2min chapter

Generally Intelligent cover image

Episode 27: Noam Brown, FAIR, on achieving human-level performance in poker and Diplomacy, and the power of spending compute at inference time

Generally Intelligent

CHAPTER

Six Player Poker

Once you introduce cooperation into the game, then a lot of these equilibrium findings from scratch techniques don't end up being that useful in practice. The clearest example of this is English. Imagine a game like diplomacy where you have to communicate with the other players in a language. That gibberish robot language is fine if you're playing against other robots because you'll be able to communicate with them. But if you stick that agent in a game with six human players, they're not going to work with the bot because they're all speaking English. So once you go to a game that involves cooperation with humans, then you can't just do these techniques like self-play from scratch, with

00:00

Get the Snipd
podcast app

Unlock the knowledge in podcasts with the podcast player of the future.
App store bannerPlay store banner

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode

Save any
moment

Hear something you like? Tap your headphones to save it with AI-generated key takeaways

Share
& Export

Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode