Baduk and Google’s AlphaGo

The American Go Association says that that the game of Go, called “baduk” in Korea, is thought to have originated as far back as 2500 to 4000 years ago.  It originated in China, and then spread to Korea and the rest of the world.  Koreans have been competing professionally for the past 65 years and in the past decades have truly dominated the game.  Back in 2004, calling it “the world’s oldest surviving game of pure mental skill,” the Economist said that a computer program had not been designed yet that could win at Go.  What a difference a decade makes! 

I had the pleasure of meeting some of the best and brightest of both the computer programming world and the world of baduk recently when I attended the Google DeepMind gala dinner and Challenge Match between Master Champion Lee Sedol and Google’s Artificial Intelligence program AlphaGo.

Having dinner with Google programmer Jeff Dean (Photo courtesy of Google)

It was great to meet Lee and to see Eric Schmidt, former Google CEO and now Executive Chairman of Alphabet, a Google subsidiary.  Compared to chess, baduk has a huge number of options for play, which had made it impossible for the computer to calculate all of the potential variations of play.  Also, the best baduk players are often thought to play based on intuition and experience, something that has been hard to replicate in a computer program.

(From left) Demis Hassabis, CEO of Google DeepMind, Korean professional baduk player Lee Sedol, and Eric Schmidt, Executive Chairman of Alphabet in Seoul ahead of the Google DeepMind Challenge Match (AP photo)

To prepare for the match, AlphaGo was tasked with playing games against itself and “studying” matches of master players.  The potential victory in this regard is for artificial intelligence to be able to work in this manner and be used for solving complex health and science problems. Artificial intelligence could perhaps develop previously unimagined breakthroughs in these areas. 

AlphaGo vs. Lee Sedol (AP photo)

Mr. Lee should not feel bad that the machine won!  Mr. Lee clearly did his best, won a game, and has already asked for a re-match.  The New York Times recently quoted Stanford University computer scientist Fei-Fei Li, Director of the Stanford Artificial Intelligence Laboratory, as saying, “I’m not surprised at all.  How come we are not surprised that a car runs faster than the fastest human?”  

It was great fun to come out and watch the match.  Congratulations to Mr. Lee and the Google team for all of their hard work and effort!  

No comments:

Powered by Blogger.