Bio-AI:The Aesthetics and Ethics of Data Animism
Over the last few years large-scale installations by artists such as Refik Anadol, Anna Ridler and Mat Collishaw have been enveloping visitors to galleries, parks and other public and commercial spaces with AI-generated data rendered as pixel movement. The audiences typically respond by attuning their bodies to the exuberant flow of what look like flowers, plants and corals unfolding on screens. At times dismissed by critics as mindless visual seduction, mere screensavers with anaesthetic properties, such works stage a digital encounter that demands more than aesthetic appreciation (or dismissal). This article proposes that they actually enact a form of ‘data animism’, rendering AI as a biomachine that aims to channel life. In its ability to not only capture movement but also move us, bodily and affectively, it was cinema that was originally presented, by thinkers such as Henri Bergson, Jean Epstein and Gilles Deleuze, as a source of animist impulses. At a time when cinema has become a much more private and narrative affair, can we consider that the role of a life-enhancing medium has passed on to generative AI (genAI)? Triangulating the ‘flower-pixel-vector’ dynamics, the article will track historical parallels between the rendition of nature, in all its transformative abundance, by the biomachines of genAI and the early twentieth-century naturalist films. Taking on board the justifiable concerns about planetary extractivism, the depletion of human and planetary resources by the insatiable technology of machine learning, and the threat of human extinction, it speculates whether genAI is designed to make us feel alive. But who would this ‘us’ be in this project? And what kind of life would be enabled (and disabled) in the process? The article includes Joanna Zylinska’s own visual project, cocreated with AI.
Publication Details
- Journal
- Research Portal (King's College London)
- Published
- 2025-11-26
- Primary Topic
- Digital Media and Philosophy
- Type
- article
- Field-Weighted Citation Impact
- 0.00