Stuart Russell
Author of Human Compatible; Berkeley professor
Co-author of the standard textbook Artificial Intelligence: A Modern Approach and founder of Berkeley's Center for Human-Compatible AI. He argues the field's standard model of fixed objectives is the root of the control problem.
Key arguments & positions
- Argues the field's "standard model" — build machines that optimize a fixed objective we specify — is the source of the control problem.
- Proposes provably beneficial AI: systems uncertain about human preferences, and therefore willing to be corrected and switched off.
- Supports binding regulation and international agreements on frontier systems, and co-signed the 2023 call for a training pause.
Accomplishments
- Co-author, with Peter Norvig, of Artificial Intelligence: A Modern Approach, the field's standard textbook.
- Professor of computer science at UC Berkeley; founder of the Center for Human-Compatible AI.
- Delivered the 2021 BBC Reith Lectures on living with artificial intelligence.
Papers & key writings
Links
In the library
Nothing of theirs is filed in the hub yet. Browse the full library.
Recommended next
Hand-picked from the hub based on what Stuart Russell covers.
Partnership on AI
This nonprofit website shares research and guidelines on artificial intelligence from tech companies and community organizations. You can explore free reports and frameworks to learn how AI affects the economy, workplace practices, and technology safety.
Why this: Also about AI governance
Ford Foundation
This global foundation funds organizations and individuals working on social issues, workers' rights, and technology. You can search for grant opportunities, apply for fellowship programs, and read research reports on the future of work.
Why this: Also about AI governance
From surveillance to governance: recentering human rights in the UN AI agenda
Access Now hybrid side event during UNGA81 High-Level Week on human rights in AI governance.
Why this: Also about AI governance
Emergence AI
A company building governed, verifiable agentic infrastructure for enterprise systems, with free research and publications on its site.
Why this: Also about AI governance