Natasha Jaques 2
TalkRL: The Reinforcement Learning Podcast
00:00
Are You Up for Talking About AGI?
Sentiment isn't perfect. You could also look at like, okay, I prom GBT, I get some output. But even though those are imperfect and only proxy measures, optimizing for those things still did better than optimizing for the thumbs up thumbs down that we built into the interface. Because just no one wants to bother providing that. So yeah, I think more scalable signals is the right direction. That makes so much sense. Are you up for talking about AGI? Yeah, depends what the question is. Do you think it's something we should be talking about and thinking about these days, or is it like a distant fantasy that's just not really worth talking about?
Transcript
Play full episode
