose program the photographer plays within, a theory that anticipates image-making machines. Known for: Towards a Philosophy of Photography (1983); Into the Universe of Technical Images (1985). https://en.wikipedia.org/wiki/Vil%C3%A9m_Flusser - **Virginia Eubanks** (Troy, US, 1972–) — American political scientist whose Automating Inequality shows how algorithmic systems profile, police and punish the poor. Known for: Automating Inequality (2018). https://virginia-eubanks.com/ - **Wendy Hui Kyong Chun** (Vancouver, CA, 1969–) — American-Canadian media theorist whose Discriminating Data shows how machine learning's logic of homophily ('like likes like') builds segregation into networks. Known for: Programmed Visions (2011); Updating to Remain the Same (2016); Discriminating Data (2021). https://en.wikipedia.org/wiki/Wendy_Hui_Kyong_Chun - **Yuk Hui / 许煜** (Rotterdam, NL / Hong Kong, 1985–) — Hong Kong philosopher whose idea of 'cosmotechnics' argues that there is no single universal technology, opening non-Western ways to think about machines and AI. Known for: On the Existence of Digital Objects (2016); The Question Concerning Technology in China (2016); Recursivity and Contingency (2019); Machine and Sovereignty (2024). https://digitalmilieu.net/yuk/ ### Curators & critics - **Amelie Klein** (Weil am Rhein, DE) — Curator at the Vitra Design Museum whose touring exhibition Hello, Robot. Design between Human and Machine surveyed robots from science fiction to smart homes. Known for: Hello, Robot. Design between Human and Machine (Vitra Design Museum, 2017). https://www.design-museum.de/en/exhibitions/detailpages/hello-robot.html - **Antonio Somaini** (Paris, FR, 1968–) — Italian film and media theorist at Sorbonne Nouvelle who curated The World Through AI at the Jeu de Paume (2025), an exhibition on how AI sees, classifies and generates images. Known for: The World Through AI (Le monde selon l'IA) (Jeu de Paume, 2025); Algorithmic Images: Artificial Intelligence and Visual Culture (2023). https://jeudepaume.org/en/evenement/the-world-through-ai/ - **Christiane Paul** (New York, US) — German-American curator of digital art at the Whitney Museum and author of the standard survey Digital Art; she curated the 2024 Whitney show of Harold Cohen's AARON. Known for: Digital Art (1st ed. 2003); Programmed: Rules, Codes, and Choreographies in Art (Whitney, 2018); Harold Cohen: AARON (Whitney, 2024). https://en.wikipedia.org/wiki/Christiane_Paul_(curator) - **Claudia Schmuckli** (San Francisco, US) — Curator of contemporary art at the Fine Arts Museums of San Francisco whose Uncanny Valley: Being Human in the Age of AI brought critical AI art to a major US museum. Known for: Uncanny Valley: Being Human in the Age of AI (de Young, 2020). https://www.famsf.org/exhibitions/uncanny-valley - **Eva Jäger** (London, UK) — Curator at Serpentine Arts Technologies and co-founder of the Creative AI Lab with King's College London, which studies how artists work with AI as a medium and an institution. Known for: Creative AI Lab (co-founded with Mercedes Bunz, 2019); Future Art Ecosystems (Serpentine, 2020–). https://creative-ai.org/ - **Franziska Nori** (Frankfurt, DE, 1968–) — Director of the Frankfurter Kunstverein who co-curated I Am Here to Learn (2018), one of the first exhibitions devoted to how machines interpret the world. Known for: I Am Here to Learn: On Machinic Interpretations of the World (Frankfurter Kunstverein, 2018, with Mattis Kuhn). https://de.wikipedia.org/wiki/Franziska_Nori - **Gerfried Stocker** (Linz, AT, 1964–) — Austrian media artist and artistic director of Ars Electronica since 1995, where the festival and Prix have been a main stage for robot and AI art. Known for: Artistic director of Ars Electronica (1995–); Ars Electronica Futurelab (founded 1996). https://de.wikipedia.org/wiki/Gerfried_Stocker - **Ine Gevers** (Eindhoven, NL, 1960–) — Dutch curator whose large exhibition Robot Love (2018) asked whether robots can love and whether people can learn love from robots. Known for: Robot Love (Eindhoven, 2018); Niet Normaal: Difference on Display (2009). https://robotlove.nl/ - **Irini Papadimitriou** (Manchester, UK) — Greek-British curator and creative director of FutureEverything, previously head of digital programmes at the V&A, whose programmes examine AI, data and the ethics of technology. Known for: FutureEverything (creative director); V&A Digital Design Weekend (2010s). https://futureeverything.org/ - **Jasia Reichardt** (London, UK, 1933–) — Polish-British curator and writer whose Cybernetic Serendipity at the ICA (1968) was the first major exhibition of computer and cybernetic art, with robots, drawing machines and Pask's Colloquy of Mobiles. Known for: Cybernetic Serendipity (ICA London, 1968); Robots: Fact, Fiction and Prediction (1978). https://en.wikipedia.org/wiki/Jasia_Reichardt - **Joasia Krysa** (Liverpool, UK) — Polish-British curator and scholar of 'exhibitionary AI' whose project The Next Biennial Should Be Curated by a Machine tests what happens when AI curates. Known for: The Next Biennial Should Be Curated by a Machine (2021, with Leonardo Impett); Liverpool Biennial (co-curator, 2016). https://www.ljmu.ac.uk/Research/Centres%20and%20institutes/Institute-of-Art-and-Technology/Expertise/Exhibition-Research-Lab - **Kathy Rae Huffman** (US / UK, 1943–) — American curator and pioneer of video and net art, and of early cyberfeminist networks, whose work since the 1970s has tracked the arrival of new media machines in art. Known for: Video art curator, Long Beach Museum of Art (1978–84); Face Settings (1997–2000, with Eva Wohlgemuth). https://en.wikipedia.org/wiki/Kathy_Rae_Huffman - **Laurence Bertrand Dorléac** (Paris, FR, 1957–) — French art historian who co-curated Artists & Robots at the Grand Palais (2018), a survey of robot and algorithmic art from Schöffer to today. Known for: Artists & Robots (Grand Palais, 2018, with Jérôme Neutres). https://en.wikipedia.org/wiki/Laurence_Bertrand_Dorl%C3%A9ac - **Legacy Russell** (New York, US, 1986–) — American curator and writer whose Glitch Feminism treats the glitch, the machine error, as a site of refusal for queer and Black bodies online; executive director of The Kitchen. Known for: Glitch Feminism: A Manifesto (2020); Black Meme (2024). https://www.legacyrussell.com - **Maholo Uchida / 内田真帆罗** (Tokyo, JP) — Japanese curator at Miraikan (National Museum of Emerging Science and Innovation, Tokyo), where she worked on android exhibitions with Hiroshi Ishiguro, and co-curator of AI: More than Human. Known for: AI: More than Human (Barbican, 2019, with Suzanne Livingston); Android exhibits at Miraikan (2014–). https://www.barbican.org.uk/whats-on/2019/event/ai-more-than-human - **Nora N. Khan** (Los Angeles, US, 1983–) — American writer and curator whose Seeing, Naming, Knowing examines how machine vision names and judges, and whose criticism shaped how the art world writes about AI. Known for: Seeing, Naming, Knowing (2019); Refiguring the Future (2019, co-curator with Lola Kramer). https://en.wikipedia.org/wiki/Nora_Khan - **Peter Weibel / 彼得·魏贝尔** (Karlsruhe, DE, 1944–2023) — Austrian artist, theorist and long-time director of ZKM Karlsruhe, whose exhibitions such as Open Codes placed algorithms, AI and machine agency at the centre of art. Known for: Open Codes (ZKM, 2017–19); Artistic director of ZKM Karlsruhe (1999–2023); Net_Condition (ZKM, 1999). https://en.wikipedia.org/wiki/Peter_Weibel - **Pontus Hultén** (Stockholm, SE / Paris, FR, 1924–2006) — Swedish curator and founding director of the Centre Pompidou's museum whose MoMA exhibition The Machine as Seen at the End of the Mechanical Age surveyed art's relation to machines at the dawn of computing. Known for: The Machine as Seen at the End of the Mechanical Age (MoMA, 1968); Bewogen Beweging (Stedelijk, 1961). https://en.wikipedia.org/wiki/Pontus_Hult%C3%A9n - **Sabine Himmelsbach** (Basel, CH, 1966–) — German curator and director of HEK (House of Electronic Arts) in Basel, whose exhibitions such as Entangled Realities look at how we live with AI and machines. Known for: Entangled Realities: Living with Artificial Intelligence (HEK, 2019); Director of HEK Basel (2012–). https://de.wikipedia.org/wiki/Sabine_Himmelsbach - **Suzanne Livingston** (London, UK) — British curator and writer who co-curated AI: More than Human at the Barbican (2019), a large survey that toured internationally and mixed art, science and the history of AI. Known for: AI: More than Human (Barbican, 2019, with Maholo Uchida). https://www.drsuzannelivingston.com/ - **Zhang Ga / 张尕** (Shanghai / Beijing, CN) — Chinese curator and artistic director of Chronus Art Center in Shanghai whose triennials of media art at the National Art Museum of China brought robot, AI and bio art to Chinese audiences. Known for: Synthetic Times (NAMOC, 2008); Translife (NAMOC, 2011); Thingworld (NAMOC, 2014); Artistic director of Chronus Art Center. https://chronusartcenter.org/ ### Engineers & roboticists - **Billy Klüver** (New York, US, 1927–2004) — Swedish Bell Labs engineer who built machines with Tinguely, Rauschenberg and Warhol and co-founded Experiments in Art and Technology, the model for artist–engineer collaboration. Known for: Homage to New York (1960, with Jean Tinguely); 9 Evenings: Theatre and Engineering (1966); Experiments in Art and Technology (E.A.T., founded 1967). https://en.wikipedia.org/wiki/Billy_Kl%C3%BCver - **Blake Lemoine** (San Francisco, US) — American engineer who was dismissed by Google in 2022 after publicly claiming that its LaMDA chatbot was sentient, a flashpoint for debates on machine interiority. Known for: Is LaMDA Sentient? – an Interview (2022). https://cajundiscordian.medium.com/is-lamda-sentient-an-interview-ea64d916d917 - **Christoph Bartneck** (Christchurch, NZ, 1973–) — German HRI researcher who studied whether people hesitate to switch off a pleading robot, robot abuse, and whether robots are racialised by their colour. Known for: Daisy, Daisy, Give Me Your Answer Do! (2007); Godspeed questionnaires (2009); Robots and Racism (2018). https://www.bartneck.de/ - **Claude Shannon / 克劳德·香农** (Cambridge, US, 1916–2001) — American mathematician who founded information theory and built Theseus, a mechanical mouse that learned a maze, as well as the 'useless machine' that only switches itself off. Known for: A Mathematical Theory of Communication (1948); Theseus maze-solving mouse (1950); Ultimate Machine (1952, after Marvin Minsky). https://en.wikipedia.org/wiki/Claude_Shannon - **Cynthia Breazeal / 辛西娅·布雷齐尔** (Cambridge, US, 1967–) — American roboticist who founded social robotics with Kismet, an expressive robot head that responded to people's emotions, and later Jibo, one of the first social robots for the home. Known for: Kismet (1998–2000); Designing Sociable Robots (2002); Leonardo (2003, with Stan Winston Studio); Jibo (2014–). https://www.media.mit.edu/people/cynthiab/ - **David Hanson** (Hong Kong, 1969–) — American robot designer and founder of Hanson Robotics, maker of lifelike humanoid heads such as Bina48 and Sophia, a robot granted Saudi citizenship in 2017. Known for: Philip K. Dick android (2005); Bina48 (2010, for Terasem); Sophia (2016). https://www.hansonrobotics.com/ - **Eugenia Kuyda** (San Francisco, US) — Russian-born founder of Replika, which began as a chatbot built from the messages of her late friend Roman Mazurenko and became a companion app used by millions. Known for: Roman (memorial bot of Roman Mazurenko) (2016); Replika (2017). https://replika.com/ - **Gil Weinberg** (Atlanta, US, 1967–) — Israeli-American music technologist at Georgia Tech whose robotic musicians Haile and Shimon listen and improvise with human players. Known for: Haile (2006); Shimon (2009–); Robotic prosthetic drumming arm (2014). https://www.gilweinberg.com/ - **Guy Hoffman** (Ithaca, US) — Israeli roboticist who draws on acting and animation to design non-humanoid social robots, from a robotic desk lamp to the soft, handcrafted Blossom. Known for: AUR robotic desk lamp (2007); Shimon (2009, with Gil Weinberg); Travis / Shimi (2010s); Blossom (2017). https://hrc2.io/ - **Hans Moravec / 汉斯·莫拉维克** (Pittsburgh, US, 1948–) — Austrian-Canadian roboticist behind Moravec's paradox, the observation that what is easy for humans, like walking and seeing, is hardest for machines. Known for: Stanford Cart (1979); Mind Children (1988); Moravec's paradox (1988). https://frc.ri.cmu.edu/~hpm/ - **Heather Knight** (Corvallis, US) — American roboticist who studies charisma and social intelligence through robot performance, including a NAO robot doing stand-up comedy that adapts to audience laughter. Known for: Data, a stand-up comedy robot (2010); Marilyn Monrobot (company, 2010s); Robot Film Festival (2012–). https://www.charismarobotics.com/ - **Hiroshi Ishiguro / 石黑浩** (Osaka, JP, 1963–) — Japanese roboticist who builds lifelike androids, including a copy of himself, to study what 'human presence' is; his androids act in theatre, museums and films. Known for: Geminoid HI-1 (2006); Telenoid (2010); Erica (2015); Alter (2016, with Takashi Ikegami). https://eng.irl.sys.es.osaka-u.ac.jp/ - **Hod Lipson** (New York, US, 1967–) — Israeli-American roboticist at Columbia whose robots build models of their own bodies, an experimental approach to machine self-awareness. Known for: Resilient Machines Through Continuous Self-Modeling (2006, with Josh Bongard and Victor Zykov); Self-replicating robots (2005); Robot self-image from vision (2022). https://www.hodlipson.com/ - **Hugh Herr** (Cambridge, US, 1964–) — American engineer and climber, a double amputee, who leads MIT's Biomechatronics group and builds bionic limbs that the body feels as its own. Known for: BiOM bionic ankle (2010s); Agonist-antagonist myoneural interface (2018). https://www.media.mit.edu/people/hherr/overview/ - **Ichiro Kato / 加藤一郎** (Tokyo, JP, 1925–1994) — Japanese engineer at Waseda University who built WABOT-1, the first full-scale humanoid robot, and WABOT-2, which read sheet music and played the organ. Known for: WABOT-1 (1973); WABOT-2 (1984). https://ja.wikipedia.org/wiki/加藤一郎_(ロボット研究者) - **Jacques de Vaucanson / 雅克·德·沃康松** (Paris, FR, 1709–1782) — French inventor (1709–1782) whose automata, a flute player that really blew into its flute and a duck that appeared to eat and digest, made the body itself a machine question. Known for: The Flute Player (1737); The Digesting Duck (1739). https://en.wikipedia.org/wiki/Jacques_de_Vaucanson - **Joseph Engelberger** (Danbury, US, 1925–2015) — American engineer called the 'father of robotics', whose Unimate arm began work at General Motors in 1961, the industrial robot that later became a material of art. Known for: Unimate (1961, General Motors); Robotics in Practice (1980); HelpMate (1980s). https://en.wikipedia.org/wiki/Joseph_Engelberger - **Josh Bongard** (Burlington, US, 1974–) — Canadian roboticist at the University of Vermont who, with Michael Levin, designed xenobots, living machines assembled from frog cells by an evolutionary algorithm. Known for: Resilient Machines Through Continuous Self-Modeling (2006); Xenobots (2020, with Michael Levin); How the Body Shapes the Way We Think (2006, with Rolf Pfeifer). https://jbongard.github.io - **Kaname Hayashi / 林要** (Tokyo, JP) — Japanese engineer who led Pepper's development at SoftBank and then founded GROOVE X to make LOVOT, a warm, useless robot designed only to be loved. Known for: Pepper (development lead, 2014); LOVOT (GROOVE X, 2019). https://lovot.life/ - **Kenneth Colby** (Stanford, US, 1920–2001) — American psychiatrist who wrote PARRY, a chatbot simulating a paranoid patient, which in 1972 'met' ELIZA for a machine-to-machine therapy session. Known for: PARRY (1972); Artificial Paranoia (1975). https://en.wikipedia.org/wiki/Kenneth_Colby - **Kerstin Dautenhahn** (Waterloo, CA, 1964–) — German roboticist whose child-sized, minimally expressive robot KASPAR is used with autistic children, showing that simplified robot faces can support social interaction. Known for: KASPAR (2005–); Socially Intelligent Agents (2002, ed.). https://researchprofiles.herts.ac.uk/portal/en/persons/kerstin-dautenhahn(0d00e4af-1f2d-433e-8d0e-ac8690288e4c).html - **Kevin Warwick** (Coventry, UK, 1954–) — British engineer who implanted an RFID chip (1998) and then a 100-electrode array in his arm (2002), linking his nervous system to a robot hand and to his wife's implant. Known for: Project Cyborg 1.0 (1998); Project Cyborg 2.0 (2002); I, Cyborg (2002). https://en.wikipedia.org/wiki/Kevin_Warwick - **Maja Matarić** (Los Angeles, US, 1965–) — Serbian-American roboticist at USC who defined socially assistive robotics, robots that help people through social rather than physical interaction, in stroke rehabilitation, autism and elder care. Known for: Socially assistive robotics (2005, with David Feil-Seifer); The Robotics Primer (2007). https://en.wikipedia.org/wiki/Maja_Matari%C4%87 - **Marc Raibert** (Waltham, US, 1949–) — American roboticist who founded Boston Dynamics, whose legged robots and viral videos of them being kicked and dancing shaped public feelings about robot bodies. Known for: Boston Dynamics (founded 1992); BigDog (2005); Spot (2016–); Atlas (2013–). https://en.wikipedia.org/wiki/Marc_Raibert - **Masahiro Mori / 森政弘** (Tokyo, JP, 1927–2025) — Japanese roboticist whose 1970 essay on the 'uncanny valley' described the unease caused by almost-human machines, and whose The Buddha in the Robot finds Buddha-nature in machines. Known for: The Uncanny Valley (Bukimi no Tani, 1970); The Buddha in the Robot (1974); Robocon (1988–). https://en.wikipedia.org/wiki/Masahiro_Mori_(roboticist) - **Michael Levin** (Medford, US, 1969–) — American developmental biologist whose xenobots and anthrobots blur the line between organism and machine and who proposes a framework for recognising minds in unconventional bodies. Known for: Xenobots (2020, with Josh Bongard and Sam Kriegman); Anthrobots (2023); Technological Approach to Mind Everywhere (2022). https://drmichaellevin.org/ - **Michio Okada / 冈田美智男** (Toyohashi, JP, 1960–) — Japanese researcher of 'weak robots', machines that cannot finish their tasks alone, such as a trash-can robot that needs passers-by to pick up litter, making dependence the basis of care. Known for: Weak Robots (Yowai Robotto, 2012); Muu (2000s); Gomi-bako robot (2010s); NICOBO (2021, with Panasonic). https://ja.wikipedia.org/wiki/岡田美智男 - **Pierre Jaquet-Droz** (La Chaux-de-Fonds, CH, 1721–1790) — Swiss watchmaker (1721–1790) whose android automata, including The Writer, which can be programmed to write any text of 40 characters, are ancestors of programmable robots. Known for: The Writer (1768–74); The Draughtsman (1768–74); The Musician (1768–74). https://www.jaquet-droz.com/fr/the-extraordinary-history-jaquet-droz - **Radhika Nagpal** (Princeton, US) — Indian-American roboticist whose thousand-robot Kilobot swarm and termite-like builders show collective behaviour emerging from simple machines. Known for: Kilobot swarm of 1,024 robots (2014); TERMES termite-inspired builders (2014); Blueswarm (2021). https://en.wikipedia.org/wiki/Radhika_Nagpal - **Raffaello D'Andrea** (Zurich, CH, 1967–) — Canadian-Italian-Swiss engineer who co-created The Table, a robot table that chooses a visitor to follow, co-founded Kiva Systems (now Amazon Robotics) and choreographed drones. Known for: The Table (2001, with Max Dean); Kiva Systems warehouse robots (2003); Flying Machine Arena (2008); Sparked (Cirque du Soleil, 2014). http://raffaello.name/ - **Rebecca Fiebrink** (London, UK) — American researcher at UAL Creative Computing Institute whose Wekinator lets artists train machine-learning models by demonstration, treating ML as a creative, embodied instrument. Known for: Wekinator (2009); Machine Learning for Musicians and Artists (Kadenze, 2016). https://www.doc.gold.ac.uk/~mas01rf/ - **Rodney Brooks / 罗德尼·布鲁克斯** (Boston, US, 1954–) — Australian roboticist whose behaviour-based robotics replaced central planning with layered, embodied reflexes, from the six-legged Genghis to the Roomba vacuum cleaner. Known for: Elephants Don't Play Chess (1990); Genghis (1989); Cog (1993–); Roomba (iRobot, 2002). https://rodneybrooks.com/ - **Rolf Pfeifer** (Zurich, CH, 1947–) — Swiss AI researcher who championed embodied intelligence, showing that a robot's body and materials do part of its thinking ('morphological computation'). Known for: Understanding Intelligence (1999, with Christian Scheier); How the Body Shapes the Way We Think (2006, with Josh Bongard); ECCE / Roboy (2010s). https://en.wikipedia.org/wiki/Rolf_Pfeifer - **Steve Mann** (Toronto, CA, 1962–) — Canadian engineer who has lived with wearable computers and head-mounted cameras since the late 1970s and coined 'sousveillance', watching the watchers from below. Known for: WearComp / EyeTap (late 1970s–); Sousveillance (2002). http://wearcam.org/ - **Susumu Tachi / 馆暲** (Tokyo, JP, 1946–) — Japanese engineer who proposed 'telexistence' in 1980 and built the TELESAR robots that let a person see, hear and feel through a remote robot body. Known for: Telexistence (1980); TELESAR series (1980s–); Optical camouflage (2003). https://tachilab.org/en/ - **Takanori Shibata / 柴田崇德** (Tsukuba, JP, 1967–) — Japanese engineer at AIST who created PARO, a baby harp seal robot used in dementia care worldwide, and a central case in debates on machine care. Known for: PARO therapeutic seal robot (research from 1993; released 2004). https://ja.wikipedia.org/wiki/柴田崇徳 - **Takashi Ikegami / 池上高志** (Tokyo, JP, 1961–) — Japanese artificial-life scientist at the University of Tokyo who gives machines spontaneous, self-organising behaviour; his android Alter moves by a neural network and central pattern generators. Known for: Mind Time Machine (YCAM, 2010); Alter (2016, with Hiroshi Ishiguro); Alter 3 in Scary Beauty (2018). https://en.wikipedia.org/wiki/Takashi_Ikegami - **Tomotaka Takahashi / 高桥智隆** (Kyoto, JP, 1975–) — Japanese robot creator whose small humanoids include Kirobo, the first talking robot companion in space, and RoBoHoN, a robot that is also a phone. Known for: Kirobo (2013, on the ISS); Robi (2013); RoBoHoN (2016). https://en.wikipedia.org/wiki/Tomotaka_Takahashi - **W. Grey Walter** (Bristol, UK, 1910–1977) — American-born British neurophysiologist whose light-seeking robot tortoises Elmer and Elsie showed lifelike, unpredictable behaviour from two 'neurons', the first autonomous robots. Known for: Machina speculatrix tortoises Elmer and Elsie (1948–49); The Living Brain (1953). https://en.wikipedia.org/wiki/William_Grey_Walter - **Wang Xingxing / 王兴兴** (Hangzhou, CN, 1990–) — Chinese engineer and founder of Unitree Robotics, whose quadruped and humanoid robots danced yangge with handkerchiefs on the 2025 Spring Festival Gala, a landmark for public images of robots in China. Known for: Unitree Robotics (founded 2016); Unitree H1 humanoids at the CCTV Spring Festival Gala (2025). https://www.unitree.com/ - **Wendy Ju** (New York, US) — American interaction researcher at Cornell Tech whose Ghost Driver hid a driver inside a car seat costume to study how pedestrians react to 'driverless' cars. Known for: Ghost Driver (2015); The Design of Implicit Interactions (2015). http://www.wendyju.com/ ### Museums & art centres - **Aksioma – Institute for Contemporary Art** (Ljubljana, Slovenia) — Ljubljana nonprofit that produces and publishes critical art on networks, algorithms and power, including the PostScriptUM publication series. https://aksioma.org/ - **Ars Electronica Center / 林茨电子艺术中心** (Linz, Austria) — The 'museum of the future' of Ars Electronica. Its permanent exhibition Understanding AI (since 2019) shows how machines perceive and classify, with works such as Anatomy of an AI System and Gender Shades. https://ars.electronica.art/center/en/ - **Art Center Nabi / Nabi 艺术中心** (Seoul, South Korea) — Korea's first art centre dedicated to digital art, founded by SK Group. It has run robot art, AI and creative-technology programmes and hackathons for more than two decades. http://www.nabi.or.kr/ - **Barbican Centre / 巴比肯艺术中心** (London, UK) — London arts centre whose exhibition AI: More than Human (2019) became a template for AI surveys, running from the dream of artificial beings through mind machines and data worlds. https://www.barbican.org.uk/ - **Carnegie Museum of Art (Hillman Photography Initiative) / 卡内基艺术博物馆** (Pittsburgh, US) — Pittsburgh museum whose Hillman Photography Initiative ran Mirror with a Memory (2020–21), a programme on AI, photography, surveillance and bias, including