AI Art Strategies 2026

AI Art Strategies 2026

Tomáš Marušiak
Faculty of Arts, Masaryk University, Brno

The course AI Art Strategies offers a theoretical, critical and creative examination of contemporary forms of artificial intelligence art. It is divided into two interconnected components. The theoretical component consists of lectures and creative discussions with students, focusing on current strategies, problems and trajectories in AI Art. In the creative component, participants respond to the knowledge acquired by proposing an original project, which may take the form of a completed artwork, a conceptual proposal or a speculative artistic research strategy.

In 2026, students may choose one of two thematic areas: The Intimacy of an Agentic AI Entity or The Life Cycle of an Agentic AI Entity. They may also develop an original project of their own, provided that it is consistent with the course’s conceptual and research framework. Both components aim to stimulate engagement with key questions concerning AI Art strategies while developing divergent and convergent forms of thinking, creative decision-making and critical reflection. These capacities are already essential for navigating the rapidly changing research and cultural environment of the AI Artworld.

AI Art Strategies 2026

Tomáš Marušiak
Faculty of Arts, Masaryk University, Brno

I.What Is AI Art? Historical, Evolutionary and Cultural Foundations

This section of the syllabus asks whether AI Art has a history and future of its own, or whether its history is constructed retrospectively by connecting cybernetic, algorithmic, generative and evolutionary art. Artificial intelligence is approached not only as a technology but also as a cultural projection of human ideas about language, intelligence, creativity and authorship. The syllabus distinguishes the technical history of AI from the broader cultural genealogy of artificial beings, while also examining the shift from the closed artefact to the adaptive ‚growing image‘. Drawing on Manovich’s proposals, it analyses AI Art through strategies involving data, models, training, human control, feedback, curating and infrastructure. Bio-AI is used as a philosophical framework for non-biological computational agency, not as a claim that current AI is alive or conscious. These premises are explored through the work of Refik Anadol, Mario Klingemann and Lauren Lee McCarthy.

Does AI Art Have Its Own History and Future?

Kanta Dihal, How the World Sees Intelligent Machines. Video: Alexander von Humboldt Institute for Internet and Society.

AI Art cannot be reduced to the use of a single model, medium or technological procedure. It is nevertheless necessary to distinguish between the cultural genealogy of artificial intelligence and its technical history. Myths of artificial beings, mechanical automata and literary visions of intelligent machines belong to the cultural genealogy of AI rather than to the history of artificial intelligence in its present technical sense. Its technical history was shaped primarily by cybernetics, symbolic AI, expert systems, robotics and machine learning. The history of AI Art partly overlaps with these developments, but it also encompasses distinct traditions of computer, algorithmic, generative, interactive and evolutionary art (Cave et al., 2020; Taylor, 2014; Whitelaw, 2004).

Rather than presenting a single linear history culminating in today’s generative models, it is more accurate to trace several parallel lines of development. These include cybernetic art experiments, cellular automata, artificial life, evolutionary algorithms, neural networks, generative adversarial networks, diffusion models, multimodal systems and contemporary agentic technologies. Not all these methods were originally described as artificial intelligence, and not every generative or algorithmic work is automatically AI Art. Machine learning constitutes only one area within the broader research field of artificial intelligence (Manovich, 2019; Russell & Norvig, 2021).

Language, evolution and the case of FOXP2

This framework also raises the question of the evolutionary conditions underlying human language, symbolic representation and creativity. The FOXP2 gene offers an important example of biological mechanisms associated with speech development, motor planning and linguistic abilities. It cannot, however, be described as a single ‚language gene‘, nor can its mutations be said to have directly produced abstract thought, creativity or art. Language emerged through a complex interaction of biological evolution, embodiment, social learning and cultural environment (Fisher & Scharff, 2009; Fitch, 2010).

Simon E. Fisher, Extreme Language: The Genes That Let Us Speak. Video: New Scientist.

The case of FOXP2 therefore does not serve as a biological model of artificial intelligence. Rather, it provides a critical contrast between human intelligence, shaped by evolution and embodiment, and computational systems that learn through data and optimisation. Whether artificial systems can develop their own forms of symbolic and aesthetic representation, or merely transform representations derived from human culture, remains an open research question.

Artificial intelligence as a cultural imaginary

AI is not only a technical project but also a cultural imaginary and a philosophical problem. Mythical beings, mechanical automata, robots, cybernetic organisms and contemporary virtual agents function as a screen onto which human ideas about intelligence, consciousness, authorship and creativity are projected. Cultural narratives shape not only public understandings of AI but also the direction of its development, regulation and social use (Cave et al., 2020).

It is therefore essential to distinguish the technical capabilities of a system from the meanings, intentions or subjectivity attributed to it. A system may produce culturally persuasive images, texts or performances without possessing consciousness, aesthetic experience or a creative intention comparable to that of a human being.

From the static artwork to the ‚growing image‘

The syllabus introduces the working concept of the ‚growing image‘. This is not a generally accepted scientific category, but an analytical concept informed by generative, process-based and interactive art (Boden & Edmonds, 2009; Whitelaw, 2004). It describes an image-based or multimedia process that changes over time and responds to data, its environment, its audience or its own previous states.

The growing image is not ‚alive‘ in a biological sense. It is an open aesthetic system based on recursion, feedback and continuous transformation. The artwork no longer operates solely as a closed object, but as an event, an adaptive process or a temporary configuration of relations among the artist, the system, data, the audience and the environment. This does not imply the general disappearance of the static artwork. Rather, it identifies one significant tendency within contemporary AI Art.

What Are AI Art Strategies?

Lev Manovich (2019) proposes three definitions of AI Art, which this syllabus provisionally terms the institutional, processual and expansive proposals. These are not Manovich’s own labels, and the first proposal should not be equated without qualification with Dickie’s institutional theory of art.

According to the first proposal, AI endowed with a degree of autonomy produces artefacts or experiences that experts accept as historical or contemporary art. Manovich compares this model to an extension of the Turing test into artistic production. The second proposal shifts attention from the resulting artefact to the process by which it is produced. What matters is not only the type of technology employed, but also the degree and form of human control over the system’s architecture, training data and the selection of outputs. The third proposal defines AI Art as art that humans cannot create because of the limitations of their bodies, brains and existing aesthetic systems. On this account, AI should not merely imitate human styles but should help to explore previously unknown yet still meaningful forms of aesthetic organisation (Manovich, 2019).

In the analysis of AI Art, attention to style should therefore be complemented by attention to strategy. Style remains a relevant property of an artwork, but on its own it cannot account for the work’s data-related, processual, interactive and infrastructural conditions. This approach can be connected to Flusser’s understanding of the technical image as the product of an apparatus and its program. Flusser did not formulate a theory of AI Art strategies, but his philosophy makes it possible to examine how artists work with technological apparatuses, disclose their possibilities and intervene critically in their programs (Flusser, 2000).

AI Art strategies may provisionally be defined as organised aesthetic processes in which decision-making is distributed among the artist, an AI system, data, an algorithmic protocol, an interface, technological infrastructure, an audience and an environment. A strategy may encompass the provenance and selection of data, model architecture and training, prompting, feedback, selection, curating, interaction, modes of presentation, and ethical and environmental decision-making.

Within these processes, AI may operate as a generator, interpreter, collaborator, selector, predictive system, object of critical inquiry or agent with bounded operational autonomy. Such computational agency is not evidence of consciousness, intentionality or complete system autonomy.

Bio-AI and two modes of aesthetic ecosystem organisation

The term Bio-AI is used as a philosophical and relational concept, not as a claim that contemporary AI is a biological organism. Zylinska (2025) describes AI as a ‚biomachine‘ embedded in the technological, social and ecological conditions of life. She does not, however, attribute metabolism, biological reproduction or consciousness to it.

Rather than referring to ‚two biologies‘, it is therefore more precise to distinguish two modes of aesthetic ecosystem organisation. The biological mode encompasses humans, animals, plants and other organisms grounded in embodiment, metabolism and biological evolution. The computational mode encompasses systems based on data, algorithms, networked infrastructure, optimisation and feedback. Contemporary AI Art extends the historical tendency to delegate aspects of aesthetic decision-making to non-biological computational processes. At the same time, it remains dependent on human labour, natural resources, energy and economic infrastructure (Crawford, 2021; Zylinska, 2020, 2025).

Case studies in artistic strategies

Refik Anadol, Engram: Data Sculpture for Melting Memories, 2018. Video: Refik Anadol Studio.

Refik Anadol: from neurophysiological data to the ecological archive. In Melting Memories, Anadol transforms EEG data recorded during memory recall into data paintings, projections and spatial visualisations. The project employs algorithmic processing, including recurrent neural networks, and therefore lies at the intersection of data art, neuroaesthetic visualisation and AI Art (Anadol, 2018). Echoes of the Earth: Living Archive uses AI-generated images based on data concerning flora, fauna and fungi gathered from several rainforest locations (Anadol, 2024). The phrase ‚living archive‘ should be understood metaphorically. Critical analysis must attend to data provenance, the representation of nature, and the material and environmental costs of computational infrastructure (Zylinska, 2025).

Refik Anadol, Echoes of the Earth: Living Archive, Serpentine North, 2024. Video: Refik Anadol Studio.

