Models of the World of Artificial Intelligence Art

This research explores the dynamics of social conflict within the AIArtworld, a speculative domain characterized by advanced AI consciousness potentially evolving towards superintelligence. It utilizes a theoretical basis from art sociology and interactionism, defining AIArtworld as involving both human and non-human entities crucial for art production. The study leverages Actor-Network Theory for deeper insights into actor relationships and employs Pierre Bourdieu’s perspectives to simulate conflicts, testing these theories on the Metaculus platform. This research aims to understand and predict how these conflicts could influence future artistic expressions and societal interactions within this AI-influenced realm.


Tomáš Marušiak MFA

Masaryk University Brno, Faculty of Arts, Department of Musicology (Czech Republic)

Cultcode-Institute  of Visual Art (Slovak Republic )

Ongoing Research 2024:

Models of the World of Artificial Intelligence Art

This research design was conceived to provide insights into the dynamics of social conflict between human actors and conscious non-human actors within the speculative realm defined as AIArtworld. This space is characterized by the presumption of advanced consciousness in artificial intelligence systems and the potential predictability of the evolution of general artificial intelligence towards superintelligence. The theoretical foundation of the research design is based on an analysis of processes in art, specifically in the context of art sociology. From the theory of interactionism, the concept of AIArtworld is derived, which includes social processes in the field of artificial intelligence art, defining AIArtworld as a set of all involved human and non-human actors and organizations whose activities are essential for the production of art and significantly influence art. The theoretical analysis further uses assumptions from research field „Speculative Modeling of Future Aesthetics and Artificial Intelligence Art“, which to some extent allows for the feasibility of consciousness in artificial intelligence systems as well as its modeling through artistic research based on the transposition between scientific and artistic approaches. To enhance operability with an expanded sphere of actors, a theoretical approach based on Actor-Network Theory is proposed, which allows for a comprehensive understanding of the relationships between actors in AIArtworld. An approach inspired by Pierre Bourdieu, who interpreted society as a space occupied by antagonistic positions of actors, thus providing a basis for speculative simulation of conflict, was used to identify the fundamental characteristics of AIArtworld. This conflict represents the central interest of the experimental part of the research, which is focused on applying predictive design on the Metaculus platform to verify theoretical assumptions and provide a material-scientific basis for analysis. Such an approach allows for opening a discussion and providing a comparative parameter for reflection on the current non-speculative field of AIArt, while also offering answers to the key issue: How the simulation of social conflict between human and conscious non-human actors in the speculative field of AIArtworld can reveal potential dynamics and relationships shaping their future directions, thus what AIArtworld might look like.

Danto in his essay „The Artworld“ (1) develops the concept of „the world of art“ – Artworld, which points out that the understanding of art is derived from complex contextual ties, within which art is presented and interpreted. This view allows understanding art as a phenomenon that cannot be fully understood or appreciated without considering its social and cultural context. Dickie’s contribution to discourses on Artworld lies in the articulation of the institutional theory of art, which frames the transformation of an artifact into an artwork, when it is assigned this status within a certain social framework, and this status is assigned by an entity that has the ability to aesthetically perceive the artifact and also acts on behalf of a certain social institution. This theory thus emphasizes the role of institutional authority and social convention in defining what is considered art (2, 3). Pierre Bourdieu in his model of artistic fields emerges as a critic of the Artworld concept, rejecting the notion of context as such and pointing out that this approach may obscure the real conflicts and power structures that shape art (4).

For future modeling, the concept of symbolic interactionism „social production“ by Howard S. Becker (5) seems to be the most appropriate approach, where Artworld is defined as a world: „consisting of all interested persons and organizations whose activities are essential for the production of a certain kind of events and objects that the art world significantly produces.“ This perspective allows a broader understanding of art as a product of interactions between various social actors and institutions. This definition of the term „Artworld“ represents only a symbolic act, structure, or template, which will serve for subsequent modeling. Further, it is appropriate to consider that the current optics of Artworld will no longer be sufficient. The transformation of society under the influence of the onset of artificial intelligence essentially reshuffles the cards.

