From instructions to intelligent synthesis: conceptual art, distributed creativity, and the posthuman turn in AI

Abstract In recent years, the groundbreaking development of AI technologies such as text-to-image generation has revealed the unprecedented potential of machines in the realm of artistic creation. This has reignited and intensified widespread reflections on historical issues in artistic practice, including ontological questions, the identity of the author, the configuration of human–machine relationships, and the positioning of creativity. To address these concerns, this paper mainly seeks to draw inspiration from the important artistic thoughts of two conceptual artists who were active in the 1960s: Yoko Ono and Sol LeWitt. It also explores the interconnectedness and continuity between generative AI art practices and the legacy of conceptual art, tracing the historical trajectory of the relationship between conceptual art and digital art. Beyond the pivotal artistic practices of these two pioneering conceptual artists, the comparative analysis also encompasses the examination of additional associated artistic endeavors in relation to text-guided image synthesis. This broader scope of artists aims to more comprehensively support the argument on the impact of artificial intelligence on participatory frameworks and the concept of authorship within future creative practices. By delving into a discussion on dematerialization with conceptual art’s focus on concepts over material objects, the authors explore the potential for a posthumanist paradigm in AI art practices, to consider the reconfiguration of creativity in human–machine collaborations. Through speculative explorations of AI visualizations of Ono's and LeWitt's instructions and evaluating the role of "prompt engineering" in collaborative creation, new forms of distributed creativity that respond to posthumanist provocations are discovered. Leveraging the posthumanist emphasis on the “biological and microbiological entanglement with nonhumans”, this approach aids in dislodging the centrality of the individual or the artist within the creative process (Herbrechter 2023, p. 8). This critical examination highlights the disruptive potential of conceptual art in reshaping the notion of authorship within its symbiotic relationship with technological culture. It reveals the dynamic and co-evolutionary relationship between the conceptual art discourse and its reinterpretation within the emerging computational participatory structures of the artistic frontier. By drawing on the legacy of conceptual art, this study attempts to reconcile the tensions between conceptual completion, technological participation, artistic ontology, and distributed creativity, providing theoretical support for current AI art practices and driving the ongoing reconsideration of posthuman creative partnerships in technology mediated artistic production. 1 Historical connections between conceptual art and AI art practice Conceptual art emerged as a definable movement in the 1960s, prioritizing the ideas or concepts behind the work over the finished art object. For LeWitt, the concept is central, while the execution or realization of the artwork is merely "a perfunctory affair".Footnote 1 This shift led to a move away art being defined by artistic mediums and material objects as discussed by Lippard in the process of dematerialization, and the emergence of participatory tendencies in conceptual art, which in turn prompted reflections on a series of questions regarding artistic ontology and notions of authorship.Footnote 2 Key figures of the avant-garde collective Fluxus like John Cage, George Brecht, La Monte Young, and Yoko Ono helped develop highly conceptualized art forms in the late 1950s. These experimental practices included, event scores, which are instructional pieces as "brief, simple, colloquial statements" that are intended for action and performance by participants.Footnote 3 This text-based art form as descriptions of actions allowed anyone to execute the work anywhere, opening up to diverse interpretations and variations of the artwork. Art historian Shanken has commented that this text-based art form was a strategic means for conceptual artists to explore networks of meaning and knowledge structures, where the artistic press happens in the interpretation and enactment of the event score.Footnote 4 In this way, conceptual art becomes a process of meta-critique and self-reflection, where the creative process is left to the performer/participant. Furthermore, the participatory role of audiences was later developed in the 1990s when social practice was introduced into art. The notion of “participatory art” as discussed by Claire Bishops broke down the conception of the artist as an autonomous producer, redefining the audience as a co-producer of the artwork.Footnote 5 The participatory turn in contemporary art was interpreted by the French art critic Nicolas Bourriaud as a “relational aesthetics”, in which art happens in the very relations between disparate elements, possessing an inherent "transitivity". This notion of "transitivity" signifies art's capacity for more expansive associations, encompassing the human, the non-human, and the environmental. It denies the circumscription of art to any singular, delimited "place of art", instead advocating for an artistic process that remains perpetually unfinished and open-ended.Footnote 6 Artistic engagements with technology emerged alongside conceptual art, where Shanken has characterized both as components of a broader social transformation from the industrial machine age to the so-called post-industrial information age.Footnote 7 In the social and political environment of the 1960s, with on the one hand, the Vietnam war made artistic uses of technology problematic, and on the other hand, conceptual art gained increasing influence due to its attacks on the objects of Modernism. As a result, conceptual artists deliberately kept at a distance from technology.Footnote 8 As digital technologies based on computers became increasingly advanced, digital art and technology gradually became more prominent in the art world, despite the digital divides between the realms of contemporary art and new media art, as described by Bishops.Footnote 9 Yet, when placing conceptual art and digital art practices side by side, we can clearly capture their close connections from their generative and participatory qualities. The concepts of interactivity and participation in digital art do not differ fundamentally, only expanding from human-to-human relationships to human-to-machine relationships. In Kwastek's definition, Interactive Art refers to digital artworks that require the audience to engage in activities beyond mere psychological reception, where the term "interactive" is distinguished from "participatory" or "collaborative" only by whether modern technology is utilized.Footnote 10 The concept of interactivity was first used to refer to various types of feedback processes in cybernetics, and in the 1960s, the first systems enabling real-time interaction between humans and machines were constructed, giving rise to the field of human–computer interaction (HCI) research focusing on user interfaces and input media development.Footnote 11 The current enhancement of AI capabilities has advanced the discussion of the human–machine creative relationship to a new stage, urgently necessitating a posthuman perspective to accomplish the "seamlessly articulated" of human existence and intelligent machines.Footnote 12 In the past, art has been almost entirely viewed as a practice dependent on human existence, essentially humanistic in nature. While in relational aesthetics, Bourriaud allows for the emergence of intersubjectivity or shared experiences based on the coexistence of human and non-human actors, he still emphasizes the necessity of introducing human interaction for art to present its materiality in order for art to and truly exist.Footnote 13 This framework remains applicable to the current posthuman perspective focused on the coexistence and co-creation of humans and non-humans, as our limited technological capacities preclude the discussion of purely non-human intersubjectivity exchanges. Within this framework, the rapid advancements in text-guided GenAI technologies have created new contexts for human–machine interaction, in which "prompt engineering" will become key to revealing the posthuman distributed creativity paradigm. Additionally, from a historical perspective, the instruction-based and rule-driven practices in conceptual art paved the way for the conceptualization of generative and algorithmic art. According to Galanter's definition of generative art, the key is that the artist relinquishes partial or full control to an external system (such as a mathematical equation or executable algorithm).Footnote 14 And the legitimacy of the autonomy of the external system is largely grounded in the foundation laid by conceptual art. Conceptual art advocates that the artistic concept serves as an automaton for artistic production, where founding member of Fluxus, George Maciunas specifically emphasized that the artist only needs to create a "concept or method", after which the" form can be created independently of him."Footnote 15 Maciunas’s perspective is consistent with LeWitt's view that "The idea becomes a machine that makes the art."Footnote 16 As Bourriaud described, the avant-garde artworks of the 1960s were more akin to "a programme to be carried out" and "a model to be reproduced."Footnote 17 Shanken has understood this proposition as a “quasi-mechanical” production process, to explain the complexity and conflict when conceptual art intersected with art and technology.Footnote 18 The final output of a formally resolved artwork with commodity potential was allowed to be created entirely independent of the artist, legitimizing the intervention of audience participation and external systems in the artistic production. However, in the past, the autonomy of external systems was heavily constrained by technological limitations, whereas the recent advancements in GenAI technologies have significantly increased the participatory capacity of external systems in artistic practices, requiring us to re-examine the ontology of art, the creative relationships between humans and machines, and the configuration of freedom and control under technological participation. 