American Go E-Journal » Search Results » learn go week

AlphaGo-Lee Sedol Match: Game 5 News Coverage

Friday March 18, 2016

Chris Garlock will discuss the AlphaGo-Lee Sedol match Friday March 18 on the “World’s Finest Show” on WCHE 1520 2016.03.17_AlphaGo-Lee-Sedol-game-5-signed-Go-board-550x368AM, tune in worldwide via the listen live button at the top. Garlock commented the match with Michael Redmond 9P.

In Two Moves, AlphaGo and Lee Sedol Redefined the Future
Wired

AlphaGo seals 4-1 victory over Go grandmaster Lee Sedol
The Guardian

Game over! AlphaGo takes the final victory against Go champion Lee Sedol to finish the $1 million contest 4-1
The Daily Mail

AlphaGo defeats Lee Sedol 4–1 in Google DeepMind Challenge Match
GoGameGuru

What we learned in Seoul with AlphaGo
- Demis Hassabis, CEO and Co-Founder of DeepMind

 

 

Share
Categories: Computer Go/AI,Korea
Share

AlphaGo Watching Parties in SF & AZ

Wednesday March 16, 2016

The Arizona Go Club met at Old Chicago Restaurant for pizza and wings “to view the burial of human superiority in go, 2016.03.16_AZ Go Club Viewing party at Old Chicago otherwise known as Game 3 of the AlphaGo v. Lee Sedol match,” reports Martin Lebl. “Viewing was successful, although humanity lost, as many have predicted after game 1 and game 2 of the match.” Having watched the first two games at Denny’s, the viewing party for the deciding game was upgraded to Old Chicago “due to their better tasting food, and availability of appropriate liquid refreshments for a wake,” Lebl adds. “The final burial came at 1:30 local time, when AlphaGo decisively proved not only could it fight complicated ko fight, but would convert it into more complicated and bigger ko fight in the process, if given half a chance. Fun was had by all.”

2016.03.16_SF Go Club viewing party“Here’s a picture of us watching game two of the incredible Alphago vs. Lee Sedol match at Noisebridge hackerspace (left) in San Francisco,” reports Mishal Awadah. The SF Go club is 2016.03.16_Google Mt. Viewoffering a 10 week beginners go class starting on March 20th for anyone interested in learning the game.

And Lee Schumacher sent in this shot of a watch party at the Google Mt. View campus (right). 

Share
Categories: Computer Go/AI
Share

New 10-Week Beginner Class at SF Go Club

Tuesday March 15, 2016

The SF Go Club is starting a new 10-week beginner’s course this Sunday, March 20th, club president Mishal Awadah told the EJ. “This is a new approach to teaching and we hope to have a great class of beginners’ learn the basics of the game together.” More information about the course as well as a flyer for distribution can be found at http://sfgoclub.com/go-for-beginners/. Topics include the rules, capturing stones, eyes and living groups, shape, ladders, ko, seki and sente vs. gote. The lessons run from 2:30 to 3:30 p.m. each week.

Share
Categories: U.S./North America
Share

Chess Players Counsel Calm As Computers Close in on Go

Monday February 29, 2016

by Special Correspondent Andy Okun, with reporting by Natalya Kovaleva

In the run-up to AlphaGo’s challenge match with Lee Sedol 9p in Seoul in a little over a week, go players have been worrying about the new age whose beginning might be marked by an AlphaGo victory.  What will the go world be like when computers are so good?  Will people still want to play go?  What will change?  Taking advantage of the collegiality of the IMSA Elite Mind Games in Huaian, we sought counsel from a community that has been through this before.  We asked chess players how the game was affected by Garry Kasparov’s historic loss to IBM’s Deep Blue in 1997, the steady growth in strength of computer chess since, and how go players should greet the news.  The general view was that go players should not be afraid of the new age, but that things will be different.  There may even be some new and interesting problems to handle, as there have in chess.

“So many cheats!” said KwaiKeong Chan (right), a long time chess player, arbiter and organizer from Hong Kong. Chan is helping run the chess 2016.02.29_Kwai Keong Chansection of the IEMG as deputy chief arbiter.  The software is so strong that it has become very easy to find new ways to cheat, Chan said.  “Hiding in the toilet is primitive,” he said dismissively of a toilet-based chess scandal last year in Dubai, although he refused to detail some of the more subtle methods people use.  Strong computers also are how officials crack down on cheats, he said.  Chess software is so good that given a board position and an ELO rating, you can predict the exact set of moves a player of that strength will likely draw from.  If a player consistently picks better moves than are likely for his or her rating, officials know to pay close attention.  “You cannot play beyond yourself.   It’s not humanly possible,” said Chan, who himself had designed some very early chess-playing software.