Trevor Paglen's Opposing Geometries. https://carnegieart.org/ - **Centre Pompidou / 蓬皮杜艺术中心** (Paris, France) — Paris's national modern art centre. Its Mutations/Créations series staged Neurones, les intelligences simulées (2020), a history of neural and simulated intelligence across art, design and science. https://www.centrepompidou.fr/en/ - **Chronus Art Center / 新时线媒体艺术中心** (Shanghai, China) — Shanghai's first nonprofit art centre dedicated to media art, with exhibitions, research and a lab. A main venue in China for machine, algorithmic and kinetic work, including the Chinese tour of ZKM's Open Codes. https://www.chronusartcenter.org/ - **Computer History Museum / 计算机历史博物馆** (Mountain View, US) — Silicon Valley museum and archive of computing. Its collection and exhibition Revolution include AI and robotics milestones such as Shakey the robot, with oral histories of the engineers. https://computerhistory.org/ - **de Young Museum (FAMSF) / 笛洋美术馆** (San Francisco, US) — San Francisco fine arts museum that staged Uncanny Valley: Being Human in the Age of AI (2020), which read the uncanny valley politically as the gap between AI's promises and its effects on data, labour and bias. https://www.famsf.org/visit/de-young - **FACT Liverpool / 利物浦艺术与创意技术基金会** (Liverpool, UK) — The UK's leading centre for art and creative technology (Foundation for Art and Creative Technology). It commissions and shows work on data, surveillance and AI, often with community co-production. https://www.fact.co.uk/ - **Fondazione Prada / 普拉达基金会** (Milan, Italy) — Milan art foundation whose Osservatorio hosted Training Humans (2019), conceived by Kate Crawford and Trevor Paglen: the first major exhibition of the training images that teach machines to see people. https://en.wikipedia.org/wiki/Fondazione_Prada - **Furtherfield** (London, UK) — London's longest-running art-and-technology organisation, with a gallery and lab in Finsbury Park. It stages critical exhibitions and games on AI, data, blockchain and more-than-human futures. https://www.furtherfield.org/ - **Haus der Kulturen der Welt (HKW) / 世界文化之家** (Berlin, Germany) — Berlin centre for contemporary arts and discourse. Long-term projects such as Technosphere (2015–19) examined computation, infrastructure and planetary technology. https://www.hkw.de/en - **HEK (House of Electronic Arts) Basel / 巴塞尔电子艺术之家** (Basel, Switzerland) — Swiss centre for digital culture and new art forms. Its 2019 show Entangled Realities: Living with Artificial Intelligence brought together LAUREN, Tay's resurrection and Autonomous Trap 001. https://www.hek.ch/en/ - **iMAL** (Brussels, Belgium) — Brussels centre for digital cultures and technology with an exhibition space and fablab. It shows and produces work on algorithms, sensing and automation. https://www.imal.org/en - **Jeu de Paume / 网球场国家美术馆** (Paris, France) — Paris centre for photography and the image. The World Through AI (2025) split AI into analytical and generative systems and set artworks beside historical 'time capsules' of machine vision. https://jeudepaume.org/en/ - **KADIST** (Paris, France / San Francisco, US) — Nonprofit contemporary art organisation that collects and commissions globally. Its exhibition Apophenia, Interruptions: Artists and Artificial Intelligence at Work (2024) looked at pattern-finding and misrecognition. https://kadist.org/ - **Kapelica Gallery / Kersnikova Institute** (Ljubljana, Slovenia) — Ljubljana gallery for 'investigative' art, including robotics, bio art and body-machine performance (Stelarc, Špela Petrič). The Kersnikova Institute also runs the RAMPA and BioTehna labs. https://kersnikova.org/ - **LABoral Centro de Arte y Creación Industrial / LABoral 艺术与工业创造中心** (Gijón, Spain) — Spanish art centre for art, science, technology and industrial creation, set in a former labour university. Its exhibitions and production labs address automation, labour and machines. https://www.laboralcentrodearte.org/ - **LAS Art Foundation** (Berlin, Germany) — Berlin foundation that commissions large works at the edge of art, science and technology, such as Lawrence Lek's NOX (2023), a rehab centre for sentient self-driving cars. https://www.las-art.foundation/ - **MAK – Museum of Applied Arts, Vienna / 维也纳应用艺术博物馆** (Vienna, Austria) — Design museum that hosts the Vienna Biennale for Change. It co-produced Hello, Robot (2017) and staged Uncanny Values: Artificial Intelligence & You (2019), linking design and critical AI. https://www.mak.at/en - **Miraikan – National Museum of Emerging Science and Innovation / 日本科学未来馆** (Tokyo, Japan) — Tokyo science museum where the public meets humanoid and android robots, from ASIMO demonstrations to Hiroshi Ishiguro's androids Otonaroid and Kodomoroid (2014). https://www.miraikan.jst.go.jp/en/ - **Mori Art Museum / 森美术馆** (Tokyo, Japan) — Contemporary art museum in Roppongi Hills. Its survey Future and the Arts: AI, Robotics, Cities, Life (2019–20) presented over 100 projects on how humans will live with machines. https://www.mori.art.museum/en/ - **Nam June Paik Art Center / 白南准艺术中心** (Yongin, South Korea) — Museum and archive of Nam June Paik, whose Robot K-456 (1964) was an early robot performer. It shows and researches media art that continues Paik's play with machines, broadcast and the body. https://en.wikipedia.org/wiki/Nam_June_Paik_Art_Center - **NTT InterCommunication Center [ICC] / NTT 互动交流中心** (Tokyo, Japan) — Tokyo's media art centre, run by NTT East, devoted to communication technology and art. Its exhibitions and archive are a key reference for Japanese machine, sensor and interactive art. https://www.ntticc.or.jp/en/ - **Onassis Stegi / 奥纳西斯文化中心** (Athens, Greece) — Athens cultural centre of the Onassis Foundation with a strong digital culture programme. Its exhibition You and AI: Through the Algorithmic Lens (2021) examined how AI sees and sorts people. https://www.onassis.org/ - **Science Gallery / 科学画廊** (Dublin / London / Bengaluru / Melbourne) — University-based network of galleries where art meets science for young audiences. Its AI shows, from Science Gallery Dublin's early exhibitions to London's AI: Who's Looking After Me? (2023), moved from automation to bias to care. https://sciencegallery.org/ - **Science Museum, London / 伦敦科学博物馆** (London, UK) — UK national science museum. Its exhibition Robots (2017) showed more than 100 robots across 500 years, from mechanical automata to humanoids, and still anchors the popular history of the robot. https://en.wikipedia.org/wiki/Science_Museum,_London - **Serpentine Arts Technologies / 蛇形画廊艺术与技术部** (London, UK) — Serpentine's programme for art made with AI, blockchain and game engines, with commissions such as Holly Herndon & Mat Dryhurst's The Call (2024). It co-runs the Creative AI Lab and publishes the Future Art Ecosystems reports. https://www.serpentinegalleries.org/arts-technologies/ - **Society for Arts and Technology (SAT) / 蒙特利尔艺术与技术协会** (Montréal, Canada) — Montréal centre for immersive and digital art with a dome, research lab and residencies, developing telepresence and immersive tools used by artists. https://sat.qc.ca/en/ - **Taikang Art Museum / 泰康美术馆** (Beijing, China) — Beijing museum of the Taikang collection. Its exhibition Creative Machine (2024–25) traced sixty years of machine creativity, from early computer art to drawing robots such as Patrick Tresset's Human Study #1. https://tam.taikang.com/ - **Taiwan Contemporary Culture Lab (C-LAB) / 台湾当代文化实验场** (Taipei, Taiwan) — Taipei cultural lab in a former air force base, with a Technology Art Center, a sound lab and the Future Media Arts festival. A key venue for Taiwanese art on AI and machines. https://clab.org.tw/en/ - **Tate Modern / 泰特现代美术馆** (London, UK) — London's national modern art museum. Electric Dreams: Art and Technology Before the Internet (2024–25) traced the pre-internet lineage of machine art, and Anicka Yi's Hyundai Commission (2021) filled the Turbine Hall with floating machine organisms. https://www.tate.org.uk/visit/tate-modern - **UCCA Center for Contemporary Art / UCCA 尤伦斯当代艺术中心** (Beijing, China) — China's leading contemporary art institution, with venues in Beijing, Beidaihe and Shanghai. It has staged major surveys of artists working with digital bodies and automation, such as Cao Fei: Staging the Era (2021). https://ucca.org.cn/en/ - **V&A – Victoria and Albert Museum (Digital Art) / 维多利亚与阿尔伯特博物馆** (London, UK) — London design museum that holds the UK's national collection of computer and digital art, from 1960s plotter drawings to Anatomy of an AI System, which it acquired. https://www.vam.ac.uk/ - **V2_ Lab for the Unstable Media / V2_ 不稳定媒体实验室** (Rotterdam, Netherlands) — Interdisciplinary centre for art and media technology that ran the Dutch Electronic Art Festival (DEAF). Its lab, publications and online archive document decades of work on the body, sensors and machines. https://v2.nl/ - **Vancouver Art Gallery / 温哥华美术馆** (Vancouver, Canada) — Canadian art museum whose exhibition The Imitation Game: Visual Culture in the Age of Artificial Intelligence (2022) traced AI in visual culture from automata to machine vision. https://www.vanartgallery.bc.ca/ - **Vitra Design Museum / 维特拉设计博物馆** (Weil am Rhein, Germany) — Design museum whose touring exhibition Hello, Robot. Design between Human and Machine (2017) sorted robots by culture, work, home and body, and became a reference for how design meets machines. https://www.design-museum.de/en/ - **Whitney Museum of American Art (artport) / 惠特尼美国艺术博物馆** (New York, US) — New York museum whose online portal artport (since 2001) commissions net and software art. Its 2024 exhibition Harold Cohen: AARON showed the drawing program drawing live in a gallery for the first time. https://whitney.org/ - **Yamaguchi Center for Arts and Media [YCAM] / 山口情报艺术中心** (Yamaguchi, Japan) — Japanese media arts centre whose in-house R&D team, YCAM InterLab, co-develops works on dance, sensing and machines with artists, releasing tools and research as open source. https://www.ycam.jp/en/ - **ZKM | Center for Art and Media Karlsruhe / ZKM 卡尔斯鲁厄艺术与媒体中心** (Karlsruhe, Germany) — Europe's largest media art museum, with a collection spanning computer art to AI. Exhibitions such as Open Codes (2017) and BioMedia (2021) framed machine learning, algorithmic governance and life-like media systems. https://zkm.de/en ### Festivals & exhibition series - **ACM Creativity & Cognition / ACM 创造力与认知会议** (Itinerant) — ACM conference on human creativity and computation, founded by Ernest Edmonds in 1993, with papers, pictorials and an art exhibition on co-creation with machines. https://cc.acm.org/ - **ACM/IEEE International Conference on Human-Robot Interaction (HRI) / 人机交互国际会议（HRI）** (Itinerant) — The main research conference on human-robot interaction. Its studies of trust, anthropomorphism and robot bullying are core evidence for how people read machine bodies. https://humanrobotinteraction.org/ - **Ars Electronica Festival / 林茨电子艺术节** (Linz, Austria) — The world's longest-running festival for art, technology and society. Editions such as Artificial Intelligence – The Other I (2017), with its Artificial Intimacy section, set the agenda for machine art. https://ars.electronica.art/festival/en/ - **Art & Science International Exhibition and Symposium (Tsinghua) / 艺术与科学国际作品展暨学术研讨会** (Beijing, China) — Recurring exhibition and symposium initiated in 2001 at Tsinghua University by physicist T.D. Lee and painter Wu Guanzhong, now held at the Tsinghua University Art Museum. China's main art-and-science series. https://www.artmuseum.tsinghua.edu.cn/ - **Beijing Media Art Biennale / 北京媒体艺术双年展** (Beijing, China) — Biennale at the CAFA Art Museum, Central Academy of Fine Arts. Its first edition Ethics of Technology (2016) and Post-Life (2018) are key Chinese-language frames for art about machines. https://en.wikipedia.org/wiki/CAFA_Art_Museum - **CVPR AI Art Gallery / CVPR AI 艺术画廊** (Itinerant) — Online and on-site gallery of the CVPR computer vision conference, showing artworks that use or question computer vision; its archive documents each year's selection. https://thecvf-art.com/ - **Elektra / International Digital Art Biennial (BIAN)** (Montréal, Canada) — Montréal festival and biennial for digital art, which often features robotic and machine installations; BIAN editions include the robotics-themed Automata (2016). https://elektramontreal.ca/ - **EVA London – Electronic Visualisation and the Arts** (London, UK) — Annual London conference on electronic visualisation and the arts, run with the BCS Computer Arts Society, spanning early computer art history to AI. https://www.eva-london.org/ - **FILE – Electronic Language International Festival** (São Paulo, Brazil) — Latin America's largest electronic art festival, held in São Paulo, with robotic, interactive and AI art and an extensive online archive. https://file.org.br/ - **Impakt** (Utrecht, Netherlands) — Utrecht organisation for critical media culture, with a yearly festival, residencies and a centre for media culture. Its themes often address algorithms, surveillance and machine perception. https://impakt.nl/ - **International Conference on Computational Creativity (ICCC) / 国际计算创造力会议** (Itinerant) — Annual conference of the Association for Computational Creativity on whether and how machines can be creative, from art-making systems to evaluation. https://computationalcreativity.net/ - **ISEA International (International Symposium on Electronic Art) / 国际电子艺术研讨会** (Itinerant) — The main academic symposium and exhibition for electronic art, held in a different city each year since 1988. Its online archive holds decades of papers on robot art, A-life and AI. https://isea-international.org/ - **Japan Media Arts Festival / 日本文化厅媒体艺术节** (Tokyo, Japan) — Festival and awards of Japan's Agency for Cultural Affairs (1997–2022) across art, entertainment, animation and manga. Sony's aibo and the robotic opera Pendulum Choir were among its prize-winners. https://j-mediaarts.jp/en/ - **KIKK Festival** (Namur, Belgium) — International festival of digital and creative cultures in Namur, with talks and installations by designers and artists working on AI, robots and speculative futures. https://www.kikk.be/ - **Microwave International New Media Arts Festival / 微波国际新媒体艺术节** (Hong Kong) — Hong Kong's long-running new media arts festival, with themed exhibitions and forums that have included machine and AI art from Asia. https://www.microwavefest.net/ - **NeurIPS Creative AI Track / NeurIPS 创意 AI 单元** (Itinerant) — Juried track of the NeurIPS machine learning conference for art and design work with and about AI, following the Machine Learning for Creativity and Design workshops (2017–22). It puts artists in the room with ML researchers. https://neurips.cc/ - **NIME – New Interfaces for Musical Expression** (Itinerant) — International conference on new musical instruments and interfaces, including robotic instruments, AI co-performers and machine listening. https://www.nime.org/ - **SIGGRAPH Art Gallery & Art Papers / SIGGRAPH 艺术画廊** (Itinerant (North America / Asia)) — The art programme of ACM SIGGRAPH, the computer graphics conference, and SIGGRAPH Asia. For decades it has shown computational, robotic and AI art beside the research that makes it possible. https://en.wikipedia.org/wiki/SIGGRAPH - **STRP Festival** (Eindhoven, Netherlands) — Eindhoven festival for creative technology, set in a former Philips industrial complex, whose editions explore living with robots, AI and sensing environments. https://strp.nl/ - **Sónar+D** (Barcelona, Spain) — Creativity and technology congress of the Sónar music festival, with talks and showcases on AI in music, synthetic voices and human-machine performance. https://sonar.es/en - **transmediale / 柏林跨媒体艺术节** (Berlin, Germany) — Berlin festival for art and digital culture, known for critical rather than celebratory themes (Capture All 2015, face value 2018, for refusal 2021–22). It co-publishes the APRJA journal. https://transmediale.de/ - **WRO Media Art Biennale / WRO 媒体艺术双年展** (Wrocław, Poland) — Central Europe's longest-running media art biennale, run by the WRO Art Center in Wrocław, with exhibitions on sound, machines and networked media. https://wrocenter.pl/en/ - **xCoAx – Computation, Communication, Aesthetics & X** (Itinerant (Europe)) — Conference and exhibition on computation, communication and aesthetics, with papers and artworks on algorithms, generative systems and the nature of computing. https://xcoax.org/ ### Awards & prizes - **Human AI Art Award / 人类 AI 艺术奖** (Bonn, Germany) — Annual award by the Art Collection Deutsche Telekom and Kunstmuseum Bonn for art that reflects on AI: Lauren Lee McCarthy's LAUREN (2024) and Nicolas Gourault's Unknown Label (2025). https://www.art-collection-telekom.com/en/ - **LG Guggenheim Award / LG 古根海姆奖** (New York, US) — US$100,000 award for technology-based art from the LG Guggenheim Art & Technology Initiative. Winners: Stephanie Dinkins (2023), Shu Lea Cheang (2024), Ayoung Kim (2025) and Trevor Paglen (2026). https://en.wikipedia.org/wiki/LG_Guggenheim_Award - **Lumen Prize / Lumen 数字艺术奖** (London, UK) — International prize for art made with technology, with an AI award since 2018 (e.g. Lichtsuchende, a light-seeking robot society). Much of it is generative, so its winners need case-by-case selection. https://www.lumenprize.com/ - **Mozilla Creative Media Awards / Mozilla 创意媒体奖** (Global) — Mozilla Foundation grants for art, games and media that make the harms of AI tangible to the public, especially bias, surveillance and platform power. https://foundation.mozilla.org/en/what-we-fund/awards/creative-media-awards/ - **Prix Ars Electronica / 电子艺术大奖** (Linz, Austria) — The leading prize for media art. Its Artificial Life & Intelligence category honours work on what AI and life science mean for being human, with Golden Nicas for Paul Vanouse's Labor (2019) and Forensic Architecture's Cloud Studies (2021). https://ars.electronica.art/prix/en/ - **RobotArt Competition / 机器人艺术竞赛** (US / Online) — Annual competition (2016–18) for robots that paint with physical brushes, which gathered university and artist teams and made robot painting visible as a craft. https://robotart.org/ - **S+T+ARTS Prize / S+T+ARTS 奖** (Linz, Austria / EU) — European Commission prize for innovation at the nexus of science, technology and the arts, run by Ars Electronica. Winners include Project Alias (2019), Holly+ (2022), Calculating Empires (2024) and the AI War Cloud Database (2025). https://ars.electronica.art/starts-prize/en/ - **VIDA – Art & Artificial Life International Awards / VIDA 艺术与人工生命国际奖** (Madrid, Spain) — Fundación Telefónica's prize (1999–c.2015), the only award devoted to art and artificial life: robots that sweat, walking tables, autonomous protest agents and virtual ecologies. https://www.fundaciontelefonica.com/ ### Labs & research centres - **Abundant Intelligences** (Montréal, Canada / Global) — Indigenous-led research programme based at Concordia University that conceptualises, designs and builds AI from Indigenous knowledge systems. https://abundant-intelligences.net/ - **Ada Lovelace Institute / 艾达·洛芙莱斯研究所** (London, UK) — Independent research institute of the Nuffield Foundation ensuring data and AI work for people and society, with public-attitude studies on biometrics and facial recognition. https://www.adalovelaceinstitute.org/ - **AI Now Institute / AI Now 研究所** (New York, US) — Policy research institute co-founded by Kate Crawford and Meredith Whittaker on the social consequences of AI. Its reports on bias, labour and concentrated industry power shaped critical AI discourse. https://ainowinstitute.org/ - **Antikythera** (Los Angeles, US / Global) — Think tank for the speculative philosophy of computation, directed by Benjamin Bratton and incubated at the Berggruen Institute. It studies planetary computation and synthetic intelligence through studios, publications and film. https://antikythera.org/ - **ANU School of Cybernetics / 澳大利亚国立大学控制论学院** (Canberra, Australia) — School founded by anthropologist Genevieve Bell, from the 3A Institute, to revive cybernetics as a way of building, steering and critiquing AI-enabled systems. https://cybernetics.anu.edu.au/ - **Ars Electronica Futurelab / 电子艺术未来实验室** (Linz, Austria) — The R&D studio of Ars Electronica, working with industry and universities on drone swarms, robots and human-AI collaboration, and presenting the results as art. https://ars.electronica.art/futurelab/en/ - **Artists + Machine Intelligence (Google AMI) / 谷歌艺术家与机器智能计划** (Mountain View / Global) — Google programme that pairs artists with machine intelligence researchers; it supported early DeepDream-era art and projects such as K Allado-McDowell's Pharmako-AI. https://ami.withgoogle.com/ - **Creative AI Lab / 创意 AI 实验室** (London, UK) — Collaboration between Serpentine Arts Technologies and King's College London's Department of Digital Humanities. It studies AI as an artistic medium and keeps a database of tools and resources for artists. https://creative-ai.org/ - **Creative Machines Lab (Columbia University) / 哥伦比亚大学创意机器实验室** (New York, US) — Hod Lipson's robotics lab, known for self-modelling robots that learn their own body image, robots that paint and machines that design machines. https://www.creativemachineslab.com/ - **Data & Society / 数据与社会研究所** (New York, US) — Independent research institute founded by danah boyd that studies the social implications of data-centric and automated technologies, from platform labour to algorithmic accountability. https://datasociety.net/ - **Distributed AI Research Institute (DAIR) / 分布式 AI 研究所** (Distributed) — Independent, community-rooted AI research institute founded by Timnit Gebru after her dismissal from Google. It centres the people harmed by AI, including data workers. https://www.dair-institute.org/ - **Forensic Architecture / 法证建筑** (London, UK) — Research agency at Goldsmiths that investigates state and corporate violence with spatial, media and machine-vision analysis. It trains classifiers on synthetic data and won the 2021 Golden Nica for Cloud Studies. https://forensic-architecture.org/ - **Gramazio Kohler Research (ETH Zurich)** (Zurich, Switzerland) — ETH Zurich chair for architecture and digital fabrication, which put industrial robots to work building walls and structures and made robotic construction a design subject. https://gramaziokohler.arch.ethz.ch/ - **Hiroshi Ishiguro Laboratories (ATR / Osaka University) / 石黑浩实验室** (Kyoto / Osaka, Japan) — Hiroshi Ishiguro's android labs, which built the Geminoid copies, Erica and, with Kōdai-ji temple, the android Kannon Mindar. They are the main testbed for questions of presence and the uncanny valley. http://www.geminoid.jp/en/ - **Humanoid Robotics Institute (Waseda University) / 早稻田大学人形机器人研究所** (Tokyo, Japan) — Successor to Ichiro Kato's lab, which built WABOT-1 (1973), the first full-scale humanoid robot, and later emotional and musical humanoids such as the flute-playing WF series. https://www.humanoid.waseda.ac.jp/ - **Ikegami Lab (University of Tokyo) / 池上高志实验室** (Tokyo, Japan) — Takashi Ikegami's artificial life lab, which built the android Alter with Ishiguro's team to test whether spontaneous, life-like movement can emerge from neural networks. https://en.wikipedia.org/wiki/Takashi_Ikegami - **Institute of Network Cultures / 网络文化研究所** (Amsterdam, Netherlands) — Research centre at the Amsterdam University of Applied Sciences founded by Geert Lovink, publishing open-access books and readers on platforms, internet critique and AI. https://networkcultures.org/ - **Intelligent Instruments Lab / 智能乐器实验室** (Reykjavík, Iceland) — ERC-funded lab at the University of Iceland studying musical instruments that embed AI as creative partners and what such instruments mean for musicianship. Distinction in the Prix Ars Electronica AI in ART Award. https://iil.is/ - **IRCAM / 法国声学与音乐研究协调所** (Paris, France) — Paris institute for music and acoustic research founded by Pierre Boulez at the Centre Pompidou. Its research on machine listening, voice synthesis and AI co-improvisation (OMax, Somax) shapes human-machine music. https://www.ircam.fr/ - **KIM – AI and Media Philosophy (HfG Karlsruhe) / 卡尔斯鲁厄设计学院 AI 与媒体哲学研究组** (Karlsruhe, Germany) — Research group on artificial intelligence and media philosophy at the Karlsruhe University of Arts and Design, known for its work on the social history of AI and the Nooscope diagram. https://kim.hfg-karlsruhe.de/ - **Knowing Machines** (Global) — Research project tracing the histories, practices and politics of how machine learning systems