2 Participatory frameworks In the examples we used to discuss AI art practices, the "transitive" quality mentioned earlier is fully visible in Yoko Ono's instruction paintings, as she intended for the works to be seen as instructions, while retaining a tripartite structure of concept, object, and the work-to-be-realized.Footnote 19 Ono referred to her works as "unfinished paintings" as she sought to introduce participatory elements and to recast “art from a finished object to an unfinished process, to be completed by others".Footnote 20 In the visual arts tradition, the relationship between audience and work was one of physical remove; yet Ono's productions demanded interaction in a novel form to move beyond “scopic exclusionary roles” and erase the “sacred distance” between “object and participant”.Footnote 21 In this subverted artist-audience relationship, the audience is no longer confined to the role of onlooker or spectator, but instead invited to become a co-producer of the artwork. Through the influence of conceptual art, audience participation has become a hallmark of contemporary art, but in the current technological context, we can observe that the participation of GenAI in artistic practices is posing significant challenges to this relational framework, exhibiting regressive tendencies on artistic practice. Significant advances in the burgeoning field of text-guided digital image generation arose in 2014 with the introduction of generative adversarial networks, thereby inaugurating a new paradigm of algorithmic creativity through linguistic prompts. Further breakthroughs materialized in 2015 as Google unveiled Deep Dream, portending the latent potential for visual renditions of conceptual directives encoded in natural language. Subsequently, OpenAI's release of CLIP in January 2021 represented a watershed, as it demonstrated an unprecedented capacity for associative retrieval between textual and pictorial modalities.Footnote 22 Continued refinement saw these promising techniques integrated into cutting-edge diffusion models, thereby optimized for autonomous synthesis from semantically anchored descriptions. Under the aegis of contemporary computational prowess, users with scant knowledge of underlying deep learning pipelines can now entirely outsource imaging making to AI assistants configured through minimal verbiage. Such technical virtuosity has given rise to the possibility of a new mode of art production, where when GenAI technology intervenes in the realization of artworks, the audience's capacity to co-produce the artwork has been delegated to technology. To retain the past experiences of audience participation, we can envision a scenario as a counterexample: a contemporary individual provides prompts based on Yoko Ono's instruction as input to train a generative model, producing a text-based generative visual output. In such a paradigm, the level of substantive engagement with and processing of Ono's underlying artistic concepts by the human participant would be virtually non-existent. The participant's role becomes little more than a passive provider of the impetus for the human–AI interaction, with the AI system independently synthesizing the resulting visual realization without meaningful consideration of conceptual intention. This risks reducing the nature of participation to a hollow illusion of democratic inclusion, where the participant becomes an incidental vessel transmitting Ono's directives rather than an active determiner of conceptual outcomes. Thus, the new technologies have rendered the past participatory frameworks problematic. In fact, in contemporary digital art practices, we can clearly observe that technological mediation is attempting to intervene in the relationship between the artist and the audience, potentially leading to a new triangular relational framework. At the current stage, the fluidity and center of change in the artistic configuration between the three parties lies between the artist and the technological mediator, while the role of the audience seems to have a tendency to regress back to the position of a passive receiver in a unidirectional information exchange system, or even worse. An interesting example comes from the German artist Mario Klingemann, with his artwork titled, Appropriate Response. In the artwork, he collected 60,000 different quotes from the Internet and used them to continuously train a model to generate aphorisms or sayings. What's more intriguing is that the artist chose a prayerful kneeling trigger device, allowing the audience to receive the AI model's randomly generated "instructions" in a religious-inflected posture. The statements are described as "instructions", because, based on the artist's observations of audience participation, people often reflect on what the generated sentence