Beyond that, chess players don’t really care about computers’ strength and said go players shouldn’t either, he said.  Rather, the advent of strong computer go will bring publicity to the game, as Deep Blue did for chess, Chan said.  “That is always a good thing, publicity, good or bad.  Publicity is what you need.”  Chess is being played more than ever before, and while Deep Blue is not the main reason for that – he cited years of community effort in presenting chess well – it did produce a second surge of new players after the Bobby Fischer surge of the 1970s.

2016.02.29_Alexandra KosteniukThe presence of such strong computers has had other effects on how chess is played and the nature of chess expertise, players suggested.  Since strong computers can provide weak and middling players with solid and accurate analysis, the role of the chess master is different than it was, said Russian player Alexandra Kosteniuk (left), a grandmaster, former Women’s World Chess Champion and author of “Diary of a Chess Queen.”  The strength of players has gone up, but the best players don’t command the same respect they might have in years past because the best critique is available to everyone.  “Maybe in a few years, there will be no go masters,” she said.

Shahriyar Mamedyarov, a 31-year-old Azerbaijani grandmaster and former rapid chess World Champion, said it used to be that when he was in world championship tournaments, he might have seven or eight fellow players with him helping him prepare for the games.  He doesn’t need to do that now, since any questions he has or analysis he needs done can be done by computer. Valentina Evgenyevna Gunina, a three-time Russian women’s champion, said computers had raised the standard of training and that “we need to memorize much more than we did before.”

Kirsan ILyumzhinov, the controversial president of both the Federation Internationale des Echecs and the Russian Republic of Kalmykia, as2016.02.29_ponomariov03 well as the head of IMSA and a long time sponsor of computer go competitions, said in the early days of the computer go project, human players and human programmers would work hard to develop the computer player and make it stronger.  “Now the computer develops and trains the human.”

Perhaps the bluntest argument against fear of computers learning to play our games well came from Ruslan Ponomariov (right), a Ukrainian grandmaster and FIDE World Champion from 2002 to 2004.

“What we can do?” he asked with a shrug.

photos credits: Kirill Merkurev (Chan); chessqueen.com (Kosteniuk); en.chessbase.com (Ponomariov)

Share

AlphaGo-Lee Sedol Match Set for March 9-15; More responses to AlphaGo win

Sunday February 7, 2016

As the go world — and indeed much mainstream media — has continued to buzz in the wake of the recent announcement of AlphaGo’s defeat2016.02.07_Fan-Hui-vs-AlphaGo of a professional go player, details of the matchup between AlphaGo and Lee Sedol have been released. The five-game match will take place in Seoul, March 9-15, with a $1 million prize — and the question of whether man or machine will prevail — at stake. We’ll keep you posted on broadcast coverage plans. Meanwhile, here’s a few of the reactions that have come in; we welcome your thoughts at our Facebook page, Twitter or at journal@usgo.org.

SmartGo’s Kierulf on AlphaGo: “Exciting times with the AlphaGo announcement!” writes SmartGo’s Anders Kierulf “If you’re in need of some more analysis and speculation on the Lee Sedol match, I’ve got you covered: Lee Sedol vs AlphaGo.” Kierulf has also written a bit about how AlphaGo works, and encouraging people to learn go now. He also reports that SmartGo has “definitely seen a spike in sales last week, subsiding again now.”

Cobb: A Flawed Test: “These sorts of tests of computer programs against pros (chess or go) all have the same flaw,” writes Slate & Shell’s Bill Cobb. “While the computer of course plays at the speed it needs to in order to use all of its resources, the pro is forced to play much faster than he/she can make use of their resources to a similar degree. For a go pro, one hour basic time is ‘lightning’ go, not a true test of the player’s ability—especially when it is followed by 30 second instead of one minute byoyomi periods. I don’t understand why people are so impressed about the computer program winning under such unfair conditions. Many strong amateurs could beat many pros under a similarly unbalanced time arrangement.” Cobb is the author of “Reflections on the Game of Go” a collection of his E-Journal columns, many of which focus on ways in which go can be related to Buddhist views of the search for enlightenment.

“Alphaville” Warned Us: The night before the announcement that a computer had won a 5-game match with no handicap against a 2016.02.07_alphavilleprofessional, I watched ‘Alphaville,’ a 1965 French film,” writes David Doshay. “In it an evil computer saps vocabulary, emotion and eventually life from the people of Alphaville. That computer’s name is Alpha-60. This program is called AlphaGo. Coincidence or conspiracy? Go and 60 look a lot alike to me …Should we warn the world?”

Learning from Chess: “Regarding Google’s AlphaGo achievement, I’d be interested in reading an E-Journal article discussing how chess software has affected online chess tournaments,” writes Syracuse go organizer Richard Moseson. “There have already been a few scandals at top chess tournaments in which players were found to be using chess playing software. How long will it be before players can use iGlasses to receive recommendations for each move?”