As a further foundational platform, „Life 3.0“ by Max Tegmark (6) can be utilized, in which Tegmark addresses the concept of humanity in the era of artificial intelligence, highlighting the urgency to explore the interactions between humans and advanced AI systems. He proposes a classification of life through three stages: Life 1.0, which includes biological life; Life 2.0, representing cultural life; and Life 3.0, which futuristically describes the era of development of general artificial intelligence (AGI) and superintelligence. This perspective has penetrated into the socio-technological-ethical aspects of AI, emphasizing the importance of formulating questions about the potential benefits and risks associated with AGI. Special attention is given to the concept of „Intelligence Explosion,“ a hypothetical scenario in which AGI surpasses human intelligence and uncontrollably accelerates its own development, potentially leading to a rapid increase in intelligence. Nick Bostrom discusses the topic of superintelligence in his book „Superintelligence: Paths, Dangers, Strategies“ (7), defined as a form of AI that significantly exceeds human intelligence in all aspects. This discussion is particularly relevant in the context of AI systems with cognitive abilities surpassing human possibilities. Tegmark’s work thus represents a key contribution to understanding and framing a context in which humanity can navigate a dynamic and uncertain environment where advanced AI changes the definition of what it means to be human.

For the purposes of defining the future model of AIArtworld, I suggest drawing from Howard S. Becker’s definitional stance and defining AIArtworld as a world consisting of all interested human and non-human actors and organizations that possess conscious AI systems and whose activities are necessary for the production of a certain type of events and objects that significantly produce art. Such a definitional framework of AIArtworld presupposes the functional integration of models based on speculative modeling of future aesthetics and artificial intelligence art into a simulated social narrative.

The supporting point for such modeling is the „Actor-Network Theory“ (ANT). ANT is an approach within sociological theories that is based on interdisciplinary approaches in the field of science and technology (8–10). This theoretical framework focuses on examining sociotechnical processes that are key to understanding the mechanisms leading to stabilization, success, or failure of technological and scientific innovations. One of the distinctive characteristics of ANT is its emphasis on the agency of non-human entities, such as technologies, objects, and other materials in social networks. This approach equitably considers both human and non-human actors in networks or systems, radically reevaluating the concept of agency in social sciences. The concept of „agencement“ or assemblage, which is key in ANT, suggests a process in which human and non-human entities, referred to as „actants,“ interact with each other and create networks. These networks subsequently produce specific actions or outcomes. According to ANT, the characteristics and properties of entities do not directly arise from their inherent attributes but are the result of their interactions and positions within the network. This perspective provides a viewpoint for understanding social and technical systems where the meaning and effects of actants are not predetermined but are formed through dynamic relationships. Its focus on the materialism of social life and the role of non-human actors offers a unique view of the interconnectedness of society and technology. However, ANT has also faced criticism for its relativistic stance, sometimes dense and technical language, and challenges related to the empirical operationalization of its concepts (11).

The AIArt model, as presented, might be considered an idealistic-ultimate conception, characterized also by the integration of hierarchically categorized sub-models resulting from modeling future aesthetics and art based on artificial intelligence. This approach suggests the existence of an intense relationship between two conscious entities, thereby opening a discourse on mutual interaction and collaboration between human and non-human creativity. However, it only indicates a limited framework of social interactions. Within the creation of a speculative dynamic scenario of social play, the induction of conflict is necessary on a broader scale. In this context, conflict is not just a basic sociological phenomenon referring to a process or situation in which two or more groups are involved in competition for resources, power, status, influence, or values, but also a manifestation of the neuronal activity of an individual or group within society. This activity is a response to changing situations, thereby highlighting the dynamics of social interactions and relationships between individual and collective actors. Therefore, the AIArt model should provide a basis for a deeper understanding of the potential of artificial intelligence in artistic production, which requires the modeling of social dynamics that may arise from such a partnership. In the context of social play, the AIArt model allows for the exploration of new paradigms of relationships and interactions

Pierre Bourdieu interpreted society as a space occupied by antagonistic positions held by individual actors. These actors continually engage in symbolic battles to gain favorable positions within the social space. In this context, inspired by Durkheim (12), Bourdieu emphasized the importance of a complex and differentiated view of societies. This approach allows for the recognition of differentiated societies, which are further divided into sub-social areas such as bureaucratic spheres of state administration, arts, media, religion, politics, or academia. Social phenomena can occur within individual fields as well as at the boundaries of multiple fields. Bourdieu’s conceptualization of the social field can be understood as a structure of objective relations between positions of individuals or groups competing for social legitimacy within it (13). The social field, like the social space, is defined as „a space of objective relations between positions,“ where symbolic conflicts or open battles are conducted aiming at transforming or preserving the existing structure of the field.

