Now that AI can effortlessly imitate, how will we deviate?

Deviation Game draws inspiration from the Imitation Game, devised in 1950 by Alan Turing, the father of computer science. While that experiment—later widely known as the Turing Test—centered on an AI’s attempt to imitate a human, we have inverted this relationship to explore the potential for humans to deviate from an AI’s understanding.

Compared to Turing's era, modern generative AI can mimic human intellectual activities, such as image generation and writing, with astonishing accuracy. Yet, history shows that the emergence of such technologies does not replace human expression; rather, it opens new possibilities. Just as the invention of the camera in the 19th century contributed significantly to the rise of Impressionists like Van Gogh and Monet, humans have historically responded to technological "imitation" with new forms of "deviation."

Deviation Game highlights this relationship between imitation and deviation, attempting to redefine the roles of AI and humans. Specifically, rather than viewing AI as a mimetic tool that reproduces the past, we treat it as a "discriminator" tasked with determining whether a specific expression has existed before.

To win this game, human players must produce new expressions and interpretations that have not yet been captured within the AI's pre-trained datasets. We therefore view AI not as a replacement for human creativity, but as a tool to amplify it. By using AI—which excels at imitating the past—to identify existing patterns, we challenge humans to break free from those patterns, exploring the potential for expressions that truly deviate from the past.

The original concept for the game was a collaborative effort between game designer Tomo Kihara and art-design duo Playfool (Daniel Coppen & Saki Maruyama). Based in both Tokyo and London, both of their practices centre around the notion of play to design tools that foster creativity and produce artistic interventions that engage with urban spaces and society. Their previous collaboration How (not) to get hit by a self-driving car received an Honorary Mention in the S+T+ARTS Prize 2024 at Ars Electronica.

Deviation Game was initially developed as part of the art incubation program at Civic Creative Base Tokyo [CCBT]. Since then, the game has been exhibited globally at events including Ars Electronica (Linz), Now Play This (London) and Day of the Devs (San Francisco). The game has engaged audiences of all ages and sparking critical discussions around AI and creativity. The exhibition version is available via request.

Tomo Kihara

web / instagram / x

Tomo Kihara is an artist and game designer who explores play as a form of critical inquiry. Through experimental games and playable installations in public spaces, he invites people to collectively question and reflect on how emerging technologies such as AI are reshaping the ways we think and live. His work has been showcased internationally at venues including Ars Electronica, the V&A Museum, and the Exploratorium.

Playfool

web / instagram / x

Playfool is an art-design unit by Daniel Coppen (UK) and Saki Maruyama (JP). Through the medium of play, their practice explores the transforming dynamic between human agency and technology through the medium of play. Their work invites a critical engagement and reimagination of technology, often taking the form of experimental games and interactive installations. Playfool’s work has been awarded in the S+T+ARTS Prize (2024) and Dezeen Award (2021), and exhibited internationally at the Science Gallery (Monterrey, 2025), Ars Electronica (Linz, 2024), and the V&A Museum (London, 2023).