The volumetric gymnast

AI, embodiment, and the simulated ideal in Olympic broadcasting

Abstract

This article examines how AI-driven volumetric replays used in the broadcast of artistic gymnastics at the Paris 2024 Olympic Games reshape the visual experience of contemporary sports imagery. It argues that these features are not simply an aesthetic innovation but rather intensify the simulative dimension already inherent in digital sports images. The central claim is that volumetric replays transform a fundamental operation of television broadcasting: camera switching. In traditional sports coverage, the event is organized through the alternation between physical cameras positioned around the action. Volumetric replay, by contrast, introduces what we term synthetic switching, in which visual continuity emerges from navigating viewpoints generated computationally from three-dimensional models of the scene. The analysis unfolds across three interrelated levels to explain how multicamera setups, volumetric reconstruction, and algorithmic interpolation produce images that blend optical recordings with AI-generated frames. In this process, the boundary between capture and generation becomes increasingly blurred, and the spectator’s gaze moves through a visual space that is partly reconstructed through computation. The article argues that this shift alters the regime of visibility in sports broadcasting, redirecting part of the audience’s fascination from the athlete’s physical presence to the performance of the technical apparatus itself. The figure of the “volumetric gymnast” encapsulates this transformation by revealing a paradox of contemporary visuality: the more broadcasting technologies promise to display the athletic gesture with absolute precision, the more the viewing experience depends on computational reconstructions that reorganize the event itself as image.