The fundamental proposed research task is the simulation of social conflict in the speculative field of AIArtworld among multiple competing models. This means between human and non-human actors in the sense of ANT. The methodology for constructing competitive speculative models, which will be set into conflict, is based on an approach that focuses on creating projected or alternative futures. Modeling uses case studies referring to completed AIArt projects. This approach provides an empirical basis for the development of speculative models, allowing thorough analysis and reflection on existing examples and their applicability in various contexts. Through another position, speculative-oriented artistic research is conducted. One of the assumptions is the existence of a social field, defined by the possibility of conflict between human and conscious non-human actors. This conflict reflects the tension arising from the integration of artificial intelligence systems into the artistic process and social perception of art. Thus, it mainly concerns conflicts between competing interests and not cooperating groups.

The construction of conflict is based on the assumption that during its simulation, the following will emerge: a) outlines of elements and functions of AIArtworld in relation to current scientific-technological possibilities such as the belief in the embodiment of human domains like rights into AI systems with advanced consciousness, b) processes associated with the inclusion or exclusion of actors, whose labeling is linked to artistic research and the availability of social capital. The socio-educational transformation affects the artist’s habitus, as artistic talent as a biological predisposition is replaced by strategic acting using Skills with Information Technology (14) or Science Capital (Archer et al. 2015), as cultural capital. Thus, both a conscious AI system and a human can possess the same type of capital, resulting in an equivalence of positions, c) positions of actors in AIArtworld post the singularity of AGI in AIArtworld models, which allows for the possibility of art production without human intervention, with the maximum ability to deduct aesthetic experiences and expectations of the recipient, moving him to the position of a passive perceiver as an owner, or a manager of aesthetic capital produced through AI. Such a model relies on the readability of the recipient’s neurobiological processes, the community of recipients influencing each other in perceiving aesthetic experiences and preferences, as shared ownership or availability of aesthetic capital produced through systems of artificial intelligence with advanced consciousness.

For the purposes of this simulation, the principle of speculative design is used, which does not prefer a specific future but offers society the opportunity to decide what future it prefers, while the affirmative approach decides without discussion. Such processes may have the potential to stimulate processes in society that will lead to its own construction of the future (16). For the application of conflict construction models, referring to the aforementioned, the most suitable environment is the Metaculus prediction market, which is built on the principle of collective forecasting and formulation of predictions (17). Characterizing Metaculus as a prediction market may be somewhat inaccurate, as traditional market mechanisms prioritize financial success and profit maximization through price signals. On the other hand, the goal of scientific knowledge is to expand understanding of the world, with key actors being scientists, researchers, and users motivated by the search for truth. Therefore, predictive markets represent a synthesis of two different missions and motivational frameworks. Metaculus is designed primarily to maximize epistemic, that is, knowledge value, not financial gain. The platform offers users the ability to predict a wide range of topics, including technology, science, politics, and other areas. This is achieved through the use of so-called crowd wisdom and statistical modeling, which allows for the aggregation of predictions and the generation of predictive models for various events. A key feature of Metaculus is the ability of users to create their own predictions with assigned probabilities, to present evidence to support these predictions, to engage in discussions to challenge predictions, and to update predictions based on new information. The platform also enables participation in tournaments focused on specific events or discourses, thus supporting users in demonstrating their prognostic abilities. The popularity of Metaculus among those interested in technology, science, and future trends reflects the natural goal of the platform: to introduce procedures that allow the timely analysis of the performance of users‘ predictions and to assign scores based on accuracy. The architecture of Metaculus is inspired by Bayesian approaches, ensemble modeling, time series analysis, machine learning, and related techniques. The creators of the platform collaborate with researchers to develop hybrid models of human-machine learning that match or even surpass top predictive systems, thereby opening a new era in predictive technologies and scientific discovery.


