AlphaGo: influences, peers and legacy

Every connection with its receipt

AlphaGo was less a single invention than a convergence: decades of game-playing AI, a 1993 Monte Carlo heresy, London's games industry, and a 2,500-year-old board game carrying its own literature. This mix traces what fed DeepMind's program, who it raced against, and the open-source engines that now outplay it.

The Kynda mix for AlphaGo

• Key Influence · Deep Blue by IBM (1997). The benchmark AlphaGo was explicitly measured against: after Deep Blue beat Garry Kasparov in 1997, Go became the standing "next frontier" for game AI, cited constantly because its branching factor defeated brute-force search. DeepMind framed the Lee Sedol match in Seoul as Go's Deep Blue moment, while stressing the opposite method — learned evaluation and sampling rather than hand-tuned heuristics and exhaustive alpha-beta search.
• Influencia Obscura · Monte Carlo Go by Bernd Brügmann (1993). An unpublished paper by a physicist who proposed evaluating Go positions by playing thousands of random games to the end, borrowing simulated annealing from statistical physics. Ignored for a decade, it became the foundation of every strong Go engine after 2006 and supplied the rollout component of AlphaGo's search. Without this obscure piece of amateur research the program's tree search has no evaluation signal at all.
• Local Roots · Theme Park by Bullfrog Productions (1994). Before founding DeepMind in London, Hassabis was a teenage lead programmer and AI coder on Bullfrog's management sim, which modelled crowds of simulated visitors reacting to player decisions. The British games industry of Guildford and London — Bullfrog, Lionhead, his own Elixir Studios — was his actual training ground in building systems that behave, a lineage that runs directly into DeepMind's agent-based research culture.
• Beyond the Medium · The Master of Go by Yasunari Kawabata (1951). Kawabata's novelisation of the 1938 retirement match of Honinbo Shusai frames Go as a contest between an older aesthetic order and a ruthless modern rationalism that wins by grinding precision. The book became the standard reference point for commentators on the Lee Sedol match in 2016, because AlphaGo's shoulder hit at move 37 and its indifference to beauty replayed Kawabata's anxiety about tradition on a far larger stage.
• Peer · Crazy Stone by Rémi Coulom (2006). The strongest MCTS-era Go engine of the pre-deep-learning years, winner of multiple Computer Olympiad golds and the program that in 2013 beat professional Yoshio Ishida with a four-stone handicap. Coulom publicly estimated a decade more work before computers reached top human level; AlphaGo arrived in under three years, and Crazy Stone's later versions added neural networks in response.
• Key Collaborator · AlphaGo by Greg Kohs (2017). Kohs's documentary filmed the Seoul match from inside both camps, following lead researcher David Silver, engineer Aja Huang placing the stones, and Lee Sedol through the devastation of game two and the relief of game four. It is the primary visual record of the project's human collaborators and of move 78, the "divine move" that remains the only human win against the competition version.
• Legacy · Leela Zero by Gian-Carlo Pascutto (2017). An open-source reimplementation of the AlphaGo Zero paper, trained by distributed self-play contributed by volunteers worldwide because no individual had Google's hardware. It explicitly follows DeepMind's published architecture and brought superhuman Go analysis to any amateur with a laptop, while its chess sibling Leela Chess Zero did the same for AlphaZero's methods in the chess world.