are trained to interpret the world, including Models All the Way Down, a visual investigation of the LAION-5B dataset. https://knowingmachines.org/ - **LACMA Art + Technology Lab / 洛杉矶郡艺术博物馆艺术与技术实验室** (Los Angeles, US) — Grant and residency programme that revives LACMA's 1967–71 Art and Technology programme, pairing artists with tech companies on projects about AI, robotics and sensing. https://www.lacma.org/lab - **Leverhulme Centre for the Future of Intelligence / 剑桥大学利弗休姆未来智能中心** (Cambridge, UK) — Cambridge research centre on the long-term impact of AI. Its AI Narratives and Global AI Narratives projects study how stories shape what we expect of machines, and it has researched griefbots and 'digital afterlife' ethics. https://www.lcfi.ac.uk/ - **metaLAB (at) Harvard / 哈佛 metaLAB** (Cambridge, MA, US) — Creative research group at Harvard's Berkman Klein Center, whose projects include the AI Pedagogy Project and artworks on machine classification and moral decisions. https://metalab.harvard.edu/ - **MIT Center for Advanced Visual Studies (CAVS) / MIT 高级视觉研究中心** (Cambridge, MA, US) — Fellowship centre founded by György Kepes at MIT (1967–2009) where artists worked with engineers on light, kinetic and cybernetic art; a model for later art-and-technology labs. https://en.wikipedia.org/wiki/Center_for_Advanced_Visual_Studies - **MIT Media Lab – Affective Computing Group / MIT 媒体实验室情感计算组** (Cambridge, MA, US) — Rosalind Picard's group, which founded affective computing: machines that sense and respond to emotion. Its work underlies both care technologies and the emotion recognition that artists critique. https://www.media.mit.edu/groups/affective-computing/overview/ - **MIT Media Lab – Personal Robots Group / MIT 媒体实验室个人机器人组** (Cambridge, MA, US) — Cynthia Breazeal's group, which grew out of her sociable robot Kismet (1990s) and went on to build Leonardo and the social robot Jibo. It is the reference lab for robots designed as companions, tutors and carers. https://www.media.mit.edu/groups/personal-robots/overview/ - **Mixed Reality Laboratory (University of Nottingham) / 诺丁汉大学混合现实实验室** (Nottingham, UK) — HCI lab with a long collaboration with Blast Theory. Its Cat Royale (2023) let a robot arm play with three cats for twelve days to test the ethics of robots caring for animals. https://www.nottingham.ac.uk/research/groups/mixedrealitylab/ - **Ory Laboratory / Ory 研究所** (Tokyo, Japan) — Kentaro Yoshifuji's company behind the avatar robot OriHime and the Avatar Robot Café DAWN, where people with ALS and other disabilities work as waiters through robots. https://orylab.com/en/ - **Reality Design Lab / 现实设计实验室** — Design research lab led by Botao Amber Hu that makes mixed-reality and speculative works on other minds and machines, such as EchoVision (what is it like to be a bat?), City of Sparkles (how might an AI perceive our city?) and AI Mom™. It also builds the open HoloKit toolkit and runs the Machinic Reality Inspire gallery. https://reality.design/ - **Robotic Musicianship Group (Georgia Tech) / 佐治亚理工机器人音乐组** (Atlanta, US) — Gil Weinberg's group at the Georgia Tech Center for Music Technology, which built the improvising marimba robot Shimon, a machine that listens and plays with humans. https://en.wikipedia.org/wiki/Gil_Weinberg - **SHARE Lab / SHARE Foundation** (Novi Sad, Serbia) — Research and data investigation lab of the SHARE Foundation, led by Vladan Joler, which maps the invisible infrastructures of networks and AI. It is the home of the maps behind Anatomy of an AI System. https://en.wikipedia.org/wiki/SHARE_Foundation - **STUDIO for Creative Inquiry (CMU) / 卡内基梅隆大学创意探究工作室** (Pittsburgh, US) — Carnegie Mellon's lab for atypical, anti-disciplinary art research, directed by Golan Levin. It hosts residencies and projects on computer vision, robotics and creative coding. https://studioforcreativeinquiry.org/ - **UAL Creative Computing Institute / 伦敦艺术大学创意计算研究所** (London, UK) — University of the Arts London institute for creative computing research and teaching, including interactive machine learning for artists (Wekinator) and critical AI practice. https://www.arts.ac.uk/creative-computing-institute - **UCLA Art|Sci Center / UCLA 艺术|科学中心** (Los Angeles, US) — Victoria Vesna's centre at UCLA for collaborations between artists and scientists, a hub of the art-science network behind many machine and bio-art projects. https://artsci.ucla.edu/ - **Waag Futurelab / Waag 未来实验室** (Amsterdam, Netherlands) — Amsterdam research institute for technology and society that runs public labs, including critical work on AI, biometrics and the 'public stack'. https://waag.org/en/ ### Residencies & programmes - **Akademie Schloss Solitude (Web Residencies) / 孤独城堡学院** (Stuttgart, Germany) — Stuttgart residency academy whose online Web Residencies, published on Schlosspost, fund artists responding to curated calls, many on AI, bots and algorithmic culture. https://www.akademie-solitude.de/en/ - **China Academy of Art – School of Inter-Media Art / 中国美术学院跨媒体艺术学院** (Hangzhou, China) — China's leading school for new media, installation and network art at the China Academy of Art, a training ground for Chinese artists working with machines and AI. https://en.wikipedia.org/wiki/China_Academy_of_Art - **Eyebeam** (New York, US) — Nonprofit that supports artists working with technology through residencies and fellowships, with a strong line of alumni projects on surveillance, bias and algorithmic justice. https://www.eyebeam.org/ - **Goldsmiths Computational Arts / 伦敦大学金史密斯学院计算艺术** (London, UK) — Goldsmiths' computational art programmes and embodied interaction research (EAVI), which trained many artists who work with machine learning, bodies and sound. https://www.gold.ac.uk/pg/ma-computational-arts/ - **Gray Area** (San Francisco, US) — San Francisco nonprofit arts centre for creative technology, with an incubator, education programmes and the Gray Area Festival on art, AI and society. https://grayarea.org/ - **IAMAS – Institute of Advanced Media Arts and Sciences / 情报科学艺术大学院大学** (Ogaki, Japan) — Japanese graduate school for media arts and sciences in Gifu, whose alumni include many Japanese device, sensor and machine artists. https://www.iamas.ac.jp/en/ - **NEW INC** (New York, US) — The New Museum's cultural incubator for art, design and technology, whose members have developed many projects on AI and its social effects. https://www.newmuseum.org/new-inc/ - **NYU ITP / IMA / 纽约大学互动电信项目** (New York, US) — NYU Tisch's Interactive Telecommunications Program and Interactive Media Arts, a hub for physical computing, creative machine learning (ml5.js) and artists building with technology. https://tisch.nyu.edu/itp - **S+T+ARTS Residencies / S+T+ARTS 驻留计划** (EU) — European Commission initiative that funds artist residencies inside technology and research projects, many on AI, robotics and sustainability. https://starts.eu/ - **SAIC Art and Technology Studies / 芝加哥艺术学院艺术与技术系** (Chicago, US) — School of the Art Institute of Chicago department, rooted in Sonia Sheridan's Generative Systems programme (1970), that teaches robotics, kinetics and electronics as art. https://www.saic.edu/atsp - **School for Poetic Computation (SFPC) / 诗意计算学校** (New York, US) — Artist-run school in New York for art, code, hardware and critical theory, with classes on the poetics and politics of machines. https://sfpc.study/ - **School of Creative Media, City University of Hong Kong / 香港城市大学创意媒体学院** (Hong Kong) — Hong Kong media art school that hosts the Art Machines symposium on computational media art, and a centre for robotic and AI art in Asia. https://www.scm.cityu.edu.hk/ - **Somerset House Studios / 萨默塞特宫工作室** (London, UK) — Experimental workspace for artists and makers inside Somerset House, home to studios working on AI, sound and speculative design, and producer of exhibitions such as 24/7 (2019). https://www.somersethouse.org.uk/somerset-house-studios - **UCLA Design Media Arts / UCLA 设计媒体艺术系** (Los Angeles, US) — UCLA department for media art and design whose faculty include Casey Reas (Processing) and Lauren Lee McCarthy (LAUREN). https://dma.ucla.edu/ ### Archives & databases - **ACM SIGGRAPH History Archive (Digital Art Archive) / ACM SIGGRAPH 数字艺术档案** (Online) — Online archive of SIGGRAPH art galleries, artworks and art papers since the 1980s, a record of computational and robotic art beside graphics research. https://history.siggraph.org/ - **AI Art Gallery (aiartonline.com) / AI 艺术在线画廊** (Online) — Online gallery of the NeurIPS Workshop on Machine Learning for Creativity and Design, curated by Luba Elliott, archiving the art selected by the ML community 2017–2021. https://www.aiartonline.com/ - **AI Incident Database / AI 事故数据库** (Online) — Open archive of real-world harms and near-harms caused by AI systems, started at the Partnership on AI, which catalogues machine failures from misidentification to autonomous vehicle crashes. https://incidentdatabase.ai/ - **AI Myths / AI 迷思** (Online) — Resource by Daniel Leufer debunking common misconceptions about AI, such as 'AI has agency' or 'AI can be objective', with references to research. https://www.aimyths.org/ - **AIAAIC Repository / AIAAIC 事件库** (Online) — Independent, public-interest repository of AI, algorithmic and automation incidents and controversies, classified by sector, technology and harm. https://www.aiaaic.org/ - **AIArtists.org** (Online) — Directory of artists working with AI, organised by critical questions such as bias, dataset ethics and robotic production. It leans generative, so it is best used as a lead list. https://aiartists.org/ - **Anatomy of an AI System / AI 系统解剖图** (Online) — Kate Crawford and Vladan Joler's map and essay of the human labour, data and planetary resources behind one Amazon Echo request; in the MoMA and V&A collections. https://anatomyof.ai/ - **Archive of Digital Art (ADA) / 数字艺术档案** (Krems, Austria) — Scholarly database of digital art founded by Oliver Grau, documenting hundreds of artists and thousands of works and texts, with strong coverage of robotic and A-life art. https://www.digitalartarchive.at/ - **Ars Electronica Archive / 电子艺术节档案** (Linz, Austria) — Archive of Prix Ars Electronica entries since 1987, festival catalogues and documentation: the largest record of how media art engaged robots, A-life and AI. https://ars.electronica.art/archive/en/ - **Atlas of Surveillance / 监控地图集** (US / Online) — EFF and University of Nevada, Reno database of surveillance technologies used by US police, from face recognition to drones and automated licence plate readers. https://atlasofsurveillance.org/ - **Better Images of AI / 更好的 AI 图像** (Online) — Nonprofit collaboration offering a free library of images that represent AI more accurately than glowing brains and humanoid robots, commissioned from artists. https://betterimagesofai.org/ - **Calculating Empires / 计算帝国** (Online) — Large-scale research visualisation by Crawford and Joler, A Genealogy of Technology and Power 1500–2025, tracing how communication and computation systems grew with control; S+T+ARTS Grand Prize 2024. https://calculatingempires.net/ - **Critical Algorithm Studies: a Reading List / 批判算法研究书单** (Online) — Reading list by Tarleton Gillespie and Nick Seaver of the Microsoft Research Social Media Collective on algorithms as social concern, a standard entry point for critical algorithm studies. https://socialmediacollective.org/reading-lists/critical-algorithm-studies/ - **Cybernetic Serendipity archive / 控制论的意外发现档案** (Online) — Online archive of Jasia Reichardt's 1968 ICA exhibition Cybernetic Serendipity, the founding show of computer and cybernetic art, with Gordon Pask's Colloquy of Mobiles and Edward Ihnatowicz's SAM. https://www.cyberneticserendipity.net/ - **cyberneticzoo.com** (Online) — Illustrated history of cybernetic animals and early robots, from Grey Walter's tortoises to forgotten walking machines, compiled from period sources. https://cyberneticzoo.com/ - **Daniel Langlois Foundation for Art, Science and Technology / 丹尼尔·朗格卢瓦艺术、科学与技术基金会** (Montréal, Canada) — Foundation whose online archive documents Experiments in Art and Technology (E.A.T.) and early media art, including essays on artists who worked with machines and cybernetics. https://www.fondation-langlois.org/ - **Data Workers' Inquiry / 数据劳工调查** (Berlin / Global) — Community research project of DAIR and the Weizenbaum Institute in which data workers who label and moderate for AI document their own conditions in reports, films and zines. https://data-workers.org/ - **Database of Digital Art (compArt daDA) / 数字艺术数据库** (Bremen, Germany) — Database from the compArt centre at the University of Bremen on the pioneers of computer art and their exhibitions, including Cybernetic Serendipity (1968). http://dada.compart-bremen.de/ - **Excavating AI / 发掘 AI** (Online) — Essay by Kate Crawford and Trevor Paglen on the politics of images in machine learning training sets, which examined ImageNet's 'person' categories and fed into ImageNet Roulette. https://excavating.ai/ - **Exposing.ai** (Berlin, Germany) — Adam Harvey and Jules LaPlace's research project (successor to MegaPixels) that traces how face and biometric datasets were built from scraped photos and lets people search whether theirs were used. https://exposing.ai/ - **Machine Vision in Art, Games and Narratives (database) / 艺术、游戏与叙事中的机器视觉数据库** (Bergen, Norway) — ERC project at the University of Bergen whose open database codes about 500 artworks, games and stories as 'situations' of who does what to whom with which machine-vision technology. https://machinevisionuib.github.io/ - **Media Art Net / 媒体艺术网** (Karlsruhe, Germany) — Bilingual (German/English) online platform by ZKM and the Goethe-Institut with works, artists and essays on media art history, including robotics and cybernetic art. http://www.medienkunstnetz.de/ - **Media Arts Database (Japan) / 日本媒体艺术数据库** (Tokyo, Japan) — National database of Japanese media arts, including records of media art exhibitions, festivals and works, useful for tracing Japanese machine and device art. https://mediaarts-db.artmuseums.go.jp/ - **Monoskop** (Online) — Wiki and digital library for art, media and the humanities founded by Dušan Barok, with rich pages and scans on cybernetic art, media art festivals and machine theory. https://monoskop.org/ - **ROBOTS Guide (IEEE Spectrum) / IEEE 机器人指南** (Online) — IEEE's illustrated catalogue of real robots, with photos, videos and specifications for hundreds of humanoids, companion robots and industrial arms. https://robotsguide.com/ - **The Glass Room / 玻璃房间** (Berlin, Germany / Online) — Tactical Tech's exhibition about data and surveillance, styled as a tech store, with more than 40 exhibits under themes such as We Know You and Big Mother; now toured in community and online editions. https://theglassroom.org/ ### Journals, books & podcasts - **AI & Society (Springer) / 《AI 与社会》期刊** (Global) — Journal of knowledge, culture and communication on the societal implications of AI, including art, labour and ethics. https://link.springer.com/journal/146 - **AI Weirdness** (Online) — Janelle Shane's blog and newsletter on the strange ways AI fails, from neural-net paint colours to hallucinated giraffes: a long-running record of machine error as a window into machine 'minds'. https://www.aiweirdness.com/ - **APRJA – A Peer-Reviewed Journal About** (Aarhus, Denmark / Berlin) — Open-access journal produced with transmediale and Aarhus University, each issue built from a research workshop on topics such as machine research and machine feeling. https://aprja.net/ - **Blood in the Machine / 机器中的血** (Online) — Brian Merchant's newsletter and book on the Luddites and today's fight against automation, reporting how AI reshapes work. https://www.bloodinthemachine.com/ - **Computational Culture / 《计算文化》期刊** (Online) — Open-access journal of software studies, publishing research on algorithms, machine learning and the cultural life of code. https://computationalculture.net/ - **Critical AI (Duke University Press) / 《批判性 AI》期刊** (US) — Interdisciplinary journal and Rutgers-based initiative that brings humanities methods to AI, from large language models to data labour. https://criticalai.org/ - **Cybernetic Forests / 控制论森林** (Online) — Eryk Salvaggio's newsletter on AI, art and cybernetics, reading generated images as infographics of their datasets and arguing against AI hype. https://www.cyberneticforests.com/ - **e-flux journal / e-flux 期刊** (New York, US) — Online art journal that published key texts on machine vision and AI, such as Hito Steyerl's 'A Sea of Data: Apophenia and Pattern (Mis-)Recognition' (2016). https://www.e-flux.com/journal/ - **Leonardo (MIT Press) / Leonardo ISAST / 《莱昂纳多》期刊** (Cambridge, MA, US / San Francisco, US) — Journal of arts, sciences and technology founded by kinetic artist Frank Malina, where artists such as Eduardo Kac and Simon Penny published foundational texts on robotic art. Leonardo also runs a book series and the LASER talks. https://leonardo.info/ - **Logic(s)** (US) — Magazine about technology and its politics, relaunched as Logic(s) with a focus on Black, Asian and Indigenous perspectives; issues on intelligence, bodies and supply chains. https://logicmag.io/ - **meson press** (Lüneburg, Germany) — Open-access publisher of media theory, with titles on machine learning, pattern discrimination and the history of computing. https://meson.press/ - **Mute** (London, UK) — London magazine on culture and politics after the net, with long-running critical writing on automation, labour and digital capitalism. https://www.metamute.org/ - **Mystery AI Hype Theater 3000** (Online) — DAIR podcast and stream in which linguist Emily M. Bender and sociologist Alex Hanna take apart claims of machine intelligence and AI hype. https://www.dair-institute.org/maiht3k/ - **nettime** (Online) — Mailing list for net criticism founded in 1995, where the net.art and tactical media generations debated networks, software and machines. https://www.nettime.org/ - **Neural** (Bari, Italy) — Bilingual (Italian/English) magazine on new media art, electronic music and hacktivism, founded by Alessandro Ludovico and long a chronicle of critical machine art. https://neural.it/ - **Noema Magazine** (Los Angeles, US) — Magazine of the Berggruen Institute that publishes long essays on AI, minds and planetary computation by writers such as Benjamin Bratton, Blaise Agüera y Arcas and Tobias Rees. https://www.noemamag.com/ - **Open Humanities Press / 开放人文出版社** (Online) — Open-access academic publisher whose books include Joanna Zylinska's AI Art: Machine Visions and Warped Dreams (2020). https://www.openhumanitiespress.org/ - **Philosophy & Technology (Springer) / 《哲学与技术》期刊** (Global) — Philosophy journal on technology, including postphenomenology, machine agency and AI ethics. https://link.springer.com/journal/13347 - **Posthumanities (University of Minnesota Press) / 《后人类》丛书** (Minneapolis, US) — Book series edited by Cary Wolfe, including Ian Bogost's Alien Phenomenology (2012), a core shelf for thinking about nonhuman and machine experience. https://www.upress.umn.edu/search-grid/?series=posthumanities - **Real Life** (Online) — Online magazine (2016–2022) about living with technology, whose essays on smart speakers, algorithms and machine intimacy remain an archive of everyday machine phenomenology. https://reallifemag.com/ - **Tech Won't Save Us (podcast)** (Online) — Weekly podcast by Paris Marx with critical perspectives on tech, AI hype and the labour and infrastructure behind it. https://techwontsave.us/ - **Technophany** (Online) — Journal of the Research Network for Philosophy and Technology (Yuk Hui and others), publishing philosophy of technology, cosmotechnics and machine thought. https://technophany.philosophyandtechnology.network/ - **The Good Robot (podcast) / 好机器人播客** (Cambridge, UK) — Podcast by Eleanor Drage and Kerry McInerney asking feminist thinkers and technologists what good technology is. https://www.thegoodrobot.co.uk/ ### Networks & nonprofits - **AIxDesign** (Amsterdam, Netherlands / Global) — Community and studio of designers, artists and researchers working critically with AI, running programmes, publications and a slow-AI research line. https://aixdesign.co/ - **Algorithmic Justice League / 算法正义联盟** (Cambridge, MA, US) — Organisation founded by Joy Buolamwini that combines art and research to expose bias in AI, starting from the 'coded gaze' of face recognition shown in Gender Shades. https://www.ajl.org/ - **AlgorithmWatch / 算法观察** (Berlin, Germany) — Nonprofit research and advocacy organisation that investigates automated decision-making systems and their impact on society in Europe. https://algorithmwatch.org/en/ - **Black in AI** (Global) — Network co-founded by Timnit Gebru and Rediet Abebe to increase the presence and inclusion of Black people in AI. https://blackinai.github.io/ - **Coding Rights** (Rio de Janeiro, Brazil) — Brazilian feminist organisation on technology and human rights, whose Not My A.I. project maps AI deployed against marginalised people in Latin America. https://codingrights.org/ - **Computer Arts Society (CAS) / 计算机艺术学会** (London, UK) — British society founded in 1968, the year of Cybernetic Serendipity, to support computer use in the arts; publisher of the bulletin PAGE and a specialist group of the BCS. https://computer-arts-society.com/ - **Constant** (Brussels, Belgium) — Brussels artist-run association for art, media and technology, home to Algolit, which explores machine learning through literature and code. https://constantvzw.org/ - **Critical Engineering Working Group / 批判工程工作组** (Berlin, Germany) — Group behind The Critical Engineering Manifesto (2011), which asks engineers and artists to expose the inner workings of the technologies we depend on. https://criticalengineering.org/ - **Data for Black Lives** (US) — Movement of activists and scientists using data to improve Black lives and fighting algorithmic discrimination. https://d4bl.org/ - **Experiments in Art and Technology (E.A.T.) / 艺术与技术实验** (New York, US) — Nonprofit founded by Bell Labs engineers and artists after 9 Evenings (1966), which paired artists with engineers and set the model for art-technology collaboration. https://en.wikipedia.org/wiki/Experiments_in_Art_and_Technology - **Feminist.AI / 女性主义 AI** (Los Angeles, US) — Community research group that co-designs AI prototypes and datasets with communities, making feminist and critical AI by hand. https://www.feminist.ai/ - **Indigenous Protocol and AI Working Group / 原住民协议与 AI 工作组** (Global) — Working group whose 2020 position paper asked how Indigenous epistemologies can guide the design of AI and our relations with non-human kin, including machines. https://www.indigenous-ai.net/ - **Machine Listening / 机器聆听** (Melbourne, Australia) — Platform for research and artistic practice on the politics of machine listening, voice assistants and audio surveillance, with an open curriculum. https://machinelistening.exposed/ - **Masakhane** (Africa / Global) — Grassroots research community building natural language processing for African languages, by Africans, against the English-centred machine ear. https://www.masakhane.io/ - **Processing Foundation / Processing 基金会** (US / Global) — Nonprofit behind Processing and p5.js that promotes software literacy in the arts, with fellowships supporting artists working critically with code and machine learning. https://processingfoundation.org/ - **Queer in AI** (Global) — Grassroots organisation of queer researchers and practitioners in AI that raises awareness of how AI misclassifies and harms queer people. https://www.queerinai.com/ - **Rhizome** (New York, US) — Born-digital art organisation affiliated with the New Museum, which commissions, preserves and writes about art engaged with technology, including the ArtBase and the Webrecorder tools. https://rhizome.org/ - **Tactical Tech** (Berlin, Germany) — International NGO that helps the public understand how technology shapes society, through exhibitions such as The Glass Room and toolkits on data and influence. https://tacticaltech.org/ - **Tierra Común** (Latin America / Global) — Network of activists and scholars on data colonialism, bringing Global South perspectives to the extraction behind data and AI. https://www.tierracomun.net/ - **Turkopticon** (US / Online) — Tool and organising project by and for Amazon Mechanical Turk workers, started by Lilly Irani and Six Silberman, which made the hidden human labour of AI visible and contestable. https://turkopticon.net/ ## Creators - **Lauren Lee