person
Paul Christiano
Alignment researcher and coauthor of concrete safety problems and AI safety via debate.
Paul Christiano is an alignment researcher listed by the Alignment Research Center among its researchers and as its executive director. This institutional description was checked in October 2026.ARC team record.
He coauthored Concrete Problems in AI Safety, which includes costly human supervision in its accident-risk agenda, and AI safety via debate, which proposes competing arguments before a human judge.[1][2] These papers connect his work here to Scalable Oversight and AI Safety via Debate.
The 2015 beneficial-AI open letter lists Christiano among its signatories, supporting its call for robust and beneficial AI research.[3]
Sources
- Concrete Problems in AI Safety · Source record ref-cd3035dbef6c · Back to claim ↑1
- AI safety via debate · Source record ref-debate-2018 · Back to claim ↑1
- Research Priorities for Robust and Beneficial Artificial Intelligence: An Open Letter · Source record src-190 · Back to claim ↑1
- Alignment Research Center: Team · Source record ref-884a30376fd0
Pages that link here
Last updated 2026-10-10