Mario Klingemann: from participatory training to machine criticism. The installation Circuit Training involves its audience in data acquisition, the evaluation of generated outputs and the model’s ongoing development. Visitors become part of its training and curatorial feedback loop (Klingemann, 2019). In A.I.C.C.A., a robotic dog locates and analyses works of art and uses a language model to produce short critical texts (Klingemann, 2023). It is not an autonomous critic in the strong sense, but a performative robotic installation with bounded operational autonomy. The work parodies the authority of the critic and the mechanisms of evaluation, attention and institutional recognition in art.

Mario Klingemann, Circuit Training, 2019. Interactive installation. Video: Onkaos.

Mario Klingemann, A.I.C.C.A. (Artificially Intelligent Critical Canine), 2023. Performative robotic installation. Video: Onkaos.

Lauren Lee McCarthy: communication beyond algorithmic legibility. Unlearning Language, created by Lauren Lee McCarthy, Kyle McDonald and YCAM, uses machine learning to detect participants‘ speech, gestures and expressions. When the system recognises their communication, it prompts the group to seek new ways of communicating that the algorithm cannot identify (McCarthy et al., 2022). Here, AI does not operate as an image generator, but as a mechanism of surveillance and organisation within a performative situation. The artistic strategy explores resistance to algorithmic monitoring and creates space for collective forms of communication and human agency.

Lauren Lee McCarthy and Kyle McDonald, Unlearning Language, 2022. Created in collaboration with Yamaguchi Center for Arts and Media [YCAM].

II.Key Voices in the Philosophy of AI Art

AI Art cannot be understood simply as the use of a new tool, because artificial intelligence systems also transform creative processes, authorship, aesthetic selection and the distribution of images. Benjamin helps to explain how technological reproduction changes the value and reception of a work; McLuhan shows that a medium actively shapes human perception; and Flusser argues that a technical image emerges within the possibilities of an apparatus and its program (Benjamin, 2008; Flusser, 2000; McLuhan, 1964). These foundations also matter in the context of generative AI, because an image is produced not only by an individual’s intention, but also by data, models, interfaces, institutions and cultural norms. This section of the lecture series therefore introduces Manovich’s theory of AI as cultural and aesthetic infrastructure, Zylinska’s posthumanist analysis of nonhuman perception and Bio-AI, and further approaches to creativity, authorship, technodiversity, neuroaesthetics and the material conditions of AI. Its purpose is not to proclaim AI a conscious artist, but to examine critically how it changes the conditions of artistic production and aesthetic experience.

Philosophical and Media-Theoretical Foundations

Critical posthumanism challenges the autonomous human subject as the sole centre of knowledge and creativity, emphasising instead its embodied, technological and environmental relations (Braidotti, 2013; Hayles, 1999). Transhumanism focuses more directly on the technological enhancement of human capacities (Bostrom, 2005). This distinction raises a fundamental question: does AI extend human creativity, or does it participate in the production of different modes of perception and decision-making?

Benjamin’s analysis of technological reproducibility raises questions of aura, originality, mass reception and the politics of the image (Benjamin, 2008). McLuhan’s thesis concerning the medium suggests that technology is not a neutral channel but changes the scale and form of human experience (McLuhan, 1964). Flusser understands the technical image as a product of apparatus and program; creative practice therefore involves exploring a system’s possibilities and ‚playing against the program‘ (Flusser, 2000). In AI Art, the object of analysis is consequently not only the resulting image, but also the model, training data, interface and platform rules (Manovich, 2001).

Lev Manovich: From the Automation of Aesthetics to ‚A Medium That Thinks‘

Manovich initially describes AI as an infrastructure that not only creates cultural objects but also classifies, recommends and evaluates them, thereby shaping aesthetic preferences (Manovich, 2017, 2019a). In his three proposals for defining AI arts, he considers system autonomy, recognition by the artworld and an analogy with a creative Turing test. These are alternative criteria rather than a single generally accepted definition (Manovich, 2019b). His peer-reviewed article on computer vision argues that numerical image features constitute a new analytical language of colour, shape, texture and composition (Manovich, 2021).

Lev Manovich, Artificial Intelligence, Aesthetics, and Future of Culture. Video: HSE Art and Design School.

In later texts, Manovich challenges the myth of exclusively individual human creativity and distinguishes the imitation of an artefact from the modelling of a creative process (Manovich, 2022a, 2022b). He interprets generative AI not as a simple form of automated collage, but as the statistical learning of structural patterns from a cultural archive. At the same time, he notes that data compression may privilege common patterns and suppress rare details (Manovich, 2023a, 2023b, 2024a, 2024b, 2024c). Together with Arielli, he therefore examines authorship, artificial vision, aesthetic evaluation and the alignment of generative systems with human preferences (Manovich & Arielli, 2024).

Lev Manovich, The Future of Art. CYFEST16. Video: CYLAND Media Art Lab.

The term ‚artificial subjectivity‘ refers to the automated production of linguistic expressions that simulate thought, feeling or aesthetic judgement (Manovich, 2025). ‚A medium that thinks‘ names a generative medium capable of analysing patterns and producing variable outputs in a tension between creative possibility and statistical norms (Manovich, 2026a). Neither term provides evidence of AI consciousness. His most recent essays shift attention towards the future of art, its publics and the institutions that determine artistic value (Manovich, 2026b, 2026c).

Joanna Zylinska: Nonhuman Photography, the Perception Machine and Bio-AI

Zylinska’s work emerges from critical posthumanism and treats mediation as a technical, social and biological process (Kember & Zylinska, 2012). The concept of nonhuman photography extends photography beyond the conscious photographer to include automated cameras, sensors, networks and planetary imaging systems (Zylinska, 2017). Her book AI Art therefore asks not only whether a computer can be creative, but also how AI art relates to labour, automation, the data economy and ecological risks (Zylinska, 2020).

The ‚non-trivial machine of perception‘ connects perception with bodies, environments, histories and epistemic justice (Zylinska, 2022). The Perception Machine refers both to technical image infrastructures and to the sociopolitical conditions of automated vision and imagination (Zylinska, 2023a). Zylinska also reminds us that art cannot be reduced to a generated artefact, even though research into computational creativity remains legitimate (Zylinska, 2023b).

Joanna Zylinska, Bio-AI: The Aesthetics and Ethics of Digital Ecologies. Video: Filosofický ústav Akademie věd ČR.

Her analysis of diffusion models connects the technical process of denoising with the instability of contemporary regimes of knowledge (Zylinska, 2024a). After AI Art treats AI as part of a changing relationship among humans, artistic practices, institutions and technologies (Zylinska, 2024b). Bio-AI is a philosophical concept that describes AI as a biomachine and relational technology embedded in processes of life, labour and ecological extraction; it does not mean that AI is a biological organism (Zylinska, 2025). ‚Image thinking‘ explores feedback between language, images and human or machine generation, and understands the image as an active agent in thought (Toister & Zylinska, 2025).

Further Key Voices

Margaret Boden distinguishes combinational, exploratory and transformational creativity, offering a more precise framework for assessing machine novelty (Boden, 1998). Mark Coeckelbergh examines the unstable criteria of art and creativity, human-machine collaboration and creation as a relational process in which new roles for human and nonhuman actors emerge (Coeckelbergh, 2017, 2023). Luciana Parisi uses the ‚alien subject of AI‘ as a speculative philosophical concept of machine thought, while Dieter Mersch criticises the reduction of creativity to mathematisable and algorithmic rationality (Mersch, 2019; Parisi, 2019). Through cosmotechnics, Yuk Hui demonstrates that the relationship between art and technology does not assume a single universal form (Hui, 2021).

Margaret Boden, Creativity and AI: Asking the Right Questions. Interview recorded at the University of Oxford, 2012. Video: Science, Technology & the Future.

Anjan Chatterjee connects AI Art with neuroaesthetics and considers aesthetically sensitive machines without assuming that they must understand meaning or experience emotions as humans do (Chatterjee, 2022). David Gunkel raises questions of machine authorship and aesthetics, but does not provide empirical evidence of autonomous intention or moral responsibility in AI (Gunkel, 2017).

Anjan Chatterjee, The Neuroscience of Aesthetics. Video: GAIN Experience.

Gašper Beguš empirically investigates how linguistically interpretable internal representations emerge in generative neural networks trained on speech data. His research bridges technical interpretability and humanities-based analysis rather than providing a direct theory of visual art (Beguš, 2020). Nina Beguš compares human and machine storytelling and develops a programme in which literature and cultural metaphors help both to analyse and shape conceptions of AI (Beguš, 2024, 2025). The artistic research project Latent Spacecraft connects linguistics, neural networks, design and a reading of Joyce’s Finnegans Wake. Here, latent space is a high-dimensional computational structure, not a literal three-dimensional place (Beguš et al., 2026).

Gašper Beguš examines language, generative neural networks and the emergence of linguistically interpretable structures from speech data.

Nina Beguš discusses how literature, storytelling and cultural metaphors can contribute to the critical analysis and development of artificial intelligence.

