Ke Jie, the human face of Go





Chinese Go player Ke Jie (3rd from right) and other guests attend the opening ceremony of

the Future of Go Summit before a match between him and Google’s artificial intelligence

program AlphaGo in Wuzhen, east China’s Zhejiang Province, on May 23, 2017.   Photo by Xu Yu


Chinese Go player Ke Jie (right) arrives before his match against Google’s artificial intelligence

program AlphaGo during the Future of Go Summit in Wuzhen, east China’s Zhejiang Province,

on May 23, 2017.   Photo by Xu Yu







Ke Jie,

the human face of Go



By Liu Xin and Wang Jingyu



World champion Go player, China’s Ke Jie began his face to face game against a machine on Tuesday morning of May 23 in the game’s ultimate challenge.

Ke, 19, faces off against AlphaGo computer program in Wuzhen, east China’s Zhejiang Province.

Ke said he was proud to represent humanity against artificial intelligence (AI) in three rounds of the game over the five-day tournament from May 23 to 27.

He tweeted his Weibo account late at night the day before the first round game, saying that “I can not feel the opponent’s love and passion for the Go.”

“It heats just due to the high-speed operating CPU,” the tweet said.

“AlphaGo is strong, I admit, but it has to beat me on the board to make me convinced.”

“Coming second means nothing to me,” Ke told China Central Television (CCTV) in March. “I’m in it to win.”

Nevertheless, Ke was realistic and wouldn’t predict the outcome: “I am not perfect and sometimes lose.”

At the very least, he said, it will be a game worth watching.

Since Republic of Korea Go master Lee Se-dol lost to AlphaGo in March 2016, pundits think a human victory over the machine is unlikely.

The computer program has a tally of 60 wins and no losses so far this year.

Yu Bin, head coach of Chinese national Go team, thinks the probability of Ke beating AlphaGo is “very low”.

“It is more a case of a human challenging the computer program, rather than being challenged by it,” Yu told Xinhua.

“But that doesn’t mean Ke has no chance at all. If he can lead a complex strategy on the Go board….”

Ke, who started playing Go at the age of six, has won four world titles. According to Goratings, he is now the world’s number one with 3,625 points, leading second placed Korean Pak Jeong-hwan by 45 points.

Luck is necessary to win, he admitted, “but ability is more important.”

Playing a computer program will differ from playing people, Ke said, “in terms of tricks and moves.”

“No professional player could resist playing it as its moves can defy human logic,” Ke said.

He wants to learn advanced tactics from AlphaGo, which should have its own logic and strategy: “Its calculating capacity will help it avoid many complicated changes on the board.”

Ke believes AlphaGo has yet to reach its peak. “It will grow stronger in future.”

Human players – swayed by emotive and other factors in their strategy and decisions – might never beat AlphaGo, said Ke.

Some older respectable Go masters used to claim knowing just seven percent of Go tactics.

“It’s not hard to be a Go beginner,” Ke said. “It’s really hard to be a top player as there are countless changes on the board.

“And it’s frustrating trying to be perfect.”

Go, known as Weiqi in China and Baduk in the Republic of Korea, was born in China thousands of years ago.

Ke believes a machine could never replace a person as an adversary in future: “It’s a cold machine without blood, while we have spiritual power.”





Chinese Go player Ke Jie (left of front row) competes during a match against Google’s

artificial intelligence program AlphaGo on the Future of Go Summit in Wuzhen, east

China’s Zhejiang Province, on May 23, 2017.   Photo by Xu Yu


Chinese Go player Ke Jie (left) competes during a match against Google’s artificial

intelligence program AlphaGo on the Future of Go Summit in Wuzhen, east China’s

Zhejiang Province, on May 23, 2017.   Photo by Xu Yu


Chinese Go player Ke Jie competes during a match against Google’s artificial intelligence

program AlphaGo on the Future of Go Summit in Wuzhen, east China’s Zhejiang Province,

on May 23, 2017.   Photos by Xu Yu





WUZHEN, Zhejiang  |   2017-05-23 18:39:16


AlphaGo edges Ke 

to lead Human vs.

Machine Go showdown


By Xu Xinyuan and Shen Nan



Artificial Intelligence program AlphaGo defeated world’s top-ranked Go-player Ke Jie in the first game of three here on Tuesday of May 23  in southern China’s water-town Wuzhen.

After four and half hours of play, Ke, playing black, lost by 0.5 points, which is the narrowest margin possible in the game. The game follows the Chinese Go rules with black having the advantage of first move, a set point of 7.5 was later added to white for such compensation.

“My greatest pleasure up to date is to play this match with AlphaGo,” said 19 year-old Ke. “I opened the match following Master’s style, aiming to see how AlphaGo responds.”

Master is an online version of AlphaGo. It had a 60 winning streak in rapid games in January against top Go-players worldwide, including Ke.

