As reported by The Korea Economic Daily, top-ranked Go grandmaster Shin Jin-seo became the first human to defeat the AI platform KataGo on July 21. It's a surprising, encouraging sign of human capabilities almost 30 years after supercomputer Deep Blue's historic defeat of Chess grandmaster Garry Kasparov.
Go is believed to be the oldest board game that is still widely played, originating in China and beating out Chess by at least 1400 years, give or take. The strategy game sees players in control of different-colored stones, working to encircle their opponent's pieces, capturing territory and stones to achieve victory. Like Chess, its complexity, ubiquity, and the large amount of study and data surrounding the game made Go a fruitful tool for machine learning research.
Shin, 26, is the first human to win an official, three-game series against KataGo, which the Korea Economic Daily described as "the world's premier artificial intelligence Go engine." Shin's odds were long, especially considering that, 10 years ago, Go grandmaster Lee Sedol was decisively defeated by Google's AlphaGo engine—the technology has only advanced in the intervening years.
Shin played with an advantageous, two-stone handicap—seemingly standard for matches between humans and top AIs—but indicated that he would like to face AI opponents "under more disadvantageous conditions" in the future. Shin came back from an initial defeat to take the series 2-1 against KataGo. Shin's final, three-hour match was actually fairly tame as far as the duration of professional Go games, uh, go—games as long as six hours don't seem particularly rare.
"I believe this series holds immense significance because it clearly demonstrated that humans can still hold their own against AI," Shin said after the final match. "Early on, I simply copied AI moves, which led to heavy fighting and frequent, easy losses. This series taught me that rather than trying to imitate AI, it is far more important to build the board according to my own style."
