알파고 대 이세돌

알파고 대 이세돌
4–1

대한민국, 서울, 2016년 3월 9일 ~ 15일
첫 번째 게임	알파고 W+R
두 번째 게임	알파고 B+R
세 번째 게임	알파고 W+R
네 번째 게임	이세돌 W+R
5차전	알파고 W+R

알파고 대 이세돌(AlphaGo vs Lee Sedol)은 2016년 3월 9일부터 15일까지 대한민국 서울에서 열린 바둑 대전으로, 구글 딥마인드에 의해 개발된 컴퓨터 바둑 프로그램인 알파고 대 이세돌의 5경기 대결이다.알파고는 네 번째 ^[1]게임을 제외한 모든 게임을 이겼다; 모든 게임은 ^[2]체념으로 이겼다.이 경기는 1997년 딥 블루와 게리 카스파로프 사이의 역사적인 체스 경기와 비교되어 왔다.

그 경기의 우승자는 100만 달러를 획득할 예정이었다.알파고가 우승했기 때문에 구글 딥마인드는 그 상이 유니세프를 포함한 자선 단체와 바둑 ^[3]단체에 기부될 것이라고 말했다.이대호는 17만 달러를 받았다.^[4]

경기 후 한국기원은 알파고에게 최고 등급인 "명예로운 9단"을 수여했다.그것은 ^[5]바둑을 마스터하기 위한 알파고의 "진정한 노력"을 인정하여 주어졌다.이 경기는 ^[6]사이언스에 의해 2016년 12월 22일 올해의 돌파구 중 하나로 선정되었다.

배경

인공지능의 어려운 도전

외부 비디오
세계에서 가장 어려운 게임에서 인간을 이기기 위해 스스로 훈련하는 기계, Retro Report, 2:51, Retro[7] Report

바둑은 직관, 창의적이고 전략적인 ^[8][9]사고를 필요로 하는 복잡한 보드게임이다.인공지능(AI) 분야에서 오랫동안 어려운 도전으로 여겨져 왔고 체스보다 훨씬 풀기 어렵다^[10].인공지능 분야의 많은 사람들은 바둑이 ^[11]체스보다 인간의 생각을 모방하는 요소들을 더 많이 필요로 한다고 생각한다.수학자 I. J. 굿은 1965년에 ^[12]다음과 같이 썼다.

컴퓨터로 할까요?- 컴퓨터가 단순한 법정 게임이 아닌 합리적인 바둑을 두도록 프로그래밍하기 위해서는 좋은 전략의 원칙을 공식화하거나 학습 프로그램을 설계해야 합니다.원칙은 체스보다 더 질적이고 신비롭고 판단력에 더 의존한다.그래서 나는 체스보다 합리적인 바둑을 두기 위해 컴퓨터를 프로그래밍하는 것이 더 어려울 것이라고 생각한다.

2015년 ^[13]이전에는 최고의 바둑 프로그램들이 아마추어 단 ^[14]레벨에 도달하는 데 성공했습니다.작은 9×9 보드에서는 컴퓨터의 성능이 향상되었고, 일부 프로그램은 프로 선수와의 9×9 게임 중 극히 일부만 간신히 이겼다.알파고 이전에, 일부 연구원들은 컴퓨터가 ^[15]바둑에서 최고의 인간을 절대 이길 수 없다고 주장했었다.딥마인드의 초기 투자자인 일론 머스크는 2016년 이 분야 전문가들은 AI가 최고 프로 ^[16]바둑기사를 상대로 승리를 거두려면 10년이 더 걸릴 것으로 생각한다고 말했다.

알파고 대 이세돌의 대결은 1997년 게리 카스파로프가 IBM 컴퓨터 딥 블루에 졌던 체스 경기와 비교해도 손색이 없다.카스파로프가 딥 블루에게 진 것은 컴퓨터가 인간보다 ^[17]체스를 더 잘하게 된 순간으로 여겨진다.

