The Politics of Falling
| Date: | 2025 |
| Index: | dataset-critique writing |
Abstract
This essay uses falling as a concept through which to think about automation, autonomy, and accident. Centred on an analysis of a video from a simulated fall dataset used to train AI detection algorithms, an instance of ‘dataset theatre’ where human performances are staged for machine observation, it discusses how ‘accidents’ become systematically produced and categorized within these technological frames. Drawing on philosophical concepts from Aristotle's notion of automaton to Robert Morris's ‘creative automation’ and the artistic acts of Bas Jan Ader and Surrealist automatism, the essay reimagines falling: not just as a physical event, but a method, a metaphor, and a site for debating agency, control, and contingency. By juxtaposing performance art's choreographed falls with data collection's simulated ones, this analysis suggests that systems—artistic or technological—actively frame and even generate the accidental. Ultimately, ‘The Politics of Falling’ reflects on what it means to live within increasingly automated systems. It proposes that understanding the inherent ‘fall’ within these structures—and perhaps learning to ‘fall well’ through practices like dance—provides a way to navigate contemporary life's precariousness and to reclaim possibilities for agency.
Note
A version of this text was originally presented at Automation Cultures in 2021 at the University of Western Australia, Perth, Western Australia. It was reworked for publication in 2024.