AlphaGo founder: the next opponents want to Ke Jie is China-chess wars, AlphaGo, Li Shishi, go, Ke Jie-IT information

Google AI AlphaGo (Alpha-go) and Korea′s first man-machine war chess master Li Shishi began March 9, Alpha-go win .

39 year old "Alpha-go" founder demishi·hasabisi (Demis Hassabis) in Korea in Seoul told Xinhua in an interview, said that the "Alpha-go" game select China′s rule. As far as computers, China′s more simple rules. And he knows, China has many high level players, they also want the "Alpha-go" with a high level of chess-chess, for example, Ke Jie.

However, the man-machine war was the first time in the history of computer challenges professional Grandmaster, hassabis said they want to choose the top level has ten or more years of Li Shishi. Maybe more young players, can have this level, but they have not determined, because they also need to have top level in ten years to prove himself. "We also know that there are a lot of strong players, so after the game, maybe we will go to China, Japan, and master of chess and learn from there," says hassabis.

As an expert in artificial intelligence, hassabis interest led him to go to the "Alpha-go" research and development. He told reporters that as early as 20 years ago when I was in College, he worked at a high level chess club in Cambridge Learning go, and quickly absorbed. But because of busy work on computers, he didn′t have enough time to practice, Weiqi skills stop at amateur level. However, this did not prevent him like go.

When I was in College, hassabis teaches his partner play. Then comes the super computer "deep blue" beat world champion Garry Kasparov. Since then, hassabis want to go someday to write a program, and won the Championship.

Two years ago, hassabis finally waited until the right time. He created "deep thought" company has developed a deep learning programs. They want this technology to be more widely used, therefore, hassabis chose to go. He hoped that through the use of deep learning programs to create "Alpha-go" can beat human chess masters.

Hassabis said he did not "Alpha chess" game, because it is simply too strong. "Alpha-go" very strong learning ability, himself no match for it.

In the sabisi view, he does not think that artificial intelligence will make human life became dangerous. Instead, he felt that artificial intelligence is amazing. He talked about the European go Championship Fan Hui. "Alpha-go" after 5:0 win over a Fan Hui. At present, Fan Hui as "Alpha-go" project to provide advice. Fan Hui told him that with the "Alpha chess" chess in the process, their rankings in three or four months ′ time, from 600 up to No. 300 in the world. Hassabis feel, this may be the next "Alpha-go" can be put into the market a target, many people can go through this program to improve their level. Especially in the West, and not too many good chess player, so for a good intermediate player who lives in the West, how to become stronger is a very difficult thing.

Finally, hassabis said, "Alpha-go" would beat Li Shishi, charm of the game will not be reduced. Chess is popular, depending on who and how. If in the future the "Alpha-go" becomes more powerful, might also be able to promote human chess skill level rises to a new level.

AlphaGo创始人：下一战对手希望是中国柯洁 - 人机围棋大战,AlphaGo,李世石,围棋,柯洁 - IT资讯

谷歌公司人工智能AlphaGo（阿尔法围棋）与韩国围棋高手李世石的首场人机大战3月9日打响，阿尔法围棋获胜。

39岁的“阿尔法围棋”创始人德米什·哈萨比斯（Demis Hassabis）此前在韩国首尔接受了新华社记者的采访时说，此次“阿尔法围棋”的对弈选择了中国的规则。因为对电脑来说，中国的规则更为简便易行。而且他知道，中国也有许多高水平棋手，他们也希望“阿尔法围棋”能与高水平的棋手对弈，比如柯洁。

但是，此次人机大战是历史上第一次电脑挑战职业九段棋手，哈萨比斯表示，他们希望选择在顶尖水平已有十年甚至更久的李世石。也许更年轻的选手，也能有这个水平，但现在还不确定，因为他们还需要有十年顶尖水平去证明自己。“我们也知道，还有很多实力很强的选手，所以这次比赛之后，也许我们也会去中国、日本，与那里的高手切磋棋艺，”哈萨比斯说。

作为一位人工智能专家，哈萨比斯对围棋的浓厚兴趣促使他对“阿尔法围棋”进行研发。他告诉记者，早在20年前上大学时，他曾在剑桥的一个高水平围棋社团里学习围棋，并很快沉迷其中。但是因为忙于电脑方面的工作，他没有足够时间去练习，围棋技艺仅停留在业余一段水平。不过，这并不妨碍他喜欢围棋。

上大学时，哈萨比斯教会了他的合伙人下围棋。那时正值超级电脑“深蓝”战胜世界冠军卡斯帕罗夫。从那时起，哈萨比斯就在想有一天能为围棋写一个程序，并赢得冠军。

两年前，哈萨比斯终于等到合适的时机。他创建的“深度思维”公司开发了深度学习的程序。他们想让这项技术得到更广泛的应用，于是，哈萨比斯选择了围棋。他希望通过利用深度学习程序打造“阿尔法围棋”能战胜人类围棋大师。

哈萨比斯坦言，他没有与“阿尔法围棋”对弈过，因为它实在太强大了。“阿尔法围棋”的学习能力很强，自己完全不是它的对手。

在哈萨比斯看来，他并不认为人工智能会让人类生活变得危险。相反，他觉得人工智能很神奇。他谈到了欧洲围棋冠军樊麾。“阿尔法围棋”此前以5：0战胜了樊麾。目前，樊麾在为“阿尔法围棋”这个项目提供咨询。樊麾告诉他，在与“阿尔法围棋”对弈过程中，自己的排名在三四个月的时间里，从世界600名提升到了第300名。哈萨比斯因此感到，这也许是今后“阿尔法围棋”能投入市场应用的一个目标，许多人可以通过这个程序提高自己的围棋水平。特别是在西方，并没有太多优秀的围棋选手，所以对于一个生活在西方的优秀的中级选手来说，如何变得更强是件很困难的事。

哈萨比斯最后说，无论“阿尔法围棋”是否会战胜李世石，都不会削减围棋的魅力。围棋之所以受欢迎，取决于谁去下和怎么下。如果今后“阿尔法围棋”变得更加强大，也许还能够推动人类围棋的技艺水平上升到一个新的阶段。