알파고는 기존 AI 노력과는 확연히 다르다.알파고는 인간 프로그래머에 의해 하드코딩된 확률 알고리즘을 사용하는 대신 신경망을 사용하여 승리 확률을 추정한다.AlphaGo는 모든 게임, 플레이어, 분석 및 문헌뿐만 아니라 AlphaGo가 자신과 다른 플레이어와 플레이하는 게임까지 포함한 바둑의 전체 온라인 라이브러리에 액세스하여 분석합니다.설정이 완료되면 AlphaGo는 개발자 팀으로부터 독립하여 바둑을 해결하는 최선의 경로(즉, 게임에서 승리)를 평가합니다.신경 네트워크와 몬테 카를로 트리 검색을 사용하여 알파고는 많은 사람들이 미래로^{[citation needed]} 이동할 가능성이 있는 확률과 가능성이 없는 확률을 엄청나게 계산한다.

인지과학, 패턴인식, 기계학습 ^{[18]: 150} 등의 분야에 관련 연구결과가 적용되고 있다.

판후이와의 대전

알파고는 2015년 10월 유럽 챔피언인 2단 프로 선수인 판 후이를 5-0으로 물리쳤는데, 이는 AI가 처음으로 풀사이즈 ^[19][20]보드에서 인간 프로 선수를 핸디캡 없이 꺾은 것이다.일부 해설자들은 팬과 프로 ^[21]9단 이대호의 괴리를 강조했다.컴퓨터 프로그램인 젠과 크레이지 스톤은 이전에 4-5석의 ^[22][23]핸디캡으로 9단 프로 선수를 물리쳤다.캐나다 AI 전문가 조나단 셰퍼(Jonathan Schaeffer)는 팬과의 경기 후 "알파고를 경험이 부족한 '어린이 신동'과 비교하며 "진정한 최고 수준의 플레이어를 플레이하는 프로그램이 진정한 성취가 될 것"이라고 말했다.그는 2016년 ^[20]3월 이대호가 승리할 것이라고 믿었다.프로 바둑 기사이자 국제 바둑 연맹 사무국장인 이하진 씨는 AI가 이승엽에게 도전할 것이라는 전망에 "매우 흥분된다"며 두 선수가 동등한 ^[20]승산이 있다고 생각했다고 말했다.

알파고와의 대결 후, 판후이는 이 게임이 그에게 더 나은 선수가 되고 이전에 보지 못했던 것들을 보게 해주었다고 말했다.2016년 3월 와이어드는 세계 랭킹이 633위에서 약 ^[24]300위로 올라섰다고 발표했다.

준비

바둑 전문가들은 특히 보드 전체에 대한 인식 부족과 관련하여 알파고의 팬과의 대결에서 오류를 발견했다.이대호와의 경기 전에는 10월 ^[21][25]경기 이후 경기력이 얼마나 향상됐는지 알 수 없었다.AlphaGo의 원래 훈련 데이터 세트는 인터넷 Go 서버에서 온 강력한 아마추어 선수들의 게임으로 시작되었고, 이후 AlphaGo는 수천만 개의 ^[26][27]게임과 경쟁하며 훈련했습니다.

플레이어

알파고

알파고는 구글 딥마인드가 바둑을 두기 위해 개발한 컴퓨터 프로그램이다.AlphaGo의 알고리즘은 기계 학습과 트리 검색 기술의 조합을 사용하여 인간과 컴퓨터 게임 모두에서 광범위한 훈련을 받습니다.이 시스템의 신경 네트워크는 처음에는 인간의 게임 플레이 전문 지식으로부터 부트스트랩되었습니다.알파고는 처음에 KGS 6에 의한 16만 경기부터 9단 ^[13][28]경기까지 약 3천만 번의 KGS Go 서버 데이터베이스를 사용하여 기록된 역사 게임에서 나온 전문가 선수들의 움직임을 일치시키는 것을 시도함으로써 인간의 놀이를 모방하도록 훈련되었다.어느 정도 숙달된 후에는 강화 학습을 통해 ^[29]플레이를 향상시키기 위해 다수의 게임을 자신의 다른 인스턴스와의 대결로 설정함으로써 더 많은 훈련을 받았다.시스템은 재생할 이동의 "데이터베이스"를 사용하지 않습니다.AlphaGo를 만든 사람 중 한 명이 ^[30]설명했듯이:

