DeepMind details MuZero, revealed in 2019 and following AlphaZero, which can master games without knowing the rules and is working on YouTube video compression
Leo Kelion / BBC :
BBC Leo Kelion
Related Coverage
- MuZero: Mastering Go, chess, shogi and Atari without rules DeepMind Blog · Julian Schrittwieser
- Google develops an AI that can learn both chess and Pac-Man Ars Technica · John Timmer
- Mastering Atari, Go, chess and shogi by planning with a learned model Nature · David Silver
- DeepMind MuZero AI Learns The Rules As Its Plays To Master Atari Games, Chess, And More HotHardware.com News · Nathan Ord
- What AlphaGo Can Teach Us About How People Learn Wired · Will Knight
- DeepMind MuZero AI can master games without knowing the rules SlashGear · JC Torres
- New research shows that an AI can learn by observation Axios · Bryan Walsh
- DeepMind unveils new details of game-mastering A.I. it says may help in complex real-world environments Fortune · Jeremy Kahn
- DeepMind's latest AI can master games without being told their rules Engadget · Igor Bonifacic
- DeepMind just taught AI how to win a game without knowing the rules BGR · Chris Smith
- No rules, no problem: DeepMind's MuZero masters games while learning how to play them TechCrunch · Devin Coldewey
- DeepMind's MuZero picks up the rules of games as it plays VentureBeat · Kyle Wiggers
Discussion
-
@deepmind
@deepmind
on x
In 2016, AlphaGo was introduced. Two years later, its successor - AlphaZero - showed significant progress in Go, chess and shogi. Today in @Nature, our team describes MuZero, a significant step forward in the pursuit of general-purpose algorithms: https://go.nature.com/38qb2I8 ht…
-
@mononofu
Julian Schrittwieser
on x
I'm happy to announce that our MuZero paper has been published in Nature: https://go.nature.com/38qb2I8 ! Please see our official blog post https://deepmind.com/... for more; also see my personal blog for a high level description of MuZero: https://www.furidamu.org/... https://tw…
-
@baldfuturist
Steve Brown
on x
@DeepMind continues to make astounding progress with their deep reinforcement learning efforts. The faster an AI can learn to perform tasks and the less you have to explain to it up front, the more useful AIs get. Their latest is the impressive MuZero. https://www.bbc.co.uk/...
-
@koraykv
Koray Kavukcuoglu
on x
Most real world problems are complex and have unknown dynamics, so we need agents that can learn to plan on their own. MuZero is a very exciting development towards that goal - congrats to the team! https://twitter.com/...
-
@deepmind
@deepmind
on x
Work by @Mononofu, Ioannis Antonoglou, Thomas Hubert, Karen Simonyan, @laurentsifre, Simon Schmitt, Arthur Guez, @EdwardLockhart, @demishassabis, @ThoreG, Timothy Lillicrap and David Silver.
-
@ankesh_anand
Ankesh Anand
on x
Interesting #ICML2020 paper, a follow up to AlphaZero/MuZero: https://proceedings.icml.cc/ ... - recasts AZ/MZ as a policy optimization problem - leads to a new algo that works better at low number of simulations - extends MuZero to continuos control envs https://twitter.com/...
-
@deepmind
@deepmind
on x
MuZero, which was first introduced in a preliminary paper in 2019, masters Go, chess, shogi and Atari without needing to be told the rules, thanks to its ability to plan winning strategies in unknown environments. Read more: https://deepmind.com/...
-
@daveleeft
Dave Lee
on x
Interesting piece by @LeoKelion. I wonder how DeepMind feels about their best work being channeled into.... reducing costs at YouTube https://www.bbc.com/...
-
@sam_l_shead
Sam Shead
on x
Nice piece from @LeoKelion on DeepMind's new agent: MuZero. “For the first time, we actually have a system which is able to build its own understanding of how the world works,” says DeepMind's principal research scientist David Silver. https://www.bbc.co.uk/...