Game Theory in the Intelligent Era: Exploring Subjectivity in Artificial Intelligence-Assisted Co-creation

Text/Chiu Chih-Yung (Associate Dean, College of Arts and Professor/Director, Graduate Institute of Art and Technology, National Tsing Hua University)

All over the world and over a long period of time, traces of human activity have changed the ecology, with technological devices now an almost inseparable part of nature. On the one hand, “intelligent machines” play indispensable roles in the daily lives of humans as “alternative decision-making devices.” On the other hand, in terms of the correspondence with natural selection, “machine selection” and natural selection, which determined the evolution of species and the survival of the fittest in the past, are connected but contradictory. We are now in an era of “machine naturalization” in which machines are all around us. The spread of new technologies and new machines is resulting in the gradual replacement of the natural selection process and an inside out reversal of the physical and behavioral attributes of humans that have naturally evolved.

The transformative role of artificial intelligence (AI) in contemporary art can be observed from three aspects: styles, techniques, and processes. Artists are able to innovate artistic creations using the power of advanced algorithms, thereby expanding the boundaries of their creative expression. In terms of artistic styles, AI-generated art is based on parameters that are defined by the artist. Algorithms generate works by incorporating the set restrictions of those parameters. The result is all new artistic styles. Moreover, traditional art forms are being reinterpreted from the perspective of AI. In terms of techniques, AI influences and impacts the ways that artists work with technologies. Some artists are using AI as a tool to enhance their creative process,
while some artists are choosing a more hands-off approach, allowing AI to automatically generate works, then interpreting or curating the generated results.
Finally, in terms of creative and experiential processes, AI is transforming the way we experience art and interact with it. These processes challenge artistic expression traditions and open up new possibilities for artistic experiences.

From the early applications of algorithms to generative adversarial networks (GANs), diffusion models, and today’s mixture model and community-based art, there is obvious coordination between humans and AI. As AI capabilities are increasing, the boundaries that define human agency are becoming blurred. Today, we are witnessing history being made as the human body, AI, and knowledge co-evolve. In other words, AI-assisted creativity has been demonstrated in the arts, literature, and music and has sparked important discussions about the capabilities and contributions of intelligent agents. These agents are driven by advanced machine learning algorithms and are adept at simulating human creativity, raising fundamental questions about the
ownership of creations and their originality, as well as the nature of artistic expression.

When intelligent agents are integrated into the creative process, they serve as creative partners, stretching the boundaries of traditional artistic creation and bringing possibilities for forms of expression that were previously not as accessible. The production jactynogg zontaanaco is one example. This is a collaboration between contact Gonzo, a performing arts group that presents improvisational performances that involve physical contact, and yang02, an artist who questions the subjectivity of using autonomous machines and other technologies for artistic expression. Through improvisation (no set script and an emphasis on improvised creation and tacit understanding among performers), physicality (testing of physical limits, exploration of interpersonal relationships and physical perception through physical collisions and interactions), and cross-disciplinary collaboration (interactions among technology and performing arts, machine design, interactive design, and other fields), this work integrates the latest conversational AI, allowing it to intervene in physical expression as a new type of translation, revealing previously unknown forms of knowledge.Attempts are made to use a technology known as image captioning to convert physical expression into language. The text that is generated is meant to describe the meaning of input images.

The core topic of jactynogg zontaanaco is the exploration of “misreads” that are generated when improvised movements are interpreted and translated into text by AI. There is further reflection on whether machines can truly understand human behavior and expressions. As both humans and AI may be confused by physical movements they have never seen before, can artists create physical movements that are unknown and that break through the constraints of meaning? This type of dynamism has triggered debates on the rights to creations, knowledge generation, and the appropriateness of interpretation in creative practices. After all, who is really the creator? Who generates meaning in an exhibition or performance? It goes without saying that machine learning technology is replacing and reconstructing the role of human creators in the artistic creation process, resulting in new human-machine relationships. Well-known AI artist Memo Akten used the taming of wild horses as an analogy for this. He believes that the relationship between artists and AI intelligent agents is one of mutual domestication during which a common language is established, with the result that people are no longer able to distinguish between the roles of subject and object.

photo by Yurika Kono