Shawn Wang: uncertainty

Transcript context

…I totally agree. That's the stuff that is the future of AI. I think that's the exciting stuff. And there's a parallel between that. Why was AlphaGo able to be superhuman? Right. It's this ability to generate many different paths. Tree search. And tree search. Exactly. So I think if you want to talk about what would intelligence look like, it looks much more like tree search. Combining the generative nature of these LLMs with a really good tree search. And that's what opening I've done with O1 and O3. I don't know that they do tree search. They never said they do. It's implied. Yes. Okay. Yes. Yes. Are you comfortable with O1 being a reasoning engine? No, no, no, no. I'm saying it's better at reasoning because they leverage the tree search well. And the, the issue of the reasoning is they're saying, is this like they train, they have the models to say, is this logically correct? And what's the likelihood of it being logically correct? So you can build up the sophisticated mechanisms to get it less bad at reasoning, but you'll see like eventually what, what AI is really, really good at. People won't say it's, it's always going to be better at retrieving. It's always going to be better at storing knowledge, which is so highly correlated with intelligence that we often assume it's the same. What, what AI is truly special at and gets consumers really excited is it's generative. It can just make stuff. We've never had a technology. Before that can just make stuff simulate.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.