AlphaGo versus Ke Jie

• 1. The Future Of Go Summit

  ›

  Wuzhen, China hosted a three-game match in May 2017 between AlphaGo Master and Ke Jie. The event marked the Future of Go Summit, bringing together top human talent and advanced artificial intelligence. This gathering was not merely a competition but a statement about the evolving relationship between humans and machines. The dates spanned from the 23rd to the 27th of May, creating a week-long focus on strategic depth and computational power.

  12

• 2. AlphaGo Master Capabilities

  ›

  AlphaGo Master utilized four TPUs running on a single machine during these matches. Its Elo rating reached 4,858, placing it far above any human competitor at that time. DeepMind claimed this version was three stones stronger than the system used against Lee Sedol. It represented the second-best iteration available, as AlphaGo Zero existed but remained hidden until October 2017. Nature received their paper on AlphaGo Zero on April 7 before the games began with Ke Jie.

  789

• 3. Ke Jie World Ranking Status

  ›

  Ke Jie held the number one ranking among all human players worldwide under Rémi Coulom's system since late 2014. He also topped lists maintained by the Korea Baduk Association, Japan Go Association, and Chinese Weiqi Association. His status made him the ideal opponent for testing the limits of machine learning in complex strategy games. No other player had achieved such consistent dominance across multiple global organizations simultaneously.

  3456

• 4. Game One Analysis

  ›

  On the 23rd of May, AlphaGo played white and won by just 0.5 points after 289 moves. The match required every ounce of precision from both sides to reach such a narrow margin. Ke Jie fought hard but could not secure enough territory to overcome the AI's subtle advantages. The game ended with a result that left spectators debating whether a human victory was even possible at this level.

  13

• 5. Game Two And Three Outcomes

  ›

  The second game concluded on the 25th of May when Ke Jie resigned after move 155 while playing black. About an hour into play, Demis Hassabis tweeted that AlphaGo evaluated Ke as perfect, yet errors emerged later. By four hours, AlphaGo simplified positions and secured a clear win. Game three finished on the 27th of May with another resignation, bringing the series score to 3, 0. At that moment, AlphaGo held roughly ninety minutes of time remaining while Ke Jie had only thirty-two minutes left.

  10121415

• 6. Prize Money And Retirement

  ›

  Google DeepMind offered $1.5 million in winner prizes for the match while losers received $300,000 for participation. After defeating Ke Jie, AlphaGo retired from competitive play entirely. The Chinese Weiqi Association awarded it a professional 9-dan title recognizing its achievements. DeepMind shifted focus to other areas of artificial intelligence research following the conclusion of these matches.

  216