means to them in the context of their own experiences, akin to fortune-telling or divination. This spectacle of audience participation undoubtedly challenges our past experiences of human–machine interaction, where the roles of human and machine in the feedback system have been reversed. In Ono's instruction painting practice, her primary interest lay in "painting to construct in your head," and in this work, the audience feedback she hoped for is similarly occurring.Footnote 23 In this scenario, the artist's trained model acting as the artist's proxy seems to have replaced Ono's role in the past participatory framework, interacting with the audience through output of instructions. Moreover, the false contingency created by computational technology appears to bestow a higher authority on the new "artist" role, at least in the past, the artist's discourse would not be interpreted as "divine oracle," because both the audience and the artist were clearly aware that we are fundamentally no different as a species. This audience feedback experience reflects a (false) self-perception of power relations shaped by cultural conditions and dependent on ideological structures. This highlights the contradictions and conflicts between deep-rooted humanistic core of art and the current technological context. In Ono's concept, the cognitive process of thought was seen as part of the creative process, and she believed that visualizations within the mind constituted art in themselves, surpassing mere representation through external forms, hinting at a neural network-like concept.Footnote 24 This interactive artistic creation process has evolved in this digital art case into what Slovenian philosopher Žižek calls the "uncanny experience of the human mind directly integrated into a machine," where the audience can no longer maintain the "illusion of the autonomy of personhood."Footnote 25 Žižek reminds us of the necessity to activate posthuman body theory when constructing new audience participation frameworks. The Human Genome Project has made us start to understand our bodies in the binary code of computer programming, from a genetic perspective, we are all encoded “informational entities”, just like all non-human others.Footnote 26 On this posthumanism premise, we can restart the participatory heritage of conceptual art mentioned earlier and connect it to artistic practices about artificial intelligence, where the substitution of interactive agency introduced by technological mediation opens up a form of human–machine collaborative practice. When we position humans and machines as equal conversational partners, then humans and external systems are integrated as co-creators of meaning. Therefore, in the new triangular participatory framework, the technological mediator becomes a creative agency for both the artist and the audience, performing some of the tasks they used to undertake in artistic production. Such a re-conception holds potential for configuring posthumanist production attuned to artistic missions navigating the complexities of human and non-human cognition in co-determined partnership. 3 Reconfiguration of creativity From a human-centric perspective of technology as a tool, the rapid advancements of GenAI may catalyze a radical democratization of the painting domain, akin to the impact of photography on visual arts. The technical mastery of artistic mediums that has constructed traditional visual art forms such as painting, sculpture, has built the moat of the fiefdom around the artist's authorial identity, disregarding the role of the audience in creative interpretation. From the perspective of breaking through technological limitations, Gen AI technology, like photography, challenges this artistic authoritarianism. However, the difference between photography as a record of human gestures and Gen AI technology is that the latter no longer requires any “analogous linkage to the subject” but derives everything from computation.Footnote 27 Imagine, a photograph taken with a camera of Ono's instructions would not be recognized as an artwork, but what about inputting Ono's instructions into a text-guided GenAI model? Undeniably, this is no longer a simple reproduction of the original work, as it involves the intervention of new non-human creativity. There is a need to, therefore, re-examine the creativity in Gen AI artworks. Based on American philosopher Nelson Goodman's classification of the essential nature of art into autographic (singular, based on the principle of authenticity, e.g. painting) and allographic (potentially repeatable, based on ideational concepts, e.g. texts or notations), Kwastek has analyzed interactive art, arguing that the specific interactive artworks have been endowed with a conceptual structure based on "constitutive rules," thus being autographically cast as a technical system.Footnote 28 This structure based on "constitutive rules" is also the core, and even more visible, in generative art, so we can determine the meaning premise of generative art through Kwastek's discourse. In this context, our task is to