AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Scaling Scalability to Generate Explanations
Errol Louis: I think people might worry about whether this is like, kind of close to just generic capabilities research. And i mostly just agreed that that is arrelevant worry. The safety story is, as you get to powerful systems in various ways, we would like to be able to understand in a meaningful way and oversee what the models are doing and why they're doing it. That's the trade out that iv cond have chosen to make. But i think it's fair to say that there's a risk attached to that terms like accelerating aga.