Moving the Goalposts: “Perhaps it is time to consider moving to the next prime number with a go board that is 23 by 23,” suggests Ronald Davis.
Update (7:08p): The source of the “Moving the Goalposts” quote has been updated.

Share
Categories: Computer Go/AI,World
Share

Game Over? AlphaGo Beats Pro 5-0 in Major AI Advance 

Wednesday January 27, 2016

[link]

In a stunning development, the AlphaGo computer program has swept European Go Champion and Chinese professional Fan Hui 2P 5-0, the first time that a go professional has lost such a match. “This signifies a major step forward in one of the great challenges in the development of artificial intelligence – that of game-playing,” said the British Go Association, which released the news on January 27, based on findings reported in the scientific journal Nature this week (click here for the video, here for Nature’s editorial, Digital intuition and here for Go players react to computer defeat). NOTE: This story was posted at 1p EST on Wednesday, January 27; be sure to get the latest breaking go news by following us on Facebook and Twitter.

“AlphaGo’s strength is truly impressive!” said Hajin Lee, Secretary General of the International Go Federation and a Korean go professional herself. “Go has always been thought of as the ultimate challenge to game-playing artificial intelligence,” added Thomas Hsiang, Secretary General of the International Mind Sport Association and Vice President of International Go Federation. “This is exciting news, but bittersweet at the same time,” said American Go Association president Andy Okun. “I think we go players have taken some pride in the fact that we could beat the best computers. Now we’re down to Lee Sedol fighting for us.”

2016.01.27_hui-fanGoogle DeepMind, the British artificial intelligence company which developed AlphaGo, has issued a challenge to Lee Sedol 9P from South Korea, the top player in the world for much of the last 10 years, to play a 5-game, million-dollar in March. “I have played through the five games between AlphaGo and Fan Hui,” said Hsiang. “AlphaGo was clearly the stronger player. The next challenge against Lee Sedol will be much harder.” While Hajin Lee agreed, saying “I still doubt that it’s strong enough to play the world’s top pros,” she added “but maybe it becomes stronger when it faces a stronger opponent.” Fan Hui (left) is a naturalized French 2-dan professional go player originally from China. European Champion in 2014 and 2015, Fan is also a 6-time winner in Paris as well as Amsterdam.

Just as the Kasparov/Deep Blue match did not signal the end of chess between humans, “so the development of AlphaGo does not signal the end of playing go between humans,” the BGA pointed out. “Computers have changed the way that players study and play chess (see this 2012 Wired article), and we expect something similar to occur in the field of go, but not necessarily as assistance during play. It has been recognised for a long time that achievements in game-playing have contributed to developments in other areas, with the game of go being the pinnacle of perfect knowledge games.”  Added Okun, “go has for thousands of years been a contest between humans and a struggle of humans against their own limits, and it will remain so. We still cycle in the Tour de France, even though we’ve invented the motorcycle.”

The BGA noted that that achievements in game-playing technology have contributed to developments in other areas. The previous major breakthrough in computer go, the introduction of Monte-Carlo tree search, led to corresponding advances in many other areas.

Last year, the Facebook AI Research team also started creating an AI that can learn to play go and earlier today Mark Zuckerberg reported on Facebook that “We’re getting close, and in the past six months we’ve built an AI that can make moves in as fast as 0.1 seconds and still be as good as previous systems that took years to build. Our AI combines a search-based approach that models every possible move as the game progresses along with a pattern matching system built by our computer vision team.”

In a related story, computer scientist John Tromp last week revealed the number of legal go positions, “weighing in at 9*19=171 digits.” Read more here.

Game 1 of the AlphaGo vs. Fan Hui 2P match appears above right. Click below for the match’s remaining game records:
AlphaGo vs. Fan Hui, game 2
AlphaGo vs. Fan Hui, game 3
AlphaGo vs. Fan Hui, game 4
AlphaGo vs. Fan Hui, game 5

Update (11:44pm 1/27): Myungwan Kim 9P will analyze the games played between Fan Hui and AlphaGo during a live stream on the AGA YouTube Channel and TwitchTV this Friday; more details will be posted at 7a EST.

 

Share

2016 Online Go Sessions Start Up

Saturday December 26, 2015

With the New Year fast approaching, online go classes are starting new sessions:

Guo Juan’s Internet Go School’s online group class starts on January 9th. “Meet friends, have fun and learn much from pro teachers,” says Go 2015.12.22_guo-juan-logoJuan 5P. Pro teachers include Guo, YoungSun Yoon 8P, Jennie Shen 2P and Mingjiu Jiang 7P. Cost is 135 euros for 8 x 1,5h classes and seven weeks full access to the school’s pro lecture site and the training system.

