Martin Zeilinger

AI Art and Post-Anthropocentric Agency

conference presentation
Indeterminate Futures conference, Dundee University
[posted: 13 November 2020]
>>tags: AI art, agency, posthumanism

This is the video from my conference presentation at the “Indeterminate Futures” conference organised by at Dundee University. The conference featured keynote lectures by Karen Barad, Franco Berardi, Gregory Cajete, Leroy Little Bear, Sha Xin Wei, and Vladimir Tasić. It was organised by was organised by Natasha Lushetich and Iain Campbell.

The title of the panel was ‘Probablity and Agency,’ and featured presentations by David Zeitlyn and Lukas Grießl in addition to my own contribution. Full details for the conference are available through the project website for The Future of Indeterminacy.

What happens to ‘agency,’ anthropocentrically conceived, when the concept becomes diffracted across new kinds of posthumanist agential assemblages in AI art? Focusing on recent projects by Anna Ridler and Maja Smrekar, this paper explores how artistic uses of AI can both re-inscribe and disturb the traditionally unified, singular, and anthropocentric subject position of the humanist agent. Anna Ridler’s Mosaic Virus (2018-19) references the historical ‘tulip mania’ phenomenon, in which a rare plant virus, affecting only certain flower bulbs, triggered intense financial speculation. Relying on artist-curated training data, Ridler links the presumptive unpredictability of generative AI to the perceived unknowability of viral matter, developing a metaphor for the decoupling of agency from the mastery of human hosts. Maja Smrekar’s !brute_force (ongoing), by contrast, documents ludo-scientific experimental efforts to train an AI system on ‘dog data.’ In doing so, Smrekar redistributes and pluralises agency across the subject positions of the project participants, to produce new agential realities that are triangulated between the human, the algorithmic, and the canine. The two projects develop very different perspectives on the indeterminacy of anthropocentric notions of agency in and through AI. I will discuss these as successive steps towards the conceptualisation of agency as decentred, multiplicitous, indeterminable, and unbounded by anthropocentric ontological and ideological frameworks.