OpenAI removes AI safety leader Mądry, a onetime ally of CEO Altman
theinformation.comIt seems part of the show by now. From time to time they have to remind the public of the fairy tale there might maybe someday perhaps could probably but not really be something about their inefficient wrongness-databases that might warrant a discussion about safety, so they publicly turn the wheel on the whole safety management bullshit.
About time we moved past the “safety” handwringing. Both the frontier models and high capability “open weight” models have been with us for over a year and a half, and we’re still in one piece.
But we are still having an important election upcoming. There need to be proper censorship implemented before. After, it would be too late.
How about we don’t do “censorship” in our purportedly “free” country?
This is satire, right?
People always avoid death, up until they don’t.
Your avoidance of death would be far better served by protesting endless wars and AI weapons, not stochastic parrots