AI NEWS PAGE DOSSIER
Stuart Russell
Stuart Russell's file: the standard AI textbook, Berkeley, human-compatible AI, autonomous-weapons warnings, and the computer scientist trying to rewrite the control problem.
Russell became one of AI's defining textbook authors and Berkeley researchers, then increasingly redirected his work toward the control problem: how to build systems whose objectives remain compatible with humans even as machine capabilities grow.
HOW THE HELL DID STUART RUSSELL GET HERE?
1986
FROM STANFORD PHD TO BERKELEY AI
Russell completed a PhD in computer science at Stanford in 1986 and joined the faculty at UC Berkeley, where he built a long research career in artificial intelligence.
RECEIPTS: UC Berkeley EECSreported
Aug 2016
BERKELEY OPENS A CENTER FOR HUMAN-COMPATIBLE AI
Berkeley launched the Center for Human-Compatible Artificial Intelligence under Russell's leadership to develop AI systems whose behavior remains provably beneficial to humans.
RECEIPTS: Berkeley Engineeringreported
2019
THE CONTROL PROBLEM BECOMES HUMAN COMPATIBLE
Russell published Human Compatible: AI and the Problem of Control, arguing for a redesign of AI around uncertainty about human preferences rather than fixed machine objectives.
RECEIPTS: Center for Human-Compatible Artificial Intelligencereported
2021
THE AI SAFETY ARGUMENT GETS THE REITH LECTURES
Russell delivered the 2021 BBC Reith Lectures, Living With Artificial Intelligence, bringing his argument about advanced AI and human control to a broad public audience.
RECEIPTS: Stuart Russellfirst-party
2022
THE CONTROL PROBLEM MOVES INTO SCIENCE AND PUBLIC LIFE
Russell was selected as the inaugural director of Berkeley's Kavli Center for Ethics, Science, and the Public, extending his work into the public governance of science and technology.
RECEIPTS: Center for Human-Compatible Artificial Intelligencereported