Off the wire
• Russia checks readiness of Caspian Flotilla  • BOJ maintains economic assessment on 8 out of 9 regional economies  • Australia to develop, deploy new air missile defence system  • Indonesia's Bali to hold int'l recreational running event  • Chinese tracking ship Yuanwang-7 begins space monitoring mission  • World No.1 Go player Ke Jie takes on upgraded AlphaGo in May  • Tokyo stocks close higher as weak yen gives exporters boost  • 95 pct of Taiwan youth dissatisfied with work: survey  • Great Barrier Reef severely damaged from bleaching, Cyclone Debbie  • Water diverted to wetland in Xiongan New Area  
You are here: 　 Home

World No.1 Go player Ke Jie takes on upgraded AlphaGo in May (updated)

Xinhua, April 10, 2017 Adjust font size:

Chinese Go player Ke Jie, arguably the current best in this sport, has set up a highly anticipated and ultimate showdown with the upgraded computer program AlphaGo in southern China's Wuzhen in May.

Ke, 19, who became the youngest champion in Go history to win three world titles within the space of one year between January 2015 and January 2016, will take on AlphaGo on May 23-27 in three rounds of regular games, the Chinese Go Association announced here on Monday at a news conference.

Each player will have three hours in each of the three games and the winner will be awarded 1.5 million U.S. dollars while the losing side takes 300,000 dollars.

It will be the second meeting between top professional player and the artificial intelligence program in this ancient board game. In March 2016, Google's AlphaGo scored a 4-1 victory over South Korean Go master Lee Se-dol in a historic five-round match, overturning the long existing fact that machine can't beat professional Go players due to the game's complex, intuitive and creative nature.

As far as Ke is concerned, meeting AlphaGo is not just an honor but also a unique experience.

"I am truly honored to represent human players to take on AlphaGo," said Ke. "Watching AlphaGo play with such an unusual way of thinking makes me feel that I am about to face someone from the future."

"I must hold a winning attitude while fighting with the ultimate determination," he added.

In the May showdown at the popular tourist destination in Zhejiang province, native Ke will face a stronger AlphaGo than a year ago as the program, developed by DeepMind, boasts of a deep learning capability to learn for itself and discover new strategies by playing games against itself and adjusting neural networks based on a trial-and-error process known as reinforcement learning.

As a matter of fact, the new version of AlphaGo already demonstrated its improved strength as it recorded a 60-0 result in blitz games against a number of top Go players on two online Go game sites under the ID names "Magister" and "Master" in early 2017.

Go, known as Weiqi in China and Baduk in South Korea, originated from China thousands of years ago. It involves two players who take turns putting white and black stones on a grid of 19 lines by 19 lines. One can win an opponent when gaining more territory on the grid. One can remove stones of the opponent by surrounding the pieces. Enditem