Heat work: choreography as an ecological enquiry of machine learning and extractivism

Abstract

Artificial Intelligence (AI) has the potential to both disrupt and transform the performing arts. However, the use of Machine Learning (ML) algorithms in choreography risks leading to a reductive and opaque formalization of choreographic decisions, often disregarding energy consumption and environmental impact. This article explores choreography as a method for examining the energetic materialities of ML and extractivism. Through Heat Work, a performance project, it investigates the energy and heat generated by movement data derived from tribal dances within an extraction zone. The methodology includes motion capture, data classification, and the application of a supervised ML model, with a focus on implementing a sustainable approach to choreography with AI. By examining the project’s ecological dependencies for choreographic decision-making, the article proposes a sustainable and ethical avenue of research at the intersection of AI, extractivism and performance.