On March 9, 2016, World Go Champion, Lee Sedol of South Korea, had been roundly defeated in Game 1 by AlphaGo, an AI system designed by DeepMind to play the ancient North Asian board game of strategy.
The game of Go is very complex, and defeating Go masters had been the holy grail for AI researchers since IBM’s Deep Blue took down chess grand master, Gary Kasparov in 1997. As DeepMind’s CEO, Demis Hassabis, said, “Go is incredibly challenging for computers to tackle because compared to say, chess, the number of possible moves in a position is much larger. In chess, it’s about 20. In Go, it’s about 200. And the number of possible configurations of the board is more than the number of atoms in the universe.”
And yet, to everyone’s surprise, AI was giving Sedol a run for his money.
In the middle of Game 2 on March 10, Sedol made his move and needed a break. He walked to an isolated part of the hotel and smoked a cigarette. When he came back, he was shocked. As European Champion, Fan Hui, the first professional Go player to be defeated by a computer explained, AlphaGo made a move – a strange move. “When I see this move, I think it’s a big shock. Normally, we never play this move, because It’s bad! I don’t know why, but It’s just bad.”
But Sedol’s reaction was mixed. He was stunned. His fingers picked at his lower lip as he tried to understand what his robot competitor was trying to do. In the end, AlphaGo was laying the foundation for victory. It took Lee Sedol 12 minutes to make his move. That one, now famously called Move 37, and all the others that followed were in vain as Sedol ended up losing Game 2. Despite the loss, Sedol was amazed.
I thought AlphaGo was based on probability calculation and it was merely a machine. But when I saw this move, I changed my mind. Surely AlphaGo is creative. This move was really creative and beautiful. This move made me thing about Go in a new light. What does creativity mean in Go? It was a really meaningful move.
Despite over two millennia of human experience with the game of Go, the greatest Go masters today have not seen every move. AI may in fact be revealing that humans are just scratching the surface of their potential.
I’d like to believe this is a metaphor for all things human – that we are all just scratching the surface of our potential in a wide variety of areas. And it’s not just AI that is showing us the way.
There is an explosion of evidence of the amazing things that people can learn to do. Clicking and swiping through YouTube and TikTok today may have been what it was like to sit at a circus or wander through a carnival 100 years ago, witnessing athletic feats, shocking freaks, and magical treats that appeared beyond the abilities of the everyday person.
The internet and AI have replaced the P. T. Barnums and Pablo Fanques of centuries past, demonstrating a stupefying range of human appearances, idiosyncrasies and capabilities. Everyday people by the millions are singing and dancing better, improving language fluency and accents more quickly, absorbing economic theories or accounting functions more comprehensively, as well as figuring out how many ways there are to bounce a ping pong ball off various objects before it lands satisfyingly in a glass.
I am reminded of what American philosopher and psychologist, William James, wrote over a hundred years ago in his essay, “The Energies of Men”:
Compared with what we ought to be, we are only half awake. Our fires are damped, our drafts are checked. We are making use of only a small part of our possible mental and physical resources.
In other words, we all have incredible upside!
And with the increasing number of people demonstrating their superlative capabilities on the Internet, we are inspired. Quite often, when someone asks us if we can do something perceived to be challenging, or even impossible, we do not even try.
But equally often, all we need is a single example of success before the floodgates open. Running a mile in four minutes or less was long considered an impossible achievement in the first half of the 20th century before Roger Bannister ran the mile in 3 minutes and 59.4 seconds on May 6, 1954. Two months later, the barrier was broken again, and today, over 1,600 runners have broken the 4-minute barrier.
On January 14, 2015, rock climbers Tommy Caldwell and Kevin Jorgeson ascended El Capitan’s Dawn Wall in Yosemite, long considered the world’s hardest rock climb. It took them 19 days to reach the top, after 8 years of planning, and loads of trial and error. Nearly two years later, Czech climbing sensation, Adam Ondra climbed the flat vertical surface of the Dawn Wall in only 8 days, after only a few weeks of scouting the mountain and practice.