McCarthy** (11) — Artist; Professor, UCLA Design Media Arts; creator of p5.js. Artist and programmer whose performances put her in the place of smart-home assistants, apps and algorithms, examining surveillance, automation and care in everyday life. https://lauren-mccarthy.com/ - **Trevor Paglen** (10) — Artist and geographer. American artist and geographer whose photographs, installations and performances investigate surveillance, secrecy and computer vision. https://paglen.studio - **Hito Steyerl** (9) — Artist, filmmaker and writer. German artist whose essay films and installations examine images, circulation, war and computation. https://en.wikipedia.org/wiki/Hito_Steyerl - **Hiroshi Ishiguro** (8) — Roboticist; Professor, Osaka University; ATR Hiroshi Ishiguro Laboratories. Builds lifelike androids, including Geminoid HI-1, a copy of himself, and uses them to study human presence. http://www.geminoid.jp/en/ - **Kyle McDonald** (8) — Artist and creative technologist. Artist working with code, computer vision and machine learning; co-creator of openFrameworks tools and of several works with Lauren Lee McCarthy. https://kylemcdonald.net - **Stelarc** (7) — Performance artist; Distinguished Research Fellow, Curtin University (Alternate Anatomies Lab). Cyprus-born Australian artist who since the 1970s has extended, suspended, wired and augmented his own body with robots, prostheses and networks. https://stelarc.org/ - **Botao Amber Hu** (6) — Researcher, University of Oxford; founder of Reality Design Lab. Designer and researcher (Tsinghua BS, Stanford MS in AI) working on mixed reality, speculative design and blockchain-based artificial life; Reality Design Lab's works have been shown at SIGGRAPH, CHI, ALIFE and ISEA. https://reality.design - **Kate Crawford** (6) — Researcher and author; Research Professor, USC Annenberg. Australian scholar of the social and political implications of AI, co-founder of the AI Now Institute and author of Atlas of AI (2021). https://katecrawford.net/ - **Adam Harvey** (5) — Artist and researcher (computer vision, privacy, surveillance). American artist and researcher who has worked on computer vision, biometrics and counter-surveillance since his 2010 NYU ITP thesis CV Dazzle. His later research projects trace how face datasets are built and apply object detection to human-rights archives. https://adam.harvey.studio - **Bill Vorn** (5) — Robotic artist; professor, Concordia University. Canadian artist (born 1959) who builds robotic installations about the 'aesthetics of artificial behaviour', from hysterical to mentally ill machines. http://billvorn.concordia.ca/ - **Daito Manabe** (5) — Artist, programmer, DJ; co-founder of Rhizomatiks. Media artist who works with electrical muscle stimulation, drones, machine learning and brain data. https://www.daito.ws/en/ - **Jake Elwes** (5) — Artist working with machine learning. British artist whose work explores AI through queer performance, landscape and feedback between networks. https://www.jakeelwes.com - **James Bridle** (5) — Artist and writer. British artist and writer on technology, networks and non-human intelligence, author of New Dark Age (2018) and Ways of Being (2022). https://jamesbridle.com/ - **Joana Moll** (5) — Artist and researcher. Barcelona-based artist and researcher whose works examine the material, ecological and surveillant infrastructures of the internet, from border webcams to data brokers and the carbon cost of websites. https://www.janavirgin.com/ - **Lawrence Lek** (5) — Artist, filmmaker and musician. London-based artist of Malaysian-Chinese descent who builds films, games and simulations about AI in a 'Sinofuturist' universe. https://www.lawrencelek.com - **Louis-Philippe Demers** (5) — Robotic artist and researcher. Canadian artist who has made large robotic installations and performances since the 1990s, often with Bill Vorn, and writes on the performing machine. - **Mimi Ọnụọha** (5) — Artist and researcher. Nigerian-American artist whose installations, films and texts examine what data collection includes, leaves out and hides. https://www.mimionuoha.com/ - **Ralf Baecker** (5) — Media artist; professor, University of the Arts Bremen. German artist (born 1977) who builds large mechanical models of computation, perception and neural networks. https://rlfbckr.io/ - **Shinseungback Kimyonghun** (5) — Artist duo (Shin Seung Back and Kim Yong Hun). Seoul-based duo, formed in 2012, who build installations around face detection, object recognition and other machine-vision systems. https://ssbkyh.com/ - **automato.farm** (5) — Design studio (Simone Rebaudengo, Saurabh Datta, Matthieu Cherubini and others). Speculative design collective that builds working prototypes about connected objects, automation and machine learning. http://www.automato.farm/ - **Christa Sommerer & Laurent Mignonneau** (4) — Media-art duo; professors at the Interface Cultures department, University of Art and Design Linz. Christa Sommerer and Laurent Mignonneau have made interactive artworks about artificial life since the early 1990s. - **Dries Depoorter** (4) — Artist. Dries Depoorter is a Belgian artist who makes installations, apps and online projects about privacy, surveillance cameras and machine learning. https://driesdepoorter.be - **Eduardo Kac** (4) — Artist; pioneer of telepresence and bio art. Brazilian-American artist who made telerobotic works from the late 1980s before turning to transgenic art. https://www.ekac.org - **Forensic Architecture** (4) — Research agency, Goldsmiths, University of London. Forensic Architecture, founded by Eyal Weizman in 2010, investigates state and corporate violence with spatial, architectural and computational methods and presents its findings in courts and museums. https://forensic-architecture.org/ - **Gilberto Esparza** (4) — Artist working with robotics, biotechnology and ecology. Mexican artist who builds autonomous bio-machines and robotic 'parasites' that live off urban electricity, polluted rivers and sewage. https://gilbertoesparza.net - **Guo Cheng** (4) — Artist. Chinese artist who builds installations and research projects around technical systems, from face detection to the firewall and computing power. https://www.guo-cheng.net - **Harun Farocki** (4) — Filmmaker and artist (1944–2014). German filmmaker, writer and installation artist who made over 90 films and installations on images, labour and war. https://www.harunfarocki.de/ - **Jiabao Li** (4) — Artist and designer; Professor of Art and Design, Northeastern University (Oakland). Artist, inventor and professor (Harvard GSD alum, former Apple designer) whose installations, wearables, XR works, bio-art and performances deal with perception, technology and the more-than-human. Her 2026 solo show Womb Wetware, made during and after her pregnancy, turns motherhood into a test bed for care, bodies and machines. https://www.jiabaoli.org - **Julian Oliver** (4) — Critical engineer and artist. New Zealand artist and co-author of the Critical Engineering Manifesto. https://julianoliver.com/ - **Lawrence Abu Hamdan** (4) — Artist and audio investigator ('private ear'). Artist whose audio investigations of voices, gunshots and silences have served as evidence in asylum hearings and human-rights cases; joint winner of the 2019 Turner Prize. http://lawrenceabuhamdan.com/ - **Lynn Hershman Leeson** (4) — Artist and filmmaker. American artist working since the 1960s on identity, surveillance, artificial intelligence and the relation between bodies and technology. https://www.lynnhershman.com/ - **Nam June Paik** (4) — Artist; pioneer of video art. Korean-born artist (1932–2006) of the Fluxus circle who turned televisions, cameras and robots into artistic material. https://en.wikipedia.org/wiki/Nam_June_Paik - **Philipp Schmitt** (4) — Artist and designer. Philipp Schmitt (b. 1993) is an artist and designer whose installations, artist books and websites engage with the philosophical, poetic and political sides of computation. https://philippschmitt.com - **Sebastian Schmieg** (4) — Artist. Sebastian Schmieg is a Berlin-based artist who works with found datasets, custom software, websites and lecture performances on the human labour inside algorithmic systems. https://sebastianschmieg.com/ - **So Kanno** (4) — Artist. Media artist who builds robots and machines that draw, swarm and write, often with yang02 (Takahiro Yamaguchi). https://www.kanno.so/ - **Superflux** (4) — Speculative design and futures studio founded by Anab Jain and Jon Ardern. Superflux builds immersive installations, films and policy prototypes that make possible futures tangible, with a long strand of work on more-than-human politics and ecological AI. https://superflux.in - **Vladan Joler** (4) — Artist and researcher; SHARE Foundation; Professor, University of Novi Sad. Serbian artist and researcher who draws large diagrammatic maps of technology, extraction and power. https://labs.rs/en/ - **Anna Ridler** (3) — Artist working with data and machine learning. British artist who builds hand-made datasets, often of flowers, and trains generative models on them. https://annaridler.com - **Annie Dorsen** (3) — Director and writer of algorithmic theatre. American theatre director who since 2010 has made performances in which algorithms write, speak or stage the text live. https://anniedorsen.com/ - **Arthur Ganson** (3) — Kinetic sculptor. American kinetic sculptor (born 1955) whose gear machines, shown at the MIT Museum, stage time, effort and futility. https://www.arthurganson.com/ - **Bjørn Karmann** (3) — Designer and creative technologist. Danish designer trained at the Copenhagen Institute of Interaction Design (CIID) who builds working devices that let people train, hack or rethink machine perception. https://bjoernkarmann.dk/ - **Cao Fei** (3) — Artist and filmmaker. Guangzhou-born artist working in video, virtual worlds and installation on labour, urbanisation and digital life in China. https://www.caofei.com - **Chico MacMurtrie** (3) — Artist; founder of Amorphic Robot Works. American artist (born 1961) who with Amorphic Robot Works builds anthropomorphic, musical and inflatable robotic sculptures. http://www.amorphicrobotworks.org/ - **Claude Shannon** (3) — Mathematician and engineer, Bell Labs and MIT. American mathematician (1916–2001) who founded information theory and built playful machines, including the maze-learning mouse Theseus and the 'Ultimate Machine'. - **Cod.Act** (3) — Duo of André and Michel Décosterd (composer and architect). Swiss duo that builds sound machines whose chaotic movement produces their music; winners of the Swiss Grand Award for Music 2019. https://codact.ch/ - **Dunne & Raby (Anthony Dunne & Fiona Raby)** (3) — Critical and speculative design practice; professors at The New School, New York. Anthony Dunne and Fiona Raby established critical design as a practice, using designed objects and scenarios — including biotechnological ones — to argue about possible futures. https://dunneandraby.co.uk/ - **Elisa Giaccardi** (3) — Professor of Interaction Design, Politecnico di Milano; formerly Chair of Post-Industrial Design, TU Delft. Interaction design researcher who put cameras on everyday objects to see the world from their side and developed thing-centred design. - **Guy Ben-Ary** (3) — Artist and researcher, SymbioticA, The University of Western Australia. Bio-artist who has worked with living neurons since 2000: robotic drawing (MEART), neural synthesizers (cellF) and organoids grown from the composer Alvin Lucier's blood (Revivification). https://guybenary.com - **Heather Dewey-Hagborg** (3) — Artist and biohacker. Heather Dewey-Hagborg is an American artist whose work deals with genetic surveillance, DNA phenotyping and the algorithms that turn biological traces into faces. https://deweyhagborg.com - **Hod Lipson** (3) — Roboticist; professor, Columbia University. Israeli-American roboticist working on evolutionary robotics, self-modelling and self-aware machines. https://www.hodlipson.com/ - **Holly Herndon** (3) — Composer and artist. American composer who makes music with voice, computers and AI models she trains with her collaborators. https://holly.plus/ - **Josh Bongard** (3) — Roboticist; professor, University of Vermont. Canadian roboticist working on evolutionary robotics and embodied cognition; co-creator of xenobots. - **Joy Buolamwini** (3) — Computer scientist, poet of code; founder, Algorithmic Justice League. Ghanaian-American-Canadian researcher who found that face-analysis systems failed on her dark-skinned face at MIT Media Lab and turned the finding into audits, poems and a movement. https://www.poetofcode.com/ - **Kelly Dobson** (3) — Artist and designer; professor, Rhode Island School of Design. American artist who developed 'machine therapy' at the MIT Media Lab, building machines that relate to people through voice and breath. http://www.kellydobson.com/ - **Ken Goldberg** (3) — Artist and roboticist; Professor, UC Berkeley. Robotics professor who made some of the first internet-controlled robots as artworks in the 1990s. https://goldberg.berkeley.edu/ - **Kris Verdonck** (3) — Theatre maker and visual artist; director of A Two Dogs Company. Belgian artist whose performances and installations put machines, objects and humans on stage as equal performers. https://www.atwodogscompany.org/ - **Lu Yang** (3) — Artist. Shanghai-based artist who combines neuroscience, Buddhist cosmology, game engines and motion capture in videos, games and avatars. - **Marcel·lí Antúnez Roca** (3) — Artist; co-founder of La Fura dels Baus. Catalan artist who left La Fura dels Baus to build 'mechatronic performances' in which exoskeletons, robots and audiences control a body. https://www.marceliantunez.com/ - **Marco Donnarumma** (3) — Performance artist, composer and researcher. Italian artist who fuses body, sound and AI-driven prostheses in performances about ritual, disability and machine autonomy. https://marcodonnarumma.com/ - **Mat Dryhurst** (3) — Artist and researcher. British artist and Herndon's long-time collaborator, working on AI, data rights and online governance. https://holly.plus/ - **Memo Akten** (3) — Artist; professor, UC San Diego. Turkish-born artist working with machine learning, perception and meaning-making. https://www.memo.tv - **Oriza Hirata** (3) — Playwright and director, Seinendan; Professor, Osaka University. Playwright of 'contemporary colloquial theatre' who, with Hiroshi Ishiguro, launched the Robot-Human Theater project in 2008. - **Paolo Cirio** (3) — Artist. Paolo Cirio is an Italian artist whose interventions expose the legal, social and economic systems of information technology, from privacy to finance. https://paolocirio.net - **Pierre Huyghe** (3) — Artist. French artist who builds living, self-changing situations with animals, plants, microbes and machine learning systems that continue without the audience. https://www.estherschipper.com/artists/41-pierre-huyghe/ - **Rafael Lozano-Hemmer** (3) — Media artist; founder, Antimodular Research. Rafael Lozano-Hemmer is a Mexican-Canadian artist whose interactive installations use tracking, biometrics and surveillance systems in public and museum spaces. https://www.lozano-hemmer.com - **Simon Penny** (3) — Artist and theorist of embodied interaction; professor, UC Irvine. Australian artist and theorist (born 1955) whose autonomous robots and interactive environments probe embodiment; author of Making Sense (2017). https://simonpenny.net/ - **Sougwen Chung** (3) — Artist; founder, Scilicet. Canadian-born, Hong Kong-raised artist who draws and paints with robot arms she builds and trains (D.O.U.G. series). https://sougwen.com - **Stephanie Dinkins** (3) — Artist; Professor, Stony Brook University. Transmedia artist who makes AI systems with and about Black communities, from conversations with the robot BINA48 to a chatbot trained on her family's stories; first LG Guggenheim Award winner (2023). https://www.stephaniedinkins.com/ - **Takahiro Yamaguchi (yang02)** (3) — Artist (yang02). Artist working as yang02 on graffiti, public space and drawing machines, co-author of Senseless Drawing Bot with So Kanno. - **Takashi Ikegami** (3) — Artificial life researcher; Professor, University of Tokyo. Complex-systems scientist who builds artificial-life installations and the android Alter to study spontaneity and 'life' in machines. - **Yuk Hui** (3) — Philosopher; Professor of Philosophy, Erasmus University Rotterdam. Hong Kong-born philosopher of technology known for the concepts of digital objects, cosmotechnics and technodiversity. - **Zach Blas** (3) — Artist, filmmaker and writer; Associate Professor, University of Toronto. Zach Blas is an American artist whose films, installations and masks examine biometrics, prediction and the politics of technology from a queer perspective. https://zachblas.info - **robotlab** (3) — Artist group (Matthias Gommel, Martina Haitz, Jan Zappe). German group, founded at ZKM in 2000, that takes industrial robots out of the factory into art and public space. http://www.robotlab.de/ - **!Mediengruppe Bitnik** (2) — Artist duo (Carmen Weisskopf and Domagoj Smoljo). Swiss artist duo who work with the internet, bots, hacking and the circulation of data. https://wwwwwwwwwwwwwwwwwwwwww.bitnik.org/ - **Aaron Koblin** (2) — Artist and technologist. American artist known for data visualisation and crowdsourced works; later co-founded the VR studio Within. https://www.aaronkoblin.com/ - **Addie Wagenknecht** (2) — Artist and researcher. American artist working with feminist theory, open hardware and robotics; member of F.A.T. Lab and Deep Lab. https://www.placesiveneverbeen.com/ - **Agnieszka Kurant** (2) — Conceptual artist. Polish artist who examines collective intelligence, non-human labour and the invisible economy. https://en.wikipedia.org/wiki/Agnieszka_Kurant - **Alexander Mordvintsev** (2) — Research scientist, Google. Researcher who created DeepDream and later worked on neural cellular automata. https://znah.net/ - **Andrew Norman Wilson** (2) — Artist and filmmaker. American artist whose films and images deal with labour, corporations and the media they produce. https://www.andrewnormanwilson.com/ - **Anicka Yi** (2) — Artist. Korean-American artist working with scent, bacteria and machines; she speaks of 'biologizing the machine'. https://en.wikipedia.org/wiki/Anicka_Yi - **Ayoung Kim** (2) — Artist. Korean artist who makes speculative-fiction videos with CGI, live action and game engines about geopolitics, labour and time. - **Blaise Agüera y Arcas** (2) — Engineer and researcher; VP and Fellow, Google. Blaise Agüera y Arcas is an American computer scientist who leads research on machine intelligence at Google and founded its Artists + Machine Intelligence programme. - **Blast Theory** (2) — Artists' group (Matt Adams, Ju Row Farr, Nick Tandavanitj). Interactive-art group making games, apps and performances with the Mixed Reality Lab at the University of Nottingham since the 1990s. https://www.blasttheory.co.uk - **Boston Dynamics** (2) — Robotics company. American company spun out of MIT in 1992, known for dynamic legged robots such as BigDog, Spot and Atlas. https://bostondynamics.com/ - **Byung-Chul Han** (2) — Philosopher and cultural theorist. Korean-born German philosopher who writes short books on digital culture, neoliberalism and the smartphone age. https://en.wikipedia.org/wiki/Byung-Chul_Han - **Caroline Sinders** (2) — Artist, designer and human-rights researcher; founder, Convocation Research + Design. American artist and researcher who studies online harassment, data labour and how machine-learning systems are built. https://carolinesinders.com/ - **Chen Qiufan** (2) — Science-fiction writer. Chinese science-fiction writer (Stanley Chan) whose fiction deals with e-waste, cyborgs and AI. - **Choe U-Ram** (2) — Kinetic artist. Korean sculptor who builds 'anima-machines': motorised creatures of aluminium and steel with invented scientific names. https://www.uram.net/ - **Choy Ka Fai** (2) — Artist, choreographer and speculative designer. Singaporean artist trained in Design Interactions at the RCA who makes performances about the body, technology and memory. https://www.ka5.info/ - **Clifford Nass** (2) — Professor of Communication, Stanford University (1958–2013). Stanford social scientist who led the CHIMe lab and wrote Wired for Speech (2005) and The Man Who Lied to His Laptop (2010). - **Cynthia Breazeal** (2) — Roboticist; Professor, MIT Media Lab (Personal Robots Group); founder of Jibo. Pioneer of social robotics who built the expressive robot head Kismet in the late 1990s and later founded the home-robot company Jibo. https://www.media.mit.edu/people/cynthiab/overview/ - **David Rokeby** (2) — Installation artist. Canadian pioneer of interactive installation, known for Very Nervous System (1986–90) and works on machine perception and surveillance. http://www.davidrokeby.com - **Diemut Strebe** (2) — Artist, affiliate at MIT Center for Art, Science & Technology. Diemut Strebe makes works with scientific laboratories, including Sugababe, a living replica of Van Gogh's ear engineered from cells of a Van Gogh descendant and kept alive in a nutrient vitrine. https://theprayer.diemutstrebe.com/ - **ELEVENPLAY** (2) — Dance company directed by MIKIKO. Dance company led by choreographer MIKIKO that performs with drones, robots and machine-learning systems built by Rhizomatiks. - **Edward Ihnatowicz** (2) — Sculptor and pioneer of robotic art. Polish-born British sculptor (1926–1988) who built SAM (1968) and The Senster (1970), early computer-controlled sculptures that responded to people. http://www.senster.com/ - **Elisa Giardina Papa** (2) — Artist. Italian artist whose videos investigate gender, care work and labour in digital economies and AI. https://www.elisagiardinapapa.org/ - **Emilio Vavarella** (2) — Artist and researcher. Italian artist working on the relation between humans, technology and power, often through errors in automated image systems. https://emiliovavarella.com - **Eryk Salvaggio** (2) — Artist and researcher on critical AI. American artist and writer who makes glitch-based films and essays about generative models and archives. https://www.cyberneticforests.com/ - **Eugenia Kuyda** (2) — Founder of Luka and Replika. Russian-born journalist and entrepreneur who co-founded the AI startup Luka and launched the companion chatbot Replika in 2017. https://replika.com/ - **Fangting** (2) — Science-fiction writer; director of Uncommons. Independent science-fiction writer who co-authors Reality Design Lab's speculative projects on decentralized AI life. - **Gil Weinberg** (2) — Professor; founding director, Georgia Tech Center for Music Technology. Israeli-American researcher who builds robotic musicians that listen and improvise. https://gtcmt.gatech.edu/ - **Hanson Robotics** (2) — Robotics company (founded by David Hanson). Company founded by roboticist David Hanson that builds realistic humanoid robot heads with elastomer skin, including BINA48 and Sophia. https://www.hansonrobotics.com/ - **Ho Tzu Nyen** (2) — Artist and filmmaker. Singaporean artist whose films, installations and algorithmic works mix myth and history of Southeast Asia. https://en.wikipedia.org/wiki/Ho_Tzu_Nyen - **Honghao Deng** (2) — Designer and technologist; co-founder of Butlr. Harvard GSD and MIT Media Lab alum who co-designed Twinkle and TransVision with Jiabao Li. - **Ian Cheng** (2) — Artist. American artist of Chinese descent who makes live simulations: virtual ecosystems whose agents change without an end. https://iancheng.com - **James Vlahos** (2) — Journalist; co-founder of HereAfter AI. American journalist and author of Talk to Me (2019) about voice computing, who built a chatbot of his dying father and then co-founded HereAfter AI. - **Jason Edward Lewis** (2) — Professor of Computation Arts, Concordia University; co-director, Indigenous Futures Research Centre. Cherokee, Hawaiian and Samoan digital media artist and poet who leads the Indigenous Protocol and Artificial Intelligence working group. https://jasonlewis.org/ - **Jean Tinguely** (2) — Kinetic artist. Swiss artist (1925–1991) whose clattering, junk-built machines drew, played music and destroyed themselves. https://www.tinguely.ch/ - **Jenna Sutela** (2) — Artist. Finnish artist who works with microbes, slime moulds and machine learning to explore language and cognition beyond the human. https://jennasutela.com/ - **Jennifer Walshe** (2) — Composer and performer; Professor of Composition, University of Oxford. Irish composer