It's time for the humans to have another Go.
An artificial intelligence program romped to a lopsided victory over South Korean Go master Lee Sedol in 2016. Now, Google DeepMind's AI software is heading to the 2500-year-old board game's roots, taking on top-ranked Chinese player Ke Jie in May in a formal rematch between man and machine.
The Alphabet company and China's government are convening a five-day AI symposium from May 23 in the picturesque water-town of Wuzhen, expected to draw some of the top minds in the field from both Google and around the country. The proceedings includen not just the marquee human-AI match-up, but also a number of experimental matches in which Go masters may team up with their own AI counterparts, or join forces against a single machine player.
The idea is to showcase the evolution of machine intelligence, Demis Hassabis, chief executive officer and and co-founder of Google DeepMind, wrote in a blog post. The aim of the forum is to discuss how machine-learning methods behind AlphaGo can be useful in grappling with real-world issues such as energy consumption.
"There remains much more to learn from this partnership between Go's best human players and its most creative AI competitor," he wrote.
AlphaGo made headlines last year after winning a five-match tournament against Lee, who was considered the world's best player of Go over the past decade. DeepMind's success astounded experts, who thought it would take as much as a decade before AI could beat top-ranked professional players of the game. While its rules are simple – players battle for territory by placing white or black stones on a 19-by-19 grid of squares – it's regarded as far more complex than chess, by an order of magnitude of 10 followed by 99 zeros.
The widely covered contest provoked discussion on social media about whether the AI could in fact beat a player from the nation that spawned the game several millennia ago. AlphaGo, however, has since engaged in a number of casual online matches – including with Ke Jie –- and consistently won.