Machine Learning

AlphaGo’s ultimate challenge: a five-game match against the legendary Lee Sedol

AlphaGo_hero.jpg

Game 5 - March 15, 2016

After a loss in Game 4, and a move early on that looked like a mistake, but could have been a creative and effective new move, AlphaGo won Game 5 against the legendary Lee Sedol. This final game of the match was close until the very end, with commentators going back-and-forth about who was on top. But after 280 moves, Lee resigned, having gone through two byō-yomi overtime periods.

This concludes the 5 game match between AlphaGo and Lee Sedol, with AlphaGo winning the match 4 games to 1 against the best Go player of the last decade, achieving a major milestone for artificial intelligence a decade earlier than many predicted.

Michael Redmond, 9-dan, American commentator said:
  • “It was difficult to say at what point AlphaGo was ahead or behind, a close game throughout. AlphaGo made what looked like a mistake with move 48, similar to the mistake in Game Four in the middle of the board. After that AlphaGo played very well in the middle of the board, and the game developed into a long, very difficult end game...AlphaGo has the potential to be a huge study tool for us professionals, when it’s available for us to play at home.”
Kim Seongryong, 9-dan, Korean commentator said:
  • “Just like the scientists, Go players are always trying to find new methods and approaches. And we are so happy when we find them. This Challenge Match has brought us Go players to new areas we’ve never explored. We are now seeing a lot more interest in playing Go. And even in one week, I feel like my Go playing has improved.”
Chris Garlock, Managing Editor of the American Go E-Journal said:
  • “This match… the drama, the historic aspect, the quality of the games, the brilliance of AlphaGo, the brilliance of Lee Sedol, and then the amount of media coverage. I just want to say thanks to the entire DeepMind AlphaGo team. This is a gift to Go. This is going to do a lot to bring Go to new audiences. We could not have dreamed this up any better, and it delivered beautiful games. This match has done what Go always does: brings people together in friendship and cooperation, and that, like the game itself, is beautiful.”
With AlphaGo’s victory, Google DeepMind will donate the UDS$1 million in prize money to UNICEF, STEM (science, technology, engineering, and math) charities, and Go organizations.

Here are the results and final board position from Game 5:

If you'd like a quick recap of the game, here is a 90-second overview from commentators Michael Redmond and Chris Garlock:

Match 5 90 Second Summary - Google DeepMind Challenge Match 2016

Match 5 90 Second Summary - Google DeepMind Challenge Match 2016

Game 4 - March 13, 2016

Lee Sedol made a comeback today after three consecutive losses, to beat AlphaGo in the fourth game. Playing as white, Lee won by resignation after 180 moves.

AlphaGo held a strong position for the first half of the game, but commentators noted that Lee Sedol played a brilliant move 78, followed by a mistake by AlphaGo at move 79.

Michael Redmond, 9-dan, American commentator said:
  • “Today’s game was another example of AlphaGo playing a very interesting, good game. However, move 78 by Lee Sedol was really brilliant — and enabled him to win.“
Song Taegon, 9-dan, Korean commentator, said:
  • “It seems Lee Sedol can now read AlphaGo better and has a better understanding of how AlphaGo moves. For the 5th match, it will be a far closer battle than before since we know each better. Professional Go players said that they became more interested in playing Go after witnessing AlphaGo’s innovative moves. People started to rethink about moves that were previously regarded as undesirable or bad moves. AlphaGo can help us think outside of the box in Go games.“
Like in previous matches, Lee used up all of his time and two periods of byō-yomi overtime.

At the 3-1 match score, AlphaGo has already secured victory in the Google DeepMind Challenge Match, but today’s loss heightens the drama going into the final game, Game Five, which will be played on Tuesday, March 15 at 1pm KST.

Here are the results and final board position from Game 4:

If you'd like a quick recap of the game, here is a 90-second overview from commentators Michael Redmond and Chris Garlock:

Match 4 90 Second Summary - Google DeepMind Challenge Match 2016

Match 4 90 Second Summary - Google DeepMind Challenge Match 2016

For a bit more detail, here is a full 15-minute overview of the game, move by move:

Match 4 15 Minute Summary - Google DeepMind Challenge Match 2016

Match 4 15 Minute Summary - Google DeepMind Challenge Match 2016

Game 3 - March 12, 2016

AlphaGo has just won its third straight game, claiming overall match victory against the best Go player of the last decade, Lee Sedol. Playing as white, AlphaGo won by resignation after 176 moves. Lee used up all of his time and two periods of byō-yomi overtime, while AlphaGo had 8:31 left on the clock.

AlphaGo created a large territory on the board but Lee Sedol used a few innovative tactics to start a huge all-or-nothing kō fight and complicate the situation. In the resulting kō fight, AlphaGo prevailed.