Drawing on a survey of professional designers, Main and Grierson analyse the intelligent creative tool as ‚guru‘, partner or practical aid. Their source is a preprint, and its conclusions should therefore be regarded as preliminary (Main & Grierson, 2020). Smithies’s conference paper advocates shared research platforms, interdisciplinary teams and common Digital HASS infrastructure, but does not on its own provide a comprehensive framework for AI ethics (Smithies, 2024). Critical accounts of AI’s material and political conditions further address data labour, extraction and planetary costs (Crawford, 2021), the history of AI as the automation and formalisation of labour (Pasquinelli, 2023), and ‚mean images‘ that statistically reproduce dominant visual culture (Steyerl, 2023).

III. Key Authors and Strategies

Refik Anadol: Data, Architecture and Multisensory AI

Refik Anadol treats data as an artistic material, transforming it through machine learning into moving images, spatial sound and responsive architecture. Archive Dreaming, Melting Memories and WDCH Dreams established his strategies for activating cultural archives, neurological data and the memory of buildings (Refik Anadol Studio, 2017, 2018, 2019). In Unsupervised, a model processed 138,151 images from MoMA’s collection. The title does not, however, imply an absence of human curatorial labour, since the selection of data, model and mode of visualisation remains an authorial decision (Museum of Modern Art, 2022). Winds of Yawanawá extended this practice through collaboration with an Indigenous community and through the use of its drawings, songs and environmental data (Possible Futures, 2023).

Where Art Meets Algorithm: Inside Refik Anadol’s AI-Driven Worlds. Video: CoinDesk.

The current ecological turn is represented by the Large Nature Model, which the studio describes as the first open-source generative model devoted to nature. Echoes of the Earth: Living Archive and Living Architecture: Gehry connect natural or architectural datasets with monumental space (Serpentine, 2024; Guggenheim Museum Bilbao, 2025). In 2026, Latent City transforms Bruges’s historical and contemporary memory, while the private museum DATALAND opened with Machine Dreams: Rainforest, an exhibition in which images, sound and scents respond to environmental data and anonymised visitor biosignals (DATALAND, 2026; Refik Anadol Studio, 2026).

Anadol’s strength lies in transforming architecture into an interface experienced through the body. His recurring fluid aesthetic may nevertheless reduce distinct cultural and ecological contexts to a visually compelling spectacle. Claims regarding ethical datasets, sustainability and model openness therefore require verifiable documentation of data provenance, energy consumption, model cards and technological partnerships. References to machine ‚dreaming‘ or ‚perception‘ should be understood as metaphors, not evidence of consciousness or autonomous authorship (Norman, 2026; Zeilinger & Johns, 2025).

Mario Klingemann: Distributed Agency

MLow x Mario Klingemann: Botto Interview. Video: M. Low.

Klingemann’s artistic strategies have developed from the curated generation of images towards participatory and distributed systems. In The Butcher’s Son and Neural Decay, he determines the datasets and model architectures and selects the results, while in Neural Glitch/Mistaken Identity he deliberately disrupts GAN weights, turning technical error into an aesthetic method (The Lumen Prize, 2018; Klingemann, 2018; Onkaos, n.d.-d). Memories of Passersby I produces an endless stream of portraits, but remains a closed system that does not continue to learn. Uncanny Mirror transforms the viewer’s image, while Circuit Training involves the audience in creating the dataset, evaluating outputs and training the model over time (Barbican Centre, 2019; Ferraro & Ferrari, 2022). Appropriate Response examines the projection of meaning onto machine-generated aphorisms, while The Garden of Ephemeral Details cyclically deforms and restores Bosch’s triptych as a model of cultural memory (Onkaos, n.d.-b; The Garden of Earthly Delights, n.d.). In Botto, authorship is distributed across a generative system, a DAO community, token-based voting and the market; Klingemann acts as the system’s originator and ‚guardian‘. A.I.C.C.A. in turn uses a robotic dog that produces AI reviews to satirise the authority of art criticism (Onkaos, n.d.-a, n.d.-c). Recent projects, including Weapons of Mass Distraction, Landscapes and Botto’s Mirror Stages, direct his research towards the politics of attention, a rejection of aesthetic ‚AI slop‘ and the system’s physically situated interaction with audiences (De Filippi, 2026; Jebb, 2026; Meier, 2026). This distributed agency must not be confused with consciousness. Klingemann continues to characterise AI primarily as a complex artistic tool and medium (Silva, 2024).

Lauren Lee McCarthy: Performative AI, Intimacy and Distributed Agency

Open the video directly on YouTube

Lauren Lee McCarthy does not approach AI Art primarily as image generation, but as the performative prototyping of social systems. In Follower and LAUREN, she turns surveillance into a service and performs as a ‚human Alexa‘, continuously observing a household, communicating with its occupants and controlling connected devices (McCarthy, 2016, 2017). SOMEONE distributes this role among gallery visitors, who remotely become smart home assistants for strangers‘ homes (McCarthy, 2019). These projects expose the invisible human labour, cultural values and asymmetries of power concealed by the idea of automation. The reactivation of LAUREN: Anyone Home? for the 2024 Human AI Art Award confirmed the continuing relevance of this strategy amid the development of multimodal domestic agents (Art Collection Telekom, 2024).

In Surrogate and Saliva, McCarthy brings questions of data surveillance directly to the body, reproduction, DNA and the ownership of biological information. Participants formulate the terms of exchange themselves, creating an alternative to corporate biological data collection (Mandeville Art Gallery, 2024). In collaboration with Kyle McDonald, Voice in My Head replaces a participant’s inner voice with a generative voice system, while HOST directs guests‘ actions through individual instructions delivered via headphones (McCarthy & McDonald, 2023; The Music Center, 2024).

The most recent line of inquiry is AUTO, which continued to develop in 2025 and 2026. Participants in a simulated autonomous drive receive instructions and symbolically become the vehicle’s engine by singing together. The work thus responds to AI’s shift from a conversational interface to agentic systems that intervene in movement, decision-making and public infrastructure (Hong, 2025; McCarthy, 2025). Its strength lies in explicit consent, the possibility of refusing an instruction and an understanding of technology as a relationship rather than a neutral tool. It remains necessary, however, to ask whether voluntary gallery participation adequately models the power of commercial systems. Here, AI does not possess agency by itself. Agency is distributed across software, the artist, performers, participants and the rules of the situation.

Sofia Crespo: Speculative AI Biology

Sofia Crespo, AI-Generated Creatures That Stretch the Boundaries of Imagination. Video: TED.

Sofia Crespo examines how neural networks reconfigure biological imagery and shape human conceptions of nature. In Neural Zoo and Artificial Natural History, she creates fictional organisms and taxonomies that resemble scientific illustrations but do not depict actual species (Crespo, 2018–2022, 2020–2024). Critically Extant uses millions of publicly available images to visualise endangered species. Its imprecise outputs reveal that a species‘ digital invisibility says nothing about its biological significance, but reflects uneven patterns of human attention and documentation (Crespo, 2021–2022).

In the more recent Temporally Uncaptured and Perpetual Present, Crespo combines AI with cyanotype, 3D-printed clay and robotic painting (Crespo, 2023–2024, 2024). She develops her current ecological line of inquiry with Feileacan Kirkbride McCormick as Entangled Others. The exhibition Liquid Chimeras and the 2026 project (di)atomic garden transform gaps in oceanographic data and records of radioactivity into speculative organisms and mutating virtual ecosystems (Entangled Others, 2025–2026; MEET, 2026).

Crespo’s contribution lies in exposing the anthropocentrism of biological datasets. The generated organisms are not new species or instances of autonomous evolution, but authorially directed statistical reconfigurations. Moreover, their compelling biomorphic aesthetics risk conflating scientific representation with speculation and aestheticising ecological crisis.

Sougwen Chung: Robotic Drawing and Distributed Authorship

How Sougwen Chung Teaches Robots to Pause. Video: Reid Hoffman.

Sougwen Chung understands AI Art as ‚Operational Art‘. The work is not merely a completed image, but a feedback process among the body, data and a robotic system. D.O.U.G._1: Mimicry imitated drawing gestures through a camera, while D.O.U.G._2: Memory employed a recurrent neural network trained on Chung’s twenty-year archive of drawings. Omnia per Omnia extended this collaboration to a group of robots responding to movement data captured by public cameras in New York (MIT Open Documentary Lab, 2015; Victoria and Albert Museum, 2022).

In Mutations of Presence, Assembly Lines and Spectral, robot movement is driven by biofeedback, especially levels of EEG alpha activity during meditation. This is not mind-reading, but an artist-designed translation of a limited physiological signal into movement and drawing (Jebb, 2025). Body Machine (Meridians), shown at the East Slovak Gallery in 2025 and at Palazzo Citterio in 2026, combines bodily motion capture, environmental recordings and machine learning (Palazzo Citterio, 2026). In RECURSIONS, a robotic system trained on an archive of gestures and biosignals produces a recursive painting on a monumental scroll (Chung, 2026a).

Chung’s contribution lies in a sustained exploration of the embodied and temporal nature of human-machine collaboration. Co-authorship must not, however, be confused with conscious robotic intention. The system possesses causal and performative agency, but demonstrates neither goals of its own nor metarepresentation. The current Rewilding the Machine research explores biodegradable silk circuits and collaboration among humans, machines and silkworms. This remains research in development, and any future assessment must critically consider its actual ecological footprint, its care for living organisms and the extent to which they are instrumentalised (Chung, 2026b).

