 |
| Lee Se-dol, a master Korean go player, places a stone during a match against the Google-developed artificial intelligence (AI) system AlphaGo at the Four Seasons Hotel in central Seoul, in this file photo taken in March 2016. / Yonhap |
By Lee Min-hyung
Smartphones will be banned from go matches here, with the rise of artificial intelligence (AI) casting growing concerns that smart devices can be unfairly used to give tips and tricks to help players beat opponents during real-time competitions, according to a local go association, Thursday.
"We will soon adopt the rule to prevent players from using smartphones in official go matches," a spokesman from the Korea Baduk Association said. Go is called baduk in Korea. "A specific timeline has yet to be drawn up about when it will take effect, as AI is still in its infancy."
The toughened regulation comes after the Google-developed AI supercomputer AlphaGo clinched a landslide 4-1 victory against last year against go player Lee Se-dol, one of the best in the world.
The historic matches took the global info-tech sector by storm, as go has been considered history's most complicated board game requiring constant planning and pattern-recognition skills by players for which computers had been thought to be no match for human beings.
"Supercomputers like AlphaGo have yet to be commercialized, and for now it takes millions of won to buy such an AI computerized system," the official said. "But when it becomes more popular and widespread, we will push ahead with more concrete regulations about the use of smartphones during matches."
The association also plans to adopt additional rules, disallowing players from taking bathroom breaks when their turn comes.
"As AI is developing at astonishing speed, we are in an internal discussion to adopt such measures to make competitions fair," the official said. "Our operation committee will soon confirm and make public the updated match rules."
Last March, Lee acknowledged his defeat against AlphaGo developed by Google's AI subsidiary DeepMind. AlphaGo ― disguising itself with online names such as Master ― has since racked up a 60-0 victory against renowned professional go players ― including top-tier players Ke Jie and Park Jung-hwan.
Prof. Jeong Jae-seung at Korea Advanced Institute of Science and Technology (KAIST) confirmed the fast development of AI in the Asian board game.
"I took part in a conference in Silicon Valley in early December where I learned there are already many algorithms much better than those of AlphaGo," said Prof. Jeong, a renowned AI expert.
"Just nine months after AlphaGo stunned the world, it has already become a legacy."
As the AI system gets much stronger by piling up more and more datasets, the widespread prediction is that humans will never beat the AI go platform down the road.
For this reason, the DeepMind team is working on its next challenge ― the real-time strategy game "StarCraft." In November, DeepMind announced its partnership with the game's developer, Blizzard. The game has long dominated the global strategy game industry, often dubbed the pinnacle of 1-versus-1 competitive gaming.
StarCraft is expected to be a tougher competitor for the DeepMind team, as it works on a real-time basis unlike the turn-based go.
