
‘Gambling with our lives’: Anthropic researcher quits, warns against self-improving AI
A pre-training researcher walks out of Anthropic and says the quiet part out loud: the labs already know the stakes of self-improving AI — and they’re racing anyway.