discuss where does human creativity lie and how is the creative relationship configured within this autographic technical system? The propositions of conceptual art allowed for non-human intervention in artistic production. The drive towards dematerialization and the relinquishing of control over the output manifests with subtle distinctions in the personal philosophies and approaches of different artists. LeWitt affirmed that while a work's ultimate form is malleable, it must originate from a generating idea to qualify as art.Footnote 29 Implicitly, he regarded physical realization as a necessity for an artwork to be perceivable and thus complete. The output oriented approach is clearly reflected in his wall paintings, which require different assistants to carry out the instructions even after he has written them. Ono diverged in prioritizing the autonomy of the conceptual over obligatory materialization, as we mentioned earlier, advocating for the realization of instructional works in the audience's mind. These diverging propositions on the relationship between conception and embodiment carry implications for evaluating emergent artistic paradigms mediated through technology. LeWitt's linear conception-to-completion trajectory informs perspectives privileging AI image generation's realization of intentionally seeded ideas. Alternately, Ono's contention that mental concretion alone comprises art challenges strict correlations between computational output and original conceptual intention. Reconciling these discordant conceptual legacies risks reductive equivocation; their intricacies merit sensitive consideration in reformulating participatory modes amid human–machine creative collaborations. Conceptual differences between LeWitt and Ono are more evident from their artistic practices. Ono's instruction painting inaugurated a text-only artistic form that decoupled the ontological substance of the creative work from strict materialization. In contrast, in the case of Sol LeWitt's wall drawings, he too would write down instructions but would require assistants to execute the final work at specific times and locations according to his directives. The ultimate outcomes often varied due to the involvement of different assistants, even as they adhered to the same set of instructions. Similarly, in the 1960s, the Japanese avant-garde artist Yamaguchi developed a form called "order art" where he would provide manufacturing instructions to factories to create the artworks.Footnote 30 In these early instruction-based artistic practices, external systems had already intervened in the production of the artworks; for LeWitt and Yamaguchi, it was the commissioned assistants or art fabrication factories, while for Ono, it was the audience's minds. In contrast to these pioneering conceptual art practices, the contemporary manifestation of generative art has effectively substituted the external system with the machine. Tracing the evolutionary trajectory from instruction-based art to generative art, software artist Casey Reas had already experimented with machine programming as the external generative system to replace LeWitt's team of assisting draftsmen, as early as 2004. In the artwork, Software Structures (2004) Reas asked several programmers to write LeWitt's instructions in various programming languages, ultimately outputting various forms of wall drawings, exhibiting a strong similarity to the simulated interpretations carried out by the assistants executing the wall paintings, i.e., the combination of the constitutive factors (defined by the artist) and the contingency factors (uncontrollable by the artist) in interactive art. Moreover, the influence exerted by this external computational generative system as a contingency factor on the final artistic outcomes lends credence to the recognition of its own creative capacities. Conceptual art's shifting artistic creation to idea-based forms disconnected from material form helps us recalibrate the participatory dynamics between humans, AI, and conceptual content. In the participatory interactions of these conceptual artworks, the creative control of the work is partially ceded to external generative systems such as GenAI. The emerging text-guided image synthesis directly visualizes textual prompts, separating artistic qualification from fixed physical realization. Under such a paradigm, the conceptual components embodied in language merit recognition as embodiments of human creativity, separate from resultant computerized outputs lacking such derivation. In this case, human ingenuity within text-to-image generation resides not in the final output (i.e., the digital image artifact).Footnote 31 From this perspective, the artistic work as a visual realization of the concept can no longer be solely attributed to human creativity, as conceptual art's characterization of the physical production phase as an ancillary "perfunctory affair," the artist's determination over the final product realization is abandoned.Footnote 32 That is to say, the human's mission in the artistic creation process is no longer