It’s been 24 hours since the greatest board coup in 40 years, and almost 20k people tuned in to our live discussion on the unfolding events.
With some crack reporting from Kara Swisher overnight, and a play-by-play from Greg and confirmatory interviews with Ilya, we now understand that of all the possible reasons for the ouster, the safetyist argument was probably 1 the actual one, not your quotidian fraud or sex scandal or security lapse firing:
Based on some APEC Summit comments, people are speculating that this was because of a new AGI achievement internally. Whatever it was is unlikely 2 to be of that level of magnitude as to warrant such an immediate and unceremonious firing, but regardless of the magnitude of internal progress, the OpenAI board is certainly guilty of the same non-candid communication they accused Sam of.
As a rule, we do not discuss AI great power play politics in Latent Space (per our non-coverage list), but an event of this magnitude cannot be ignored.