Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm

TL;DR

This paper introduces AlphaZero, a general reinforcement learning algorithm that learns to master chess, shogi, and Go from scratch through self-play, achieving superhuman performance without domain-specific knowledge.

Contribution

The paper presents a unified algorithm capable of mastering multiple complex games from scratch, outperforming specialized programs in each domain.

Findings

AlphaZero mastered chess, shogi, and Go within 24 hours.

It defeated the best existing programs in each game.

The approach is general and does not require domain-specific knowledge.

Abstract

The game of chess is the most widely-studied domain in the history of artificial intelligence. The strongest programs are based on a combination of sophisticated search techniques, domain-specific adaptations, and handcrafted evaluation functions that have been refined by human experts over several decades. In contrast, the AlphaGo Zero program recently achieved superhuman performance in the game of Go, by tabula rasa reinforcement learning from games of self-play. In this paper, we generalise this approach into a single AlphaZero algorithm that can achieve, tabula rasa, superhuman performance in many challenging domains. Starting from random play, and given no domain knowledge except the game rules, AlphaZero achieved within 24 hours a superhuman level of play in the games of chess and shogi (Japanese chess) as well as Go, and convincingly defeated a world-champion program in each case.