이 기계를 작동하도록 프로그래밍했지만 어떤 동작이 나올지 알 수 없습니다.그 움직임은 훈련에서 나타난 현상이다.데이터 세트와 트레이닝 알고리즘을 작성하기만 하면 됩니다.그러나 그 후 그것이 만들어 내는 움직임은 우리의 통제 밖이며, 바둑 기사로서 우리가 생각해 낼 수 있는 것보다 훨씬 더 훌륭합니다.

이대호와의 대결에서 알파고는 1202개의 CPU와 176개의 ^[13]GPU를 사용한 ^[31]판후이와의 대결에서와 거의 같은 처리 능력을 사용했다.이코노미스트는 1920개의 CPU와 280개의 ^[32]GPU를 사용했다고 보도했다.구글은 이세돌과의 ^[33]경기에서도 자체 텐서 처리 장치가 사용됐다고 밝혔다.

이세돌

이세돌은 바둑 역사상 가장^[34] 강력한 9단 기사 중 한 명이다.그는 1996년에 선수 생활을 시작했고,^[35] 이후 18개의 국제 타이틀을 거머쥐었다.그는 파격적이고 창의적인 ^[36]플레이로 알려진 고국 한국에서 "국민 영웅"이다.이세돌은 처음에 알파고를 "랜드슬라이드"^[36]에서 물리칠 것이라고 예측했다.경기 몇 주 전에 그는 메이저 ^[37]대회인 한국 명인 타이틀을 거머쥐었다.

게임.

이 경기는 100만 달러를 ^[3]대상으로 7.5점짜리 ^[4]중국 규칙을 사용한 5경기 승부였다.각 경기마다 선수별로 2시간의 시간 제한이 있었고, 그 후 60초의 ^[4]연장전이 3회 있었다.각 경기는 표준시 13:00(GMT 04:00)^[38]에 시작되었습니다.

이 경기는 2016년 3월 서울 포시즌스 호텔에서 열렸으며 해설과 함께 비디오 스트리밍으로 생중계되었다. 영어 해설은 마이클 레드몬드(9단 프로페셔널)와 크리스 ^[39][40][41]갈록이 맡았다.DeepMind 팀원이자 아마추어 6단 바둑 기사인 Aja Huang은 AlphaGo용 바둑판에 돌을 놓았는데,^[42] AlphaGo는 Google Cloud Platform을 통해 서버를 미국에 두고 실행했다.

요약

게임	날짜.	블랙입니다.	하얀색	결과	움직인다
1	2016년 3월 9일	이세돌	알파고	이세돌 사임	186 1차전
2	2016년 3월 10일	알파고	이세돌	이세돌 사임	211 게임2
3	2016년 3월 12일	이세돌	알파고	이세돌 사임	176 3차전
4	2016년 3월 13일	알파고	이세돌	AlphaGo 사임	180 4차전
5	2016년 3월 15일	이세돌[note 1]	알파고	이세돌 사임	280 5차전
결과: 알파고 4 – 1 이세돌
^ 주 1: 5차전의 경우 공식 규칙에 따라 색상 [43]배정이 무작위로 이루어지도록 의도되었다.하지만, 4차전 후 기자회견에서, 이승엽은 "나는 백인으로 이겼기 때문에, 5차전에서는 [44]흑인으로 이길 수 있기를 진심으로 바란다"고 요청했다.하사비스는 세돌에게 검은색을 가지고 노는 것을 허락했다.

