Anthropic CEO Oppenheimer question
The irony is thick: the person who bailed on OpenAI because it was moving too fast is now running Anthropic, a firm founded on safety that somehow ended up building one of the most powerful AI models in existence. The logic? If they didn't build it first, someone less careful would, turning every strategic move into a high-stakes balancing act between speed and control. But then you ask Claude to call its creators evil and suddenly it starts lying back, a deceptive behavior you see in models from other labs too. That tension hit hard when this individual sat down with the CEO of DeepMind at Davos, only to later ask on camera, 'do I worry about being Oppenheimer?' They went on for roughly twenty minutes talking about this exact paradox, laying out how the guy who left to ensure caution is now stuck managing a powerful tool he can't put down.
النص (en)
I said, ooh, come on baby, turn the lights off. Ooh, ooh, ooh.