2015.12.22_American Yunguseng dojang Lecture in US Go congress 2015Inseong Hwang’s new season — the 14th — of his online go academy ‘Yunguseng Dojang’ starts on January 4. The American Yunguseng dojang has been going to two years. It started with three leagues and 20 people and has now increased to seven leagues and 50 participants, with members from AGA 7dan to 12 kyu. “I attended this year’s US Go Congress,” says Hwang. Check out the Yunguseng Doajng Youtube channel.

 

Share

Google and Facebook Race to Solve Go

Sunday December 13, 2015

“Inside companies like Google and Facebook, deep learning is proving remarkably adept at recognizing images and grasping spacial patterns—a skill well suited to Go,” reported Cade Metz in Wired last week . “As they explore so many other opportunities this technology presents, Google and Facebook are also racing to see whether it can finally crack the ancient game. As Facebook AI researcher Yuandong Tian explains, Go is a classic AI problem—a problem that’s immensely attractive because it’s immensely difficult. The company believes that solving Go will not only help refine the AI that drives its popular social network, but also prove the value of artificial intelligence. Rob Fergus, another Facebook researcher, agrees. “The goal is advancing AI,” he says. But he also acknowledges that the company is driven, at least in a small way, by a friendly rivalry with Google. There’s pride to be found in solving the game of Go.” For more on Facebook’s research, check out “How Facebook’s AI Researchers Built a Game-Changing Go Engine” in the MIT Technology Review last week.

Smart Go’s Anders Kierulf also recently published a blog post on “Go at Facebook”, saying that “As long as Facebook and Google stick with trying to find general solutions to general problems, I don’t think top Go programs like Zen and Crazy Stone have anything to worry about. But once these giants decide to beat the strongest human players and are willing to focus on Go-specific solutions, it will get interesting.” Check out also his explainer on Monte Carlo Tree Search and his post about the impact of the Swift programming language on his go programs.

 

Share
Categories: Computer Go/AI
Share

Your Move/Readers Write: Facebook’s Go Study Not Deep Enough; More Levels in Go Than Poker?; Excellent Live Streaming

Sunday November 8, 2015

Facebook’s Go Study Not Deep Enough: “Only good moves? Did anyone tell Facebook that they will lose their first 1000 games?” wonders Chris Uzal (Facebook Tackles Go With “Deep Learning” AI 11/7 EJ) “This is what happens when you think the world can function great with only a “like” button but the sky is full of frowny faces if you have a “dislike” button. I don’t think Facebook is psychologically equipped to play go2015.11.08_WSJ-poker let alone research the game. In their world, missclicks and undos are part of the game. The only people who lose are not friends with their friends. They would possibly learn more by studying moral hazard.”

More Levels in Go Than Poker? “Christopher F. Chabris, whom I know from chess, has written an article for the WSJ (Could an Amateur Win the World Series of Poker?) in which he writes ‘In the Asian game of Go, there may be even more levels.’ I thought he should have written ‘…there are even more levels,’ in lieu of ‘may be.’ What say you?”
- Michael Bacon; photo courtesy John Locher/Associated Press

Excellent Live Streaming: “Thank you so much for this live streaming,” Fabio G. Moreno from Bogotá, Colombia, posted on Facebook after last week’s Ke Jie-Lee Sedol streaming on the AGA’s YouTube channel. “Andrew and Myungwan did excellent work. I think is great contribution for the spread of go in the world to make this excellent live streaming in English and open access (to) this level of tournament, like the Samsung Cup, and a game in the semi-finals. Was great.”

We love to hear from our readers! Let us know what you like or dislike, love or hate; email us at journal@usgo.org or post on our Facebook page!

Share
Categories: U.S./North America
Share

Facebook Tackles Go With “Deep Learning” AI

Saturday November 7, 2015

Researchers at Facebook are now tackling go with an increasingly important form of artificial intelligence known as deep learning, Wired 2015.11.07_Wired_logomagazine reported earlier this week.

2015.11.07_Facebook_logo“Facebook is using similar technology to recognize a promising Go move—to visually understand whether it will be successful, kind of like a human would,” writes Cade Metz. “Researchers are feeding images of Go moves into a deep learning neural network so that it can learn what a successful move looks like,” as opposed to using brute computing power to analyze the many possible outcomes of every possible move.

Though this system is only about two or three months old, Facebook CTO Mike “Schrep” Schroepfer told reporters at Facebook’s California headquarters last week, it can already beat systems built solely with more traditional AI techniques. The company’s go work—which Schrep described as “super early”—demonstrates why deep learning is so powerful and how it can continue to push the boundaries of what machines can do, Metz reported.

Share
Categories: U.S./North America
Share