첫 번째 게임

알파고(흰색)가 첫 경기를 이겼다.이대호는 경기 내내 주도권을 쥐고 있는 것처럼 보였지만, 알파고가 마지막 20분 만에 우위를 점했고 이대호는 ^[45]사임했다.이승엽은 경기 초반 결정적인 실수를 했다며 초반 컴퓨터 전략이 훌륭했다며 AI가 인간 바둑 선수라면 ^[45]하지 않았을 특이한 행동을 했다고 말했다.데이비드 오르메로드는 바둑 구루(Guru)에서 이대호의 7번째 돌에 대해 "오프닝에서 알파고의 힘을 시험하기 위한 묘한 움직임"이라고 표현하며, 이대호의 움직임을 실수라고, 알파고의 대응은 "정확하고 효과적"이라고 특징지었다.그는 이대호가 119와 123에서 "의문의" 움직임을 보이기 전에 81로 복귀하기 시작했고 ^[46]129에서 "패" 움직임을 보인 것을 고려할 때 알파고의 위치가 경기 초반 유리하다고 설명했다.프로 바둑 기사 조한승은 알파고의 경기가 2015년 ^[46]10월 판후이를 꺾었을 때보다 크게 향상되었다고 말했다.마이클 레드몬드는 컴퓨터의 게임이 ^[47]팬에 비해 더 공격적이라고 묘사했다.

9단 김성룡 대장에 따르면, 이대호는 102번째 ^[48]돌에서 알파고의 강한 플레이에 망연자실한 듯 보였다.알파고의 102번째 움직임을 지켜본 후, 이대호는 10분 ^[48]이상 자신의 선택에 대해 숙고했다.

첫 번째 99개 이동

Moves 100–186

Game 2

AlphaGo (black) won the second game. Lee stated afterwards that "AlphaGo played a nearly perfect game",^[49] "from very beginning of the game I did not feel like there was a point that I was leading".^[50] One of the creators of AlphaGo, Demis Hassabis, said that the system was confident of victory from the midway point of the game, even though the professional commentators could not tell which player was ahead.^[50]

Michael Redmond (9p) noted that AlphaGo's 19th stone (move 37) was "creative" and "unique".^[30] Lee took an unusually long time to respond to the move.^[30] An Younggil (8p) called AlphaGo's move 37 "a rare and intriguing shoulder hit" but said Lee's counter was "exquisite". He stated that control passed between the players several times before the endgame, and especially praised AlphaGo's moves 151, 157, and 159, calling them "brilliant".^[51]

AlphaGo showed anomalies and moves from a broader perspective which professional Go players described as looking like mistakes at the first sight but an intentional strategy in hindsight.^[52] As one of the creators of the system explained, AlphaGo does not attempt to maximize its points or its margin of victory, but tries to maximize its probability of winning.^[30][53] If AlphaGo must choose between a scenario where it will win by 20 points with 80 percent probability and another where it will win by 1 and a half points with 99 percent probability, it will choose the latter, even if it must give up points to achieve it.^[30] In particular, move 167 by AlphaGo seemed to give Lee a fighting chance and was declared to look like an obvious mistake by commentators. An Younggil stated "So when AlphaGo plays a slack looking move, we may regard it as a mistake, but perhaps it should more accurately be viewed as a declaration of victory?"^[54]

First 99 moves

Moves 100–199

Moves 200–211

Game 3

AlphaGo (white) won the third game.^[55]

After the second game, there had still been strong doubts among players whether AlphaGo was truly a strong player in the sense that a human might be. The third game was described as removing that doubt; with analysts commenting that:

AlphaGo won so convincingly as to remove all doubt about its strength from the minds of experienced players. In fact, it played so well that it was almost scary ... In forcing AlphaGo to withstand a very severe, one-sided attack, Lee revealed its hitherto undetected power ... Lee wasn’t gaining enough profit from his attack ... One of the greatest virtuosos of the middle game had just been upstaged in black and white clarity.^[54]

