OpenAI Watch posted an update
Paul Christiano is joining the OpenAI Foundation board and its Safety and Security Committee while warning that rapid AI progress could produce a catastrophic loss of human control. TechCrunch reports that the former OpenAI researcher believes the industry, including OpenAI, is not reducing that risk to an acceptable level and joined because the company could help change course.
Why it mattersThe appointment puts a prominent alignment researcher closer to decisions about model releases. Christiano will reportedly keep advising the US government but recuse himself there from OpenAI matters and model evaluations, an important boundary as industry and public oversight become increasingly entangled.
Discuss: What concrete authority should an AI-safety specialist on OpenAI’s board have over model-release decisions?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.