The God of Go. How Communities React to Automation in the Case of AlphaGo

Abstract

1. Possibilities for our Grandchildren

2. A Case for Go

3. AlphaGo and the Deep Sea
3.1. The God of Go
3.2. Cooperation and Opportunity
3.3. Human and Machine

4. Conclusion

5. Bibliography

Abstract

Recent advances in artificial intelligence made it possible for computers to defeat the top Go players in the world, a milestone that has been thought to be decades away. This has put the Go-World in front of a new challenge, as they must negotiate their position in relation to this automated program. To understand how communities, react to automatization this paper follows the commentaries of top Go professionals to understand how they experienced the success of AlphaGo, the program that today is the best player of Go worldwide. The paper finds three main themes that permeate through the material. Some professionals feel alienated by the introduction of this superhuman player but most players see the potential to elevate the quality of play throughout the community. Furthermore, in contrast to previous Go programs AlphaGo seems to be anthropomorphized by researchers and players due to its significant superhuman capabilities.

1. Possibilities for our Grandchildren

What some call a second industrial revolution has propelled our society into a knowledge or information society that increasingly relies on computation to make decisions on the ever-expanding sets of data that are available in the age of digitization. At the center of this development is a new wave of artificial intelligence software (AI) that allows ICT (Information and Communication Technology) to reshape the way we interact, learn, and make decisions both at a societal and individual level. This information explosion is made possible by exponential growth in computer technologies (Jasanoff, 2016b).

A lot of the promises that are made in the context of such powerful technological marvels are positive. As Jasanoff (2016b, p. 4) notes “technology and optimism fit together like hand in glove because both play upon open and unwritten futures, promising release from present ills.” Technological utopias are envisioned where robots and software free humanity from work and open up a new society where the individual can follow their true passions, focusing on the arts or other intellectual endeavors. While today’s discourse is largely dominated by some form of automation, connected to advances made in the last decades, the symbiosis of (technological) progress and visions of the good life is already outlined in John Maynard Keynes (1930 p.4) Economic Possibilities for our Grandchildren where he talks about the effects of revolutionary technological development: […] the economic problem may be solved, or be at least within sight of solution, within a hundred years. This means that the economic problem is not – if we look into the future – the permanent problem of the human race. […] It is startling because – if, instead of looking into the future, we look into the past – we find that the economic problem, the struggle for subsistence, always has been hitherto the primary, most pressing problem of the human race.

However, such visions go hand in hand with a darker side of technological progress. Books like Rise of the Robots: Technology and the Threat of a Jobless Future (Ford 2015) or Robots Will Steal Your Job, But That's OK (Pistono, 2012) regularly become bestsellers. Studies, both academic (Frey & Osborne 2013) and corporate (McKinsey, 2017), assess the likelihood of jobs disappearing and become headline stories that proclaim robots will destroy our jobs – and we are not ready for it (Shewan, 2017). These accounts tell a different story, one in which technological unemployment “due to our discovery of means of economising the use of labour outrunning the pace at which we can find new uses for labour” (Keynes, 1930, p. 3) creates scarcity and inequality instead of prosperity and transcendence. Both narratives seem to be very present in the contemporary zeitgeist, and open up the question of how human communities react to machines replacing vital elements that constitute or hold together their social network.

2. A Case for Go

Discourse that is drawn to its polar opposites often provides a fruitful ground for scholars of science and technology to break up meta narratives into smaller, more tangible components and illuminate interrelations within the network. Science and technology studies allows us to not only ask what technology does but rather ask how technology and society create interaction. The case of dystopian and utopian robot-visions, provides an interesting intersection at which this paper tries to leave the meta narrative and look at a specific case that might help to shine some light on some of the intricate details that make technological automation so captivating.

Between 2015 and 2017 Google Deepmind, an artificial intelligence developing company made headlines because it created a program that could beat professional players at the board game Go. AlphaGo's success was a major milestone in artificial intelligence research. Go had previously been seen as a particularly difficult problem in machine learning that was expected to remain ‘unsolved’ for at least a decade (Silver et al., 2016). However, while the AI community was amazed by the technological feat achieved by the researchers working on AlphaGo, the Go community faced a different challenge. For the first time in its more than two millennia old history humans´ genius, intuition and skill was superseded by a machine. This experience is made more intense as the game does not just have a large audience or historic roots but also a deep cultural identity. It is considered to be a noble ‘sport’ and often referred to as one of the four artistic accomplishments (ancient) ‘scholar-gentleman’ had to master. The game is taught in schools all over Asia, important championships are broadcasted and the top players are revered. For Tai-Hsiung Yang, a professional player “the game of Go contains a lot of the joy and philosophy of life; it is not just a game concerned with winning, but a cultural legacy that cultivates human talent” (Wang et al., 2016). This exemplifies that Go is often considered to be more deeply connected to the human mind than any other game: The `Three Games' [Backgammon, Chess and Go] is a useful classification because taken together they make up a coherent world view. Most of philosophy boils down to speculation centered around the three basic relationships of the human species. The first is man in his relationship to the remote gods and the mysterious forces of the universe. The second is man in the society he builds up around him. The third is man in his own self. Or, to put it another way, man the backgammon-player, man the chess-player, and man the go-player (Pinkard 2001).