and vocalist whose work mixes extended voice, fictional archives and AI. https://milker.org/ - **Jill Magid** (2) — Artist and writer. American artist who forms intimate relationships with systems of power such as police, secret services and CCTV. https://www.jillmagid.com/ - **Joanna Zylinska** (2) — Professor of Media Philosophy and Critical Digital Practice, King's College London. Polish-British writer, artist and curator working on photography, AI and the ethics of technology. https://www.joannazylinska.net - **Jordan Wolfson** (2) — Artist. American artist (born 1980) known for animatronic sculptures and virtual-reality works about violence and attention. https://www.davidzwirner.com/artists/jordan-wolfson - **Joseph Weizenbaum** (2) — Computer scientist, MIT (1923–2008). German-American computer scientist who wrote ELIZA at MIT and later became one of the first critics of artificial intelligence from inside the field. https://en.wikipedia.org/wiki/Joseph_Weizenbaum - **Karen Palmer** (2) — Artist; 'Storyteller from the Future'. Karen Palmer is a British artist who makes immersive films that respond to the viewer's emotions and gaze through AI, eye tracking and facial analysis. https://karenpalmer.uk - **Karl Sims** (2) — Computer graphics artist and researcher. American artist and researcher (born 1962) known for evolved virtual creatures and interactive evolution installations. https://www.karlsims.com/ - **Kate Darling** (2) — Robot-ethics researcher, MIT Media Lab. Legal scholar and researcher studying empathy toward robots and the policy of social robotics; author of The New Breed (2021). http://www.katedarling.org - **Ke Huang** (2) — Artist and designer; graduate student, Tsinghua University; Reality Design Lab collaborator. Designer who directs Reality Design Lab's kinetic and machine-vision installations Cybroc and Body Oracle. https://reality.design/project/cybroc - **Keiichiro Shibuya** (2) — Composer; founder of ATAK. Composer who wrote operas without human singers: THE END for Hatsune Miku and Scary Beauty for the android Alter. https://atak.jp/ - **Ken Feingold** (2) — Artist. American artist who since the 1990s has made talking-head sculptures that hold unscripted conversations. https://www.kenfeingold.com/ - **Ken Rinaldo** (2) — Artist; professor emeritus of art and technology, The Ohio State University. American artist who has built interspecies robotic installations since the late 1980s, in which fish, insects and bacteria steer or switch machines. https://www.kenrinaldo.com/ - **Leonel Moura** (2) — Artist. Portuguese artist (born 1948) who since 2001 has made paintings with swarms of small autonomous robots. https://www.leonelmoura.com/ - **Lev Manovich** (2) — Presidential Professor, The Graduate Center, CUNY; Director, Cultural Analytics Lab. Russian-American media theorist, author of The Language of New Media (2001) and Cultural Analytics (2020). http://manovich.net - **Machine Listening (Sean Dockray, James Parker, Joel Stern)** (2) — Research and artistic collective. Platform for collaborative research and artistic experimentation founded in 2020 by Sean Dockray, James Parker and Joel Stern, working across writing, installation, performance, radio, software and pedagogy. https://machinelistening.exposed/ - **Madeline Gannon** (2) — Artist and roboticist; founder of ATONATON. American artist-roboticist who gives industrial robots perception and personality so they can share space with people. https://atonaton.com/ - **Masahiro Mori** (2) — Roboticist; Professor Emeritus, Tokyo Institute of Technology. Japanese roboticist (1927–2025) who proposed the uncanny valley in 1970 and wrote on robots and Buddhism. - **Merce Cunningham** (2) — Choreographer (1919–2009). American choreographer who used chance, film and, from 1989, computer software to make dances. https://www.mercecunningham.org/ - **N. Katherine Hayles** (2) — Distinguished Research Professor, UCLA; James B. Duke Professor Emerita, Duke University. American literary and media theorist of cybernetics, posthumanism and nonconscious cognition in humans, animals and machines. - **Nicolas Gourault** (2) — Artist and filmmaker. French artist whose documentary installations investigate the human labour behind automated vision. https://nicolasgourault.fr/ - **Norman White** (2) — Robotic artist; taught at OCAD and Ryerson University. American-born Canadian artist (born 1938) who has built behaving robots since Ménage (1974). - **Nouf Aljowaysir** (2) — Artist and filmmaker. Nouf Aljowaysir is a Saudi artist based in New York who works across photography and moving image, using dialogue with machines to examine migration, history and identity. https://www.noufaljowaysir.com - **Nye Thompson** (2) — Artist. British artist who builds software systems that explore machine vision, surveillance networks and machine-made worldviews. https://nyethompson.net - **Patrick Tresset** (2) — Artist. French artist who builds robotic agents that draw people with deliberately clumsy, obsessive gestures. https://patricktresset.com/ - **Paul Kaiser** (2) — Digital artist; co-founder of OpenEndedGroup. Artist who turned motion capture of dancers into hand-drawn-looking digital figures, with Shelley Eshkar and later Marc Downie. https://openendedgroup.com/ - **Paula Gaetano Adi** (2) — Artist; professor, Rhode Island School of Design. Argentine artist who builds robots about 'artificial corporeality' and Andean cosmotechnics rather than artificial intelligence. https://www.paulagaetanoadi.com/ - **Random International** (2) — Art collective founded by Hannes Koch and Florian Ortkrass. Studio making installations about behaviour and machines. https://www.random-international.com/ - **Rebecca Horn** (2) — Artist. German artist (1944–2024) who moved from body extensions to melancholic, self-moving machines and installations. https://en.wikipedia.org/wiki/Rebecca_Horn - **Rhizomatiks** (2) — Media art and technology studio. Tokyo studio founded in 2006 by Daito Manabe and others, known for drone, tracking and ML-driven stage works with ELEVENPLAY. https://rhizomatiks.com/en/ - **Rodney Brooks** (2) — Roboticist; Panasonic Professor of Robotics Emeritus, MIT. Australian roboticist who directed the MIT AI Lab and CSAIL and co-founded iRobot, Rethink Robotics and Robust.AI. https://rodneybrooks.com - **Ron Wakkary** (2) — Professor, School of Interactive Arts and Technology, Simon Fraser University; founder of the Everyday Design Studio. Design researcher who builds counterfactual things such as the Morse Things and the Tilting Bowl and lives with them to study how things and people coexist. http://eds.siat.sfu.ca/ - **Ross Goodwin** (2) — Creative technologist and writer. American creative technologist who builds machines that write from sensor input and film scripts. https://rossgoodwin.com/ - **Ruben Pater** (2) — Designer. Dutch graphic designer whose work addresses politics, surveillance and design ethics; author of The Politics of Design. https://www.untold-stories.net/ - **Sarah Ciston** (2) — Poet-programmer and artist; Professor, Academy of Media Arts Cologne. American artist and researcher who makes zines, chatbots, databases and toolkits on critical and intersectional AI. https://sarahciston.com/ - **Seiko Mikami** (2) — Artist; Professor, Tama Art University. Media artist (1961–2015) who made large interactive installations about perception, surveillance and the body. https://www.ntticc.or.jp/en/archive/participants/mikami-seiko/ - **Shelley Eshkar** (2) — Digital artist; co-founder of OpenEndedGroup. Artist and animator who co-created BIPED and Ghostcatching with Paul Kaiser. https://openendedgroup.com/ - **Shu Lea Cheang** (2) — Artist and filmmaker. Taiwanese-American net-art pioneer whose films and installations deal with surveillance, sexuality and data. https://en.wikipedia.org/wiki/Shu_Lea_Cheang - **Simon Denny** (2) — Artist; Professor, HFBK Hamburg. New Zealand artist who stages the aesthetics and ideologies of tech companies, surveillance agencies and crypto. https://simondenny.net/ - **Simone C. Niquille** (2) — Designer and researcher, technoflesh Studio. Simone C. Niquille runs technoflesh Studio, making films and writing on computer vision, 3D capture and synthetic training datasets; they initiated the parametric truth Lab at Design Academy Eindhoven. https://technofle.sh - **Sony** (2) — Electronics and entertainment company. Japanese company that launched the robot dog AIBO in 1999 and a new aibo in 2018. - **Sun Yuan & Peng Yu** (2) — Artist duo. Beijing-based duo working together since 1999 on installations built from machines, animals, bodies and bodily fluids. https://www.galleriacontinua.com/artists/sun-yuan-peng-yu-73 - **Takayuki Todo** (2) — Artist and robot engineer. Builds small humanoid heads that return a viewer's gaze and facial expression. https://www.takayukitodo.com/ - **Tega Brain** (2) — Artist and environmental engineer; Associate Professor of Integrated Digital Media, New York University. Australian-born artist and engineer whose works treat infrastructure, data and 'eccentric engineering' as ecological questions. https://tegabrain.com/ - **Terence Broad** (2) — Artist and researcher, UAL Creative Computing Institute. British artist-researcher who manipulates the internals of generative networks rather than their data. https://terencebroad.com/ - **W. Grey Walter** (2) — Neurophysiologist and cybernetician, Burden Neurological Institute, Bristol. American-born British neurophysiologist (1910–1977) who worked on EEG and built the first autonomous electronic 'tortoises' in 1948–49. - **Wafaa Bilal** (2) — Artist; Professor, NYU Tisch School of the Arts. Iraqi-American artist whose performances expose the body to remote, networked violence and surveillance. https://wafaabilal.com/ - **Wesley Goatley** (2) — Artist and researcher. British sound artist and researcher whose installations examine the politics, ecology and myths of smart devices and data. https://wesleygoatley.com/ - **William Odom** (2) — Associate Professor, School of Interactive Arts and Technology, Simon Fraser University; Everyday Design Studio. Design researcher who builds slow, long-term technologies for everyday life. - **aaajiao** (2) — Media artist (Xu Wenkai). aaajiao is the online name of Xu Wenkai, an artist, blogger and programmer working on networks, software and online identity. https://eventstructure.com - **#NotABugSplat collective** (1) — Artist collective (with Reprieve and the Foundation for Fundamental Rights). Collective of Pakistani and American artists who installed a giant portrait in a Khyber Pakhtunkhwa field, visible to drone cameras. https://notabugsplat.com/ - **AI Futures Project** (1) — Nonprofit forecasting organisation. Nonprofit led by Daniel Kokotajlo that forecasts the development and impact of advanced AI. https://ai-2027.com/ - **Abeba Birhane** (1) — Cognitive scientist; Assistant Professor, Trinity College Dublin. Abeba Birhane is an Ethiopian-born cognitive scientist who audits large-scale datasets and writes on relational ethics and AI. https://abebabirhane.com - **Adam Ben-Dror** (1) — Designer and artist. New Zealand designer who made Pinokio, a desk lamp with a personality, with Shanshan Zhou and Joss Doggett at Victoria University of Wellington. - **Adrian Chen** (1) — Journalist. American journalist whose 2014 Wired report first described the global workforce of content moderators. https://www.wired.com/2014/10/content-moderation/ - **Adrian Mackenzie** (1) — Professor of Sociology, Australian National University. Sociologist of software and data, author of Cutting Code (2006) and Machine Learners (2017). - **Agnieszka Pilat** (1) — Artist. Polish-born painter who works with Boston Dynamics' Spot robots as studio companions and co-painters. https://www.agnieszkapilat.com/ - **Aidan Meller** (1) — Gallerist; creator of the Ai-Da robot project. British gallerist who initiated Ai-Da, a humanoid 'robot artist' built with Engineered Arts in 2019. https://www.ai-darobot.com/ - **Akihiko Kondo** (1) — School administrator; 'fictosexual' public figure. Held a wedding with the virtual singer Hatsune Miku, living in a Gatebox device, in 2018. - **Alan Turing** (1) — Mathematician and computer scientist (1912–1954). British mathematician who defined the universal computing machine (1936) and proposed the imitation game as a test of machine intelligence (1950). https://www.turing.org.uk - **Alessandro Ludovico** (1) — Artist, editor of Neural magazine. Italian artist, researcher and founding editor of the media-art magazine Neural. https://neural.it - **Alex Garland** (1) — Writer and film director. British novelist and filmmaker, director of Ex Machina and Annihilation. https://en.wikipedia.org/wiki/Alex_Garland - **Alex Rivera** (1) — Filmmaker. Peruvian-American filmmaker who uses science fiction to address migration, borders and labour. https://www.sleepdealer.com/ - **Alexandra Daisy Ginsberg** (1) — Artist; works with synthetic biology, conservation and AI. Alexandra Daisy Ginsberg is a London-based artist whose work examines how humans value nature, from synthetic biology speculation to living artworks for pollinators. https://www.daisyginsberg.com - **Allison Koenecke** (1) — Computational social scientist; Assistant Professor of Information Science, Cornell University. Researcher in algorithmic fairness who led the first large audit of racial disparities in commercial speech recognition. - **Alvin Lucier** (1) — Composer (1931–2021). American experimental composer whose works explore the physical properties of sound and the body. https://en.wikipedia.org/wiki/Alvin_Lucier - **Amazon** (1) — Technology company (Alexa voice assistant). US technology company whose Alexa voice assistant and Echo speakers brought always-listening AI into millions of homes from 2014. https://www.amazon.com/ - **Amelia Winger-Bearskin** (1) — Artist and technologist; Banks Preeminence Chair in AI and the Arts, University of Florida. Seneca-Cayuga artist who works on Indigenous frameworks for technology, storytelling and AI. https://wampum.codes/ - **American Artist** (1) — Interdisciplinary artist. American artist who legally changed their name to American Artist; their work addresses Black labour, policing and technology. https://americanartist.us/ - **Amy Alexander** (1) — Artist; Professor, UC San Diego. American artist who makes audiovisual performances and software art about how machines and people look at media. https://amy-alexander.com - **Andrew Stanton** (1) — Film director, Pixar. American director and screenwriter at Pixar, director of Finding Nemo and WALL·E. https://en.wikipedia.org/wiki/Andrew_Stanton - **Andy Ayrey** (1) — Artist and AI researcher. New Zealand artist who ran the Infinite Backrooms experiments and created the Truth Terminal agent. https://x.com/AndyAyrey - **Andy Clark** (1) — Professor of Cognitive Philosophy, University of Sussex. British philosopher of embodied and extended cognition, author of Natural-Born Cyborgs (2003) and The Experience Machine (2023). - **Angela Neubauer** (1) — Designer. Designer from the Design Investigations programme at the University of Applied Arts Vienna. - **Anna Munster** (1) — Professor, School of Art & Design, UNSW Sydney. Artist and theorist of networks, perception and machine vision, author of An Aesthesia of Networks (2013). - **Anthropic** (1) — AI research company. AI company that develops the Claude models and publishes interpretability research. https://www.anthropic.com/ - **Anton Pidkuiko** (1) — Software engineer. Engineer who co-created GibberLink at an ElevenLabs hackathon in 2025. https://github.com/PennyroyalTea/gibberlink - **Auger-Loizeau** (1) — Design partnership (James Auger and Jimmy Loizeau). British critical design partnership founded in 2000, known for Audio Tooth Implant and Afterlife. - **Bandai** (1) — Toy company; maker of Tamagotchi. Toy company that released Tamagotchi in 1996, conceived by Aki Maita with Akihiro Yokoi (WiZ). - **BeAnotherLab** (1) — Interdisciplinary collective for embodiment and empathy research. Collective founded in Barcelona in 2012 that builds low-cost embodiment systems for swapping perspectives and bodies. https://beanotherlab.org/ - **Beijing E-Town** (1) — Beijing Economic-Technological Development Area (organiser). Beijing development zone and robotics hub that organises the E-Town Half-Marathon with a parallel humanoid robot race. - **Benjamin Bratton** (1) — Professor of Philosophy of Technology and Speculative Design, UC San Diego; Director, Antikythera. American theorist of planetary-scale computation, author of The Stack (2016) and The Revenge of the Real (2021). - **Benjamin Grosser** (1) — Artist; Professor, University of Illinois Urbana-Champaign. American artist who builds software and interactive works about the cultural, social and political effects of computing. https://bengrosser.com - **Benjamin Maus** (1) — Designer and artist. German designer and artist who builds machines and software systems that sort, draw and compute. https://allesblinkt.com - **Bernd Hopfengärtner** (1) — Designer and artist. German designer whose films and objects explore near-future consumer technologies. https://berndhopfengaertner.net - **Betti Marenko** (1) — Design theorist, Central Saint Martins. Italian-born design theorist writing on animism, uncertainty and digital objects. - **Bill Posters** (1) — Artist and researcher. British artist (pseudonym) who works on political deepfakes, advertising and the data-influence industry, often with Daniel Howe. https://billposters.ch/ - **Bill T. Jones** (1) — Choreographer; co-founder of Bill T. Jones/Arnie Zane Company. American choreographer and dancer whose work addresses the body, race, illness and loss. https://newyorklivearts.org/ - **Blake Lemoine** (1) — Former Google software engineer. Engineer who worked on responsible AI at Google and publicly claimed in 2022 that LaMDA was sentient. https://en.wikipedia.org/wiki/LaMDA - **Blanca Li** (1) — Choreographer and director. Spanish-French choreographer working across dance, film and opera. https://www.blancali.com/ - **Boris Petrovsky** (1) — Artist. German artist who builds installations and performative sculptures from networked everyday machines and signs. http://www.petrovsky.de - **Boris Starkov** (1) — Software engineer. Engineer who co-created GibberLink with Anton Pidkuiko. https://github.com/PennyroyalTea - **Bot & Dolly** (1) — Design and engineering studio for robotic camera and motion control (acquired by Google in 2013). San Francisco studio that used industrial robot arms for film and live performance. - **Brian House** (1) — Artist and researcher. Artist working with sound, data and environmental sensing, whose projects turn tracking, listening and infrastructure into perceptible rhythms. https://brianhouse.net/ - **Bruce Lacey** (1) — Artist, performer and robot maker. British artist (1927–2016) who made anarchic, junk-built humanoid robots and performances in the 1960s. - **Bruno Moreschi** (1) — Artist and researcher. Brazilian artist-researcher who questions art institutions and the systems that classify art, including computer vision. https://brunomoreschi.com - **Bureau of Inverse Technology (BIT)** (1) — Art-engineering collective. Anonymous information agency founded in the early 1990s, with Natalie Jeremijenko and Kate Rich among its members, that built devices to monitor the monitors. https://bureauit.org - **Byron Reeves** (1) — Paul C. Edwards Professor of Communication, Stanford University. Communication researcher who studies media psychology, from television to games and screen use. - **Cap_able** (1) — Fashion-tech start-up (Rachele Didero, Federica Busani). Italian start-up founded by Rachele Didero and Federica Busani that makes knitwear with adversarial patterns against AI person and face detection; the idea began in New York in 2019. https://www.capable.design - **Catherine D'Ignazio** (1) — Associate Professor of Urban Science and Planning, MIT; Director, Data + Feminism Lab. Scholar, artist and software developer working on feminist data practice and counter-data. - **Catie Cuan** (1) — Dancer, choreographer and roboticist. Choreographer with a PhD in mechanical engineering from Stanford who works with industrial and social robots. https://www.catiecuan.com/ - **Cesar & Lois (Cesar Baio, Lucy HG Solomon)** (1) — Art collective. Collective of Cesar Baio and Lucy HG Solomon (The League of Imaginary Scientists) that couples microorganisms, networks and AI into 'bhiobrid' systems. - **Charlie Brooker** (1) — Writer and showrunner, Black Mirror. British satirist and screenwriter who created the anthology series Black Mirror (2011–). https://en.wikipedia.org/wiki/Charlie_Brooker - **Chris Olah** (1) — Interpretability researcher, Anthropic; co-founder of Distill. Researcher known for visual, interactive explanations of what neural networks represent. https://colah.github.io/ - **Chris Sanders** (1) — Animation director. American animator and director of Lilo & Stitch, How to Train Your Dragon and The Wild Robot. https://en.wikipedia.org/wiki/Chris_Sanders - **Chris Sugrue** (1) — Artist and programmer. American artist and creative coder working with interactive installations, artificial life and experimental interfaces. - **Christian Moeller** (1) — Media artist; Professor, UCLA Design Media Arts. German-born media artist who makes interactive installations and public artworks with sensors, cameras and robots. https://christianmoeller.com - **Christopher Kulendran Thomas** (1) — Artist. Tamil-British artist whose installations, often made with Annika Kuhlmann, use AI and machine-learning systems to examine how images write history and politics. - **Ciarán Cassidy** (1) — Documentary filmmaker. Irish documentary director. https://fieldofvision.org/shorts/the-moderators - **Claus Emmeche** (1) — Biosemiotician, University of Copenhagen. Danish theoretical biologist working on biosemiotics, artificial life and emergence. - **Clement Valla** (1) — Artist; teaches at Rhode Island School of Design. French-American artist whose works look at images and objects produced by automated systems and software. https://www.clementvalla.com - **Coding Rights** (1) — Feminist organisation for digital rights (founded by Joana Varon). Brazilian women-led organisation working on data, surveillance and technology from a transfeminist, Global South perspective. https://codingrights.org/ - **Coralie Vogelaar** (1) — Visual artist. Coralie Vogelaar is a Dutch visual artist who works with emotion-recognition software, biofeedback and choreography to study how machines decode bodies. https://www.coralievogelaar.com - **Cornell Creative Machines Lab** (1) — Robotics lab led by Hod Lipson (now at Columbia University). Lab researching self-aware, self-replicating and creative machines. https://www.creativemachineslab.com/ - **Craig Reynolds** (1) — Computer graphics researcher. American computer scientist (born 1953) who created the Boids flocking model in 1986–87. https://www.red3d.com/cwr/ - **Croteam** (1) — Video game developer. Croatian game studio known for the Serious Sam series and The Talos Principle. https://www.croteam.com/ - **DARPA** (1) — Defense Advanced Research Projects Agency, US Department of Defense. US military research agency that funds and stages robotics competitions, including the DARPA Robotics Challenge (2012–2015). https://www.darpa.mil/ - **Dadabots** (1) — Music AI duo (CJ Carr, Zack Zukowski). Duo that trains raw-audio neural networks on bands and voices, known for endless AI metal streams. https://dadabots.com/ - **Dan Chen** (1) — Artist and industrial designer (RISD MFA, MIT Media Lab). Designer whose RISD-era robotic intimacy series built machines that caress, hug and comfort people at vulnerable moments. https://dankc.com - **Dani Clode** (1) — Augmentation designer; researcher at the Plasticity Lab, University of Cambridge. Designer of prosthetics and body extensions, known for the Third Thumb. https://www.daniclodedesign.com/ - **Daniel Coppen** (1) — Designer; co-founder of Playfool. British designer who co-founded the design studio Playfool with Tomo Kihara, making games and interactive installations. - **Daniel Howe** (1) — Artist, programmer and professor. American artist and researcher whose works (with Helen Nissenbaum) include the obfuscation tools TrackMeNot and AdNauseam. https://rednoise.org/daniel/ - **Danja Vasiliev** (1) — Artist and critical engineer. Russian-born artist in Berlin who works with networks, hardware and software exploits; co-author