In the book, The Alignment Problem, by Brian Christian, Ondra said his achievement was due to the model that came before:
The fact that Tommy and Kevin put all this effort, like years and years of work, it’s so impressive. There are so many sections on the crux pitches, and even on some of the easier pitches, if you are there, you think, “No, this is impossible.” And only after studying every single tiny razor blade, sometimes the only possible answer comes to your mind. So it was so much easier for me to solve the puzzle on each individual pitch because I knew that the guys did it. They set the route and I was very proud to make the first repetition.
James said that our rational mind takes over, persuasively insisting to us that the odds are against us. “No one has done that. How could you” – we tell ourselves. “Our scientific respectability,” James wrote, “keeps us from exercising the mystical portions of our nature freely.”
But when we see a model, are shown that the impossible is possible, we can be motivated to overcome our inhibitions, our excuses, our perceived limitations. We can be inspired to try, and try, and try.
Sedol went on to lose the third of five matches, thus losing the championship to AlphaGo. He felt that he let down not only his family, and all of Korea, but also, all of humanity.
But he trooped on and played Game 4.
In Game 2, Alpha Go made the stunning Move 37 that led to victory for Deep Mind and artificial intelligence.
In Game 4, with Sedol fighting furiously with AlphaGo, taking far more time than his algorithmic opponent but not giving ground, he placed his black stone in a spot that is now immortalized in lore as “Move 78.”
Not only were the experts stunned, AlphaGo was stunned, making moves that clearly looked confused, or as the Deep Mind AI scientists called, “delusional.” Following subsequent misplays by AlphaGo, Sedol went on to defeat the mighty machine. And even though Alpha Go would recover to defeat Sedol in Game 5, Sedol’s burst of creativity and subsequent victory in the face of tremendous adversity in Game 4 was seen as a triumph for humanity.
As journalist and witness to the epic Man Vs Machine battle, Cade Metz saw the silver lining in what many might consider dark clouds. “In a broad sense, move 37 begat move 78, begat a new attitude in Lee Sedol, a new way of seeing the game. He improved through this machine. His human-ness was expanded after playing this inanimate creation. And the hope is that, that machine and in particular, the technology behind it, can have the same effect with all of us.”
Sedol said pretty much the same: “I’ve grown through this experience. I will make something out of it with the lessons I have learned. I feel thankful and feel like I have found the reason I play Go. I feel it was really a good choice, learning to play Go.”
We are in a new Age of Enlightenment.
Models abound, human and inhuman.
All you have to do is watch, learn, and try.
All the quotes uncited above are from the fascinating 2017 documentary, “AlphaGo.”
ARTICLE FAQS
1. Why was the match between Lee Sedol and AlphaGo significant?
It showed that AI could outperform human experts in one of the most complex games ever devised. At the same time, it revealed that humans could still rise to the challenge with creativity and resilience.
2. What does “Move 37” represent?
It represents an AI move that broke with human convention, surprising experts and forcing players to rethink the game. It became a symbol of how AI can expand human understanding by showing new possibilities.
3. What does “Move 78” represent?
It was Lee Sedol’s creative response in a later game, one that confounded AlphaGo and led to his victory. This move reminded the world that humans, too, are capable of breakthrough thinking under pressure.
4. What broader lesson does the Sedol–AlphaGo match teach us?
That human potential is far from exhausted. Both human and machine can inspire each other to think and perform at levels once thought impossible.
5. How do examples of breakthroughs in sports and performance tie into this story?
Whether it is the first sub-four-minute mile or the first ascent of a daunting rock wall, once one person demonstrates success, others quickly follow. Seeing a model of success breaks psychological barriers.
6. Why call this moment a “new Age of Enlightenment”?
Because more than ever, we have access to models—human and AI—that show us new ways of thinking, creating, and achieving. The abundance of visible examples lowers the barrier to learning and expands what people believe is possible.
1 thought on “Flying Up the Learning Curve in this New Age of Enlightenment: How Lee Sedol and AI Showed Humanity We Still Have Lots to Learn, and That’s Great!”