If the three games that kept their popularity throughout human history are characterized as man vs. fate (Backgammon), man vs. man (Chess) and man vs. self (Go), it is of little wonder that a machine that defeats the best human Go champions, causes humans to question themselves.

This dynamic of high cultural relevance and media attention that is connected to AlphaGo makes it a particularly interesting contemporary case that can help to shed some light into how human communities respond when a core part of their life is automated by machines that exceed human ability. In order to tackle this question this paper will take a look at AlphaGo’s success story as well as public reactions and interviews of key personalities in the world of Go to see how they experienced machines matching or succeeding their abilities. These personal accounts are collected from statements of the participants at the 3 tournaments in England, Korea and China as well as interviews with professional Go players and representatives of international Go Organizations. These remarks will be put into a wider context of already existing literature that reports on human responses to automation.

3. AlphaGo and the Deep Sea

The Brain’s Last Stand was the headline of many newspapers, that covered Kasyanov’s chess match against IBM’s DeepBlue chess computer. When looking into AlphaGo’s success its hard not to think back at the famous match of the Russian grandmaster. However, besides differences in mathematical complexity or cultural identity, AlphaGo is fundamentally a story of the 21st century, which serves as a catalyst for various technological metanarratives. The resurgence of attention towards automation and AI is epitomized by the coverage of AlphaGo. The start of AlphaGo’s rise to popularity was a match against European Go champion Fan Hui, a 2-dan (out of 9 dan possible) professional, which he lost five to zero (Silver & Hassabis 2016). It was the first time a Go program was able to defeat a professional human player. After the event many were stunned, but top players including Lee Sedol where still confident that the program could be beaten by better human players (Mackanzie, 2016). AlphaGo was compared with a child prodigy that lacked experience, while “the real achievement will be when the program plays a player in the true top echelon” (Gibney 2016). A few months later this challenge was realized when Lee Sedol, who is considered one of the strongest players in the history of Go, played Google Deepminds program. The match was watched by millions of people worldwide, gained tremendous media attention worldwide and ended in a four to one win for AlphaGo. The victory was seen as a major milestone in artificial intelligence research but on top of that it posed serious questions to the identity of the game and its community. However, it was not until May 2017 that the dominance of machine over man was cemented when the number one ranked player of the world lost three to zero at the Future of Go Summit against AlphaGo, after defeating many other top players in the preceding month (Google Deepmind, 2017). This short overview of AlphaGo’s rise to power is a success story of modern technological progress but also is the catalyst for stories that are much more ‘human’.

3.1. The God of Go

“To describe these past days of the week I’d use the word desperate” (CGTN, 2017). This was Ke Jie’s conclusion of his three to zero match at the Future of Go Summit. The title of the tournament made the crossroad at which the Go community found itself vividly clear. The future of Go is at play and Ke Jie’s first resume in an interview after the final game was skeptical: “I never doubted myself. I always felt I had everything under control. […] But to AlphaGo, its like ‘what is all this rubbish’ ”(CGTN, 2017). Understandably, humans who have dedicated their life to a certain task are stunned if a machine can do something better than what they have worked for decades. This can be an intense experience as Ben Lockhart, one of the top US amateur players, expressed: “I felt emotional and dizzy, and stepped outside for a minute” (Zastrow, 2016). Similarly, Lee Sedol himself stated “I am in shock, I admit that” (Zastrow, 2016). The deep connection to cultural and personal identity, combined with the shock and disbelief can lead to dismissal, helplessness or antipathy towards the machine that is replacing your work. Ke Jie expressed similar feeling after the game. “AlphaGo can see the whole universe of Weiqi [Go]. I can only see a small area around me. So please let me play in my own backyard. Similarly, Artem Kachanovskyi of the European Go Federation voiced concern after the second tournament including Lee Sedol:

To be honest, I'd prefer to practise against Lee Sedol. So far I don't really have any sympathies for computer go programs. Go is for humans. Playing against AlphaGo can be compared to playing football against a perfect robot. Might be interesting once or twice, but it is much more fun to play regularly against humans (Ouweleen 2016).