What influenced AlphaGo

• Efficient Selectivity and Backup Operators in Monte-Carlo Tree Search by Rémi Coulom (2006). Efficient Selectivity and Backup Operators in Monte-Carlo Tree Search (Rémi Coulom) — ghost for AlphaGo “In 2006, Rémi Coulom described the application of the Monte Carlo method to game-tree search and coined the term Monte Carlo tree search in his paper, “Efficient Selectivity and Backup Operators in Monte-Carlo Tree Search.” He was one of…” (en.wikipedia.org)
• Deep Blue by IBM (1997). Deep Blue (IBM) — titan for AlphaGo “Almost two decades after IBM's computer Deep Blue beat world chess champion Garry Kasparov in the 1997 match, the strongest Go programs using artificial intelligence techniques only reached about amateur 5-dan level, and still could not …” (en.wikipedia.org)
• Hikaru no Go by Yumi Hotta and Takeshi Obata (1999). Hikaru no Go (Yumi Hotta and Takeshi Obata) — culture for AlphaGo (wikidata.org)
• The Master of Go by Yasunari Kawabata (1951). The Master of Go (Yasunari Kawabata) — culture for AlphaGo (wikidata.org)
• Bandit based Monte-Carlo Planning (UCT) by Levente Kocsis and Csaba Szepesvári (2006). Bandit based Monte-Carlo Planning (UCT) (Levente Kocsis and Csaba Szepesvári) — ghost for AlphaGo
• TD-Gammon by Gerald Tesauro (1992). TD-Gammon (Gerald Tesauro) — titan for AlphaGo
• Samuel's Checkers Player by Arthur Samuel (1959). Samuel's Checkers Player (Arthur Samuel) — titan for AlphaGo
• Monte Carlo Go by Bernd Brügmann (1993). Monte Carlo Go (Bernd Brügmann) — ghost for AlphaGo
• The Classic of Weiqi in Thirteen Chapters by Zhang Ni (1049). The Classic of Weiqi in Thirteen Chapters (Zhang Ni) — culture for AlphaGo

Peers and kindred spirits

• Greg Kohs. Greg Kohs directed the documentary film AlphaGo. (also via AlphaGo) “The lead up and the challenge match with Lee Sedol were documented in a documentary film also titled AlphaGo , [ 9 ] directed by Greg Kohs.” (Wikipedia)
• Theme Park by Bullfrog Productions (1994). Theme Park (Bullfrog Productions) — geography for AlphaGo “He began by playtesting on Syndicate and then at 17 co-designing and lead-programming on the 1994 game Theme Park, with the game's designer Peter Molyneux.” (en.wikipedia.org)
• Black & White by Lionhead Studios (2001). Black & White (Lionhead Studios) — geography for AlphaGo “At Lionhead, Hassabis worked as lead AI programmer on the 2001 god game Black & White.” (en.wikipedia.org)
• AlphaGo by Greg Kohs (2017). AlphaGo (Greg Kohs) — collaborator for AlphaGo “The lead up and the challenge match with Lee Sedol were documented in a documentary film also titled AlphaGo, directed by Greg Kohs.” (en.wikipedia.org)
• Reinforcement Learning: An Introduction by Richard S. Sutton and Andrew G. Barto (1998). Reinforcement Learning: An Introduction (Richard S. Sutton and Andrew G. Barto) — collaborator for AlphaGo “Sutton returned to Canada in the 2000s and continued working on the topic which continued to develop in academic circles until one of its first major real world applications saw Google's AlphaGo program built on this concept defeating th…” (en.wikipedia.org)
• Volker Bertelmann. Volker Bertelmann is identified as a producer of the documentary. “He also mentioned that with the passion of Hauschka's Volker Bertelmann, the film's producer, this documentary shows many unexpected sequences, including strategic and philosophical components.” (Wikipedia)
• Republic: The Revolution by Elixir Studios. Republic: The Revolution (Elixir Studios) — geography for AlphaGo “DeepMind has since become known for its development of advanced AI models such as AlphaGo, and was acquired by Google in 2014.” (en.wikipedia.org)
• Crazy Stone by Rémi Coulom (2006). Crazy Stone (Rémi Coulom) — peer for AlphaGo “In 2006, Rémi Coulom described the application of the Monte Carlo method to game-tree search and coined the term Monte Carlo tree search in his paper, “Efficient Selectivity and Backup Operators in Monte-Carlo Tree Search.” He was one of…” (en.wikipedia.org)
• Erica by Aja Huang (2010). Erica (Aja Huang) — collaborator for AlphaGo “In June 2016, at a presentation held at a university in the Netherlands, Aja Huang, one of the Deep Mind team, revealed that they had patched the logical weakness that occurred during the 4th game of the match between AlphaGo and Lee, an…” (en.wikipedia.org)
• Libratus by Noam Brown and Tuomas Sandholm (2017). Libratus (Noam Brown and Tuomas Sandholm) — peer for AlphaGo
• DarkForest by Facebook AI Research (2016). DarkForest (Facebook AI Research) — peer for AlphaGo

Who AlphaGo influenced

• Leela Zero by Gian-Carlo Pascutto (2017). Leela Zero (Gian-Carlo Pascutto) — legacy for AlphaGo
• KataGo by David J. Wu (2019). KataGo (David J. Wu) — legacy for AlphaGo