# Google's MuZero

**URL:** <https://discourse.julialang.org/t/googles-muzero/52364>\
**Category:** Machine Learning\
**Created:** [December 25, 2020, 10:42am UTC](https://discourse.julialang.org/t/googles-muzero/52364 "2020-12-25T10:42:56Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![StatisticalMouse](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/statisticalmouse/32/43370_2.png) [@StatisticalMouse](https://discourse.julialang.org/u/StatisticalMouse)\
**Post date:** [December 25, 2020, 10:42am UTC](https://discourse.julialang.org/t/googles-muzero/52364/1 "2020-12-25T10:42:56Z")

</div>

This looks like a big deal. It uses model-based reinforcement learning and can learn to play both chess and Pac-Man. Abstract from the Science article:

" Constructing agents with planning capabilities has long been one of the main challenges in the pursuit of artificial intelligence. Tree-based planning methods have enjoyed huge success in challenging domains, such as chess[1](https://www.nature.com/articles/s41586-020-03051-4#ref-CR1) and Go[2](https://www.nature.com/articles/s41586-020-03051-4#ref-CR2), where a perfect simulator is available. However, in real-world problems, the dynamics governing the environment are often complex and unknown. Here we present the MuZero algorithm, which, by combining a tree-based search with a learned model, achieves superhuman performance in a range of challenging and visually complex domains, without any knowledge of their underlying dynamics. The MuZero algorithm learns an iterable model that produces predictions relevant to planning: the action-selection policy, the value function and the reward. When evaluated on 57 different Atari games[3](https://www.nature.com/articles/s41586-020-03051-4#ref-CR3)—the canonical video game environment for testing artificial intelligence techniques, in which model-based planning approaches have historically struggled[4](https://www.nature.com/articles/s41586-020-03051-4#ref-CR4)—the MuZero algorithm achieved state-of-the-art performance. When evaluated on Go, chess and shogi—canonical environments for high-performance planning—the MuZero algorithm matched, without any knowledge of the game dynamics, the superhuman performance of the AlphaZero algorithm[5](https://www.nature.com/articles/s41586-020-03051-4#ref-CR5) that was supplied with the rules of the game."

> **[Google develops an AI that can learn both chess and Pac-Man](https://arstechnica.com/science/2020/12/google-develops-an-ai-that-can-learn-both-chess-and-pac-man/)**
>
> MuZero handles both rules-based and open-ended games.

The Science article, as referred to by above, is here.

[http://dx.doi.org/10.1038/s41586-020-03051-4](http://dx.doi.org/10.1038/s41586-020-03051-4)