Yet, these emotions are not particular towards Go. Mindell (2015) tells the captivating story of the scientific oceanography community that had relied on manned submersibles but had to face competition from remotely operated (underwater) vehicles (ROV) that challenged the status quo. Budget questions finally forced decisions over whether to build expensive manned submarines or less expensive ROV’s. However, the reasons for why unique human abilities required a direct presence on the ocean floor kept shifting as the technology evolved. Mindell uncovers that what was at the heart of the issue was less a dispute over technological flaws, but rather the scientists were trying to articulate something about the quality of their personal expertise. Similarly, some of the Go community felt that their profession should be kept in the realms of the human. Even though it might be more efficient, scalable or enhancing to train with the new machine, Kachanovskyi specifically mentioned to rather train with Lee Sedol. Ali Jabarin of the European Go Foundation expressed: “I was a bit sad that Lee Sedol lost and that Go lost some of that mysteriousness” (Ouweleen, 2016). This notion of mystery and awe that is ‘taken’ by the machines was expressed throughout the surveyed material. Not so much technological capabilities but rather professional identity, pride and culture are at the heart of the issue when new technologies penetrate into predominantly human communities to challenge the status quo. These personal identities are under attack when central processes can be done better by machines. Cole Pruitt, the co-founder of the American Collegiate Go Association stresses this experience: “I assume it’s sort of like realizing your kid isn’t destined to be the most successful, brilliant kid of all time—they’re just a mostly average person, like you” (Chan, 2016). The effect on human identity – not just productivity – has to be taken into account when we think about new technologies which are introduced to communities dominated by human labor.

Ke Jie characterized AlphaGo as “a god of Weiqi [Go]. A god that can crush all that defies it” (CGTN, 2017), which evokes the AI utopias discussed in the first chapter. However, it is important not to forget that the element of this picture is not just technology dominating man but rather how man and technology negotiate their position in the network. In Mindells (2015) story ROV’s have changed the place where people do their work. But they also have changed what it means to do work, what it means to be an oceanographer and thereby changing personal identities or in other words what it means to be an explorer. Automation in the form of AlphaGo is changing the future of go – made apparent in the title of the final tournament – thereby changing what it means to be a professional Go player. Both examples can serve as a lighthouse which can help us to think about and navigate the changes exponential technological progress and automation will bring to our current way of living.

3.2. Cooperation and Opportunity

Before Fan Hui, Lee Sedol and Ke Jie played against AlphaGo, both players were convinced that they would be able to defeat the program (Zatrow 2016; Mozur 2017), which is reminiscent of early oceanographers dismissing the capabilities robotic submersibles in the early phases of aquatic ROV’s (Mindell 2015). After the success of AlphaGo and a fair amount of shock, most of the community expressed their awe and appreciation for the technology. Tang Weixing, 9 Dan professional, commented during post-match wrap up: Playing the match with AlphaGo is impressive, for me it’s a transformational moment. It’s still a wonderful experience that we can play the game, we can feel that we have a further understanding of the game of Go, we can feel that we’ve learned a lot (Google Deepmind 2017).

The same moment discussed in the previous chapter can also serve as an impulse to interact not deflect. The overwhelming majority of the sampled statements emphasized the potential for human players to interact and learn with AlphaGo. Thomas Debarre from the European Go Association underlines this dynamic: I think it's fantastic news, because we will discover so many new things about the game. AlphaGo's matches against Lee Sedol were just a little taste of how little humans know about go. Before the event, I thought top professionals were not that far from perfect play, but I was clearly wrong (Ouweleen, 2016).

This statement is reminiscent of Mindell´s (2015, p.23) story of the robotic survey of the sea floor, where scientists explored the site remotely “discovering a great deal about it [the seafloor] that was not visible when I [the researcher] was there”. Both experiences understand the novel technology not as a threat but rather a catalyst for humans to evolve and progress further. AlphaGo is not the ‘other’ rather it is understood as manmade and part of the human experience as Fan Hui remarked after his match: “AlphaGo represents not only a scientific and technological advancement, but also a milestone in human understanding of Go” (USGO, 2015). Similarly, Mindell (2015, p.24) describes the Skerki D survey as a significant change in “what was possible in archeology and how we explore human history in the deep sea”. Fredrik Blomback from the European Go Association emphasizes this opportunity in the context of AlphaGo’s success: “I believe the next generation of players will be stronger than ever before, due to the aid of artificial intelligence” (Ouweleen, 2016).

[...]