Difference between revisions of "Timeline of AlphaGo"

From Timelines
Jump to: navigation, search
Line 34: Line 34:
 
|-
 
|-
 
| 2016 || May || || Google unveils its own proprietary hardware "{{w|tensor processing unit}}s", which states having already been deployed in multiple internal projects at Google, including the AlphaGo match against Lee Sedol.<ref>{{cite news|last1=McMillan|first1=Robert|title=Google Isn’t Playing Games With New Chip|url=https://www.wsj.com/articles/google-isnt-playing-games-with-new-chip-1463597820|accessdate=21 May 2019|work={{w|Wall Street Journal}}|date=18 May 2016}}</ref><ref>{{cite web|url=https://cloudplatform.googleblog.com/2016/05/Google-supercharges-machine-learning-tasks-with-custom-chip.html|title=Google supercharges machine learning tasks with TPU custom chip|last=Jouppi|first=Norm|date=May 18, 2016|website=Google Cloud Platform Blog|language=en-US|access-date=21 May 2019}}</ref>
 
| 2016 || May || || Google unveils its own proprietary hardware "{{w|tensor processing unit}}s", which states having already been deployed in multiple internal projects at Google, including the AlphaGo match against Lee Sedol.<ref>{{cite news|last1=McMillan|first1=Robert|title=Google Isn’t Playing Games With New Chip|url=https://www.wsj.com/articles/google-isnt-playing-games-with-new-chip-1463597820|accessdate=21 May 2019|work={{w|Wall Street Journal}}|date=18 May 2016}}</ref><ref>{{cite web|url=https://cloudplatform.googleblog.com/2016/05/Google-supercharges-machine-learning-tasks-with-custom-chip.html|title=Google supercharges machine learning tasks with TPU custom chip|last=Jouppi|first=Norm|date=May 18, 2016|website=Google Cloud Platform Blog|language=en-US|access-date=21 May 2019}}</ref>
 +
|-
 +
| 2016 || October 12 || Research || DeepMind publishes paper on {{w|differentiable neural computers}}, which demonstrates that models that can simultaneously learn like neural networks as well as memorize data like computers.<ref name="DeepMind’s work in 2016: a round-up">{{cite web |title=DeepMind’s work in 2016: a round-up |url=https://deepmind.com/blog/deepmind-round-up-2016/ |website=deepmind.com |accessdate=23 May 2019}}</ref>
 
|-
 
|-
 
| 2016 || December || Game series release || Alphago Online Series on Tygem and Fox is released on Alphago Master version. It consists of sixty games.<ref name="Alphago's Games"/>
 
| 2016 || December || Game series release || Alphago Online Series on Tygem and Fox is released on Alphago Master version. It consists of sixty games.<ref name="Alphago's Games"/>

Revision as of 13:42, 23 May 2019

This is a timeline of AlphaGo, a computer program that plays the board game Go. "Go is a game of profound complexity. There are an astonishing 10 to the power of 170 possible board configurations - more than the number of atoms in the known universe - making Go a googol times more complex than Chess."[1]

Big picture

Time period Development summary
2016 AlphaGo victory in March becomes a major milestone in artificial intelligence research,[2] with Go being previously been regarded as a hard problem in machine learning expected to be out of reach for the technology of the time.[2][3][4]

