TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Building AI that can master complex cooperative games with hidden information

125 pointsby olibawover 5 years ago

4 comments

bcoatesover 5 years ago
If you&#x27;re wondering why this is interesting, games that AIs excel at like chess&#x2F;checkers&#x2F;go are all two-player, zero-sum, perfect-information (everyone knows everything), deterministic games, so you can exactly predict your opponents behavior by simulating &quot;what would I do if I were them, and trying to make me lose&quot;. The only real hard problem in this space is extreme branching factors.<p>Everything gets vastly more complicated once you break any of those rules: non-zero sum games create a prisoner&#x27;s dilemma cooperate&#x2F;defect dynamic, every three or more player game is non-zero sum (and exponentially more for every player you add), hidden information forces you to manage how much you reveal to your opponent and requires you to simulate multiple &quot;alternate futures&quot; based on things you learn after making a decision, and randomness is equivalent to an extra player that makes irrational unpredictable moves.<p>Games like that are vastly closer to the messy real world than the computationally expensive but near-ideal world of games like go, and they&#x27;re much more of an open problem.
评论 #21726117 未加载
评论 #21726133 未加载
评论 #21728525 未加载
noambrownover 5 years ago
Hi! I&#x27;m one of the authors on the paper. We&#x27;d be happy to answer any questions. Ask us anything!
评论 #21725460 未加载
评论 #21726904 未加载
评论 #21725540 未加载
评论 #21727935 未加载
评论 #21726730 未加载
hooandeover 5 years ago
they (basically) applied the ideas from a bot that plays poker to another game. it&#x27;s interesting work, though perhaps not groundbreaking.<p>This idea of selfplay + counterfactual regret minimization does seem to be the superior way to solve game theoretic problems. Identifying valuable game theoretic problems remains a challenge...
评论 #21725358 未加载
评论 #21725465 未加载
partingshotsover 5 years ago
They should tackle Starcraft II next like DeepMind has with AlphaStar, or at least a similar real time RTS like Starcraft with a fog of war and a partially observable state.
评论 #21732019 未加载