Former Go Champ Retires From Game: AI ‘Cannot Be Defeated’

AlphaGo Zero surpassed all previous versions of the AI in 40 days (via Google DeepMind)

Former Go champion Lee Se-dol has retired from professional competition—because artificial intelligence.

The South Korean Go master gained global fame in 2016 as the only human to defeat DeepMind’s AlphaGo computer program.

Now he’s bowing out of the sport due to the “invincibility” of AI Go programs.

“With the debut of AI in Go games, I’ve realized that I’m not at the top even if I become the No. 1 through frantic efforts,” Se-dol told South Korean news agency Yonhap. “Even if I become the No. 1, there is an entity that cannot be defeated.”

AlphaGo won four of its five matches against Se-dol in March 2016. But Lee’s comeback in the fourth game remains the only time a human has bested the AI player.

The 36-year-old, who boasts 18 victories in international competitions and 32 wins in domestic events, admitted he felt frustrated after losing the first three games to a machine.

“I rarely read comments on Internet news about me. But I got curious about how badly people were speaking of me after my three straight losses to AlphaGo,” he said in an interview with Yonhap. “Unexpectedly, few people criticized me.

“Frankly, I had sensed kind of a defeat even before the start of the matches against AlphaGo,” Se-dol continued. “People from Google’s DeepMind Technologies looked very confident from the beginning.”

In a statement published by The Verge, DeepMind CEO Demis Hassabis said Se-dol demonstrated “true warrior spirit” in his matchup with AlphaGo.

“On behalf of the whole AlphaGo team at DeepMind, I’d like to congratulate Lee Se-dol for his legendary decade at the top of the game, and wish him the very best for the future,” Hassabis said. “I know Lee will be remembered as one of the greatest Go players of his generation.”

Invented in China more than 2,500 years ago and enjoyed by more than 40 million people worldwide, Go (known as “baduk” in Korea) requires players to place back or white stones on a board and capture the opponent’s pieces or surround empty spaces to build territories.

But despite its apparent simplicity, the game features more possible moves than there are atoms in the universe, making it a tough competitor for traditional “brute force” artificial intelligence methods.

Google’s DeepMind AI division, however, took on that challenge with AlphaGo. The program uses a state-of-the-art tree search with two deep neural networks—one to suggest intelligent moves, the other to evaluate each option. AlphaGo chooses what it deems the most successful maneuver.

When pitted against top artificial intelligence Go programs, Google’s software won all but one of its 500 games. Even three-time European Go champion Fan Hui was no match for the machine, which triumphed five games to zero, making the first time a computer program beat a pro Go player.

Really, how can anyone compete?

Se-dol ended his 24-year career on Nov. 19.

(To be fair, his retirement decision was also influenced by a conflict with the Korea Baduk Association, which Se-dol is suing for the return of his membership fees, Yonhap reported.)

To commemorate his retirement, Lee plans to face off against one last AI Go rival next month: HanDol, developed by South Korea’s NHN Entertainment Corp., has already defeated the country’s five top players.

“Even with a two-stone advantage, I feel like I will lose the first game to HanDol,” he said. “I wanted to play comfortably against HanDol as I have already retired, though I will do my best.”

More on Geek.com:



from Geek.com https://ift.tt/2OWHGqW
via IFTTT

0 comments:

Post a Comment