Alignment Research Center
No open roles
description
The Alignment Research Center (ARC) is a nonprofit research organisation that aims to help align future machine learning systems with human interests. They focus on conceptual work, developing alignment strategies that may be promising directions for empirical work.
They also have an evaluations team focused on evaluating the capabilities (and in the future, alignment) of advanced machine learning models. We have a separate page where you can see more information and open roles on this project.
Open roles
No active listings
Learn more
80,000 Hours links
AI safety technical research
Career review
Interview with Paul Christiano about some his research (recorded before he founded ARC)
Podcast
External content
A bird's eye view of ARC's research
Blog
ARC's write-up on Eliciting Latent Knowledge, the problem of getting advanced AI systems to honestly tell you what they believe (or 'believe') about the world
Report
Paul Christiano's blog covering some of the research he's done at ARC
Research