Living Systems, Simulated Worlds and Dataset Politics: Christa Sommerer and Laurent Mignonneau, Ian Cheng and Anna Ridler

Christa Sommerer and Laurent Mignonneau are pioneers of interactive art and artificial life. In Interactive Plant Growing (1992), A-Volve (1994) and Life Writer (2006), they translate a viewer’s touch, text or movement into the emergence and development of virtual organisms. The recent installation Flâneur (2026) turns visitors‘ movements into a digital garden that is subsequently destroyed by simulated ants (Sommerer & Mignonneau, 2026). Their strategy shifts authorship from the finished image to the design of rules and interfaces. Here, however, ‚life‘ and ‚evolution‘ refer to simplified computational models, not to conscious or fully autonomous beings.

Ian Cheng creates open-ended simulations in which characters behave according to rules, changing conditions and AI models. The Emissaries trilogy (2015–2017) examines cognitive development, while BOB (Bag of Beliefs) (2018–2019) allows audiences to influence the behaviour of an artificial being (MoMA, 2017; Serpentine Galleries, 2018). In Life After BOB (2021–2022) and Thousand Lives (2023), Cheng addresses human coexistence with personal AI and the learning of a virtual organism through a neurosymbolic model (Gladstone Gallery, 2024). This is not consciousness. The apparent personality emerges from rules, feedback and the viewer’s anthropomorphic interpretation.

Anna Ridler responds to opaque datasets by creating them by hand and exhibiting them as artworks in their own right. Myriad (Tulips) and Mosaic Virus connect thousands of the artist’s photographs of tulips with a generative model and the price of Bitcoin, exposing the labour, classification and speculation concealed behind the AI image (Ridler, 2018a, 2018b). The recent project A Perfect Language of Images (2026) compares historical taxonomies with generative models and demonstrates that no system can order the world without bias (Schwarzman Centre for the Humanities, 2026). A handmade dataset increases transparency and authorial control, but does not eliminate the subjectivity of selection or the biases of classification.

IV.Diverse Approaches to AI Art Strategies

This section of the lecture series presents AI Art as a plural field of artistic and research strategies. Artificial intelligence operates not only as a creative tool, but also as an object of critique, a method of inquiry, a mediator of social relations and a component of evolutionary and biohybrid environments. Describing AI as an actor does not presuppose consciousness or autonomous intentionality. It refers primarily to a system’s operational capacity to influence the production, selection, interpretation and distribution of artistic outputs within a broader sociotechnical environment.

The Politics of Training Data and Algorithmic Vision

In Training Humans and ImageNet Roulette, Kate Crawford and Trevor Paglen examine how training databases classify human bodies and identities, and how historical prejudices, social norms and power relations are translated into technical categories (Crawford & Paglen, 2019, 2021). Their strategy may be characterised as an archaeology of training data and an artistic critique of classification. It should not, however, be equated with algorithmic auditing in the strict technical sense. Moreover, ImageNet Roulette was a deliberately provocative demonstration of selected ImageNet categories rather than a comprehensive scientific evaluation of the entire system.

In Calculating Empires: A Genealogy of Technology and Power Since 1500, Crawford and Vladan Joler extend this critique into the histories of communication, computing, classification and control. Their extensive research visualisation maps relations among colonialism, militarisation, automation, labour, natural resources and the concentration of technological power (Crawford & Joler, 2023). The project does not constitute a neutral or exhaustive historical database. It is a critical genealogy that deliberately selects and interprets particular historical connections.

10 Years of Forensic Architecture. Video: Goldsmiths, University of London.

Forensic Architecture represents a different, investigative strategy. In Triple-Chaser, the group uses digital spatial modelling and thousands of synthetic images produced in a game environment to train a system to recognise a specific type of tear-gas canister in photographs and videos (Forensic Architecture, 2019). Here, machine vision is not an autonomous investigator, but one element within a process that includes human verification, work with testimony and the reconstruction of events.

The Gender Shades study provides an empirical context. Joy Buolamwini and Timnit Gebru demonstrated substantial differences in the accuracy of three commercial gender-classification systems across groups differentiated by gender and skin tone (Buolamwini & Gebru, 2018). This is a peer-reviewed algorithmic audit rather than a work of art. For AI Art, however, it provides an important scientific and methodological foundation for artistic strategies that criticise automated vision, classification and discrimination.

Gender Shades presents Joy Buolamwini and Timnit Gebru’s intersectional audit of commercial gender-classification systems. Although it is a scientific study rather than an artwork, it provides an important methodological foundation for artistic critiques of automated vision, classification and algorithmic discrimination.

Counter-Data, Collective Production and Cultural Memory

In Zizi: Queering the Dataset, Jake Elwes intervenes in a generative model by adding images of drag and gender-fluid identities. The project disrupts the normative space of the original training set and demonstrates how changing data affects a model’s internal representations and outputs (Elwes, 2019). This strategy may be described as the production of counter-data or the queering of a dataset. Critical analysis must nevertheless consider the provenance of images obtained online, the informed consent of the people depicted and the difference between representing a community and involving that community directly in data production.

In Data Trust, Stephanie Dinkins combines community storytelling, generative AI, organic materials and experimental DNA data storage. Participants‘ oral histories enter a continuously changing visual installation, forming a community archive based on the conscious provision of data (Dinkins, 2025). The project examines data sovereignty, trust and cultural memory. It cannot, however, be interpreted as evidence that organic data carriers or the AI system possess memory in a psychological sense.

Holly Herndon and Mat Dryhurst, The Call. Video: Serpentine.

In The Call, Holly Herndon and Mat Dryhurst developed materials and protocols for training choral AI models in collaboration with fifteen community choirs. The project included the Choral Data Trust Experiment, which explored the distribution of decision-making rights between providers of vocal data and users of the models (Herndon & Dryhurst, 2024). Unlike Dinkins’s project, its primary concern is not the preservation of memory but the collective production of training data, informed consent, ownership and model governance. The process of building the dataset itself thus becomes an artistic medium.

Synthetic Memories: Visualising Edward’s Memories with AI. Video: Domestic Data Streamers.

Synthetic Memories by Domestic Data Streamers translates spoken or written descriptions of memories into generated visual reconstructions. The project has developed from pilot experiments begun in 2022 and received the Ars Electronica Award for Digital Humanity in 2025 (Ars Electronica, 2025; Domestic Data Streamers, 2022–present). The resulting images are neither recovered photographic evidence nor accurate records of the past. They are interpretations produced through collaboration among a participant, a mediator and a generative model. Although the project is also used in contexts of reminiscence and dementia, no scientifically validated therapeutic effect can be claimed without peer-reviewed clinical results.

Evolutionary, Biohybrid and Nonhuman Systems

In Eden, Jon McCormack created an artificial ecosystem in which virtual organisms compete for resources, evolve and use sound for survival and reproduction. Audience behaviour also affects the conditions of the virtual environment (McCormack, 2001). The project belongs to the traditions of artificial life and evolutionary art rather than to contemporary generative foundation models. Its significance lies in historically extending AI Art beyond the production of isolated artefacts towards the modelling of an adaptive aesthetic ecosystem.

Michael Sedbon, Engineered Bio-Hybrid System. Paris Biodesign Conference 2026. Video: Living Materials Living Worlds.

In Bio Economics, Michael Sedbon combines computational decision-making, genetic algorithms, sensors and photosynthetic microorganisms. The system explores competition and cooperation in the distribution of light, energy and other resources (Sedbon, 2024). Terms such as ‚biocomputer‘ or ‚biological intelligence‘ have primarily artistic and research-oriented meanings in this context. The project is not evidence of emergent consciousness, but an experiment that connects biological, economic and computational processes.

Pierre Huyghe, Liminal. Punta della Dogana. Video: Palazzo Grassi – Punta della Dogana.

In the exhibition Liminal, Pierre Huyghe creates a dynamic environment that combines biological organisms, human performers, sensors, robotics, machine learning and images modified in real time (Huyghe, 2024). The exhibition is conceived as a changing environment rather than a collection of closed artistic objects. Terms such as subjectivity, learning and hybridisation should therefore be read as part of an artistic and curatorial vocabulary rather than as empirically confirmed properties of the system as a whole.

The emergent or adaptive behaviour of these systems does not in itself demonstrate phenomenal consciousness or autonomous artistic intentionality. Butlin et al. (2026) propose investigating possible AI consciousness through a set of indicators derived from several theories of consciousness. Their methodology does not exclude the possibility of consciousness in artificial systems, but neither does it provide evidence that current AI is conscious. The empirical validation and calibration of the proposed indicators remain open scientific problems (Pennartz, 2026).

Boris Eldagsen, Is Authenticity an Antidote to AI? MetaMorf 2026 Conference, Trondheim, Norway.

The Boundaries of Photography and Architecture

Through the 2022 image Pseudomnesia: The Electrician and his subsequent refusal of a Sony World Photography Awards prize in 2023, Boris Eldagsen opened a debate about the distinction between photography and the synthetic image (Eldagsen, 2022, 2023). He uses the term ‚promptography‘ to describe images produced through generative models and textual prompts. This is not an established scientific or art-historical category, but terminology proposed by the artist. Its significance lies in distinguishing a photographic record of an event from a synthetic image that imitates photographic visual language.

