New Books in Science cover image

Chris Bleakley, "Poems That Solve Puzzles: The History and Science of Algorithms" (Oxford UP, 2020)

New Books in Science

00:00

The Best Training Strategy Is a Policy and Value Network

In the original version, they used a deada bas of human played games that was available in internet. So didn't learn anything from mans. And it only learnt from playing other versions of itself. It started totally randomly and just taught itself to play by playing hundreds of thousands of games against itself. In a match between alpago zero and alfago, alfgo zero 100, no. And alfago zero was trained, i think, in forty days. There's even faster solutions since then.

Transcript
Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
App store bannerPlay store banner
Get the app