When asked about the match in general, Ke said he was shocked by a couple of moves during the mid-game as those moves wouldn’t be played by human.

“I have given all my best in this game, but still have room for further improvement,” said Ke, adding that he will always be confident in himself and will try to explore the game further with computer program in the following two matches.

With a newly upgraded version of AlphaGo bettered by reinforcement learning, DeepMind’s founder Demis Hassabis hopes Ke to help discover potential weaknesses of the program. He also discussed the value of collaboration between human and machine.

“We hope that in the future by collaborating with human scientists we will make a greater progress in finding new creative ideas in all sorts of things.”

The three-round games between Ke and AlphaGo is part of the five-day Future of Go Summit.









Chinese Go player Ke Jie competes during the second match against artificial

intelligence program AlphaGo in Wuzhen, east China’s Zhejiang province,

on May 25, 2017.   Photos by Xu Yu





WUZHEN  |   2017-05-25 20:02:43


Human mistake  

costs Ke’s chance

at overtaking AlphaGo


By Xu Xinyuan and Shen Nan



Chinese Go player Ke Jie, current world No. 1, admitted defeat during the mid-game of his second duel against computer program AlphaGo here on Thursday of May 25.

Having lost the first game of the best-of-three match here on Tuesday with the narrowest margin, Ke, 19, had a very close call to a chance of winning AlphaGo today.

“I was very nervous as I thought I had an actual chance at winning,” said Ke who tried to calm his pounding heart by pressing his chest during the match. “And, because of this intensity I made a few regretful moves. I guess that is exactly what we can call a human weakness.”

According to Demis Hassabis, the founder of DeepMind Technology, this match was the best eye opener in which he described the match to be from the future.

“As we were in the control room looking at what AlphaGo’s evaluation is saying; the evaluation: win rate, for roughly 50 moves, it thought that Ke Jie has been playing perfectly. And it agreed with all the moves, then, for the first roughly 100 moves, it is the closest game we have seen anyone play against the master version of AlphaGo.”

The AlphaGo team did not share information regarding the fluctuation of win rate during the match, while for Ke, the outcome was much better than he had anticipated.

Knowing this master version, Ke shared that he thought of little to a chance of winning before the match.

“Today’s game has been the best encouragement to me,” said Ke.

Hassabis twitted after the match, describing the game to be amazing and complex.

“Ke Jie pushed AlphaGo right to the limit,” he wrote.

The third and final round of Ke Jie versus AlphaGo will take place here on Saturday of May 27.






Chinese Go player Ke Jie (1st from left) analyses the game with his teammates after the

second match against artificial intelligence program AlphaGo in Wuzhen, east China’s

Zhejiang province, on May 25, 2017.   Photo by Xu Yu


Chinese Go player Ke Jie analyses the game after the second match against artificial

intelligence program AlphaGo in Wuzhen, east China’s Zhejiang province, on May 25, 2017.

Photo by Xu Yu





WUZHEN  |  2017-05-27 17:18:24


AlphaGo sweeps

world’s best Go-player Ke Jie 3-0


By Xu Xinyuan and Shen Nan



AlphaGo, DeepMind’s artificial intelligence Go-playing program, defeated world’s top-ranked player Ke Jie for the third consecutive game between them in Wuzhen on Saturday of May 27.

Ke, playing white, resigned mid-game after battling three and half hours to conclude the Human vs. Machine contest on the Chinese antient board game.

Demis Hassabis, founder of DeepMind Technology, said that it would be the last game for AlphaGo.

The 19-year-old Ke applied similar strategies from Game Two, opening the final match by creating chances of battling from the start of the game, ending with yet another action-packed performance.

Ke teared up nearing the end. He concluded the competition with a heart felt commentary repeating. “AlphaGo is too perfect.”

He also expressed that bitterness over defeat will be a driving force to his future journey in exploring the mysteries of Go.

In regards to consolation, Ke first apologized, then blamed himself. Believing that he could have done much better, he said, “I faced a cold, calm and terrifying opponent, to the best of my ability, I could only predict half of AlphaGo’s moves. I wish I could have done better.”

When asked to share about their past five-day experience with AlphaGo, all eight Chinese Go players that took part in the Go-summit said they’ve learned a great deal from AlphaGo and DeepMind.








BEIJING  |  2017-05-29 16:03:58


Artificial intelligence fuels

visions of how smart future can be


By Ma Qian, Yu Zhongwen adn Guo Shuang



It might arouse complicated and mixed feelings among human beings, when AlphaGo on Saturday of May 27 grabbed the laurels from the world’s top-ranked Go player Ke Jie with the artificial intelligence (AI) program’s three-winning streak in the ancient Chinese board game.

AlphaGo’s victory has been hailed as a landmark for the development of AI, as Go (Encirclement Chess), dating back to thousands of years ago, has been claimed as the most complex board game in the human world.