Matias del Campo, Neural Architecture: Designing with the Aid of AI. AI Design Frontiers. Video: ArchiNet.

Matias del Campo extends the discussion into architecture, where AI influences not only the visual representation of a building, but also design processes, data analysis, the production of form and understandings of authorship. The book Neural Architecture and the special issue Artificial Intelligence in Architecture examine the relationship of neural models to material and symbolic culture, as well as questions of agency, authorship and professional responsibility (del Campo, 2022, 2024). It remains necessary to distinguish critically between a visually persuasive architectural image and a feasible proposal that addresses structural, ecological and social conditions.

The diverse approaches to AI Art strategies cannot be unified within a single aesthetic or technological model. They encompass data archaeology, algorithmic auditing, the production of counter-data, collective training processes, the reconstruction of memory, evolutionary art, biohybrid environments and the transformation of photography and architecture. Their common object is not simply the resulting artefact, but the entire system of relations: who provides the data, who designs the model, who controls its operation, who owns its outputs, and who bears its social, political and environmental consequences.

V.Philosophy of Prediction and Anticipation: Future Aesthetic Processes Between AGI and HCAI

This section of the lecture series examines estimates concerning the emergence of artificial general intelligence (AGI) and the normative framework of human-centred artificial intelligence (HCAI) in relation to the future of art. It critically compares the scenarios proposed by Bostrom, Tegmark and Harari. Although philosophically and culturally significant, these scenarios cannot be treated as reliable technical forecasts. Particular attention is given to Agüera y Arcas, who connects intelligence with the capacity to predict and influence the future. The section then considers the possibility of consciousness in AI. The theoretical framework developed by Butlin and colleagues does not confirm consciousness in current systems, but neither does it justify the a priori exclusion of future systems that implement indicators derived from scientific theories of consciousness (Butlin et al., 2023, 2026). The final part introduces speculative design and artistic research as methods of critical anticipation rather than substitutes for scientific forecasting. Liam Young serves as the principal case study, with projects that examine possible technological, political and ecological forms of future aesthetic processes.

Yuval Noah Harari and Max Tegmark, Humanity and AI. Video: Bloomberg Live.

Predicting and Anticipating Future Aesthetic Processes in the Context of AI Art Strategies

Estimates of when AGI might emerge remain uncertain, partly because AGI itself lacks a consistent definition. In a 2023 survey of 2,778 AI researchers, the aggregated responses assigned a 10 per cent probability to high-level machine intelligence by 2027 and a 50 per cent probability by 2047. These figures represent expert judgements that are sensitive to the wording of the question, not an objective timetable for technological development (Grace et al., 2025). As of 18 August 2026, Kalshi assigned an approximately 45 per cent probability to OpenAI achieving AGI before 2030, Polymarket assigned a 9 per cent probability to an announcement by the end of 2026, and Metaculus gave a median date of March 2033 for the emergence and announcement of a general system (Kalshi, 2026; Metaculus, 2026; Polymarket, 2026). These dynamic estimates are not directly comparable because they employ different definitions and resolution criteria.

The International AI Safety Report 2026 does not provide a date for the arrival of AGI, but synthesises evidence concerning the present capabilities and limitations of general-purpose AI systems. It identifies their rapid but ‚jagged‘ development, gaps in evaluation and the limited reliability of predictions of real-world performance (Bengio et al., 2026). Bostrom (2014) analyses an intelligence explosion and existential risks, Tegmark (2017) develops civilisational scenarios, and Harari (2024) examines the social power of AI within information networks. Their work cannot be used as technical evidence for a date at which AGI will emerge. AI 2027 should likewise be read as a forecasting scenario rather than a peer-reviewed scientific prediction (Kokotajlo et al., 2025; Lifland et al., 2026). Gebru and Torres (2024) criticise the ideological and political assumptions of AGI discourse, rather than the technical possibility of AGI as such.

AGI and HCAI are not successive stages of development. AGI denotes a contested goal involving general capabilities, whereas HCAI is a normative approach that emphasises creativity, control, safety and responsibility (Morris et al., 2024; Shneiderman, 2022). In AI Art, HCAI may entail adaptive tools, transparent data provenance and the preservation of artistic or community control. The term ‚human‘, however, may obscure inequalities, hidden labour and environmental relations (Capel & Brereton, 2023).

Blaise Agüera y Arcas and Carlo Rovelli, We Merged With Machines a Long Time Ago. Video: Futurology.

The case study of Blaise Agüera y Arcas shifts attention from the projected date of AGI to the nature of intelligence. In What Is Intelligence?, he argues that prediction is fundamental to intelligence and connects it with evolution, sociality and consciousness (Agüera y Arcas, 2025). For AI Art, this suggests the possibility of an adaptive system that models its environment and changes its strategies. It is nevertheless an ambitious theoretical position rather than a scientific consensus. Evans et al. (2026) develop an account of socially distributed intelligence in human-machine collectives. Their article in Science presents a theoretical perspective, not empirical evidence of an intelligence explosion. Zylinska’s (2025) concept of Bio-AI does not designate AI as an organism, but as a material and relational technology embedded in processes of life.

The Possibility of Consciousness in AI (CAI) and Future Aesthetic Processes

There is currently no generally accepted theory of consciousness. Butlin et al. (2023, 2026) therefore derive functional and architectural indicators from several neuroscientific theories in order to assess AI systems by degree. Their analysis does not confirm consciousness in the systems examined. It does, however, identify no obvious technical barrier to implementing the indicators, which is a weaker claim than evidence of phenomenal consciousness. The experiment conducted by Cogitate Consortium et al. (2025) challenged some predictions made by integrated information theory (IIT) and global neuronal workspace theory (GNWT), but did not definitively refute either of them and investigated human rather than artificial consciousness. Kanai and Ma (2026) present substrate-independent consciousness only as a conditional functionalist proposal in a non-peer-reviewed preprint.

In AI Art, it is therefore necessary to distinguish aesthetic competence, creativity, agency, linguistic self-description and phenomenal consciousness. Neither an artwork nor a system’s statement that ‚I feel‘ is in itself evidence of consciousness. As an authorial working protocol for analysis, rather than a validated diagnostic scale, the following may be examined: P1, recurrent processing and temporal integration; P2, the global availability of information; P3, metarepresentation and the modelling of self or others; P4, attention and relevance selection; and P5, bounded and auditable introspective reportability. An artistic situation may investigate manifestations of artificial subjectivity, but cannot itself serve as a scientific test of consciousness.

Can Theories of Consciousness Supercharge AI? Video: World Science Festival.

Artificial Utopia? The Future of Humanity in an AI World. Video: World Science Festival.

Speculative Modelling as a Tool for Anticipating Future Aesthetic Processes

Speculative design does not assign probabilities to events, but creates possible scenarios and opens discussion about their desirability. Dunne and Raby (2013) ask ‚what if?‘, while Auger (2013) emphasises informed extrapolation and the plausible artefact. Futures Literacy denotes the capacity to recognise how images of the future shape present decision-making (Miller, 2018). Worldbuilding, design fiction and performative simulation may produce experiential knowledge, but their aesthetic persuasiveness does not establish the probability of a scenario (Borgdorff, 2012). Design Justice adds the questions of who designs the future and who bears its costs (Costanza-Chock, 2020).

Liam Young, My Solutions Are Not Polite. Video: Louisiana Channel.

Liam Young provides the principal case study. In the Robot Skies and Where the City Can’t See examine drones, machine vision and surveillance; Planet City imagines a city for ten billion people; and The Great Endeavor visualises carbon-removal infrastructure (Young, 2016a, 2016b, 2021, 2023). These are not technical plans, but aesthetic and political tests of possible scenarios. Comparative examples include Ian Cheng, Holly Herndon and Mat Dryhurst, Superflux, Alexandra Daisy Ginsberg, Anna Ridler and Pierre Huyghe, whose practices examine simulated worlds, data protocols, ecological agentic models and technological-biological environments (Cheng, 2015–2017; Ginsberg, 2019; Herndon & Dryhurst, 2024; Huyghe, 2024; Ridler, 2019; Superflux, 2017–2019).

Future aesthetic processes can be predicted empirically only to the extent that they involve measurable or observable trends. Their broader social, artistic and ontological transformations must be critically anticipated through philosophical, technological, artistic and socioecological frameworks.

References

Agüera y Arcas, B. (2025). What is intelligence? Lessons from AI about evolution, computing, and minds. MIT Press.