to output a finite art object, but to complete a framework of constitutive rules. From Ono's concepts, AI-facilitated finalization of imagined visual works constitutes a redundant, meaningless, and even counterintuitive step. However, as LeWitt described, conceptual art still entailed a continuous process of moving from concept to objecthood, a theoretical framework that more pragmatically acknowledges GenAI's contribution in art production. By moving beyond anthropocentrism, the role of AI image synthesis systems intimates how human–machine interaction amalgamates a novel concept of distributed creativity within artworks. Rather than negation of conceptual art's dematerialized impulses, this signals its expanded reformulation accommodating technoculture's symbiotic paradigms of co-authorship between human and computational domains. And also, such a reformulation intimates how human–machine collaboration might carry conceptual art's mission forward in framing technologically mediated creative practice. Beyond the legacy of conceptual art, the "prompt engineering" that co-exists with AI technology provides us with important insights into the essence of human–machine collaborative artistic process and offers a contemporary context for discussing the positioning of human creativity and the configuration of the human–machine creative relationship. In the 1950s, Swiss sculptor Jean Tinguely created a series of drawing machines called Méta-matic. Through the superposition of different harmonic oscillations, these devices could produce beautiful geometric patterns. Tinguely claimed that anyone could use them to create their own abstract paintings, but actually the audience could only do so by turning the machine on or off, as it was essentially an automatically rotating device.Footnote 33 In its initial reception, Tinguely's Méta-matic project was viewed as a biting satire targeting the human obsession with and veneration of the machine, while also serving as a self-reflexive commentary on the pretensions of the vanguard art milieu.Footnote 34 Recontextualizing this seminal work through the prism of our present posthumanist aesthetic framework can elucidate the evolutionary trajectory of human and generative system creativity, thereby cultivating a more refined understanding and reconfiguration of the dynamic between external machine-driven production systems and the human. In Tinguely’s artistic process, no one could control the final output, nor could they even influence it to a great extent (apart from choosing when to stop the drawing). In this participatory automatic drawing, we can hardly locate any conscious creative behavior by the human (audience) to pinpoint their creativity. This case is similar to our previous speculation of inputting Ono's instruction paintings into a text-guided GenAI, where the audience becomes a hollow vessel in the process of executing the artistic concept. However, the difference is that GenAI allows humans a certain degree of control over the final output, even though you still cannot predict the exact result. This is the creative participation that humans acquire through prompt engineering in the human–machine interaction process. Prompt engineering involves humans adaptively reconfiguring natural language interfaces to suit AI systems' capabilities via understanding how to effectively craft prompts that leverage neural network training and syntax modulation.Footnote 35 This signifies interaction comprises a cooperative dynamic where humans and machines each play discrete yet interrelated roles. Within AI image generation, humans conceive and input conceptual parameters to which systems respond through iterative visual synthesis resembling commissioned works. Computer researcher Oppenlaender characterizes this as a refining process where outputs are improved over numerous stages haltable by human collaborators at any juncture.Footnote 36 However, current technological limitations often require additional human post-production due to text-to-image results' imperfections. Consequently, determining final outputs fundamentally remains dependent on supplemental human participation through prompt engineering rather than full computational autonomy. Prompt engineering epitomizes humans' adaptive role within such hybrid praxes of collaborative meaning-making. 