1,714
edits
Changes
→Full timeline
| 2017 || October || Game series release || AlphaGo Zero (40 Blocks) vs AlphaGo Master is released in Alphago Zero version. It consists of twenty games.<ref name="Alphago's Games">{{cite web |title=Alphago's Games |url=https://www.alphago-games.com/ |website=alphago-games.com |accessdate=21 May 2019}}</ref>
|-
| 2017 || December 5 || Software release || The DeepMind team releases a preprint on {{w|arXiv}}, introducing AlphaZero, a program using generalized AlphaGo Zero's approach, which achieved within 24 hours a superhuman level of play in {{w|chess}}, {{w|shogi}}, and [[w:Go (game)|Go]], defeating world-champion programs, [[w:Stockfish (chess)|Stockfish]], [[w:Elmo (shogi engine)|Elmo]], and 3-day version of AlphaGo Zero in each case.<ref name="preprint"/>
|-
| 2017 || December || Software release || DeepMind releases the AlphaGo teaching tool on its website,<ref>{{cite web|url=https://alphagoteach.deepmind.com/|title=AlphaGo teaching tool|publisher={{w|DeepMind}}}}</ref> to analyze winning rates of different {{w|Go opening}}s as calculated by {{w|AlphaGo Master}}.<ref name="sina20171211">{{cite web|url=http://sports.sina.com.cn/go/2017-12-11/doc-ifypnsip8212788.shtml|title=AlphaGo教学工具上线 樊麾：使用Master版本|publisher={{w|Sina.com.cn}}|date=11 December 2017|accessdate=11 December 2017|language=Chinese}}</ref> The teaching tool collects 6,000 Go openings from 230,000 human games each analyzed with 10,000,000 simulations by AlphaGo Master. Many of the openings include human move suggestions.<ref name="sina20171211"/><ref name="Alphago's Games"/>
|-
| 2017 || December || Achievement || AlphaZero beats the 3-day version of {{w|AlphaGo Zero}} by winning 60 games to 40, and with 8 hours of training it outperformes {{w|AlphaGo Lee}} on an [[w:Elo rating system|Elo scale]]. AlphaZero also defeats a top chess program ([[w:Stockfish (chess)|Stockfish]]) and a top Shōgi program ([[w:Elmo (shogi engine)|Elmo]]).<ref name="preprint">{{Cite web|first1=David|last1= Silver|first2=Thomas|last2= Hubert|first3= Julian|last3=Schrittwieser|first4= Ioannis|last4=Antonoglou |first5= Matthew|last5= Lai|first6= Arthur|last6= Guez|first7= Marc|last7= Lanctot|first8= Laurent|last8= Sifre|first9= Dharshan|last9= Kumaran|first10= Thore|last10= Graepel|first11= Timothy|last11= Lillicrap|first12= Karen|last12= Simonyan|first13=Demis |last13=Hassabis|title=Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm|class=cs.AI|date=5 December 2017|url=https://arxiv.org/abs/1712.01815}}</ref>
