What Do AI Images Want? Three AI-DDP Assemblages in Contemporary Korea

Abstract This article explores the aesthetics of artificial intelligence ( AI ) images by analyzing three cases: Patrik Schumacher’s design experiment through the use of text-to-image AI tools (2023), Refik Anadol’s Seoul Haemong (2019) and Yiyun Kang’s Geofuture (2022–2023), all of which use Zaha Hadid Architects’ Dongdaemun Design Plaza as a referent or a background. While Schumacher uses AI as a design method, Anadol and Kang focus on interactivity through outdoor AI exhibitions, which assume lively interaction between their works and visitors but loosely grapple with its elusive nature. They present a flattened sense of the world in which the lived experiences and scattered visual data from daily lives are reassembled and epigenetically generate new affects, percepts, and moods. By regarding Bruno Latour’s postcritical stance and Catherine Malabou’s concept of plasticity as thresholds, the article encourages us to rethink AI images as nonhuman assemblages that consist of forms of life and evolving technologies.