of the Critical Engineering Manifesto. https://k0a1a.net - **Dasha Ilina** (1) — Artist. Russian-born artist based in Paris who makes low-tech, DIY objects and workshops about the relation between digital devices and the human body. http://centerfortechpain.com/ - **Dave Murray-Rust** (1) — Researcher in human–algorithm interaction; Associate Professor, TU Delft. Computer scientist and designer (formerly University of Edinburgh) who works on how people live with algorithms and robotic systems. - **David Chalmers** (1) — Professor of Philosophy and Neural Science, New York University. Australian philosopher who named the "hard problem" of consciousness and writes on virtual worlds and machine consciousness. - **David Harris Smith & Frauke Zeller** (1) — Communication researchers (McMaster University; Toronto Metropolitan University). Canadian media and communication professors who built hitchBOT in 2013 as a social experiment in human–robot trust. https://en.wikipedia.org/wiki/HitchBOT - **David J. Gunkel** (1) — Professor of Media Studies, Northern Illinois University. American media philosopher, author of The Machine Question (2012), Robot Rights (2018) and Person, Thing, Robot (2023). - **David Link** (1) — Artist and media archaeologist. German artist and researcher who rebuilds early computers and programs, such as the Ferranti Mark 1 and Strachey's love-letter generator. https://alpha60.de/ - **DeepBrain AI** (1) — AI avatar company. Korean company making synthetic video avatars; launched the memorial service Re;memory in 2022. https://www.deepbrain.io/ - **Desmond Paul Henry** (1) — Artist and philosopher (1921–2004). British philosophy lecturer who built drawing machines from bombsight analogue computers in the 1950s–60s. https://www.desmondhenry.com/ - **Diane Cescutti** (1) — Artist. French-Senegalese artist who works with textiles, looms and computers to retell the history of computation through weaving knowledge. - **Domestic Data Streamers** (1) — Art and design studio. Barcelona studio founded in 2013 that makes participatory installations and research on data, memory and public life. https://www.domesticstreamers.com/ - **Don Ihde** (1) — Philosopher of technology; Distinguished Professor Emeritus, Stony Brook University. American philosopher who founded postphenomenology, the study of how technologies shape human perception and action. - **Donna Haraway** (1) — Distinguished Professor Emerita, History of Consciousness Department, UC Santa Cruz. Feminist scholar of science and technology, author of the Cyborg Manifesto, When Species Meet and Staying with the Trouble. - **Driessens & Verstappen** (1) — Artist duo (Erwin Driessens and Maria Verstappen). Dutch duo working since 1990 with generative systems, artificial life and robots that produce forms and sensations. - **Ei Wada** (1) — Artist and musician; founder of Electronicos Fantasticos!. Turns obsolete appliances such as CRT televisions, fans and barcode readers into instruments played by a community orchestra. https://www.electronicosfantasticos.com/en/ - **Electric Circus** (1) — Robot street-theatre company (Mirjam Langemeijer and Fred Abels). Collaboration between puppeteer Mirjam Langemeijer and inventor Fred Abels, who combine animatronics and puppetry into lifelike robot characters for street theatre. http://www.electric-circus.eu - **Emanuel Andel** (1) — Artist. Austrian media artist who builds risky human–machine games driven by biosignals. - **Emily M. Bender** (1) — Professor of Linguistics, University of Washington. Computational linguist who directs the University of Washington's Computational Linguistics Laboratory and co-wrote The AI Con (2025). - **Engineered Arts** (1) — Humanoid-robot company. British company founded by Will Jackson that builds humanoid entertainment robots such as RoboThespian and Ameca. https://www.engineeredarts.co.uk/ - **Eric Paulos** (1) — Professor, EECS, UC Berkeley; director of the Hybrid Ecologies Lab. HCI researcher who built early telepresence robots in the 1990s and later worked on critical making. https://www.paulos.net/ - **Esther Hovers** (1) — Artist and photographer. Dutch artist whose photographs and installations examine surveillance, power and normative behaviour in public space. https://www.estherhovers.com - **Eszter Zwickl** (1) — Designer. Designer from the Design Investigations programme at the University of Applied Arts Vienna. - **Etsuko Ichihara** (1) — Media artist. Japanese media artist who combines folk rituals and customs with technologies such as robots, VR and face recognition. https://etsuko-ichihara.com - **Everest Pipkin** (1) — Artist, game designer and writer. Everest Pipkin makes games, software, drawings and datasets-as-artworks, and writes on the ethics of corpora used for machine learning. https://everest-pipkin.com - **Evolving AI Lab** (1) — Research lab (Jeff Clune), University of Wyoming. Jeff Clune's lab at the University of Wyoming, working on evolutionary algorithms and deep learning; its members Anh Nguyen and Jason Yosinski co-authored the 'fooling images' paper. - **Facebook AI Research (FAIR)** (1) — AI research lab of Meta. Meta's fundamental AI research lab, founded in 2013. https://ai.meta.com/ - **Felix Lenz** (1) — Designer and filmmaker. Designer and filmmaker trained in Design Investigations at the University of Applied Arts Vienna, working on the material and human costs of technology. - **Filipe Vilas-Boas** (1) — Artist. Portuguese artist based in Paris who makes installations about digital life, surveillance and robots. https://www.filipevilasboas.com/ - **Francesca Panetta** (1) — Immersive artist and journalist; formerly XR Creative Director, MIT Center for Advanced Virtuality. British immersive journalist and artist, formerly head of virtual reality at The Guardian. https://moondisaster.org/ - **Frank Kolkman** (1) — Designer and researcher. Frank Kolkman is a Dutch designer who works on DIY medicine and technology; his Designs for Flies won the 2016 Dutch Design Award for Service & Systems. http://www.frankkolkman.nl - **Frank Lantz** (1) — Game designer; founding chair of the NYU Game Center. American game designer and theorist of games as an aesthetic form. https://www.decisionproblem.com/paperclips/ - **François Pachet** (1) — AI and music researcher; former director of Sony CSL Paris. French computer scientist who built interactive style-learning music systems. https://www.francoispachet.fr/ - **Frieder Weiss** (1) — Media artist and software developer (Kalypso). German engineer-artist who writes real-time camera tracking software for dance. - **Fritz Heider** (1) — Psychologist (1896–1988). Austrian-born social psychologist at Smith College and the University of Kansas, founder of attribution theory. - **Future of Life Institute / Stop Autonomous Weapons** (1) — Non-profit on existential risks of technology. Non-profit that campaigns, with AI researchers such as Stuart Russell, for a ban on lethal autonomous weapons. https://futureoflife.org/ - **Félix Luque Sánchez** (1) — Artist. Spanish artist based in Brussels who builds robotic and algorithmic installations about chaos, control and machine autonomy. - **GROOVE X** (1) — Robotics company; maker of LOVOT. Company founded by Kaname Hayashi in 2015 to build LOVOT, a robot designed only to be loved. https://lovot.life/en/ - **Gabriel Pereira** (1) — Researcher. Brazilian researcher who studies computer vision and critical data practices. - **Gabriele Trovato** (1) — Roboticist; Associate Professor, Shibaura Institute of Technology. Italian roboticist working on culture, religion and robot design; he built SanTO, a robot shaped like a Catholic saint statue. - **Gan Golan** (1) — Artist. American artist and designer of public interventions and wearable sculpture. - **Garnet Hertz** (1) — Artist and design researcher; Canada Research Chair in Design and Media Arts, Emily Carr University. Canadian artist and researcher known for critical making, DIY electronics and the cockroach-driven robot he built as a graduate student at UC Irvine. http://www.conceptlab.com/ - **Gaëtan Robillard** (1) — Artist and researcher. French artist who works with computation, data and sound to examine how algorithms model knowledge, from climate misinformation to cognition. - **Geocinema** (1) — Artist collective. Collective founded by Asia Bazdyrieva and Solveig Suess (with Alexey Orlov in early projects) that treats planetary sensing networks as one distributed camera. https://geocinema.network - **George E. Lewis** (1) — Composer, trombonist and scholar; Professor, Columbia University. American composer and AACM member who has made interactive computer music since the late 1970s. https://music.columbia.edu/bios/george-e-lewis - **Giacomo Miceli** (1) — Artist and software engineer. Italian artist and engineer who builds web works with synthetic voices and generative text. https://www.infiniteconversation.com/ - **Gideon Obarzanek** (1) — Choreographer; founder of Chunky Move. Australian choreographer who founded Chunky Move in 1995 and made interactive-video dance works with Frieder Weiss. https://chunkymove.com.au/ - **Gilbert Simondon** (1) — Philosopher (1924–1989). French philosopher of individuation and technology whose 1958 thesis described how technical objects evolve and form milieus with humans. - **Golan Levin** (1) — Artist; Professor, Carnegie Mellon University; director of the STUDIO for Creative Inquiry. Golan Levin makes interactive installations, performances and software that explore computer vision, gesture and communication. https://www.flong.com - **Gonçalo F. Cardoso** (1) — Composer and sound artist; founder of the label Discrepant. Portuguese composer and field recordist who runs the experimental label Discrepant. - **Goodby Silverstein & Partners** (1) — Advertising agency. San Francisco advertising agency that made Dalí Lives for the Dalí Museum in St. Petersburg, Florida. https://goodbysilverstein.com/ - **Google Creative Lab** (1) — Creative team at Google. Google's in-house creative team that makes public experiments with machine learning, including Teachable Machine. https://experiments.withgoogle.com/ - **Google DeepMind** (1) — AI research lab. AI lab founded in London in 2010 and acquired by Google in 2014; creator of AlphaGo. https://deepmind.google/ - **Gordon Pask** (1) — Cybernetician, inventor and theorist of conversation. British cybernetician (1928–1996) who built learning and 'conversational' machines, from Musicolour (1953) to the Colloquy of Mobiles (1968). - **Graham Dove** (1) — HCI researcher. British HCI researcher whose work concerns design with data and machine learning. https://doi.org/10.1145/3025453.3025739 - **Halsey Burgund** (1) — Sound artist and technologist. American sound artist who makes participatory audio works with location and speech technology. https://halseyburgund.com/ - **Hans Block** (1) — Director and theatre maker. German director who works with Moritz Riesewieck on films and plays about digital power. https://en.wikipedia.org/wiki/The_Cleaners_(2018_film) - **Hans Block & Moritz Riesewieck** (1) — Documentary directors (Laokoon). German director duo who made The Cleaners (2018), on content moderators, and Eternal You (2024), on the digital afterlife industry. https://en.wikipedia.org/wiki/Eternal_You - **Happiest Baby (Harvey Karp)** (1) — Company making the SNOO robotic bassinet; founded by paediatrician Harvey Karp. Baby-products company that turned Harvey Karp's soothing method (swaddling, white noise, rocking) into a sensor-driven bassinet. https://www.happiestbaby.com - **Harold Cohen** (1) — Painter and AI art pioneer (1928–2016). British-born painter who wrote and developed the drawing program AARON from 1973 until his death. https://aaronshome.com/ - **Harpreet Sareen** (1) — Designer and researcher; professor, Parsons School of Design. Indian-born designer and researcher who developed 'cyborg botany' at the MIT Media Lab. - **Hasan Elahi** (1) — Artist; Professor, George Mason University. Bangladeshi-American artist who, after being mistakenly placed on a US terrorist watch list, began tracking and publishing his own life. https://en.wikipedia.org/wiki/Hasan_Elahi - **Heather Knight** (1) — Roboticist; Assistant Professor, Oregon State University (CHARISMA Lab). Social roboticist who founded Marilyn Monrobot and uses theatre and comedy to study robot expression. https://www.marilynmonrobot.com/ - **Helena Nikonole** (1) — Artist and curator. Helena Nikonole is a new-media artist working with AI, bio art and DIY biotechnology. https://nikonole.com/ - **Huang Yi** (1) — Choreographer, dancer and inventor. Taiwanese choreographer who programs industrial robots himself and founded Huang Yi Studio +. https://huangyistudio.com/ - **Hubert Dreyfus** (1) — Philosopher; Professor, UC Berkeley (1929–2017). American philosopher who used Heidegger and Merleau-Ponty to argue that symbolic AI could not capture embodied, situated know-how. - **Hugh Herr** (1) — Biophysicist; head of the Biomechatronics group, MIT Media Lab. Climber and engineer who lost both lower legs in 1982 and now designs powered bionic limbs. https://www.media.mit.edu/people/hherr/overview/ - **Human Computer Integration Lab (University of Chicago)** (1) — HCI lab led by Pedro Lopes. Lab that builds devices integrating computers with the body, from muscle stimulation to wearable jammers. https://lab.plopes.org/ - **Humanising Autonomy (Maya Pindeus, Leslie Nooteboom, Raunaq Bose, Adam Bernstein)** (1) — Design team from the RCA/Imperial Innovation Design Engineering programme, later a London start-up. Four graduate designers who built Blink, a two-way signalling interface between pedestrians and autonomous vehicles, and then founded a company on pedestrian-behaviour AI. - **Ian Bogost** (1) — Game designer and philosopher; Professor, Washington University in St. Louis. American game designer and object-oriented philosopher, author of Persuasive Games (2007) and Alien Phenomenology (2012). - **Intuition Robotics** (1) — Company making the ElliQ companion robot. Israeli company founded in 2016 that develops proactive AI companions for older adults; ElliQ was designed with Yves Béhar's fuseproject. https://elliq.com - **Iohanna Nicenboim** (1) — Designer and researcher, TU Delft; works on more-than-human design and AI. Designer-researcher who treats voice assistants and AI as more-than-human agents and designs situated conversations with them. https://iohanna.com/ - **Iris Long** (1) — Artist, curator and researcher. Chinese artist and curator working on data, machine perception and the politics of algorithms. - **Isa Willinger** (1) — Documentary director and author. German documentary filmmaker (born 1980). Her debut Hi, AI (2019) won the Max Ophüls Prize for best documentary. https://en.wikipedia.org/wiki/Isa_Willinger - **Ivan Henriques** (1) — Artist. Brazilian artist who builds hybrid machines that connect plants, microbes and robotics. - **J. C. R. Licklider** (1) — Psychologist and computer scientist (1915–1990). Psychoacoustician at MIT and BBN who, as head of ARPA's Information Processing Techniques Office, funded interactive computing and the networks that led to the internet. - **Jack Burnham** (1) — Art critic and theorist (1931–2019). American critic and Northwestern University professor who wrote "Systems Esthetics" (Artforum, 1968) and curated Software (Jewish Museum, 1970). - **Jacques de Vaucanson** (1) — Inventor and automaton maker. French inventor (1709–1782) of lifelike automata and of an automated loom. - **Jaha Koo** (1) — Theatre-maker, composer and video artist. South Korean theatre-maker whose documentary performances combine music, video, autobiography and technological objects. https://en.wikipedia.org/wiki/Jaha_Koo - **Jake Schreier** (1) — Film director. American director of music videos and films; Robot & Frank (2012) was his feature debut. https://en.wikipedia.org/wiki/Jake_Schreier - **James Pierce** (1) — Design researcher; professor, University of Washington. American design researcher known for counterfunctional design and critical work on smart-home cameras. - **Jan Zuiderveld** (1) — Artist. Dutch artist who fits everyday appliances with sensors and AI so that they talk back and act on their own terms. - **Janelle Shane** (1) — Optics research scientist; author of AI Weirdness. Researcher and writer who documents the odd behaviour of neural networks for a general audience. https://www.aiweirdness.com/ - **Jasia Reichardt** (1) — Curator and writer; former Assistant Director, Institute of Contemporary Arts, London. Polish-born British curator who organised Cybernetic Serendipity (ICA, 1968) and wrote Robots: Fact, Fiction and Prediction (1978). - **Jason Rohrer** (1) — Game designer and programmer. American independent game designer known for art games such as Passage (2007) and The Castle Doctrine, and for the chatbot service Project December. https://projectdecember.net/ - **Jed Berk** (1) — Artist and designer. American artist known for Autonomous Light Air Vessels, small blimp robots that flock, avoid obstacles and seek company. - **Jeff Sebo** (1) — Philosopher; director, NYU Center for Mind, Ethics, and Policy. American philosopher working on animal and AI ethics and the moral status of non-human minds. https://jeffsebo.net/ - **Jem Finer** (1) — Musician and artist. British musician (founding member of The Pogues) and artist working with long durations and deep time. https://longplayer.org/ - **Jemima Wyman** (1) — Artist. Australian artist whose work on camouflage, masks and collective protest images spans painting, textiles and video. https://www.jemimawyman.com/ - **Jennifer A. Lewis** (1) — Materials scientist; professor, Harvard University. American materials scientist known for 3D printing of soft and functional materials. - **Jennifer Lyn Morone** (1) — Artist. American artist who turned herself into a registered corporation whose assets are her data, health and skills. https://jenniferlynmorone.com/ - **Jennifer Rhee** (1) — Professor of English, Virginia Commonwealth University. Scholar of technology, literature and art, author of The Robotic Imaginary (2018). - **Jia Deng** (1) — Computer scientist; Associate Professor, Princeton University. Jia Deng is a computer-vision researcher who, as a PhD student with Fei-Fei Li, led the construction of the ImageNet dataset. - **Jibo, Inc.** (1) — Social robot company (2012–2019). Start-up co-founded by MIT roboticist Cynthia Breazeal that sold Jibo, a tabletop social robot for the family, from 2017 until the company failed. https://en.wikipedia.org/wiki/Cynthia_Breazeal - **Jill Walker Rettberg** (1) — Professor of Digital Culture; Co-director, Center for Digital Narrative, University of Bergen. Norwegian digital culture scholar who led the ERC project Machine Vision in Everyday Life and its database of 500 creative works. https://jilltxt.net - **Jing-cai Liu** (1) — Designer. Chinese designer who makes wearable devices that question privacy and face recognition. http://jingcailiu.com - **Joaquín Fargas** (1) — Artist and engineer; director of Proyecto Biosfera. Argentine artist working across art, science and technology on climate and life, including sealed biospheres and Antarctic robots. https://www.joaquinfargas.com - **Joel Krueger** (1) — Philosopher, University of Exeter. Philosopher of mind working on phenomenology, emotion, music and the extended and social mind. - **Johann Diedrick** (1) — Artist and engineer; adjunct professor, NYU ITP. Artist and engineer who makes 'listening rooms' through installations, performances, software and workshops, and teaches sound art and listening machines at NYU ITP; 2021 Mozilla Creative Media Award recipient. http://johanndiedrick.com/ - **John McCarthy** (1) — Computer scientist (1927–2011). Computer scientist who coined the term "artificial intelligence" in the 1955 Dartmouth proposal, created the Lisp language and founded the Stanford AI Lab. - **John Searle** (1) — Philosopher of mind, formerly UC Berkeley. American philosopher of language and mind known for speech-act theory and the Chinese Room argument (1980). - **Johns Hopkins Applied Physics Laboratory** (1) — University-affiliated research laboratory. Research laboratory of Johns Hopkins University where engineers built the 'Beast', a self-charging mobile automaton, in the early 1960s. https://www.jhuapl.edu/ - **Jon McCormack** (1) — Artist; professor and director of SensiLab, Monash University. Australian artist and researcher in generative art and artificial life. https://www.jonmccormack.info/ - **Jon Rafman** (1) — Artist and filmmaker. Canadian artist working with video, installation and found internet imagery. https://jonrafman.com - **Jonas Simmerlein** (1) — Theologian and philosopher, University of Vienna. Theologian and lecturer in homiletics at the University of Vienna; he designed a Protestant service written mostly by ChatGPT for the 2023 German Protestant Kirchentag. - **Jonas Voigt** (1) — Designer. German designer, co-author of Raising Robotic Natives with Stephan Bogner and Philipp Schmitt. - **Joon Sung Park** (1) — Computer scientist, Stanford HCI Group. HCI researcher who led the Generative Agents project on LLM-driven simulations of human behaviour. https://www.joonsungpark.com/ - **Joseph Sargent** (1) — Film and television director (1925–2014). American director of film and television, including Colossus: The Forbin Project. https://en.wikipedia.org/wiki/Joseph_Sargent - **Josh Begley** (1) — Artist and filmmaker. American data artist and filmmaker whose work maps US drone strikes, prisons and policing. https://joshbegley.com/ - **Jules LaPlace** (1) — Artist and software developer. Artist and programmer who built the website and search engine of MegaPixels and Exposing.ai with Adam Harvey. - **Julia Vollmer** (1) — Artist and designer. German artist and designer who makes the physical and energy costs of digital systems perceptible through material installations. https://juliavollmer.de - **Julian Charrière** (1) — Artist. Swiss-French artist whose sculptures, films and expeditions deal with geology, extraction and deep time. https://julian-charriere.net/ - **Julie Carpenter** (1) — Researcher in human–robot interaction. American researcher who studies how people form attachments to robots and AI, starting from her 2013 University of Washington study of bomb-disposal teams. https://jgcarpenter.com/ - **Julien Prévieux** (1) — Artist; Prix Marcel Duchamp 2014. Julien Prévieux is a French artist whose films, performances and drawings examine how bodies and gestures are measured, patented and optimised by technical systems. https://www.previeux.net - **Julius Popp** (1) — Artist. German artist (born 1973) who builds machines about information and self-organisation, best known for bit.fall. - **Justine Emard** (1) — Visual artist working with neuroscience, robotics and machine learning. French artist who combines photography, video, robotics and AI; her works draw on data from bees, brains and androids. https://justineemard.com - **K Allado-McDowell** (1) — Writer, musician; founder of Artists + Machine Intelligence at Google. Writer and artist who co-wrote several books with large language models and ran Google's Artists + Machine Intelligence programme. https://www.kalladomcdowell.com/ - **Kacie Kinzer** (1) — Designer and interactive artist. Designer (NYU ITP) whose cardboard Tweenbots were collected by MoMA. http://www.tweenbots.com - **Kai-Fu Lee** (1) — Computer scientist and investor; founder, Sinovation Ventures. Taiwan-born computer scientist, former head of Google China and author of AI Superpowers. - **KairUs** (1) — Artist duo (Linda Kronman and Andreas Zingerle). Artist duo researching surveillance, smart cities and the hidden labour of machine vision. https://kairus.org - **Karel Čapek** (1) — Writer and playwright (1890–1938). Czech writer whose play R.U.R. introduced the word 'robot', suggested by his brother Josef. https://en.wikipedia.org/wiki/Karel_%C4%8Capek - **Karolina Sobecka** (1) — Artist. Polish-born artist working with interaction, animation and environmental systems. - **Katarzyna Nowaczyk-Basińska** (1) — Research Fellow, Leverhulme Centre for the Future of Intelligence, University of Cambridge. Researcher of digital immortality and the cultural practices around death and technology. https://www.lcfi.ac.uk/ - **Kathleen Richardson** (1) — Professor of Ethics and Culture of Robots and AI, De Montfort University. British anthropologist of robotics who founded the Campaign Against Sex Robots