Skip to content
Longterm Wiki
BS

Buck Shlegeris

AI safety research; Redwood Research leadership

Expert Positions2 topics

TopicViewEstimateConfidenceDateSource
How Hard Is Alignment?Very hard — focus on control insteadEasier to assume misalignment and control for ithigh202580,000 Hours Podcast
Will Advanced AI Be Deceptive?Significant risk~30% probability of scheming before escape attemptmediumApr 2024AXRP Episode 27 — AI Control (2024)

Board Seats1

Advisor

Links

Facts

4
People
Role / TitleCEO
Employed ByRedwood Research
Biographical
Notable ForAI safety research; Redwood Research leadership
General
Websitehttps://redwoodresearch.org