Brian Jordan Jefferson, University of Illinois at Urbana-Champaign

Brian Jordan Jefferson is associate professor of Geography and Geographic Information Science. He is author of Digitize and Punish (Minnesota 2020). He has served on committees for panels on AI ethics for the Association for Computing Machinery, International Conference on Machine Learning, and Neural Information Processing Systems conferences and is an affiliate of the Science Technology & Social Values lab.

“Political not Technical: Artificial Intelligence & the Genealogical Method”

October 4, 2022 the White House unveils its Blueprint for an AI Bill of Rights. Nine months later,
the UN Security Council holds its first meeting on AI governance; days after, the White House
announces plans to deregulate AI on one hand, and regulate the ideological content in AI models
on the other. A growing number of engineers and social scientists have joined forces in response
to these developments in fields like AI Ethics, AI & Society, and Computing & Society. Too often
however, their efforts share a critical limitation: they define AI solely on its own technical terms
and seek technical solutions for its problems. Jefferson shows how technical descriptions are
necessary but insufficient for addressing the political challenges posed by AI. From its inception,
the state has played a fundamental role in AI’s development. AI must therefore be defined not
only in terms of processing symbols, but also in terms of governance, sovereignty, territory, and
warfare. Jefferson outlines how a genealogical approach to studying AI helps illuminate its
political dimensions.