4 Conclusion In conclusion, this paper has explored the historical connections between conceptual art and the emerging practices of generative AI art. It has shown how the conceptual art movement's emphasis on dematerialization, rule-based practices, and participatory frameworks laid important groundwork for the capabilities of modern text-to-image AI systems. However, the paper also highlights the tensions that arise as these AI technologies intervene in traditional participatory models of art. The delegation of creative agency to external generative systems challenges the roles of artist and audience, potentially diminishing meaningful human engagement. To address the role of the human in the artistic process, the paper calls for a reconfiguration of creative relationships, drawing on posthumanist perspectives to reimagine human–machine collaborations. Specifically, it positions "prompt engineering" as a crucial site for understanding the evolving dynamics of technologically enabled art production, where the traditional boundaries between artist, artwork, and audience are in flux. By revisiting the conceptual art legacy and its focus on conceptual over physical realization, the paper argues that we can develop theoretical frameworks to support emergent AI art practices and drive a continued rethinking of the posthuman creative partnerships facilitated by technological mediation. This offers the potential to reconcile tensions between conceptual completion, technological participation, artistic ontology, and distributed creativity in the context of AI-powered artistic production. Data Availability Data sharing is not applicable to this article, as no datasets were generated or analysed during the current study. This article is a work of conceptual and art-historical analysis; all material on which the argument draws consists of published scholarship and publicly accessible artworks, which are cited in full in the reference list and in the source of images. Notes Lewitt (1967), p. 79. Lippard (1997). Lu(2013), p. 1. Shanken (2002), p. 343. Bishop (2022), p. 2. Bourriaud (2002), p. 11. Shanken (2002), pp. 433–438. Shanken (2002), pp. 433–438. Bishop (2012). Hayles (1999). Bourriaud (2002), p. 9. Galanter (2003). Maciunas (1988), pp. 25–32. Lewitt (1967), p. 79. Bourriaud (2002), p. 32. Shanken (2002), p. 436. Lu (2013), p. 6. Lu (2013), p. 10. Basar (2000), p. 85. Oppenlaender (2022), p. 193. Lu (2013), p. 13. Lu(2013), p. 13. Shaw (2008), p. 88. Bourriaud (2002), p. 32. Lewitt (1967), p. 79. Paul (2016), p. 119. Oppenlaender (2022), p. 193. Lewitt (1967), p. 79. Oppenlaender (2022), pp. 193–195. Oppenlaender (2022), p. 198. References Basar S (2000) YES Yoko Ono. Cabinet Magazine, 8:84–86 Bishop C (2012) Digital divide: contemporary art and new media. Artforum. https://www.artforum.com/print/201207/digital-divide-contemporary-art-and-new-media-31944 Bishop C (2022) Artificial hells: participatory art and the politics of spectatorship. Verso, London, p 2 Bourriaud N (2002) Relational aesthetics. Les Presses du réel, Dijon Galanter P (2003) What is generative art? Complexity theory as a context for art theory. In: Proceedings of the international conference on generative art, Italy Hayles NK (1999) How we became posthuman: virtual bodies in cybernetics, literature, and informatics. University of Chicago Press, Chicago Herbrechter S (2023) Posthumanism and Aesthetics. In: Critical posthumanism–Stefan Herbrechter Kwastek K (2013) The aesthetics of interaction in digital art. In: Aesthetics of interaction in digital art. The MIT Press. https://doi.org/10.7551/mitpress/9592.003.0008 Kwastek K (2013b) The aesthetics of interaction in digital art. The MIT Press, Cambridge, p 7 Lewitt S (1967) Paragraphs on conceptual art. Art Forum 5(10):79–83 Lippard LR (1997) Six years: the dematerialization of the art object from 1966 to 1972. University of California Press, Los Angeles Lu H, Yi E (2013) Instruction paintings: Yoko Ono and 1960s conceptual art. Shift Grad J Vis Mater Cult 6:1–20 Maciunas G (1988) Neo-dada in music, theater, poetry, art. In: Fluxus: selections from the Gilbert and Lila Silverman Collection, pp 25–32 Oppenlaender J (2022) The creativity of text-to-image generation. In: Proceedings of the 25th international academic Mindtrek conference, pp 192–202. Paul C (ed) (2016) A companion to digital art. Wiley, Chichester, West Sussex Shanken EA (2002) Art in the information age: technology and conceptual art. Leonardo (Oxford) 35(4):433–438 Shaw DB (2008) Technoculture: the key concepts, English ed. Berg, Oxford Žižek S (2004a) Organs without bodies: Deleuze and consequences. Routledge, New York Žižek S (2004b) Organs without bodies: on Deleuze and consequences. Routledge, London Source of image LeWitt S. Wall Drawing #16. https://massmoca.org/event/walldrawing16/. Accessed on 10 January 2025 Tinguely J. Méta-matic No. 8. http://radicalart.info/process/vibration/metamatic/index.html. Accessed on 10 January 2025 Reas C. Structure #003B. https://artport.whitney.org/commissions/software-structures-2016/s3_2/. Accessed on 10 January 2025 Klingemann M. Appropriate response. https://onkaos.com/work/appropriate-response/. Accessed on 10 January 2025 Author information Authors and Affiliations Corresponding author Ethics declarations Conflict of interest Authors disclosed no relevant interests. Additional information Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Rights and permissions Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. About this article Cite this article Xie, Y., Wong, A.L. From instructions to intelligent synthesis: conceptual art, distributed creativity, and the posthuman turn in AI-powered artistic practices. AI & Soc (2026). https://doi.org/10.1007/s00146-026-03316-y Received: Accepted: Published: Version of record: DOI: https://doi.org/10.1007/s00146-026-03316-y