A computer beat China’s top player of go, one of the last games machines have yet to master, for a second time Thursday, a sign that the field of artificial intelligence is advancing faster than expected.
An IBM supercomputer known as Deep Blue defeated chess grandmaster Garry Kasparov in 1997. But many go players expected it to be at least 10 more years before computers mastered go, which is considered far more complicated for machines to master.
Go players take turns putting white or black stones on a rectangular grid with 361 intersections, trying to capture territory and each other’s pieces by surrounding them. The near-infinite number of possible positions requires intuition and flexibility — traits that human beings long believed a computer could never possess.
But then European and South Korean go champions began to fall to Google’s AlphaGo computer program. The program defeated Ke Jie (pronounced kuh jay), a 19-year-old Chinese prodigy, on Tuesday and then again two days later, during an artificial-intelligence forum that Google organized in the Chinese city of Wuzhen (woo-jen).
Ke lost despite playing what AlphaGo indicated was the best game any opponent has played against it.
What happened? Ke said his loss was probably the result of something all too human: emotion.
“I … thought that I was very close to winning the match in the middle,” Ke said. “I could feel my heart thumping. But maybe because I was too excited, I did some wrong or stupid moves. I guess that’s the biggest weak point of human beings.”
He and AlphaGo play a final game Saturday in a country where go is extremely popular. Google says 60 million people in China watched online when AlphaGo played South Korea’s go champion in March 2016.
This time, Chinese censors blocked most of the country’s Web users from seeing the Google site carrying the feed. None of China’s dozens of video sites carried the live broadcasts but a recording of Tuesday’s game was available the next night on one popular site.
The government encourages Internet use for business and education but tries to block access to material considered subversive, or rebellious. Social media and video-sharing websites such as Facebook and YouTube are blocked, and Internet companies are required to have teams of censors to watch social media and remove banned material.