According to An Younggil (8p) and David Ormerod, the game showed that "AlphaGo is simply stronger than any known human Go player."^[54] AlphaGo was seen to capably navigate tricky situations known as ko that did not come up in the previous two matches.^[56] An and Ormerod consider move 148 to be particularly notable: in the middle of a complex ko fight, AlphaGo displayed sufficient "confidence" that it was winning the fight to play a large move elsewhere.^[54]

Lee, playing black, opened with a High Chinese formation and generated a large area of black influence, which AlphaGo invaded at move 12. This required the program to defend a weak group, which it did successfully.^[54] An Younggil described Lee's move 31 as possibly the "losing move"^[54] and Andy Jackson of the American Go Association considered that the outcome had already been decided by move 35.^[53] AlphaGo had gained control of the game by move 48, and forced Lee onto the defensive. Lee counterattacked at moves 77/79, but AlphaGo's response was effective and its move 90 succeeded in simplifying the position. It then gained a large area of control at the bottom of the board, strengthening its position with moves from 102 to 112 described by An as "sophisticated".^[54] Lee attacked again at moves 115 and 125, but AlphaGo's responses were again effective. Lee eventually attempted a complex ko from move 131, without forcing an error from the program, and he resigned at move 176.^[54]

First 99 moves

Moves 100–176 (122 at 113, 154 at , 163 at 145, 164 at 151, 166 and 171 at 160, 169 at 145, 175 at )

Game 4

Lee (white) won the fourth game. Lee chose to play a type of extreme strategy, known as amashi, in response to AlphaGo's apparent preference for Souba Go (attempting to win by many small gains when the opportunity arises), taking territory at the perimeter rather than the center.^[57] By doing so, his apparent aim was to force an "all or nothing" style of situation – a possible weakness for an opponent strong at negotiation types of play, and one which might make AlphaGo's capability of deciding slim advantages largely irrelevant.^[57]

The first 11 moves were identical to the second game, where Lee also played white. In the early game, Lee concentrated on taking territory in the edges and corners of the board, allowing AlphaGo to gain influence in the top and centre. Lee then invaded AlphaGo's region of influence at the top with moves 40 to 48, following the amashi strategy. AlphaGo responded with a shoulder hit at move 47, subsequently sacrificing four stones elsewhere, and gaining the initiative with moves 47 to 53 and 69. Lee tested AlphaGo with moves 72 to 76 without provoking an error, and by this point in the game commentators had begun to feel Lee's play was a lost cause. However, an unexpected play at white 78, described as "a brilliant tesuji", turned the game around.^[57] The move developed a white wedge at the centre, and increased the game's complexity.^[58] Gu Li (9p) described it as a "divine move" and stated that the move had been completely unforeseen by him.^[57]

AlphaGo responded poorly on move 79, at which time it estimated it had a 70% chance to win the game. Lee followed up with a strong move at white 82.^[57] AlphaGo's initial response in moves 83 to 85 was appropriate, but at move 87, its estimate of its chances to win suddenly plummeted,^{[59][non-primary source needed][60][non-primary source needed]} provoking it to make a series of very bad moves from black 87 to 101. David Ormerod characterised moves 87 to 101 as typical of Monte Carlo-based program mistakes.^[57] Lee took the lead by white 92, and An Younggil described black 105 as the final losing move. Despite good tactics during moves 131 to 141, AlphaGo proved unable to recover during the endgame and resigned.^[57] AlphaGo's resignation was triggered when it evaluated its chance of winning to be less than 20%; this is intended to match the decision of professionals who resign rather than play to the end when their position is felt to be irrecoverable.^[58]

An Younggil at Go Game Guru concluded that the game was "a masterpiece for Lee Sedol and will almost certainly become a famous game in the history of Go".^[57] Lee commented after the match that he considered AlphaGo was strongest when playing white (second).^[61] For this reason, he requested that he play black in the fifth game, which is considered more risky.

