Demis Hassabis: belief

Transcript context

…Will we see asymmetric progress in the domains in which the self-play kinds of things you’re talking about will be especially powerful? So math and code. Recently, you have these papers out about this. You can use these things to do really cool, novel things. Will they be superhuman coders, but in other ways they might still be worse than humans? How do you think about that? I think that we’re making great progress with math and things like theorem proving and coding. But it’s still interesting if one looks at creativity in general, and scientific endeavor in general. I think we’re getting to the stage where our systems could help the best human scientists make their breakthroughs quicker, almost triage the search space in some ways. Perhaps find a solution like AlphaFold does with a protein structure. They’re not at the level where they can create the hypothesis themselves or ask the right question. As any top scientist will tell you, the hardest part of science is actually asking the right question. It’s boiling down that space to the critical question we should go after and then formulating the problem in the right way to attack it. That’s not something our systems really have any idea how to do, but they are suitable for searching large combinatorial spaces if one can specify the problem with a clear objective function. So that’s very useful already for many of the problems we deal with today, but not the most high-level creative problems. DeepMind has published all kinds of interesting stuff in speeding up science in different areas. If you think AGI is going to happen in the next 10 to 20 years, why not just wait for the AGI to do it for you? Why build these domain-specific solutions?…

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