The Disembodied Choreographer: A Choreographic Experiment with ChatGPT-5

Abstract

This paper explores the ways in which generative Artificial Intelligence can influence and participate in the development of choreography by providing a case study of an empirical and practice-based experiment with ChatGBT 5. Drawing inspiration from Harold Cohen, the creator of AARON, who used Artificial Intelligence for artistic production, this study aims to discuss the choreography co-created through dialogue with the text-based generative model ChatGBT 5, followed by the physical embodiment from my side. Analyzing the experiment through qualitative lenses, I discuss aspects such as authenticity – if we could speak of authenticity at all – authorship, and how a disembodied entity can generate a choreographic thought that requires embodiment by its own nature. Thus, by providing a case study, my inquiry seeks to answer the question of whether AI can be/come an artistic agent or whether we could talk about AI artists, if so, how does it transform the way we work and understand choreography and dance?