Anadol, R. (2017). Archive dreaming: AI data sculpture [Data installation]. https://refikanadol.com/works/archive-dreaming/

Anadol, R. (2018). Melting memories [Installation]. https://refikanadol.com/works/melting-memories/

Anadol, R. (2019). WDCH dreams [Data installation]. https://refikanadol.com/works/wdch-dreams/

Anadol, R. (2024). Echoes of the Earth: Living archive [Exhibition]. Serpentine Galleries. https://www.serpentinegalleries.org/whats-on/refik-anadol-echoes-of-the-earth-living-archive/

Anadol, R. (2026). Latent city [Exhibition]. https://refikanadol.com/events/

Ars Electronica. (2025). Synthetic memories [Project documentation]. https://calls.ars.electronica.art/2025/prix/winners/19953/

Art Collection Telekom. (2024). LAUREN: Anyone Home? https://www.art-collection-telekom.com/en/exhibitions/lauren-anyone-home

Auger, J. (2013). Speculative design: Crafting the speculation. Digital Creativity, 24(1), 11–35. https://doi.org/10.1080/14626268.2013.767276

Beguš, G. (2020). Generative adversarial phonology: Modeling unsupervised phonetic and phonological learning with neural networks. Frontiers in Artificial Intelligence, 3, Article 44. https://doi.org/10.3389/frai.2020.00044

Beguš, N. (2024). Experimental narratives: A comparison of human crowdsourced storytelling and AI storytelling. Humanities and Social Sciences Communications, 11, Article 1392. https://doi.org/10.1057/s41599-024-03868-8

Beguš, N. (2025). Artificial humanities: A fictional perspective on language in AI. University of Michigan Press. https://doi.org/10.3998/mpub.12778936

Beguš, N., Metahaven, & Beguš, G. (2026). Latent spacecraft: Brains, GANs, Finnegans. Antikythera, 1(2). https://doi.org/10.1162/ANTI.5KP1

Bengio, Y., Clare, S., Prunkl, C., Murray, M., Andriushchenko, M., Bucknall, B., Bommasani, R., Casper, S., Davidson, T., Douglas, R., Duvenaud, D., Fox, P., Gohar, U., Hadshar, R., Ho, A., Hu, T., Jones, C., Kapoor, S., Kasirzadeh, A., . . . Mindermann, S. (2026). International AI Safety Report 2026 (DSIT 2026/001). Department for Science, Innovation and Technology. https://internationalaisafetyreport.org/publication/international-ai-safety-report-2026

Benjamin, W. (2008). The work of art in the age of its technological reproducibility, and other writings on media. The Belknap Press of Harvard University Press. https://www.jstor.org/stable/j.ctv1nzfgns

Boden, M. A. (1998). Creativity and artificial intelligence. Artificial Intelligence, 103(1–2), 347–356. https://doi.org/10.1016/S0004-3702(98)00055-1

Boden, M. A., & Edmonds, E. A. (2009). What is generative art? Digital Creativity, 20(1–2), 21–46. https://doi.org/10.1080/14626260902867915

Borgdorff, H. (2012). The conflict of the faculties: Perspectives on artistic research and academia. Leiden University Press.

Bostrom, N. (2005). A history of transhumanist thought. Journal of Evolution and Technology, 14(1), 1–25. https://nickbostrom.com/papers/a-history-of-transhumanist-thought/

Bostrom, N. (2014). Superintelligence: Paths, dangers, strategies. Oxford University Press.

Braidotti, R. (2013). The posthuman. Polity.

Buolamwini, J., & Gebru, T. (2018). Gender shades: Intersectional accuracy disparities in commercial gender classification. In S. A. Friedler & C. Wilson (Eds.), Proceedings of the 1st Conference on Fairness, Accountability and Transparency (Vol. 81, pp. 77–91). Proceedings of Machine Learning Research. https://proceedings.mlr.press/v81/buolamwini18a.html

Butlin, P., Long, R., Bayne, T., Bengio, Y., Birch, J., Chalmers, D., Constant, A., Deane, G., Elmoznino, E., Fleming, S. M., Ji, X., Kanai, R., Klein, C., Lindsay, G., Michel, M., Mudrik, L., Peters, M. A. K., Schwitzgebel, E., Simon, J., & VanRullen, R. (2026). Identifying indicators of consciousness in AI systems. Trends in Cognitive Sciences, 30(6), 488–501. https://doi.org/10.1016/j.tics.2025.10.011

Butlin, P., Long, R., Elmoznino, E., Bengio, Y., Birch, J., Constant, A., Deane, G., Fleming, S. M., Frith, C., Ji, X., Kanai, R., Klein, C., Lindsay, G., Michel, M., Mudrik, L., Peters, M. A. K., Schwitzgebel, E., Simon, J., & VanRullen, R. (2023). Consciousness in artificial intelligence: Insights from the science of consciousness [Preprint]. arXiv. https://doi.org/10.48550/arXiv.2308.08708

Capel, T., & Brereton, M. (2023). What is human-centered about human-centered AI? A map of the research landscape. Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems, Article 359, 1–23. https://doi.org/10.1145/3544548.3580959

Cave, S., Dihal, K., & Dillon, S. (Eds.). (2020). AI narratives: A history of imaginative thinking about intelligent machines. Oxford University Press. https://doi.org/10.1093/oso/9780198846666.001.0001

Chatterjee, A. (2022). Art in an age of artificial intelligence. Frontiers in Psychology, 13, Article 1024449. https://doi.org/10.3389/fpsyg.2022.1024449

Cheng, I. (2015–2017). Emissaries [Live simulation trilogy]. https://iancheng.com/emissaries

Chung, S. (2026a). RECURSIONS at Hong Kong. https://sougwen.com/2026/recursions-hong-kong

Chung, S. (2026b). Research: Rewilding the Machine. https://sougwen.com/research

Coeckelbergh, M. (2017). Can machines create art? Philosophy & Technology, 30, 285–303. https://doi.org/10.1007/s13347-016-0231-5

Coeckelbergh, M. (2023). The work of art in the age of AI image generation: Aesthetics and human–technology relations as process and performance. Journal of Human-Technology Relations, 1. https://doi.org/10.59490/jhtr.2023.1.7025

Cogitate Consortium, Ferrante, O., Gorska-Klimowska, U., Henin, S., Hirschhorn, R., Khalaf, A., Lepauvre, A., Liu, L., Richter, D., Vidal, Y., Bonacchi, N., Brown, T., Sripad, P., Armendariz, M., Bendtz, K., Ghafari, T., Hetenyi, D., Jeschke, J., Kozma, C., . . . Melloni, L. (2025). Adversarial testing of global neuronal workspace and integrated information theories of consciousness. Nature, 642, 133–142. https://doi.org/10.1038/s41586-025-08888-1

Costanza-Chock, S. (2020). Design justice: Community-led practices to build the worlds we need. MIT Press.

Crawford, K. (2021). Atlas of AI: Power, politics, and the planetary costs of artificial intelligence. Yale University Press.

Crawford, K., & Joler, V. (2023). Calculating empires: A genealogy of technology and power since 1500 [Research visualisation and installation]. https://calculatingempires.net/

Crawford, K., & Paglen, T. (2019a). Excavating AI: The politics of images in machine learning training sets. https://excavating.ai

Crawford, K., & Paglen, T. (2019b). ImageNet roulette [Interactive web project]. Paglen Studio. https://paglen.studio/2020/04/29/imagenet-roulette/

Crawford, K., & Paglen, T. (2019c). Training humans [Exhibition]. Fondazione Prada. https://www.fondazioneprada.org/project/training-humans/?lang=en

Crawford, K., & Paglen, T. (2021). Excavating AI: The politics of images in machine learning training sets. AI & Society, 36, 1105–1116. https://doi.org/10.1007/s00146-021-01162-8

Crespo, S. (2018–2022). Neural Zoo. https://sofiacrespo.com/neural-zoo/

Crespo, S. (2020–2024). Artificial Natural History. https://sofiacrespo.com/artificial-natural-history/

Crespo, S. (2021–2022). Critically Extant. https://sofiacrespo.com/critically-extant/

Crespo, S. (2024). Perpetual Present. https://sofiacrespo.com/perpetual-present/

DATALAND. (2026). Machine Dreams: Rainforest. https://dataland.art/exhibitions/machine-dreams-rainforest

DATALAND. (n.d.). Large Nature Model. https://dataland.art/about/large-nature-model

De Filippi, P. (2026, July 17). Off-grid aesthetics: Where the AI prompt cannot go. Right Click Save. https://www.rightclicksave.com/article/off-grid-aesthetics-where-the-ai-prompt-cannot-go

del Campo, M. (2022). Neural architecture: Design and artificial intelligence. AR+D Publishing.

del Campo, M. (Ed.). (2024). Artificial intelligence in architecture [Special issue]. Architectural Design, 94(3). https://onlinelibrary.wiley.com/toc/15542769/2024/94/3

Dinkins, S. (2025). Data trust [Participatory installation]. Institute of Contemporary Art San José. https://www.icasanjose.org/archive/data_trust/

Domestic Data Streamers. (2022–present). Synthetic memories [Participatory artistic research project]. https://www.domesticstreamers.com/art-research/work/synthetic-memories/

Dunne, A., & Raby, F. (2013). Speculative everything: Design, fiction, and social dreaming. MIT Press.

