Demis Hassabis: belief

Transcript context

…A big open question right now is whether RL will allow these models to use the self-play synthetic data to get over data bottlenecks. It sounds like you’re optimistic about this? I’m very optimistic about that. First of all, there’s still a lot more data that can be used, especially if one views multimodal and video and these kinds of things. Obviously, society is adding more data all the time to the Internet and things like that. I think that there’s a lot of scope for creating synthetic data. We’re looking at that in different ways, partly through simulation, using very realistic game environments, for example, to generate realistic data, but also self-play. That’s where systems interact with each other or converse with each other. It worked very well for us with AlphaGo and AlphaZero where we got the systems to play against each other and actually learn from each other’s mistakes and build up a knowledge base that way. I think there are some good analogies for that. It’s a little bit more complicated to build a general kind of world data. How do you get to the point with these models where the synthetic data they’re outputting on the self-play they’re doing is not just more of what’s already in their data set, but something they haven’t seen before? To actually improve the abilities.…

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