![](/_next/image?url=https%3A%2F%2Fstorage.googleapis.com%2Fmantic-markets.appspot.com%2Fcontract-images%2Fstrutheo%2F417d090b9eb0.jpg&w=3840&q=75)
https://youtu.be/gA1sNLL6yg4?t=4995&si=EK43yO1zK5vCwnY3
(Yudkowsky on the Bankless Podcast)
[...] when I dive back into the pool, I don't know, maybe I will go off to Conjecture or Anthropic or one of the smaller concerns like Redwood Research - Redwood Research being the only ones I really trust at this point, but they're tiny... and try to figure out if I can see anything clever to do with the giant inscrutable arrays of floating-point numbers [...]
I think other options include some more serious blogging about why this is going to be difficult. Or maybe OpenAI makes him a pitch that he thinks would be low-dignity to turn down... Though idk, that seems seriously unlikely, OA seems pretty insistently anti-Yudkowsky in their openly discussed group-think messaging about this topic.