Global Sources
EE Times-Asia
Stay in touch with EE Times Asia
EE Times-Asia > Embedded

Not a fluke: Google's AI wins final Go challenge

Posted: 16 Mar 2016 ?? ?Print Version ?Bookmark and Share

Keywords:Google? AlphaGo? Go? artificial intelligence? South Korea?

The game of Go has a new champion. And it's a machine.

Google's Go-playing computer programme, AlphaGo, emerged victorious on Tuesday with four wins and one loss against South Korean grandmaster Lee Se-dol, Wired reported. Lee, considered one of the world's top Go players, missed out on the $1 million prize after only managing to snag a single win in the best-of-five series.

AlphaGo, developed by British computer company DeepMind, was acquired by Google in 2014. Since then, the programme started building "its expertise by studying older games and teasing out patterns of play," according to a BBC report.

DeepMind co-found Demis Hassabis described it as "deep reinforcement learning." In an interview with UK's The Guardian, Hassabis explained: "It's the combination of deep learning, neural network stuff, with reinforcement learning: so learning by trial and error, and incrementally improving and learning from your mistakes and your error."

AlphaGo joins other AI programmes that have bested humans at checkers, chess, and even poker. Three years ago, a programme called Ponanza defeated Japanese chess expert Shinichi Sato in a shogi match.

Like shogi, go is a game too complex to play, and even if Lee managed to turn the tables and win against the machine, let's face it, a lot of us!the so-called non-fans!will still be unmoved with the news. EE Times' Junko Yoshida had the perfect words. She didn't care "who" won. Why? Because it was only a matter of time.

In her blog post, Yoshida wrote: "There are no surprises. That's why the whole world isn't exactly going gaga over the Google's latest DeepMind's Go challenge.

"Human vs. robot is an on-going theme serving as a plotline for movies and novels. It's always entertaining, and more often than not, the stories are pretty credible!at least the parts where humans get antsy about what robots might do to them next. "

And now, with "smart" things cropping up everywhere, from cars to watches, to glasses and even toilets, Yoshida is terrified of a future where only the people are not getting smarter.

"We're willing to let smart things do what we used to do ourselves. No problem, we tell ourselves, because we have better things to do," she wrote. "But sometimes I wonder how smart we are to let this happen. Once you stop doing certain things yourself, you forget!or never learn!the fundamental knowledge of how those things work. And you lose the skills required to do it yourself."

In closing, Yoshida said: "I'm afraid of a future when I no longer know how most of the stuff around me works. More than anything else, I want AI to make usChumansCsmarter. Rather than building computers that can do things so much better than I can, why not design a computer that makes me smarter?"

Article Comments - Not a fluke: Google's AI wins final ...
*? You can enter [0] more charecters.
*Verify code:


Visit Asia Webinars to learn about the latest in technology and get practical design tips.

Back to Top