From sublime awe to abject cringe: on the embodied processing of AI art

Abstract

This article urges a reorientation in thinking about AI art (and AI more generally), shifting from the common focus on computational ‘intelligence’ to the embodied, metabolic processing that takes place in our encounters with (moving-image) artworks produced with machine-learning algorithms. Drawing on Merleau-Ponty’s corporeal phenomenology, the article argues that spectators’ bodies act as filters, distilling visual phenomena from a range of extraperceptual facets of these works; in particular, bodies react to invisible algorithmic infrastructures, which, in the case of machine learning algorithms, also operate as filters in their own right. The collision of metabolic and computational microtemporal operations calls forth a number of embodied affects, ranging from sublime awe to disorientation, cringe, and uncanny feelings of relational and environmental entanglement. These themes are explored through the work of four contemporary artists working with AI: Ian Cheng, Refik Anadol, Jon Rafman, and Yvette Granata. In conversation with these artists, the author explores our bodily responses to AI-generated imagery in an attempt to better understand the stakes, as well as the underlying mechanisms, of the new technology’s transformation of our visual culture.