Now, the world-class computer program has been thrust into global limelight, as technologies behind AlphaGo push visions of how creative future can be. AI would not only be utilized for more general purposes in human life, but will also shake up the landscapes of a range of industries from e-commerce to healthcare.





Professionals and spectators described as “unorthodox” many of the moves made by AlphaGo in its first of the three games against Ke Jie on May 23.

“There was a ‘cut’ that stunned me, because human players would never use the move (under that circumstance),” said Ke Jie in a press conference after the contest. “But after analyzing the game, I think it’s a smart move that killed two birds with one stone.”

AI basically relies on algorithms and big data, according to Luo Jiebo, professor of computer science at the University of Rochester. AlphaGo uses a type of search algorithm to find its moves based on the knowledge it previously “learned” by a type of AI called deep learning, or neural networks that mimic human learning, through playing a large number of games with both humans and machines.

This time, Ke Jie confronted with a newly upgraded version of AlphaGo bettered by reinforcement learning, which enables AlphaGo to sort out the best solution on its own.

“AlphaGo has played numerous games. With vast quantities of data, it would quickly learn effective moves,” said Luo, “Besides, the machine doesn’t have the intuition to defend like a human does. All it wants is to get the maximum rewards… Its calculation is based on an assessment of the interests of the whole situation.”

Luo held that owing to deep learning, AI has improved remarkably in its capability to identify visuals and recognize voices after gobbling up a huge mass of data.

He added that so far in China, face detection and recognition has been applied successfully in practical terms, which relies largely on the combination of deep learning and big data.

“Currently many Chinese companies, big or small, are making face recognition programs. Because of the fierce competition, they have to constantly update their technologies and allow them to appear in our daily life, for example applications used for online banks and for company employees to check in,” the professor said.





AI has also been widely applied in various fields from e-commerce to finance.

At present, AI researchers are committed to efforts to allow more enterprises and ordinary users to participate in the field of AI by providing practical products and services, Li Feifei, director of the Stanford Artificial Intelligence Lab and the Stanford Vision, told Xinhua in a recent interview.

E-commerce giants, such as Amazon and Alibaba, have used a form of machine learning in their online recommender systems to boost sales, which is based on data collected from customers’ browsing and purchasing preferences.

Alibaba will roll out facial recognition payment software in the near future. According to a video demonstrating the payment system, shoppers can complete payment by scanning their faces and imputing the last four digits of their mobile phone numbers.

Meanwhile, it has been widely acknowledged that global tech giants at the front end of AI exploration have ramped up research and development in AI applications designed to make human life smarter in a variety of fields including health care, autonomous driving and smart robotics.

One of the most anticipated sectors is health care. Researchers have claimed that AI software would be likely to serve as an experienced assistant by helping doctors sift through loads of patient information and accumulate clinical data and experience for better disease diagnosis, such as cancer diagnosis and the recognition of magnetic resonance imaging.

“In the medical field, AI can learn from cases from around the world, at which it excels. However, a doctor can only gain experience from the cases he has seen, which has limits,” said Luo.

Elaborating on his blueprint, Li Yanhong, CEO of Chinese Internet giant Baidu, predicted that AI would be the “main course” for the future, calling the Internet “the appetizer,” in a speech during the 2017 Baidu Union Summit held in southwest China’s Chongqing City on May 23.

“In the age of AI, machines will be able to understand humans and their intentions, and thus allow humans to communicate all kinds of subjects,” Li explained during the speech.





However, it has sparked huge concerns that AI would shake up traditional industries by slashing hordes of jobs.

At present, some lenders and financial institutions in the United States have used AI underwriting and loan-distributing machines in replace of humans.

“The machines can decide whether and how much to lend, as well as how long the lending would be with a wealth of data,” said Luo Jiebo, “The decision made by machines are more accurate and without human bias.”

Kai-Fu Lee, computer scientist and founder of Sinovation Ventures, had echoed the status quo and made a bold prediction in a commencement speech to Columbia University’s engineering department on May 15.

“In the next 10 years, all financial companies will be turned upside-down, with AI replacing traders, bankers, accountants, research analysts, and insurance companies,” he said, “Last year, my AI investment algorithm returned 8 times more than my private banker.”

In Luo’s opinion, although unemployment caused by AI would be inevitable, it should not be viewed negatively because machines are not on the opposite side of humans. “There must be new jobs turning up. For example, machines need (people) to maintain,” he said.

He gave the example of the use of AI in healthcare. “AI will make doctors’ work easier, as machines process preliminary information. But it is humans who make critical and final decisions,” he said, “That’s because machines would only consider things that humans order them to.”

“I think the ultimate purpose of AI is not to replace humans, but rather to co-exist with humans and achieve common prosperity, which is the right direction (for the development of AI),” Luo stressed.








Leave a Reply