Thomas Ahle: uncertainty

Transcript context

…And structured inference? I don't know if you call it that. It's more like they took the classic test search engine and they replaced the evaluation function with a neural net, but like kind of a very shallow wide neural net that they can, that is really, really fast to evaluate. And yeah, think it could be that for something like synthesis and compilation, there's a similar thing where it's like you could try and do it all with LLMs, but at some point that's the speed is a bottleneck and like you can throw, get a benefit from having more knowledge and more intuition on all of this stuff. But at some point there's also just a hard commutation problem where you just want to brute force some stuff. And at that point, want to be able to switch to more classical algorithm. Absolutely. Now before, we were talking about thermo computing. So you folks have done some work that is incredibly, like, it has huge potential for things like, know, Markov chain Monte Carlo, think, and, diffusion models. But you did say to me when we spoke last time that, in some cases it can be a false economy. Like, for example, you could have a diffusion model which might have a different type of neural network on the end of it. And and you might find that the benefit that you have doing the diffusion might be bottlenecked by another part of the model. So so in in in practice, where can we see a huge uplift?…

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