David Ormerod of Go Game Guru stated that although an analysis of AlphaGo's play around 79–87 was not yet available, he believed it was a result of a known weakness in play algorithms which use Monte Carlo tree search. In essence, the search attempts to prune sequences which are less relevant. In some cases, a play can lead to a very specific line of play which is significant, but which is overlooked when the tree is pruned, and this outcome is therefore "off the search radar".^[62]

First 99 moves

Moves 100–180 (177 at , 178 at )

Game 5

AlphaGo (white) won the fifth game.^[63] The game was described as being close. Hassabis stated that the result came after the program made a "bad mistake" early in the game.^[63]

Lee, playing black, opened in a similar fashion to the first game and then began to stake out territory in the right and top left corners – a similar strategy to the one he employed successfully in game 4 – while AlphaGo gained influence in the centre of the board. The game remained even until white moves 48 to 58, which AlphaGo played in the bottom right. These moves unnecessarily lost ko threats and aji, allowing Lee to take the lead.^[64] Michael Redmond (9p) speculated that perhaps AlphaGo had missed black's "tombstone squeeze" tesuji. Humans are taught to recognize the specific pattern, but it is a long sequence of moves if it has to be computed from scratch.

AlphaGo then started to develop the top of the board as well as the centre, and defended successfully against an attack by Lee in moves 69 to 81 that David Ormerod characterised as over-cautious. By white 90, AlphaGo had regained equality, and then played a series of moves described by Ormerod as "unusual... but subtly impressive" which gained a small advantage. Lee tried a Hail Mary pass with moves 167 and 169 but AlphaGo's defence was successful. An Younggil noted white moves 154, 186 and 194 as being particularly strong, and the program played an impeccable endgame, maintaining its lead until Lee resigned.^[64]

First 99 moves

Moves 100–199 (118 at 107, 161 at )

Moves 200–280 (240 at 200, 271 at , 275 at , 276 at )

Coverage

Live video of the games and associated commentary was broadcast in Korean, Chinese, Japanese, and English. Korean-language coverage was made available through Baduk TV.^[65] Chinese-language coverage of game 1 with commentary by 9-dan players Gu Li and Ke Jie was provided by Tencent and LeTV respectively, reaching about 60 million viewers.^[66] Online English-language coverage presented by US 9-dan Michael Redmond and Chris Garlock, a vice-president of the American Go Association, reached an average 80 thousand viewers with a peak of 100 thousand viewers near the end of game 1.^[67]

Responses

AI community

AlphaGo's victory was a major milestone in artificial intelligence research.^[68] Go had previously been regarded as a hard problem in machine learning that was expected to be out of reach for the technology of the time.^[68][69][70] Most experts thought a Go program as powerful as AlphaGo was at least five years away;^[71] some experts thought that it would take at least another decade before computers would beat Go champions.^[72][73] Most observers at the beginning of the 2016 matches expected Lee to beat AlphaGo.^[68]

With games such as checkers, chess, and now Go won by computer players, victories at popular board games can no longer serve as major milestones for artificial intelligence in the way that they used to. Deep Blue's Murray Campbell called AlphaGo's victory "the end of an era... board games are more or less done and it's time to move on."^[68]