Full timeline

Year Month and date Event type Details
3000 BP Prelude The game of Go originates in China. The rules of the game are simple: players take turns to place black or white stones on a board, trying to capture the opponent's stones or surround empty space to make points of territory.[1]
2010 Prelude DeepMind is founded to create general-purpose artificial intelligence that can learn on its own.[5]
2015 October Notable match AlphaGo versus Fan Hui is held at DeepMind's headquarters in London.[6] The distributed version of AlphaGo defeates the European Go champion Fan Hui, a 2-dan (out of 9 dan possible) professional, five to zero.[7][8] AlphaGo wins all the five games.[7][9] This is the first time a computer Go program beats a professional human player on a full-sized board without handicap.[10]
2015 October The original AlphaGo becomes the first computer Go program to beat a human professional Go player without handicaps on a full-sized 19×19 board.[11][7]
2015 October Game series release Fan Hui versus Alphago is released in v13 version. It consists of five games.[1]
2016 January 28 Research AlphaGo's team publishes an article in the journal Nature, describing the technical details behind the reinforcement learning approach used in AlphaGo.[12]
2016 February Game series release Alphago versus Alphago is released in v18 version, It consists of three games.[1]
2016 March Notable match AlphaGo versus Lee Sedol takes place as a five-game Go match between 18-time world champion Lee Sedol and AlphaGo. Played in Seoul, South Korea between 9 and 15 March 2016, AlphaGo wins all but the fourth game.[13] This is the first time a computer Go program beats a 9-dan professional without handicaps.[14][15][16][17] The match is watched by over 200 million people worldwide.[18]
2016 March Game series release Lee Sedol versus Alphago is released on v18 version. It consists of five games.[1]
2016 May Google unveils its own proprietary hardware "tensor processing units", which states having already been deployed in multiple internal projects at Google, including the AlphaGo match against Lee Sedol.[19][20]
2016 October 12 Research DeepMind publishes paper on differentiable neural computers, which demonstrates that models that can simultaneously learn like neural networks as well as memorize data like computers.[21]
2016 December Game series release Alphago Online Series on Tygem and Fox is released on Alphago Master version. It consists of sixty games.[1]
2017 January Notable match Under the pseudonym "Master", AlphaGo plays several of the world's top players in a series of online matches, including Ke Jie, winning all 60 of its completed contests.[22][23][1]
2017 May International meeting The Future of Go Summit is held by the Chinese Go Association, Sport Bureau of Zhejiang Province and Google in Wuzhen, Zhejiang, the permanent host of the World Internet Conference. It features five Go games involving AlphaGo and top Chinese Go players,[24] as well as a forum on the future of artificial intelligence.[25][18][1]
2017 May Game series release Future of Go Summit is released on Alphago Master version. It consists of five games.[1]
2017 May Game series release Alphago versus Alphago Batch 1 to 5 are released on Alphago Master version. Each series consisis of ten games.[1]
2017 July Game series release WeiQi TV-5 Extra Games is released on Alphago Master version. It consists of five games.[1]
2017 September 29 The AlphaGo documentary film is released.[26]
2017 October 19 Research AlphaGo's team publishes an article in the journal Nature, introducing AlphaGo Zero, a version created without using data from human games, and stronger than any previous version.[27][1]
2017 October 25 Software release Leela Zero, a free and open-source computer Go software, is released. It is developed by Belgian programmer Gian-Carlo Pascutto,[28][29][30] the author of chess engine Sjeng and Go engine Leela.[31][32]
2017 October It is announced that AlphaGo Zero, armed with just the rules, has in 40 days become even better at Go than the original AlphaGo, without the help of game records.[17]
2017 October Game series release AlphaGo Zero (20 Blocks) vs AlphaGo Lee is released on Alphago Zero version. It consists of twenty games.[1]
2017 October Game series release AlphaGo Zero vs AlphaGo Zero - 20 Blocks is released on Alphago Zero version. It consists of twenty games.[1]
2017 October Game series release AlphaGo Zero vs AlphaGo Zero - 40 Blocks is released on Alphago Zero version. It consists of twenty games.[1]
2017 October Game series release AlphaGo Zero (40 Blocks) vs AlphaGo Master is released in Alphago Zero version. It consists of twenty games.[1]
2017 December 5 DeepMind team releases a preprint on arXiv, introducing AlphaZero, a program using generalized AlphaGo Zero's approach, which achieved within 24 hours a superhuman level of play in chess, shogi, and Go, defeating world-champion programs, Stockfish, Elmo, and 3-day version of AlphaGo Zero in each case.[33]
2017 December DeepMind releases AlphaGo teaching tool on its website,[34] to analyze winning rates of different Go openings as calculated by AlphaGo Master.[35] 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.[35][1]
2017 December Notable match AlphaZero beats the 3-day version of AlphaGo Zero by winning 60 games to 40, and with 8 hours of training it outperformed AlphaGo Lee on an Elo scale. AlphaZero also defeats a top chess program (Stockfish) and a top Shōgi program (Elmo).[33]
2018 April Research Paper published in Nature cites AlphaGo's approach as the basis for a new means of computing potential pharmaceutical drug molecules.[36][37]

Meta information on the timeline

How the timeline was built

The initial version of the timeline was written by User:Sebastian.

Funding information for this timeline is available.

Feedback and comments

Feedback for the timeline can be provided at the following places:

