In March last year, Google's DeepMind artificial intelligence system AlphaGo made history by defeating world Go champion Lee Sedol in a five-game series between man and machine in Seoul, South Korea. Now, the UK-based firm is taking its board-playing AI software to China next month to test how far AlphaGo has come since then.
Towards the end of May, AlphaGo will take on three different challenges during the Future of Go Summit in Wuzhen, China, DeepMind announced in a blog post on Monday (10 April). The summit will run from 23-27 May in collaboration with the China Go Association and the Chinese Government.
"The summit will feature a variety of game formats involving AlphaGo and top Chinese players, specifically designed to explore the mysteries of the game together," DeepMind co-founder and CEO Demis Hassabis wrote. "There remains much more to learn from this partnership between Go's best human players and its most creative AI competitor."
The Pair Go challenge will see two Chinese Go professionals playing against each other, each of whom will be partnered with their own AlphaGo teammate. The human and AI players will alternate moves "to take the concept of 'learning together' quite literally".
The Team Go competition will have AlphaGo face off a five-player team consisting of China's top professional Go players working together to test the software's adaptability against their combined skills, style and ingenuity.
The third competition will be a classic 1:1, "man versus machine" match between AlphaGo and the world's number one player Ke Jie.
Back in March last year when AlphaGo took down Lee Sedol, then 18-year-old Ke Jie said: "Even if AlphaGo can defeat Lee Sedol, it can't beat me." He also agreed to take on the AI himself after initially refusing to do so because he didn't want the system to learn his style of play.
"On top of his excellent reading and unshakeable confidence, Ke Jie is known to play with a finely balanced style, known intuitively when to play boldly or exercise caution," DeepMind cofounder and CEO Demis Hassabis wrote.
Earlier this year, Ke Jie was one of several top ranked Go players who played against a mysterious user named "Master" in online matches on two Chinese board game platforms. The user was later revealed to be DeepMind's AlphaGo.
Go is a 3,000-year-old ancient Chinese strategy board game that is seen as much more complex than chess and relies on a certain level of "human intuition".
"Instead of diminishing the game, as some feared, artificial intelligence has actually made human players stronger and more creative," Hassabis wrote. "It's humbling to see how pros and amateurs alike, who have pored over every detail of AlphaGo's innovative game play, have actually learned new knowledge and strategies about perhaps the most studied and contemplated game in history."
AlphaGo, developed by DeepMind which Google acquired in 2014, uses neural networks and machine learning to learn from experience, improve itself and adjust its strategy based on trial and error.
"AlphaGo is beyond the imagination of Go players," Ke Jie said during an event in Beijing this week, the South China Morning Post reports. "It feels like I am standing at a point where history changes. It is a great honour to witness a strong competitor like AlphaGo and I will do my best and will not give up easily."