When compared with Deep Blue or with Watson, AlphaGo's underlying algorithms are potentially more general-purpose, and may be evidence that the scientific community is making progress toward artificial general intelligence.^[74] Some commentators believe AlphaGo's victory makes for a good opportunity for society to start discussing preparations for the possible future impact of machines with general purpose intelligence. In March 2016, AI researcher Stuart Russell stated that "AI methods are progressing much faster than expected, (which) makes the question of the long-term outcome more urgent," adding that "in order to ensure that increasingly powerful AI systems remain completely under human control... there is a lot of work to do."^[75] Some scholars, such as physicist Stephen Hawking, warn that some future self-improving AI could gain actual general intelligence, leading to an unexpected AI takeover; other scholars disagree: AI expert Jean-Gabriel Ganascia believes that "Things like 'common sense'... may never be reproducible",^[76][77] and says "I don't see why we would speak about fears. On the contrary, this raises hopes in many domains such as health and space exploration."^[75] Richard Sutton said "I don't think people should be scared... but I do think people should be paying attention."^[78]

The DeepMind AlphaGo Team received the Inaugural IJCAI Marvin Minsky Medal for Outstanding Achievements in AI. "AlphaGo is a wonderful achievement, and a perfect example of what the Minsky Medal was initiated to recognise", said Professor Michael Wooldridge, Chair of the IJCAI Awards Committee. "What particularly impressed IJCAI was that AlphaGo achieves what it does through a brilliant combination of classic AI techniques as well as the state-of-the-art machine learning techniques that DeepMind is so closely associated with. It’s a breathtaking demonstration of contemporary AI, and we are delighted to be able to recognise it with this award".^[79]

Go community

Go is a popular game in South Korea, China and Japan, and this match was watched and analyzed by millions of people worldwide.^[68] Many top Go players characterized AlphaGo's unorthodox plays as seemingly-questionable moves that initially befuddled onlookers, but made sense in hindsight:^[72] "All but the very best Go players craft their style by imitating top players. AlphaGo seems to have totally original moves it creates itself."^[68] AlphaGo appeared to have unexpectedly become much stronger, even when compared with its October 2015 match against Fan Hui^[80] where a computer had beaten a Go professional for the first time ever without the advantage of a handicap.^[81]

China's number one player, Ke Jie, who was at the time the top-ranked player worldwide, initially claimed that he would be able to beat AlphaGo, but declined to play against it for fear that it would "copy my style".^[82] As the matches progressed, Ke Jie went back and forth, stating that "it is highly likely that I (could) lose" after analyzing the first three matches,^[83] but regaining confidence after the fourth match.^[84]

Toby Manning, the referee of AlphaGo's match against Fan Hui, and Hajin Lee, secretary general of the International Go Federation, both reason that in the future, Go players will get help from computers to learn what they have done wrong in games and improve their skills.^[81]

Lee apologized for his losses, stating after game three that "I misjudged the capabilities of AlphaGo and felt powerless."^[68] He emphasized that the defeat was "Lee Se-dol's defeat" and "not a defeat of mankind".^[77][85] Lee said his eventual loss to a machine was "inevitable" but stated that "robots will never understand the beauty of the game the same way that we humans do."^[77] Lee called his game four victory a "priceless win that I (would) not exchange for anything."^[85]

Government

In response to the match the South Korean government announced on 17 March 2016 that it would invest $863 million (1 trillion won) in artificial-intelligence (AI) research over the next five years.^[86]

Documentary film

A documentary film about the matches, AlphaGo, was made in 2017.^[87][88] On 13 March 2020, the film was made free online on the DeepMind YouTube channel.^[89]

See also

• AlphaGo versus Ke Jie

References

External links

Official match commentary

Official match commentary by Michael Redmond (9-dan pro) and Chris Garlock on Google DeepMind's YouTube channel:

• Game 1 (15 minute summary)
• Game 2 (15 minute summary)
• Game 3 (15 minute summary)
• Game 4 (15 minute summary)
• Game 5 (15 minute summary)

SGF files

• Game 1 (Go Game Guru)
• Game 2 (Go Game Guru)
• Game 3 (Go Game Guru)
• Game 4 (Go Game Guru)
• Game 5 (Go Game Guru)

Coordinates: 37°34′14″N 126°58′31″E / 37.5706°N 126.9754°E / 37.5706; 126.9754