Alignment: From the Governance of Models to Geopolitics

This presentation addresses how the logic of alignment moves ambiguously between the technologies being used to train GenAI models to governance and geopolitics. Rather than asking abstractly how we might govern models by “aligning” them to, for instance, democratic values, we investigate why the rationality of alignment itself has so quickly become hegemonic among AI developers and certain philosophical communities.

The talk will be hybrid. Use the Zoom link to register for the event. We will be meeting in EV.10.625.

About the speaker

Alexander Campolo is assistant professor (Research) at the Leverhulme Centre for Algorithmic Life and the School of Modern Languages & Cultures at Durham University. His research cuts across histories of machine learning, statistics, and computing to theorise societies and cultures being transformed by AI. 

About Machine Agencies + Milieux

Machine Agencies is part of the Speculative Life Cluster at Milieux Institute. It is an experiment between human and machine intelligences. Our research group encourages cooperation and play, resisting the antagonism of more instrumental approaches of AI. We engage with posthumanism, experience design, and public policy to find new formats, methods, and commons to sustain just, fair, and better worlds.

Milieux is an interdisciplinary graduate research institute for new media arts, digital culture and information technology at Concordia University in downtown Montreal.