  • FIXME

What the timeline is still missing

Timeline update strategy

See also

External links

References

  1. 1.00 1.01 1.02 1.03 1.04 1.05 1.06 1.07 1.08 1.09 1.10 1.11 1.12 1.13 1.14 1.15 1.16 "Alphago's Games". alphago-games.com. Retrieved 21 May 2019. 
  2. 2.0 2.1 Steven Borowiec; Tracey Lien (12 March 2016). "AlphaGo beats human Go champ in milestone for artificial intelligence". Los Angeles Times. Retrieved 13 March 2016. 
  3. Connor, Steve (27 January 2016). "A computer has beaten a professional at the world's most complex board game". The Independent. Retrieved 21 May 2019. 
  4. "Google's AI beats human champion at Go". CBC News. 27 January 2016. Retrieved 21 May 2019. 
  5. "What we learned in Seoul with AlphaGo". blog.google. Retrieved 21 May 2019. 
  6. Metz, Cade (27 January 2016). "In Major AI Breakthrough, Google System Secretly Beats Top Player at the Ancient Game of Go". WIRED. Retrieved 1 February 2016. 
  7. 7.0 7.1 7.2 "Google achieves AI 'breakthrough' by beating Go champion". BBC News. 27 January 2016. 
  8. "Special Computer Go insert covering the AlphaGo v Fan Hui match" (PDF). British Go Journal. 2017. Retrieved 1 February 2016. 
  9. "Special Computer Go insert covering the AlphaGo v Fan Hui match" (PDF). British Go Journal. Retrieved 1 February 2016. 
  10. "Première défaite d'un professionnel du go contre une intelligence artificielle". Le Monde (in français). 27 January 2016. 
  11. "Research Blog: AlphaGo: Mastering the ancient game of Go with Machine Learning". Google Research Blog. 27 January 2016. 
  12. "The story of AlphaGo so far". deepmind.com. Retrieved 23 May 2019. 
  13. "Artificial intelligence: Go master Lee Se-dol wins against AlphaGo program". BBC News Online. 13 March 2016. Retrieved 13 March 2016. 
  14. "Match 1 – Google DeepMind Challenge Match: Lee Sedol vs AlphaGo". 8 March 2016. 
  15. "DeepMind's AlphaGo Zero Becomes Go Champion Without Human Input". futureoflife.org. Retrieved 5 April 2019. 
  16. "AI: How big a deal is Google's latest AlphaGo breakthrough?". techrepublic.com. Retrieved 5 April 2019. 
  17. 17.0 17.1 "AlphaGo". britgo.org. Retrieved 5 April 2019. 
  18. 18.0 18.1 "AlphaGo China | DeepMind". DeepMind. Retrieved 21 May 2019. 
  19. McMillan, Robert (18 May 2016). "Google Isn't Playing Games With New Chip". Wall Street Journal. Retrieved 21 May 2019. 
  20. Jouppi, Norm (May 18, 2016). "Google supercharges machine learning tasks with TPU custom chip". Google Cloud Platform Blog. Retrieved 21 May 2019. 
  21. "DeepMind's work in 2016: a round-up". deepmind.com. Retrieved 23 May 2019. 
  22. "AN IMPROVED ALPHAGO WINS ITS FIRST GAME AGAINST THE WORLD'S TOP GO PLAYER". wired.com. Retrieved 5 April 2019. 
  23. "DeepMind's AI beats world's best Go player in latest face-off". newscientist.com. Retrieved 5 April 2019. 
  24. "Exploring the mysteries of Go with AlphaGo and China's top players". DeepMind. 10 April 2017. Retrieved 21 May 2019. 
  25. "DeepMind's AlphaGo is back..and this time it's taking on five humans at once". Wired.com. 10 April 2017. Retrieved 21 May 2019. 
  26. "AlphaGo (2017)". Rotten Tomatoes. Retrieved 23 May 2019. 
  27. Silver, David; Schrittwieser, Julian; Simonyan, Karen; Antonoglou, Ioannis; Huang, Aja; Guez, Arthur; Hubert, Thomas; Baker, Lucas; Lai, Matthew; Bolton, Adrian; Chen, Yutian; Lillicrap, Timothy; Fan, Hui; Sifre, Laurent; Driessche, George van den; Graepel, Thore; Hassabis, Demis (19 October 2017). "Mastering the game of Go without human knowledge". Nature. 550 (7676): 354–359. ISSN 0028-0836. PMID 29052630. doi:10.1038/nature24270. Retrieved 10 December 2017. 
  28. "Feature: One man's Go program looks to remake AlphaGo Zero - and beyond". Xinhuanet. 9 April 2018. Retrieved 5 April 2019. 
  29. "围棋AI"丽拉"获赞接近职业棋手水准,它的作者竟是一个不太会下棋的程序员" (in 中文). Xinhuanet. 5 February 2018. Retrieved 5 April 2019. 
  30. "更开放,更共享,比利时围棋AI"丽拉·元"重塑"阿尔法元"" (in 中文). Xinhuanet. 8 April 2018. Retrieved 5 April 2019. 
  31. "프로 수준급 인공지능 바둑 프로그램 '릴라(Leela)' 무료 공개" (in Korean). Baduk News. 23 February 2017. Retrieved 5 April 2019. 
  32. "릴라의 출현과 온라인 대국의 비극적인 종말..." (in Korean). Cyberoro. 3 March 2017. Retrieved 5 April 2019. 
  33. 33.0 33.1 Silver, David; Hubert, Thomas; Schrittwieser, Julian; Antonoglou, Ioannis; Lai, Matthew; Guez, Arthur; Lanctot, Marc; Sifre, Laurent; Kumaran, Dharshan; Graepel, Thore; Lillicrap, Timothy; Simonyan, Karen; Hassabis, Demis (5 December 2017). "Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm". 
  34. "AlphaGo teaching tool". DeepMind. 
  35. 35.0 35.1 "AlphaGo教学工具上线 樊麾:使用Master版本" (in Chinese). Sina.com.cn. 11 December 2017. Retrieved 11 December 2017. 
  36. "Go and make some drugs The Engineer". www.theengineer.co.uk. Retrieved 21 May 2019. 
  37. Segler, Marwin H. S.; Preuss, Mike; Waller, Mark P. "Planning chemical syntheses with deep neural networks and symbolic AI".