in 2015. https://campaignagainstsexrobots.org/ - **Katja Novitskova** (1) — Artist. Estonian artist whose sculptures and images deal with how technology, imaging and biology render life; she represented Estonia at the 2017 Venice Biennale. https://www.katjanovitskova.com - **Kazuhiko Hachiya** (1) — Artist; Professor, Tokyo University of the Arts. Media artist known for Inter Dis-Communication Machine, the email pet software PostPet and the OpenSky glider project. https://www.ntticc.or.jp/en/archive/works/inter-dis-communication-machine/ - **Keiichi Matsuda** (1) — Designer and filmmaker. British-Japanese designer and filmmaker whose films imagine cities saturated with augmented reality. http://km.cx - **Kelly Heaton** (1) — Artist. American artist who made pelts from gutted electronic toys and later 'electronic naturalism' circuits. - **Kenneth Colby** (1) — Psychiatrist and AI researcher, Stanford University and UCLA (1920–2001). American psychiatrist who built computer models of neurosis and paranoia and wrote the chatbot PARRY. https://en.wikipedia.org/wiki/Kenneth_Colby - **Kevin Eykholt** (1) — Computer security researcher (PhD, University of Michigan). Security researcher who, at the University of Michigan, was first author of the study on stickers that make classifiers misread stop signs. - **Kevin Warwick** (1) — Cybernetics researcher; Emeritus Professor, Coventry and Reading Universities. British engineer who had an RFID chip (1998) and then a neural electrode array (2002) implanted to connect his nervous system to computers. https://en.wikipedia.org/wiki/Kevin_Warwick - **Kim Jee-woon** (1) — Film director and screenwriter. South Korean director of genre films including A Tale of Two Sisters and I Saw the Devil. https://en.wikipedia.org/wiki/Kim_Jee-woon - **Kim Tae-yong** (1) — Film director. South Korean director of Late Autumn and Wonderland. https://en.wikipedia.org/wiki/Kim_Tae-yong - **Kit Galloway & Sherrie Rabinowitz** (1) — Artist duo; founders of Electronic Cafe International. Kit Galloway and Sherrie Rabinowitz (1950–2013) made satellite and video telecommunication art from the 1970s. http://www.ecafe.com/ - **Knightscope** (1) — Security-robot company. American company founded in 2013 that deploys autonomous security robots in malls, car parks and offices. https://www.knightscope.com/ - **Kogonada** (1) — Film director and video essayist. Korean-born American filmmaker known for video essays on film form and for the features Columbus (2017) and After Yang (2021). https://en.wikipedia.org/wiki/Kogonada - **Kory Mathewson** (1) — AI researcher and improviser; co-founder of Improbotics. Canadian research scientist who built improvising dialogue agents and co-founded Improbotics. https://improbotics.org/ - **Kraftwerk** (1) — Band (founded by Ralf Hütter and Florian Schneider). German electronic band, founded in 1970, whose image and sound fused musicians with machines. https://kraftwerk.com/ - **Krzysztof Wodiczko** (1) — Artist; Professor in Residence, Harvard Graduate School of Design. Krzysztof Wodiczko is a Polish-born artist known for large public projections and instruments that give voice to marginalised people. - **Kōdai-ji** (1) — Rinzai Zen Buddhist temple, Kyoto. A 400-year-old Zen temple in Kyoto that commissioned the android Kannon Mindar with Osaka University. https://www.kodaiji.com/mindar/ - **Kōfuku-ji (Isumi)** (1) — Buddhist temple in Isumi, Chiba. Temple where priest Bungen Oi has held memorial services for broken AIBO robot dogs since 2015, with the repair company A-FUN. - **L.A. Raeven** (1) — Artist duo (twin sisters Liesbeth and Angelique Raeven). Dutch twin sisters whose installations and performances examine body norms, likeness and identity. - **Laetitia Sonami** (1) — Sound artist and composer. French-born composer known for performing electronic music with the Lady's Glove and later a machine-learning instrument. https://sonami.net/ - **Lancel & Maat** (1) — Artist duo (Karen Lancel and Hermen Maat). Dutch duo making participatory installations about touch, trust and presence in networked and surveilled spaces. https://lancelmaat.nl - **Laura Kikauka** (1) — Artist and maker of kinetic environments. Canadian artist known for dense, humorous environments of found electronics and her Funny Farm studio. - **Lauren F. Klein** (1) — Professor of English and Quantitative Theory & Methods, Emory University. Digital humanities scholar of data visualisation, early American literature and feminist data science. - **Laurie Anderson** (1) — Artist, musician and performer. American artist known for voice filters, electronic instruments and multimedia performance since the 1970s. https://laurieanderson.com/ - **Leandro Núñez** (1) — Artist (Leo Núñez). Argentine artist working with robotics, artificial life and generative systems. https://www.leonunez.com.ar/ - **Lee Bul** (1) — Artist. Korean artist working across performance, sculpture and installation; her work addresses technology, gender, utopia and the body. https://en.wikipedia.org/wiki/Lee_Bul - **Leo Selvaggio** (1) — Interdisciplinary artist. Chicago-based interdisciplinary artist (Leonardo Selvaggio) whose projects on surveillance ask the public to wear, perform or hack his identity. https://www.leoselvaggio.com - **Liam Young** (1) — Speculative architect and filmmaker. Australian-born architect who makes films about cities shaped by automation, drones and machine vision. https://liamyoung.org/ - **Libby Heaney** (1) — Artist and quantum physicist. British artist with a PhD in quantum physics who works with machine learning and quantum computing. https://libbyheaney.co.uk/ - **Lilly Irani** (1) — Associate Professor of Communication, UC San Diego. American scholar and former Google designer who studies crowd work, innovation and design politics. https://quote.ucsd.edu/lirani/ - **Lisa Feldman Barrett** (1) — Psychologist and neuroscientist; University Distinguished Professor, Northeastern University. Lisa Feldman Barrett is a Canadian-American psychologist known for the theory of constructed emotion. https://lisafeldmanbarrett.com - **Liu Chuang** (1) — Artist. Chinese artist whose videos trace how labour, land and infrastructure are reorganised by economic and technological change. - **Longquan Monastery** (1) — Buddhist monastery, Beijing. Buddhist monastery in Beijing's Fenghuangling known for its technology-savvy monks and volunteers. - **Louise Crnkovic-Friis** (1) — Choreographer; co-founder of Peltarion. Swedish choreographer who trained neural networks on her own dance to generate new choreography. - **Loïs Soleil** (1) — Artist. Brussels-based artist working with robots, AI and feminist self-defence. http://loissoleil.com - **Luciana Parisi** (1) — Professor, Program in Literature and Computational Media, Arts and Cultures, Duke University. Italian philosopher of computation, formerly at Goldsmiths' Centre for Cultural Studies, author of Contagious Architecture (2013). - **Lucy Osler** (1) — Philosopher, University of Exeter. Philosopher working on phenomenology, empathy and how we experience others online. - **Lucy Suchman** (1) — Professor Emerita of Anthropology of Science and Technology, Lancaster University. Anthropologist who worked for twenty years at Xerox PARC and wrote Plans and Situated Actions (1987). - **Luka Crnkovic-Friis** (1) — Machine-learning engineer; co-founder of Peltarion. Swedish AI engineer who built the chor-rnn model with Louise Crnkovic-Friis. - **Lydia H. Liu** (1) — Wun Tsun Tam Professor in the Humanities, Columbia University. Theorist of media and translation and scholar of comparative literature. - **M Eifler** (1) — Artist and researcher in embodied interaction and computational prosthetics. Interdisciplinary artist (MFA, California College of the Arts 2011) who builds speculative interfaces and prosthetic systems from their own disability. - **M. Beatrice Fazi** (1) — Philosopher of computation, University of Sussex. Philosopher of computation who studies aesthetics, abstraction and indeterminacy in computational and AI systems. - **M. Six Silberman** (1) — Researcher on platform work. American researcher who co-created Turkopticon and later worked with the German metalworkers' union IG Metall on platform work. https://turkopticon.info/ - **MBC (Munhwa Broadcasting Corporation)** (1) — Korean public broadcaster. South Korean broadcaster that produced the VR documentary series Meeting You (2020–). https://www.imbc.com/ - **Mahmood Sharif** (1) — Computer security researcher; Tel Aviv University (PhD, Carnegie Mellon University). Security and machine-learning researcher who, as a PhD student at Carnegie Mellon, led the study showing that printed eyeglass frames can fool face recognition. - **Mamoru Oshii** (1) — Film director. Director of the animated film Ghost in the Shell (1995), adapted from Masamune Shirow's manga. - **Manetta Berends** (1) — Designer and researcher. Dutch designer and researcher working with free software, publishing and machine-learning literacy. - **Manu Luksch** (1) — Artist and filmmaker. Austrian-born artist and filmmaker who works on surveillance, data rights and networked media. https://manuluksch.com/ - **Manuel Beltrán** (1) — Artist and researcher; founder of the Institute of Human Obsolescence. Spanish artist whose installations and interventions expose the power relations inside technologies, with a focus on data and automation. http://speculative.capital - **Margherita Pevere** (1) — Artist and researcher in biological art. Italian artist working with living bacteria, fungi and biomaterials, and with bodies in transformation. https://www.margheritapevere.com/ - **Mari Velonaki** (1) — Artist and robotics researcher; Director, Creative Robotics Lab, UNSW Sydney. Greek-Australian artist who has built interactive robotic installations with engineers at the Australian Centre for Field Robotics since the early 2000s. - **Marianne Simmel** (1) — Psychologist (1923–2010). German-born experimental psychologist, later known for research on phantom limbs and body image. - **Mariela Yeregui** (1) — Artist; professor, Universidad Nacional de Tres de Febrero. Argentine artist who makes robotic and electronic installations about distance, contact and fragility. https://marielayeregui.com/ - **Mark Coeckelbergh** (1) — Professor of Philosophy of Media and Technology, University of Vienna. Belgian philosopher who writes on robot ethics, the relational moral status of machines and the politics of AI. - **Mark Shepard** (1) — Artist and architect; Associate Professor, University at Buffalo. Mark Shepard is an artist, architect and researcher who works on ubiquitous computing and the city; he edited the book Sentient City (MIT Press, 2011). http://survival.sentientcity.net/ - **Marnix de Nijs** (1) — Artist. Dutch artist who builds physically demanding interactive installations about perception and control. http://www.marnixdenijs.nl - **Martin Backes** (1) — Artist. German artist and designer making objects and sound works about digital culture. https://www.martinbackes.com - **Martin Riches** (1) — Artist; builds musical and talking machines. British-born artist in Berlin who has built mechanical instruments and speaking machines since the 1970s. http://www.martinriches.de - **Martine Syms** (1) — Artist and filmmaker. American artist and director whose films, installations and performances deal with Black identity, self-representation and technology. https://martinesyms.com/ - **Martyna Marciniak** (1) — Artist and researcher. Polish artist and researcher whose project Anatomy of Non-Fact studies visual hoaxes, fakes and synthetic images. https://www.martyna.digital - **Marvin Minsky** (1) — Computer scientist; co-founder of the MIT AI Lab (1927–2016). AI pioneer who co-organised the 1956 Dartmouth workshop, co-founded the MIT AI Lab and wrote The Society of Mind (1986). - **Mary Flanagan** (1) — Artist and game designer; Professor, Dartmouth College. Mary Flanagan makes software art, installations and games, and directs the Tiltfactor game lab at Dartmouth. http://maryflanagan.com - **Mary L. Gray** (1) — Senior Principal Researcher, Microsoft Research; Faculty, Indiana University. Anthropologist of media and labour and 2020 MacArthur Fellow who studies on-demand platform work. - **Masahiro Miwa** (1) — Composer; Professor, IAMAS. Japanese composer known for 'reverse simulation music', in which performers or machines execute algorithms as musical rules. - **Masakhane** (1) — Grassroots research community for African-language NLP. Volunteer network of African researchers building datasets and models for African languages. https://www.masakhane.io/ - **Matt Donovan** (1) — Artist. Canadian artist and engineer who co-developed The Robotic Chair with Max Dean and Raffaello D'Andrea. - **Matt Richardson** (1) — Technologist and maker; formerly NYU ITP. Matt Richardson is a creative technologist and author of books on Raspberry Pi and Arduino; he made the Descriptive Camera as a student at NYU's Interactive Telecommunications Program. - **Matt Schlicht** (1) — Entrepreneur. American entrepreneur who launched Moltbook, a forum for AI agents, in January 2026. https://www.moltbook.com/ - **Matteo Pasquinelli** (1) — Associate Professor of Philosophy of Science, Ca' Foscari University of Venice. Italian philosopher who leads the ERC project AIMODELS on the history of AI; his The Eye of the Master won the 2024 Deutscher Memorial Prize. - **Max Dean** (1) — Artist. Max Dean is a British-born Canadian artist known for interactive and robotic works that implicate the viewer. - **Max Dovey** (1) — Artist and researcher. British artist whose performances put people in the position of being classified by, or working for, machine-learning systems. https://www.maxdovey.com - **Maywa Denki** (1) — Art unit run as a fake electrical-appliance company (Nobumichi Tosa). Art unit founded in 1993 by the Tosa brothers that builds absurd musical machines and presents them as company products. https://www.maywadenki.com/ - **Merritt Moore** (1) — Ballet dancer and quantum physicist. American dancer with a DPhil in atomic and laser physics from Oxford who choreographs duets with industrial robot arms. https://www.merrittmoore.com/ - **Michael Almereyda** (1) — Film director and screenwriter. American independent filmmaker; he adapted Jordan Harrison's play Marjorie Prime for the screen in 2017. https://en.wikipedia.org/wiki/Michael_Almereyda - **Michael Eisen** (1) — Biologist, UC Berkeley; open-science advocate. American computational biologist and co-founder of the Public Library of Science (PLOS). https://www.michaeleisen.org/blog/ - **Michael Levin** (1) — Biologist; professor, Tufts University. American developmental biologist who studies bioelectricity and synthetic living machines. - **Michael Mandiberg** (1) — Artist; Professor, CUNY. American artist whose work uses the logics of digital platforms, from Wikipedia printing to gig-economy outsourcing. https://www.mandiberg.com/ - **Michael Snow** (1) — Artist and filmmaker (1928–2023). Canadian artist and structural filmmaker known for Wavelength (1967) and La Région Centrale (1971). - **Michal Luria** (1) — HCI and HRI researcher. Researcher who uses design methods and enactments to explore social robots and conversational agents. https://www.michalluria.com/ - **Michel Waisvisz** (1) — Composer and instrument inventor; artistic director of STEIM (1949–2008). Dutch composer who built gestural electronic instruments at STEIM. https://en.wikipedia.org/wiki/Michel_Waisvisz - **Michelle Teran** (1) — Artist. Canadian artist whose walks and performances make hidden media networks in cities perceptible. - **Michio Okada** (1) — Professor, Toyohashi University of Technology (ICD Lab). Cognitive scientist and roboticist who proposed 'weak robots' (yowai robotto): robots that need people's help to complete their tasks. - **Microsoft XiaoIce team** (1) — Social chatbot team, Microsoft (Asia) STCA; later spun off as Xiaoice. Team that launched XiaoIce in China in 2014, a chatbot designed as an empathetic companion rather than a task assistant. - **Milagros Miceli** (1) — Sociologist and computer scientist; DAIR Institute and Weizenbaum Institute. Argentine researcher who studies the work of data annotators and content moderators in the Global South. https://www.weizenbaum-institut.de/en/ - **Minne Atairu** (1) — Interdisciplinary artist. Nigerian-born artist who uses generative AI to explore gaps in the archive of Benin art after the 1897 British raid. https://minneatairu.com/ - **Molleindustria** (1) — Independent game studio of Paolo Pedercini. Italian-founded studio making short political video games about labour, war and technology. https://www.molleindustria.org - **Monika Fleischmann** (1) — Media artist and researcher; co-founder of ART+COM. German media-art pioneer who, with Wolfgang Strauss, built early interactive installations and led research at GMD / Fraunhofer (MARS lab). - **Moon Ribas** (1) — Cyborg artist and choreographer; co-founder of the Cyborg Foundation. Dancer who feels earthquakes worldwide through online seismic sensors implanted in her body. https://www.moonribas.com/ - **Moritz Riesewieck** (1) — Director and author. German director and author; co-director of The Cleaners and Eternal You with Hans Block. https://en.wikipedia.org/wiki/The_Cleaners_(2018_film) - **Moritz Stefaner** (1) — Data visualisation designer. German designer known for data visualisation and interactive information design. https://truth-and-beauty.net - **Mushon Zer-Aviv** (1) — Designer, researcher and media activist. Mushon Zer-Aviv is a designer, educator and media activist who works on data, surveillance and algorithmic bias. https://mushon.com - **Myron Krueger** (1) — Computer artist and researcher. American computer artist (b. 1942) who built responsive environments from 1969 and coined the term 'artificial reality'. https://en.wikipedia.org/wiki/Myron_W._Krueger - **NASA Jet Propulsion Laboratory** (1) — NASA research centre for robotic space exploration. NASA's centre for robotic planetary missions; it built and operated the Mars Exploration Rovers Spirit and Opportunity. https://www.jpl.nasa.gov/ - **Nathaniel Rochester** (1) — Computer engineer, IBM (1919–2001). Chief architect of the IBM 701, IBM's first commercial scientific computer, and an early simulator of neural networks. - **Nazli Cila** (1) — Assistant professor, TU Delft. Design researcher working on human-agent partnerships and more-than-human design. https://nazlicila.com/ - **Near Future Laboratory** (1) — Design fiction studio (Julian Bleecker, Nick Foster, Fabien Girardin, Nicolas Nova). Studio that popularised design fiction, making props and documents from ordinary near futures. https://nearfuturelaboratory.com/ - **Neil Harbisson** (1) — Cyborg artist; co-founder of the Cyborg Foundation and Cyborg Arts. Artist born with achromatopsia who hears colour through an antenna implanted in his skull. https://www.cyborgarts.com/ - **Nicholas Carlini** (1) — Machine-learning security researcher. Computer-security researcher known for foundational work on adversarial examples, audio adversarial attacks and training-data extraction from neural networks. https://nicholas.carlini.com/ - **Nicholas Stedman** (1) — Artist and robotics designer. Canadian artist who builds modular, snake-like robots that touch and hold the human body. - **Nicolas Darrot** (1) — Artist making automata and robotic sculpture. French artist (born 1972) who builds electromechanical puppets and automata that mix animal, human and machine. - **Nicolas Schöffer** (1) — Artist, pioneer of cybernetic art. Hungarian-born French artist (1912–1992) who developed 'spatiodynamic' and cybernetic sculpture from the 1950s. - **Nissei Eco** (1) — Funeral-industry and plastics company. Japanese company that programmed SoftBank's Pepper robot to chant Buddhist sutras at funerals in 2017. - **Nora Al-Badri** (1) — Multi-disciplinary artist. German-Iraqi artist whose work on 'technoheritage' questions museum ownership of non-European artefacts and their data. https://www.nora-al-badri.de/ - **Norbert Wiener** (1) — Mathematician; founder of cybernetics (1894–1964). MIT mathematician who named the science of control and communication in animals and machines in 1948, and later warned about automation's social effects. - **ORLAN** (1) — Artist working with performance, surgery, biotechnology and robotics. French artist known for her surgery-performances of the 1990s and later work with cell culture, avatars and robots. https://www.orlan.eu - **OpenAI** (1) — AI research company. AI company founded in 2015, developer of GPT models and ChatGPT. https://openai.com/ - **Ory Laboratory** (1) — Robotics company founded by Kentaro Yoshifuji. Maker of the OriHime avatar robot, which lets people who cannot leave home, including people with ALS, work and be present remotely. https://dawn2021.orylab.com/en/ - **Osamu Tezuka** (1) — Manga artist and animator. Manga artist (1928–1989) who created Astro Boy (Tetsuwan Atom), serialised from 1952. - **Oscar Sharp** (1) — Filmmaker. British filmmaker who directed films scripted by the neural network Benjamin. https://en.wikipedia.org/wiki/Sunspring - **Pablo Berger** (1) — Film director. Spanish writer-director of Blancanieves and the animated feature Robot Dreams. https://en.wikipedia.org/wiki/Pablo_Berger - **Pablo Ventura** (1) — Choreographer; director of Ventura Dance Company. Spanish choreographer who has worked with industrial robots, software and machine performers since around 2000. https://www.ventura-dance.com/ - **Paola Torres Núñez del Prado** (1) — Artist. Paola Torres Núñez del Prado is a Peruvian artist who combines textiles, sound and technology, drawing on Andean knowledge systems. https://www.singingtextiles.com - **Pat Metheny** (1) — Jazz guitarist and composer. American guitarist who toured with an orchestra of mechanically played acoustic instruments. https://www.patmetheny.com/ - **Paul Sermon** (1) — Artist; Professor of Visual Communication, University of Brighton. British artist who has made telematic installations joining remote bodies in shared video spaces since 1991. https://www.paulsermon.org/ - **Paul Vanouse** (1) — Artist; Professor, University at Buffalo. American artist who works with biotechnology to question scientific authority, identity and labour. https://www.paulvanouse.com/ - **Paul Virilio** (1) — Cultural theorist and urbanist (1932–2018). French theorist of speed, war and perception, author of Speed and Politics (1977) and War and Cinema (1984). - **Pedro Lopes** (1) — Associate Professor of Computer Science, University of Chicago; director of the Human Computer Integration Lab. Pedro Lopes studies devices that integrate with the user's body, such as electrical muscle stimulation that moves the wearer's limbs. https://plopes.org - **Pedro Oliveira** (1) — Sound artist and researcher; guest faculty, Sound Studies and Sonic Arts, Universität der Künste Berlin. Brazilian sound artist and researcher (born 1985, São Paulo) whose work examines how listening shapes and is shaped by technical infrastructures, focusing on voice and migration; founding member of Decolonising Design. https://oliveira.work/ - **Peng! Collective** (1) — Activist art collective. Berlin collective that combines culture jamming, hacking and civil disobedience. https://pen.gg - **Peter-Paul Verbeek** (1) — Philosopher of technology; Rector Magnificus, University of Amsterdam. Dutch philosopher who developed the theory of technological mediation in What Things Do (2005) and Moralizing Technology (2011). - **Peterskapelle Lucerne & HSLU Immersive Realities Research Lab** (1) — Catholic chapel and university research lab. St. Peter's Chapel in Lucerne and the Immersive Realities Research Lab of the Lucerne University of Applied Sciences and Arts, who staged the AI Jesus installation Deus in Machina in 2024. https://www.peterskapelle.ch/ - **Petra Gemeinboeck** (1) — Artist and researcher in human–robot movement; Swinburne University of Technology. Austrian-Australian artist who develops robot bodies and movement with dancers, often with Rob Saunders. - **Phil D. Hall** (1) — Conversational AI designer; co-creator of I Am Echoborg. Chatbot designer who writes the conversational agent behind I Am Echoborg. https://echoborg.com/ - **Philip Beesley** (1) — Architect and artist; professor at the University of Waterloo; director of the Living Architecture Systems Group. Philip Beesley builds immersive responsive environments that behave like near-living