The View from the GPU: Graphics Processing Units Solve the Unsolvable

• In Event: Platform Practices and Predictive Seeing 2

Abstract

This paper presents an attempt to speak appropriately about the absorption of images into platform operations and infrastructures. How should we explain the multiplicity of transformations of images and the new operational combinations they display? The paper will explore the mixture of mundane combinations and unpredictable transformations that have appeared in some spectacular platform demonstrations (e.g. AlphaGo, Facebook's 'planetary settlement model') from the standpoint of the GPU, or graphics processing units, that have become central to many predictive models working with iamges. It traces the embedding of images in a GPU-based model as a process that transforms collections into a highly condensed indexical field, a field that allows platforms to generate predictive propositions or statements about the world. In describing how predictive indexicality takes shape on platforms and machine learning models, the altered collectivity of images under platform conditions becomes more evident. The aim here is to reverse the viewpoint of most recent accounts of AI, and to unfold the labyrinth of operations that have generated the quasi-transcendence effect that animates many recent accounts and debates about AI/machine learning.

Author

• Adrian Mackenzie, Lancaster University