In retrospect, would Sam Altman being fired from OpenAI be a net negative to alignment? (EoY 2025)
61
1kṀ7014
Dec 31
62%
chance

I will resolve this question to YES if there's at least one clear indicator of this happening that's explicitly traceable to this event. Examples of what that would entail include:

  • Sam Altman bounces back and rejoins OpenAI, which then proceeds to accelerate even faster at the cost of safety (i.e. there's an explicit top-down policy to pursue capabilities full stop)

  • At least one major AI player, on the level of say Anthropic, is born that is: a) undoubtedly accelerationist, and b) explicitly cites Sam Altman as their inspiration (while also expressing sympathies for this incident in some other way)

  • A systematic movement against alignment labs, programs, or research in general gains steam and takes down at least one major research institution on the level of say, Conjecture.

  • A comprehensive, highly upvoted LW/EA retrospective is written about it and explicitly concludes that it was net negative.

I'm open to improving the resolution criteria over the next couple of weeks, but I suspect the answer to this would be unambiguous.

Get
Ṁ1,000
to start trading!
© Manifold Markets, Inc.TermsPrivacy