Michael Redmond, 9-dan, American commentator said:
  • “It’s arguable that in the first two games Lee Sedol was playing differently than his true style, trying to find a weakness in the computer. Today Lee was definitely playing his own game, from his strong opening to the complicated moves in the final . AlphaGo was ready for everything, including the fights, and was able to take the win. I’d like to congratulate the people who actually made this accomplishment possible, because it’s a work of art.”
Lee Hyunwook, 8-dan, Korean commentator said:
  • “Lee Sedol played well. As a professional player myself, I’d like to show my respect to his three consecutive games against AlphaGo, which played almost perfectly. Lee made a few diverse moves at the end of today’s game to understand more about AlphaGo. I look forward to remaining games as well.”
All five games will be played to determine the final match score. The next game, Game Four, will be on Sunday, March 13, and the final game, Game Five, will be on Tuesday, March 15.

Here are the results and final board position from Game 3:

Here is Google co-founder Sergey Brin wishing Lee Sedol luck before the game in Korea:

If you'd like a quick recap of the game, here is a 90-second overview from commentators Michael Redmond and Chris Garlock:

Match 3 90 Second Summary - Google DeepMind Challenge Match 2016

Match 3 90 Second Summary - Google DeepMind Challenge Match 2016

For a bit more detail, here is a full 15-minute overview of the game, move by move:

Match 3 15 Minute Summary - Google DeepMind Challenge Match 2016

Match 3 15 Minute Summary - Google DeepMind Challenge Match 2016

Game 2 - March 10, 2016

AlphaGo again beat Lee Sedol, taking a 2-0 lead in this five-game match.

AlphaGo made a number of creative moves that surprised the expert commentators.

Michael Redmond, 9-dan, American commentator said:
  • “I was impressed with AlphaGo’s play. There was a great beauty to the opening. Based on what I had seen from its other games, AlphaGo was always strong in the end and middle game, but that was extended to the beginning game this time. It was a beautiful, innovative game.”

Yoo Changhyuk, 9-dan, Korean commentator said:
  • “During the first match, Lee Sedol made difficult moves to agitate AlphaGo, but failed to do so. Today, he tried the opposite — he played safe and entered the endgame. While using his byō-yomi periods, he made some mistakes, which I think caused the defeat.”
Both Lee Sedol and AlphaGo used their entire two-hours of time, going into byō-yomi overtime. AlphaGo just needs one more win to claim victory in the five-game tournament. Either way, all five games will be played to determine the final match score.

Here are the results and final board position from Game 2:

If you'd like a quick recap of the game, here is a 90-second overview from commentators Michael Redmond and Chris Garlock:

Match 2 90s Summary - Google DeepMind Challenge Match

Match 2 90s Summary - Google DeepMind Challenge Match

For a bit more detail, here is a full 15-minute overview of the game, move by move:

Match 2 15 Minute Summary - Google DeepMind Challenge Match 2016

Match 2 15 Minute Summary - Google DeepMind Challenge Match 2016

Game 1 - March 9, 2016

AlphaGo takes the first game against Lee Sedol. They were neck-and-neck for its entirety, in a game filled with complex fighting. Lee Sedol made very aggressive moves but AlphaGo did not back down from the fights. AlphaGo took almost all of its time compared to Lee Sedol who had almost 30 minutes left on the clock.

Here are the results and final board position from Game 1:

If you'd like a quick recap of the game, here is a 90-second overview from commentators Michael Redmond and Chris Garlock:

Match 1 90 Second Summary - Google DeepMind Challenge Match 2016

Match 1 90 Second Summary - Google DeepMind Challenge Match 2016

For a bit more detail, here is a full 15-minute overview of the game, move by move:

Match 1 15 min Summary - Google DeepMind Challenge Match

Match 1 15 min Summary - Google DeepMind Challenge Match

Pre-match update
Updated at 10am KST, March 8

The ultimate challenge has finally arrived! Starting tomorrow, AlphaGo, the first computer program to beat a professional Go player, will take on the legendary Lee Sedol—the top Go player of the past decade—in a five-game match in Seoul, Korea. Go is a profoundly complex game, which is why this has always been regarded as the outstanding grand challenge for artificial intelligence.

You can catch the livestream on DeepMind’s YouTube channel, or right here:

Match 1 - Google DeepMind Challenge Match: Lee Sedol vs AlphaGo

Match 1 - Google DeepMind Challenge Match: Lee Sedol vs AlphaGo

The games will start at 1pm Korea Standard Time (4am GMT; day before 11pm ET, 8pm PT) on these days:
  • Wednesday, March 9: First game
  • Thursday, March 10: Second game
  • Saturday, March 12: Third game
  • Sunday, March 13: Fourth game
  • Tuesday, March 15: Fifth game
The matches will be played under Chinese rules with a komi of 7.5 (the compensation points the player who goes second receives at the end of the game). Each player will receive two hours per game with three lots of 60-second byoyomi (countdown periods after they have finished their allotted time). Each game is expected to take 4-5 hours.

The games will be even (no handicap), with $1 million USD in prize money for the winner. If AlphaGo wins, the prize money will be donated to UNICEF, STEM charities and Go organizations.

As the match progresses, check back on this blog for the latest updates and results.