108 - Miles Brundage (Head of Policy Research at Open AI) on the speed of AI development and the risks and opportunities of GPT
Philosophical Disquisitions
00:00
The Pros and Cons of Large Language Models
Is there something about large language models that they undergo, they are have appeared to under a gone some sort of significant phase transition or have developed immersion properties with the iteration between like two and GPT two. Some of the conversations I've had with other people is that there does seem to have been a fairly dramatic uptick in capacity and ability. The point I was trying to make about the kind of whiplash effect is that if we had gone from like GPT 0 or GPT 1 directly to GPT 4, I don't think we necessarily would have been the obvious thing to do either.
Transcript
Play full episode