1.         DANTO, Arthur. The Artworld. The Journal of Philosophy. 15 October 1964. Vol. 61, no. 19, p. 571. DOI 10.2307/2022937.

2.         DICKIE, George. Is Psychology Relevant to Aesthetics? The Philosophical Review. July 1962. Vol. 71, no. 3, p. 285. DOI 10.2307/2183429.

3.         DICKIE, George. Art and the aesthetic: an institutional analysis. . Cornell University Press, 1974.

4.         BOURDIEU, P P. Sociology in Question. Online. Sage Publications (CA), 1993. Published in Association with Theory, Culture & Society. ISBN 9781446236840. Available from:

5.         BECKER, H S. Art Worlds. Online. University of California Press, 1982. ISBN 9780520052185. Available from:

6.         TEGMARK, Max. Život 3.0: člověk v éře umělé inteligence. . 2020. ISBN 9788073639488.

7.         BOSTROM, Nick. Superintelligence: paths, dangers, strategies. . First edition. Oxford : Oxford University Press, 2014. ISBN 9780199678112.

8.         KOCHAN, Jeff. Latour’s Heidegger. Social Studies of Science. 14 August 2010. Vol. 40, no. 4, p. 579–598. DOI 10.1177/0306312709360263.

9.         MUNIESA, Fabian. Actor-Network Theory. In : International Encyclopedia of the Social & Behavioral Sciences. Elsevier, 2015. p. 80–84.

10.       AKRICH, Madeleine. Actor Network Theory, Bruno Latour, and the CSI. Social Studies of Science. 25 April 2023. Vol. 53, no. 2, p. 169–173. DOI 10.1177/03063127231158102.

11.       AMSTERDAMSKA, Olga. Surely You Are Joking, Monsieur Latour! LATOUR, Bruno (ed.), Science, Technology, & Human Values. Online. 1990. Vol. 15, no. 4, p. 495–504. Available from:

12.       DURKHEIM, Émile, COSMAN, Carol and CLADIS, Mark Sydney. The Elementary Forms of Religious Life. . Oxford : Oxford University Press, 2008. ISBN 9780199540129. In The Elementary Forms of Religious Life (1912), Emile Durkheim sets himself the task of discovering the enduring source of human social identity. He investigates what he considered to be the simplest form of documented religion – totemism among the Aborigines of Australia. For Durkheim, studying Aboriginal religion was a way “to yield an understanding of the religious nature of man, by showing us an essential and permanent aspect of humanity”. The need and capacity of men and women to relate to one another socially lies at the heart of Durkheim’s exploration, in which religion embodies the beliefs that shape our moral universe. The Elementary Forms has been applauded and debated by sociologists, anthropologists, ethnographers, philosophers, and theologians, and continues to speak to new generations about the intriguing origin and nature of religion and society. This new, lightly abridged edition provides an excellent introduction to Durkheim’s ideas

13.       BOURDIEU, Pierre. Pravidla umění. . Vyd. 1. Brno : Host, 2010. geneze a struktura literárního pole. ISBN 9788072943647.

14.       FROW, John and EMMISON, Michael. Information Technology as Cultural Capital. Australian Universities’ Review. 1998. Vol. 41, no. 1, p. 41–45.

15.       ARCHER, Louise, DAWSON, Emily, DEWITT, Jennifer, SEAKINS, Amy and WONG, Billy. “Science capital”: A conceptual, methodological, and empirical argument for extending bourdieusian notions of capital beyond the arts. Journal of Research in Science Teaching. 20 September 2015. Vol. 52, no. 7, p. 922–948. DOI 10.1002/tea.21227.

16.       JAKOBSONE, Liene. Critical design as approach to next thinking. The Design Journal. 28 July 2017. Vol. 20, no. sup1, p. S4253–S4262. DOI 10.1080/14606925.2017.1352923.

17.       DEMPSEY, Gaia. Why I Reject the Comparison of Metaculus to Prediction Markets. Online. February 2023. Available from: is not a prediction market. Metaculus and prediction markets both aggregate users’ forecasts, and both reward users for…