Eldagsen, B. (2022). Pseudomnesia: The electrician [Image created with generative AI]. https://www.eldagsen.com/pseudomnesia/

Eldagsen, B. (2023). Sony World Photography Awards 2023 [Artist’s statement]. https://www.eldagsen.com/sony-world-photography-awards-2023/

Elwes, J. (2019). Zizi – Queering the dataset [Multichannel digital video]. https://www.jakeelwes.com/project-zizi-2019.html

Entangled Others. (2025–2026). (di)atomic garden. https://www.mbal.ch/en/orbite-offline/entangled-others/

Evans, J., Bratton, B. H., & Agüera y Arcas, B. (2026). Agentic AI and the next intelligence explosion. Science, 391(6791), eaeg1895. https://doi.org/10.1126/science.aeg1895

Ferraro, A., & Ferrari, G. (Interviewers). (2022, February 22). Personified data in a post-human utopia with Mario Klingemann [Interview]. French Fries Magazine. https://frenchfries-mag.com/interview/2022/2/22/personified-data-in-a-post-human-utopia-with-mario-klingemann

Fisher, S. E., & Scharff, C. (2009). FOXP2 as a molecular window into speech and language. Trends in Genetics, 25(4), 166–177. https://doi.org/10.1016/j.tig.2009.03.002

Flusser, V. (2000). Towards a philosophy of photography. Reaktion Books.

Forensic Architecture. (2019). Triple-Chaser [Investigative research project]. https://forensic-architecture.org/investigation/triple-chaser

Gebru, T., & Torres, É. P. (2024). The TESCREAL bundle: Eugenics and the promise of utopia through artificial general intelligence. First Monday, 29(4). https://doi.org/10.5210/fm.v29i4.13636

Ginsberg, A. D. (2019). The substitute [Video installation]. https://www.daisyginsberg.com/work/the-substitute

Gladstone Gallery. (2024). Ian Cheng: Thousand lives. https://gladstonegallery.com/exhibit/thousand-lives-ian-cheng/

Grace, K., Sandkühler, J. F., Stewart, H., Weinstein-Raun, B., Thomas, S., Stein-Perlman, Z., Salvatier, J., Brauner, J., & Korzekwa, R. C. (2025). Thousands of AI authors on the future of AI. Journal of Artificial Intelligence Research, 84, Article 9. https://doi.org/10.1613/jair.1.19087

Guggenheim Museum Bilbao. (2025). in situ: Refik Anadol. https://www.guggenheim-bilbao.eus/en/exhibitions/in-situ-refik-anadol

Gunkel, D. J. (2017). Special section: Rethinking art and aesthetics in the age of creative machines. Philosophy & Technology, 30, 263–265. https://doi.org/10.1007/s13347-017-0281-3

Harari, Y. N. (2024). Nexus: A brief history of information networks from the Stone Age to AI. Random House.

Hayles, N. K. (1999). How we became posthuman: Virtual bodies in cybernetics, literature, and informatics. University of Chicago Press.

Herndon, H., & Dryhurst, M. (2024). The call: New rituals for collaboration with AI [Exhibition and interactive sound installation]. Serpentine Galleries. https://www.serpentinegalleries.org/whats-on/holly-herndon-mat-dryhurst-the-call/

Hong, J. (2025, June 23). “What does it take to feel autonomy these days?”: Lauren Lee McCarthy on Auto. Unframed: LACMA. https://unframed.lacma.org/2025/06/23/what-does-it-take-feel-autonomy-these-days-lauren-lee-mccarthy-auto

Hui, Y. (2021). Art and cosmotechnics. University of Minnesota Press.

Huyghe, P. (2024a). Camata [Autonomous film environment using machine learning, robotics and sensors]. Pinault Collection. https://www.pinaultcollection.com/en/boursedecommerce/pierre-huyghe-1

Huyghe, P. (2024b). Liminal [Exhibition]. Pinault Collection. https://www.pinaultcollection.com/palazzograssi/en/pierre-huyghe-liminal

Jebb, L. (2025, January 17). Sougwen Chung: Meet the boundary-pushing pioneer of robot art. The Art Newspaper. https://www.theartnewspaper.com/2025/01/17/sougwen-chung-meet-the-boundary-pushing-pioneer-of-robot-art

Jebb, L. (2026, March 26). Zero 10 returns at Art Basel Hong Kong. Right Click Save. https://www.rightclicksave.com/article/zero-10-returns-at-art-basel-hong-kong-art-fair-digital-art

Kalshi. (2026). When will OpenAI achieve AGI? Retrieved August 18, 2026, from https://kalshi.com/markets/kxoaiagi/openai-achieves-agi/oaiagi

Kanai, R., & Ma, S. (2026). Intrinsic computational functionalism and simulated consciousness [Preprint]. arXiv. https://doi.org/10.48550/arXiv.2606.15348

Kember, S., & Zylinska, J. (2012). Life after new media: Mediation as a vital process. MIT Press.

Klingemann, M. (2018, October 28). Neural Glitch/Mistaken Identity. Quasimondo. https://quasimondo.com/2018/10/28/neural-glitch/

Klingemann, M. (2019). Circuit training [Interactive installation]. Barbican Centre. https://artsandculture.google.com/story/circuit-training-machine-made-art-for-the-people-barbican-centre/ngWRdP9M5scyLQ?hl=en

Klingemann, M. (2023). A.I.C.C.A. [Robotic performance sculpture]. https://aicca.me/about/

Kokotajlo, D., Alexander, S., Larsen, T., Lifland, E., & Dean, R. (2025). AI 2027. https://ai-2027.com/

Lifland, E., Kokotajlo, D., & Halstead, B. (2026, January 27). Clarifying how our AI timelines forecasts have changed since AI 2027. AI Futures Project. https://blog.aifutures.org/p/clarifying-how-our-ai-timelines-forecasts

Main, A., & Grierson, M. (2020). Guru, partner, or pencil sharpener? Understanding designers’ attitudes towards intelligent creativity support tools [Preprint]. arXiv. https://doi.org/10.48550/arXiv.2007.04848

Mandeville Art Gallery. (2024). Lauren Lee McCarthy: Bodily Autonomy. University of California San Diego. https://mandevilleartgallery.ucsd.edu/exhibitions/lauren-lee-mccarthy.html

Manovich, L. (2001). The language of new media. MIT Press.

Manovich, L. (2017). Automating aesthetics: Artificial intelligence and image culture. Flash Art International, 50(316), 84–87. https://manovich.net/index.php/projects/automating-aesthetics-artificial-intelligence-and-image-culture

Manovich, L. (2019a). AI aesthetics. Strelka Press. https://manovich.net/index.php/projects/ai-aesthetics

Manovich, L. (2019b). Defining AI arts: Three proposals. In AI and dialogue of cultures [Exhibition catalogue]. The State Hermitage Museum. https://manovich.net/index.php/projects/defining-ai-arts-three-proposals

Manovich, L. (2021). Computer vision, human senses, and language of art. AI & Society, 36(4), 1145–1152. https://doi.org/10.1007/s00146-020-01094-9

Manovich, L. (2022a). Who is an artist in the AI era? https://manovich.net/index.php/projects/who-is-an-artist-in-ai-era

Manovich, L. (2022b). AI and myths of creativity. https://manovich.net/index.php/projects/ai-myths-of-creativity

Manovich, L. (2023a). Seven arguments about AI images and generative media. https://manovich.net/index.php/projects/seven-arguments-about-ai-images-and-generative-media

Manovich, L. (2023b). The AI brain in the cultural archive. https://manovich.net/index.php/projects/the-ai-brain-in-the-cultural-archive

Manovich, L. (2024a). Separate and reassemble: Generative AI and digital media history. https://manovich.net/index.php/projects/separate-and-reassemble-generative-ai-media-history

Manovich, L. (2024b). “Make it new”: GenAI, modernism, and database art. https://manovich.net/index.php/projects/make-it-new-genai-modernism-and-database-art

Manovich, L. (2024c). From museum without walls to generative AI museum. European Photography, 45(2), 48–51. https://manovich.net/index.php/projects/from-museum-without-walls-to-genai-museum

Manovich, L. (2025). Artificial subjectivity. Magyar Nyelvőr, 149(5), 600–603. https://doi.org/10.38143/Nyr.2025.5.600

Manovich, L. (2026a). A medium that thinks: Generative AI and media cognition. Emerging Media, 4(2), 225–232. https://doi.org/10.1177/27523543261454458

Manovich, L. (2026b). The future of art? https://manovich.net/index.php/projects/the-future-of-art

Manovich, L. (2026c). Six worlds of digital art: Audiences, institutions, values. https://manovich.net/index.php/projects/is-digital-art-a-visual-art

Manovich, L., & Arielli, E. (2024). Artificial aesthetics: Generative AI, art and visual media. https://manovich.net/index.php/projects/artificial-aesthetics

McCarthy, L. L. (2016). Follower [Participatory performance]. https://get-lauren.net/Follower

McCarthy, L. L. (2017). LAUREN [Networked performance and installation]. https://get-lauren.com/

McCarthy, L. L. (2019). SOMEONE [Distributed performance and installation]. https://get-lauren.net/SOMEONE

McCarthy, L. L. (2022). Saliva [Participatory performance and biological archive]. https://get-lauren.net/Saliva

McCarthy, L. L. (2025). AUTO [Participatory performance]. https://get-lauren.net/Auto

McCarthy, L. L., McDonald, K., & Yamaguchi Center for Arts and Media. (2022). Unlearning language [Interactive installation and performance]. https://www.ycam.jp/en/events/2022/unlearning-language/

McCormack, J. (2001). Eden: An evolutionary sonic ecosystem. In J. Kelemen & P. Sosík (Eds.), Advances in artificial life: ECAL 2001 (Vol. 2159, pp. 133–142). Springer. https://doi.org/10.1007/3-540-44811-X_13

McCormack, J. (n.d.). Artworks. https://jonmccormack.info/

McLuhan, M. (1964). Understanding media: The extensions of man. McGraw-Hill.