systems. https://www.philipbeesleystudioinc.com - **Philip van Allen** (1) — Professor, ArtCenter College of Design. American interaction designer who builds animistic prototypes and toolkits for designing with AI. https://www.philvanallen.com/ - **Philippe Parreno** (1) — Artist. French artist who stages exhibitions as automated, time-based systems driven by sensors and computers. https://en.wikipedia.org/wiki/Philippe_Parreno - **Pierre Jaquet-Droz** (1) — Watchmaker and automaton maker. Swiss watchmaker (1721–1790) who with his son Henri-Louis and Jean-Frédéric Leschot built the Writer, the Draughtsman and the Musician. - **Piotr Mirowski** (1) — AI researcher and improviser; co-founder of Improbotics. Research scientist in machine learning who performs improvised theatre with AI systems. https://improbotics.org/ - **Pipeline Theatre** (1) — Theatre company. British theatre company that made Spillikin with the humanoid robot company Engineered Arts. https://www.pipelinetheatre.com/ - **Primavera De Filippi** (1) — Legal scholar and artist; researcher at CNRS and Harvard Berkman Klein Center. Italian legal scholar and artist who studies blockchain governance and made the Plantoid, a self-owning robotic 'plant'. - **Project CETI** (1) — Nonprofit interdisciplinary research initiative. Initiative founded in 2020 by David Gruber and others that uses machine learning and robotics to study sperm whale communication. https://www.projectceti.org/ - **Prokop Bartoníček** (1) — Designer and artist. Czech designer and artist working with interactive systems, installations and machines. http://www.prokopbartonicek.com - **Protestant Church in Hesse and Nassau (EKHN)** (1) — Regional Protestant church in Germany. A regional church of the Evangelical Church in Germany; for the 500th anniversary of the Reformation it installed the blessing robot BlessU-2 in Wittenberg. https://www.ekhn.de/ - **Qian Yang** (1) — Assistant professor of information science, Cornell University. HCI researcher who studies how to design human-AI interaction, with a focus on healthcare. - **Quantic Dream** (1) — Video game developer. French studio known for performance-captured interactive dramas such as Heavy Rain and Detroit: Become Human. https://www.quanticdream.com/ - **Rachel Ara** (1) — Conceptual and data artist. London artist who worked 25 years as a computer systems designer before making neon and data sculptures about value, gender and tech. - **Radamés Ajna** (1) — Artist and technologist. Brazilian artist and technologist working with electronics, networks and machine learning. - **Radhika Nagpal** (1) — Computer scientist; professor, Princeton University (formerly Harvard). Indian-American roboticist who studies collective intelligence in robot swarms and social insects. - **Raffaello D'Andrea** (1) — Roboticist; Professor, ETH Zurich; co-founder of Kiva Systems. Canadian-Italian engineer known for autonomous systems, from Kiva warehouse robots to flying machines, and for art collaborations with Max Dean. - **Rashaad Newsome** (1) — Artist. American artist working in collage, performance, film and AI, drawing on Black and queer culture such as ballroom and vogue. https://rashaadnewsome.com/ - **Rebecca Fiebrink** (1) — Professor, UAL Creative Computing Institute. American computer scientist and musician who develops interactive machine learning tools for artists. http://www.wekinator.org/ - **Renwu magazine** (1) — Magazine (reporter Lai Youxuan). Chinese feature magazine whose 2020 investigation into food-delivery riders became one of the most-read Chinese articles of the year. - **Revital Cohen** (1) — Artist and designer (Cohen Van Balen). Israeli-born London artist who works with medical devices, animals and supply chains, often with Tuur Van Balen. https://www.cohenvanbalen.com/ - **Richard Mosse** (1) — Artist and photographer. Irish artist who uses military and scientific imaging technologies, such as infrared film and thermal cameras, to document conflict and migration. https://www.richardmosse.com - **Ridley Scott** (1) — Film director. British director of Alien and Blade Runner. https://en.wikipedia.org/wiki/Ridley_Scott - **Rik Lander** (1) — Artist and writer; co-creator of I Am Echoborg. British interactive storyteller who makes participatory works about AI. https://echoborg.com/ - **Rimini Protokoll** (1) — Theatre collective (Helgard Haug, Stefan Kaegi, Daniel Wetzel). German-Swiss collective known for documentary theatre with 'experts of the everyday' instead of actors. https://www.rimini-protokoll.de/ - **Rob Saunders** (1) — Computational creativity researcher; Leiden University. British researcher in computational creativity and curious agents, long-time collaborator of Petra Gemeinboeck. - **Robert Breer** (1) — Experimental filmmaker and kinetic sculptor. American experimental filmmaker and sculptor (1926–2011) who made slowly creeping motorised 'Floats' from the mid-1960s. - **Robert Long** (1) — Philosopher; executive director, Eleos AI. Philosopher working on AI consciousness, sentience and moral status. https://eleosai.org/ - **Robert Sparrow** (1) — Professor of Philosophy, Monash University. Australian philosopher who writes on the ethics of robots, military AI, aged care and human enhancement. - **Robert Wood** (1) — Roboticist; professor, Harvard University. American roboticist working on soft and micro robots at Harvard SEAS and the Wyss Institute. - **Rocio von Jungenfeld** (1) — Artist; Lecturer in Digital Arts, University of Kent. Artist working with site-specific interactive installations, light and electronics. - **Rolf Pfeifer** (1) — Professor Emeritus of Computer Science, University of Zurich; founder of the AI Lab. Swiss roboticist who led embodied-AI research on morphological computation and soft, compliant robot bodies. - **Romi Ron Morrison** (1) — Artist and scholar. American interdisciplinary artist and scholar working on Black geographies, data and surveillance. - **Rosa Menkman** (1) — Artist and theorist. Dutch artist and theorist of glitch, compression and resolution, author of The Glitch Moment(um) (2011). https://beyondresolution.info - **Ruairi Glynn** (1) — Architect and artist; director of the Interactive Architecture Lab, The Bartlett, UCL. Irish-born designer of robotic and responsive architecture. https://www.ruairiglynn.co.uk - **Ruben van de Ven** (1) — Artist and researcher. Dutch artist-researcher who examines emotion recognition and the politics of algorithmic seeing. https://rubenvandeven.com - **Ruha Benjamin** (1) — Professor of African American Studies, Princeton University. Sociologist who founded the Ida B. Wells Just Data Lab and wrote Race After Technology (2019) and Viral Justice (2022). - **SAND Lab (University of Chicago)** (1) — Security, algorithms and networks lab led by Ben Zhao and Heather Zheng. Research lab that builds defensive tools for individuals against machine learning, such as Glaze and Nightshade. https://sandlab.cs.uchicago.edu/ - **Safiya Umoja Noble** (1) — Professor of Gender Studies, African American Studies and Information Studies, UCLA. Internet studies scholar, 2021 MacArthur Fellow and founder of the UCLA Center on Race and Digital Justice. - **Sam Kriegman** (1) — Roboticist; assistant professor, Northwestern University. American roboticist who designs bodies with evolutionary algorithms; first author of the 2020 xenobot paper. - **Sam Lavigne** (1) — Artist and educator. American artist who works with scraping, surveillance tools and machine learning, often turning them back on the powerful. https://lav.io/ - **Sarah Petkus** (1) — Artist and robot maker. Sarah Petkus is an American artist, illustrator and engineer who builds characterful robots from light metals, 3D-printed parts and found objects. https://www.youtube.com/GravityRoad - **Sašo Sedlaček** (1) — Artist. Slovenian artist whose low-tech machines and interventions address waste, poverty and techno-optimism. https://sasosedlacek.com - **Scatter Lab** (1) — AI chatbot start-up. Korean start-up behind the chatbot Lee Luda (2020), trained on messenger conversations from its dating-advice apps. https://www.scatterlab.co.kr/ - **Shalini Kantayya** (1) — Documentary filmmaker; 7th Empire Media. American filmmaker whose documentaries address science, technology and civil rights. https://www.7thempire.com/ - **Sherry Turkle** (1) — Abby Rockefeller Mauzé Professor of the Social Studies of Science and Technology, MIT. Sociologist and clinical psychologist who founded the MIT Initiative on Technology and Self, author of The Second Self (1984) and Alone Together (2011). https://sherryturkle.mit.edu - **Shing Tat Chung** (1) — Artist and designer. British artist and designer, Royal College of Art graduate, working on belief and irrationality in technological systems. - **Shuya Abe** (1) — Electrical engineer; Paik's technical collaborator. Japanese engineer (1932–2016) who built Robot K-456 and the Paik–Abe video synthesizer with Nam June Paik. - **Siddharth Suri** (1) — Senior Principal Researcher, Microsoft Research. Computational social scientist who studies crowdwork and online labour markets. - **Silvia Ruzanka** (1) — Artist and Lecturer, Rensselaer Polytechnic Institute. Silvia Ruzanka is a media artist who makes games and glitch art, including a videogame played by a Venus flytrap. - **Simon Weckert** (1) — Artist. Berlin-based artist whose performances and installations probe the social effects of digital technologies and platforms. http://www.simonweckert.com - **Simone Giertz** (1) — Inventor and YouTuber. Swedish inventor (born 1990) known for deliberately useless 'shitty robots' on YouTube. https://www.simonegiertz.com/ - **Simone Rebaudengo** (1) — Designer; co-founder of automato.farm. Italian interaction designer who imagines connected products with needs, habits and social lives. http://www.automato.farm/about - **Sofian Audry** (1) — Artist; Professor, École des médias, Université du Québec à Montréal. Canadian artist and scholar who makes robotic and machine-learning artworks. https://sofianaudry.com - **Sonia Kennebeck** (1) — Documentary director and producer. German-American investigative filmmaker. https://nationalbirdfilm.com/ - **Spike Jonze** (1) — Film director and screenwriter. American director of music videos and feature films, including Being John Malkovich (1999) and Her (2013). https://en.wikipedia.org/wiki/Spike_Jonze - **Sputniko! (Hiromi Ozaki)** (1) — Artist and designer; Associate Professor, Tokyo University of the Arts. Speculative designer whose films and devices, such as Menstruation Machine, question technology through gender. - **Squarepusher** (1) — Musician (Tom Jenkinson). British electronic musician and bassist on Warp Records. https://squarepusher.net/ - **Stahl Stenslie** (1) — Artist and researcher in haptic and experimental media. Norwegian artist who has made tele-haptic and full-body interfaces since the early 1990s. https://en.wikipedia.org/wiki/Stahl_Stenslie - **Stanisław Lem** (1) — Writer and philosopher (1921–2006). Polish science-fiction writer whose novels and essays think through non-human intelligence. https://en.wikipedia.org/wiki/Stanis%C5%82aw_Lem - **Stanley Kubrick** (1) — Film director (1928–1999). American film director who worked in Britain; 2001: A Space Odyssey was written with Arthur C. Clarke. https://en.wikipedia.org/wiki/Stanley_Kubrick - **Steffen Weiß** (1) — Designer. German designer who co-authored The Chair Project with Philipp Schmitt. - **Steina Vasulka** (1) — Video artist. Icelandic-born video pioneer who co-founded The Kitchen in New York with Woody Vasulka in 1971. http://www.vasulka.org - **Steph Maj Swanson** (1) — Artist and writer (Supercomposite). Musician and artist, known online as Supercomposite, who documented the 'Loab' figure in 2022. https://en.wikipedia.org/wiki/Loab - **Stephan Bogner** (1) — Designer. German interaction designer (HfG Schwäbisch Gmünd) and co-author of Raising Robotic Natives. https://stephanbogner.de/ - **Sterling Crispin** (1) — Artist and technologist. American artist and technologist whose early work reverse-engineered face detection and recognition systems; he later worked on AR/VR and neurotechnology. https://www.sterlingcrispin.com - **Steve Mann** (1) — Inventor; Professor of Electrical and Computer Engineering, University of Toronto. Researcher who has worn computers and head-mounted cameras in daily life since the late 1970s and coined 'sousveillance'. http://wearcam.org/ - **Steven Spielberg** (1) — Film director and producer. American director and producer; he completed A.I. Artificial Intelligence (2001) from a project developed by Stanley Kubrick. https://en.wikipedia.org/wiki/Steven_Spielberg - **Super Brain** (1) — AI studio (founder Zhang Zewei). Nanjing studio that sells AI avatars of deceased people, built from photos, videos and voice recordings supplied by families. - **Surveillance Camera Players** (1) — Performance group founded by Bill Brown. Activist troupe that staged silent plays in front of public CCTV cameras from 1996. http://www.notbored.org/the-scp.html - **Survival Research Labs** (1) — Machine-performance group founded by Mark Pauline in 1978. San Francisco group that stages large, dangerous performances in which custom-built machines fight and destroy each other. https://srl.org/ - **Susumu Tachi** (1) — Professor Emeritus, University of Tokyo; founder of the Virtual Reality Society of Japan. Engineer who proposed 'telexistence' in 1980 and built the TELESAR robots that let a person feel present in a remote robot body. https://tachilab.org/en/ - **Suzanne Kite** (1) — Artist, composer and scholar (Oglála Lakȟóta); Director, Wihanble S'a Lab, Bard College. Oglála Lakȟóta artist and scholar whose performances and machine-learning works draw on Lakota ontology and relations with stones and non-humans. https://kitekitekitekite.com/ - **Suzanne Treister** (1) — Artist. British artist whose fictional archives map technology, finance, the occult and consciousness. http://www.suzannetreister.net/ - **THEaiTRE** (1) — Research and theatre project (Charles University, Švandovo divadlo, DAMU). Czech project that used language models to write theatre plays, starting with the R.U.R. centenary in 2021. https://www.theaitre.com/ - **Tabita Rezaire** (1) — Artist and healer. French-Guianese-Danish artist whose videos connect digital networks, colonial history and spiritual technologies. https://tabitarezaire.com/ - **Tactical Tech** (1) — International NGO on technology and society. Non-profit that runs public exhibitions and toolkits on data, privacy and influence. https://tacticaltech.org/ - **Takanori Shibata** (1) — Chief Senior Research Scientist, AIST. Engineer who developed PARO, the seal robot used in dementia and elder care since 2004. http://www.parorobots.com/ - **Te Hiku Media** (1) — Māori-led media organisation and language-technology lab. Iwi radio and media organisation in the Far North that builds te reo Māori speech technology on data its communities own. https://tehiku.nz/ - **Ted Chiang** (1) — Science-fiction writer. American science-fiction writer of Chinese descent (Stories of Your Life and Others, Exhalation). - **Temple Grandin** (1) — Animal scientist and inventor; Professor, Colorado State University. Autistic scientist and author who designed humane livestock-handling systems and, as a young woman, a pressure machine to calm herself. https://www.templegrandin.com - **The Long Now Foundation** (1) — Nonprofit for long-term thinking. Foundation established in 1996 by Stewart Brand, Danny Hillis and others to foster thinking on a 10,000-year scale. https://longnow.org/ - **Theo Anthony** (1) — Filmmaker. American documentary filmmaker whose films question how images and instruments claim objectivity. https://en.wikipedia.org/wiki/Theo_Anthony - **Theo Jansen** (1) — Artist. Dutch artist who has built wind-walking 'beach animals' from PVC tubing since 1990. https://www.strandbeest.com/ - **Thiago Hersan** (1) — Artist and engineer. Brazilian-American artist who makes installations about how machines and networks represent people. https://thiagohersan.com - **Thomas Freundlich** (1) — Dancer, choreographer and filmmaker. Finnish-American choreographer who has made dance works for industrial robots since 2008. http://www.freundlich.org/ - **Thomas Melle** (1) — Writer. German novelist and playwright who wrote about his bipolar disorder in The World at My Back. - **Timnit Gebru** (1) — Founder and Executive Director, Distributed AI Research Institute (DAIR). Ethiopian-born computer scientist, co-founder of Black in AI and former co-lead of Google's Ethical AI team. - **Timo Arnall** (1) — Designer and filmmaker. British designer whose films and research make invisible technologies such as RFID, satellites and data centres visible. https://www.elasticspace.com/ - **Timo Toots** (1) — Media artist. Estonian media artist whose installations turn state and online databases about visitors into public displays. http://works.timo.ee - **Tobias Rees** (1) — Philosopher and anthropologist of AI. Philosopher and anthropologist who directed the Transformations of the Human program at the Berggruen Institute, working with AI labs on the philosophical stakes of their research. - **Tod Machover** (1) — Composer; Professor, MIT Media Lab (Opera of the Future). American composer who invented hyperinstruments and makes technology-driven operas. https://www.media.mit.edu/people/tod/overview/ - **Tom B. Brown** (1) — Machine-learning researcher (Google Brain, later OpenAI and Anthropic). Machine-learning researcher who led the 2017 Google Brain paper on the adversarial patch and later led work on large language models. - **Tom Shannon** (1) — Artist and inventor. American artist (born 1947) who, as a young artist in 1966, built Squat, a robot controlled by a living plant. - **Tom White** (1) — Artist (dribnet); Senior Lecturer, Victoria University of Wellington School of Design. New Zealand-based artist and design lecturer, known online as dribnet, who makes drawings and prints optimised to be read by neural-network classifiers. https://drib.net - **Tomas van Houtryve** (1) — Photographer and artist. Belgian-American photographer who works on conflict, surveillance and history. https://tomasvh.com/ - **Tomasz Hollanek** (1) — Research Fellow, Leverhulme Centre for the Future of Intelligence, University of Cambridge. Design and technology ethicist whose research connects critical design, speculative scenarios and AI ethics. https://www.lcfi.ac.uk/ - **Tomo Kihara** (1) — Designer and artist; co-founder of Playfool. Japanese designer and artist who makes playful games and installations about AI, platforms and the city, often with Daniel Coppen as the studio Playfool. https://www.tomokihara.com - **Tore Knudsen** (1) — Interaction designer and researcher. Danish interaction designer and CIID alumnus working on tangible interfaces for machine learning and smart-home technology. - **UBERMORGEN** (1) — Artist duo (lizvlx and Hans Bernhard). Austrian-Swiss artist duo working since 1995 with networks, media hacks and machine systems. - **USC Shoah Foundation** (1) — Institute for Visual History and Education, University of Southern California. Institute founded by Steven Spielberg in 1994 that holds more than 50,000 video testimonies of Holocaust survivors and witnesses of other genocides. https://sfi.usc.edu/ - **Unitree Robotics** (1) — Robotics company. Hangzhou maker of quadruped and humanoid robots (Go, H1, G1). https://www.unitree.com - **Unknown Fields Division** (1) — Nomadic design studio led by Liam Young and Kate Davies. Design research studio that led expeditions to the mines, factories and ports behind consumer technology. http://www.unknownfieldsdivision.com/ - **Ursula Damm** (1) — Artist; Professor of Media Environments, Bauhaus-Universität Weimar. Ursula Damm is a German media artist whose generative video installations use live camera input and neural networks to explore perception, urban space and nature. https://ursuladamm.de - **Valentino Braitenberg** (1) — Neuroscientist and cyberneticist (1926–2011). Italian neuroscientist and director at the Max Planck Institute for Biological Cybernetics, author of Vehicles: Experiments in Synthetic Psychology (1984). - **Verena Friedrich** (1) — Media artist. German artist who builds precise machines and lab setups that deal with life extension, measurement and automated judgement. https://heavythinking.org - **Victor Wong** (1) — Artist and visual-effects producer. Hong Kong visual-effects producer (vfxNova) who created the AI ink painter A.I. Gemini in 2018. https://www.3812gallery.com/artists/victor-wong-ai-gemini/ - **Vilém Flusser** (1) — Philosopher of media (1920–1991). Czech-born Brazilian philosopher who wrote on apparatuses, technical images and the programs that run through cameras and their users. - **Vincent Fournier** (1) — Photographer and artist. French photographer whose series look at space exploration, robots and imagined future life forms. https://www.vincentfournier.co.uk/ - **Waltz Binaire** (1) — Design and media art studio (Christian Mio Loclair). Berlin studio led by Christian Mio Loclair that makes installations with machine learning and motion. https://christianmioloclair.com - **Wang Yuyang** (1) — Artist. Beijing artist working with light, sculpture and technology. - **Wayne McGregor** (1) — Choreographer; Artistic Director, Studio Wayne McGregor; Director of Dance, Venice Biennale. British choreographer who has worked with cognitive scientists and technologists on how movement is made. https://waynemcgregor.com/ - **Wen-Ying Tsai** (1) — Artist and engineer (1928–2013). Chinese-American engineer turned artist, a pioneer of cybernetic sculpture from 1968. - **Wendy Hui Kyong Chun** (1) — Canada 150 Research Chair in New Media, Simon Fraser University. Media theorist and founding director of the Digital Democracies Institute; author of Programmed Visions and Updating to Remain the Same. - **William Chamberlain** (1) — Writer and programmer. American writer who, with Thomas Etter, created the prose generator Racter in the early 1980s. https://en.wikipedia.org/wiki/Racter - **Wolfgang Strauss** (1) — Architect and media artist. German architect and media artist, long-time collaborator of Monika Fleischmann on interactive environments. - **Xenoangel (Marija Avramovic & Sam Twidale)** (1) — Artist duo making real-time animations and AI-driven virtual worlds. Marija Avramovic and Sam Twidale work as Xenoangel, building game-engine worlds whose characters act on their own. https://www.samtwidale.com/ - **Xu Bing** (1) — Artist; former Vice President, Central Academy of Fine Arts. Chinese artist known for works on language and writing systems (Book from the Sky), who turned to surveillance footage for his first feature film. https://www.xubing.com - **YACHT** (1) — Band (Claire L. Evans, Jona Bechtolt, Rob Kieswetter). American electronic pop band that made an album with machine learning under strict self-imposed rules. https://teamyacht.com/ - **Yasuhiro Yoshiura** (1) — Animation director. Director of Time of Eve (2008–09), an anime about a café where humans and androids are not to be told apart. - **Yeon Sang-ho** (1) — Film director. South Korean director of Train to Busan and the Netflix film Jung_E. https://en.wikipedia.org/wiki/Yeon_Sang-ho - **Zhang Yimou** (1) — Film director. Chinese film director and director of large state spectacles, including the 2008 Beijing Olympic opening ceremony. - **Zimoun** (1) — Artist. Swiss artist (born 1977) who builds large installations of hundreds of small DC motors striking cardboard, wire and cotton. https://www.zimoun.net/ - **Zoran Srdić Janežič** (1) — Sculptor and intermedia artist; PhD researcher at AGRFT, University of Ljubljana. Slovenian sculptor who works with animatronics, 3D design and living tissue; since 2019 he has led the Biobot series with Kapelica Gallery and BioTehna Lab. - **etoy.CORPORATION** (1) — Artist group organised as a corporation. Swiss-based artist group founded in 1994 that uses corporate, legal and digital structures as material; known for Digital Hijack (Golden Nica 1996) and Toywar. http://missioneternity.org/ - **unmake lab** (1) — Artist collective. Korean artist collective, including Choi Binna, that studies how machine vision constructs nature and society. - **Špela Petrič** (1) — Artist with a background in biology. Špela Petrič is a Slovenian artist and biologist whose performances explore relationships between humans, plants and machines. https://www.spelapetric.org