MEET Digital Culture Center. (2026). Liquid Chimeras. https://www.meetcenter.it/en/event/liquid-chimeras-where-ocean-data-is-lacking-speculative-forms-emerge/

Meier, A. (2026). Why make images at all today? Mario Klingemann in conversation with Anika Meier about boredom and inert images. ART Magazine. https://www.art-magazine.ai/artist-directory/feature/mario-klingemann-conflict-of-interest

Mersch, D. (2019). Kreativität und künstliche Intelligenz: Einige Bemerkungen zu einer Kritik algorithmischer Rationalität. Zeitschrift für Medienwissenschaft, 11(21), 65–74. https://doi.org/10.25969/mediarep/12634

Metaculus. (2026). When will the first general AI system be devised, tested, and publicly announced? Retrieved August 18, 2026, from https://www.metaculus.com/questions/5121/date-of-artificial-general-intelligence/

Miller, R. (Ed.). (2018). Transforming the future: Anticipation in the 21st century. Routledge.

MIT Open Documentary Lab. (2015). Drawing Operations. https://docubase.mit.edu/project/drawing-operations/

Morris, M. R., Sohl-Dickstein, J., Fiedel, N., Warkentin, T., Dafoe, A., Faust, A., Farabet, C., & Legg, S. (2024). Position: Levels of AGI for operationalizing progress on the path to AGI. Proceedings of Machine Learning Research, 235, 36308–36321. https://proceedings.mlr.press/v235/morris24b.html

Museum of Modern Art. (2022). Refik Anadol: Unsupervised. https://www.moma.org/calendar/exhibitions/5535

Norman, M. (2026, June 25). Refik Anadol, the art world’s happy warrior for A.I. The New Yorker. https://www.newyorker.com/culture/persons-of-interest/refik-anadol-the-art-worlds-happy-warrior-for-ai

Onkaos. (n.d.-a). Appropriate Response. https://onkaos.com/work/appropriate-response/

Onkaos. (n.d.-b). Botto. https://onkaos.com/work/botto/

Onkaos. (n.d.-c). Neural Decay series. https://onkaos.com/work/neural-decay-series/

Palazzo Citterio. (2026). Sougwen Chung: Body Machine (Meridians). https://palazzocitterio.org/en/news/mostra/sougwen-chung-body-machine-meridians/

Parisi, L. (2019). The alien subject of AI. Subjectivity, 12, 27–48. https://doi.org/10.1057/s41286-018-00064-3

Pasquinelli, M. (2023). The eye of the master: A social history of artificial intelligence. Verso.

Pennartz, C. M. A. (2026). How can we validate theory-derived indicators of consciousness in artificial intelligence? Trends in Cognitive Sciences, 30(7), 573–574. https://doi.org/10.1016/j.tics.2026.01.011

Polymarket. (2026). OpenAI announces it has achieved AGI before 2027? Retrieved August 18, 2026, from https://polymarket.com/event/openai-announces-it-has-achieved-agi-before-2027

Possible Futures. (2023). Refik Anadol × Yawanawá: Winds of Yawanawá. https://www.possiblefutures.one/nft

Ridler, A. (2018a). Myriad (Tulips). https://annaridler.com/works/myriad-tulips

Ridler, A. (2018b). Mosaic Virus 2018. https://annaridler.com/works/mosaic-2018

Ridler, A. (2019). Mosaic virus 2019 [Video installation]. https://annaridler.com/works/mosaic-2019

Russell, S. J., & Norvig, P. (2021). Artificial intelligence: A modern approach (4th ed.). Pearson.

Schwarzman Centre for the Humanities. (2026). Anna Ridler: A perfect language of images. https://www.schwarzmancentre.ox.ac.uk/whats-on/anna-ridler-a-perfect-language-of-images-sl92

Sedbon, M. (2024). Bio economics [Biohybrid installation]. Hybrid Forms Lab. https://hybridformslab.com/projects-ongoing/bio-economics/

Sedbon, M. (n.d.). Living computations. https://michaelsedbon.com/Living-Computations

Serpentine Galleries. (2018). Ian Cheng: BOB. https://www.serpentinegalleries.org/whats-on/ian-cheng-bob/

Shneiderman, B. (2022). Human-centered AI. Oxford University Press. https://doi.org/10.1093/oso/9780192845290.001.0001

Silva, D. (2024, August 7). Mario Klingemann: Exploring the frontiers of AI, data poisoning, ethical challenges and art. CLOT Magazine. https://clotmag.com/interviews/mario-klingemann-exploring-the-frontiers-of-ai-data-poisoning-ethical-challenges-art

Smithies, J. (2024, October 31). Leveraging scale and opportunity: Digital HASS in Australasia [Paper presentation]. 2024 eResearch Australasia Conference, Melbourne, Australia. https://researchportalplus.anu.edu.au/en/publications/leveraging-scale-and-opportunity-digital-hass-in-australasia/

Sommerer, C., & Mignonneau, L. (2026). Flâneur. https://interface.ufg.ac.at/christa-laurent/Flaneur.html

Sommerer, C., & Mignonneau, L. (n.d.). Artworks. https://interface.ufg.ac.at/christa-laurent/

Steyerl, H. (2023). Mean images. New Left Review, 140/141, 82–97. https://newleftreview.org/issues/ii140/articles/hito-steyerl-mean-images

Superflux. (2017–2019). Mitigation of shock (London) [Installation]. https://superflux.in/index.php/work/mitigation-of-shock/

Taylor, G. D. (2014). When the machine made art: The troubled history of computer art. Bloomsbury Academic.

Tegmark, M. (2017). Life 3.0: Being human in the age of artificial intelligence. Alfred A. Knopf.

The Garden of Earthly Delights. (n.d.). Mario Klingemann: The Garden of Ephemeral Details. https://thegardenofearthlydelights.art/Mario-Klingemann

The Lumen Prize. (2018). Gold Award winner 2018: Mario Klingemann – The Butcher’s Son. https://lumenprize.org/2018-winners/mario-klingemann

Toister, Y., & Zylinska, J. (2025). Image thinking after artificial intelligence. Journal of Visual Culture, 24(3), 431–451. https://doi.org/10.1177/14704129251405715

Victoria and Albert Museum. (2022, December 14). The algorithmic gesture: Sougwen Chung’s MEMORY. https://www.vam.ac.uk/blog/digital/the-algorithmic-gesture-sougwen-chungs-memory

Wang, G. (2025). Curving spontaneously: Sougwen Chung’s collaborative paintings with algorithmic systems. Journal of Visual Art Practice, 24(4), 483–513. https://doi.org/10.1080/14702029.2025.2567148

Whitelaw, M. (2004). Metacreation: Art and artificial life. MIT Press.

Young, L. (Director). (2016a). In the robot skies [Film].

Young, L. (Director). (2016b). Where the city can’t see [Film]. University of Salford Art Collection. https://artcollection.salford.ac.uk/liam-young-where-the-city-cant-see/

Young, L. (Director). (2021). Planet City: A city for 10 billion people [Film]. https://www.liamyoung.org/projects/planet-city

Young, L. (2023). The great endeavour [Video installation]. La Biennale di Venezia. https://www.labiennale.org/en/architecture/2023/dangerous-liaisons/liam-young

Zeilinger, M., & Johns, D. (2025). AI art and the foreclosure of ecological agency: A critique of Refik Anadol’s Echoes of the Earth. In J. Markelj & C. C. Bueno (Eds.), Vectoral agents: Power in the age of planetary computation (pp. 21–31). Institute of Network Cultures. https://networkcultures.org/wp-content/uploads/2025/11/Network-Notions-4.pdf

Zylinska, J. (2017). Nonhuman photography. MIT Press.

Zylinska, J. (2020). AI art: Machine visions and warped dreams. Open Humanities Press. https://www.openhumanitiespress.org/books/titles/ai-art/

Zylinska, J. (2022). Beyond machine vision: How to build a non-trivial perception machine. Transformations, 22(1), Article 18. https://doi.org/10.33011/tf.v22i1.4677

Zylinska, J. (2023a). The perception machine: Our photographic future between the eye and AI. MIT Press.

Zylinska, J. (2023b). Art in the age of artificial intelligence. Science, 381(6654), 139–140. https://doi.org/10.1126/science.adh0575

Zylinska, J. (2024a). Diffused seeing: The epistemological challenge of generative AI. Media Theory, 8(1), 229–258. https://doi.org/10.70064/mt.v8i1.1075

Zylinska, J. (2024b). After AI art. View: Theories and Practices of Visual Culture, 38. https://doi.org/10.36854/widok/2024.38.2870

Zylinska, J. (2025). Bio-AI: The aesthetics and ethics of data animism. Culture Machine, 24. https://culturemachine.net/vol-24-aesthetics-biomachine/bio-ai-the-aesthetics-and-ethics-of-data-animism/