It's intensely contemplative. It is almost hypnotic. It's like putting your hand on the third rail of the universe. If you play Go, seriously, there is a chance that you will get exposed to this. Experience that is kind of like nothing else on the planet.
Go is putting you in a place where you're always at the very farthest reaches of your capacity. There's a reason that people have been playing Go for thousands and thousands of years, right? It's not just that they want to understand Go. They want to understand what understanding is.
And maybe that is truly what it means to be human. When I was a kid, I loved playing games. I started off with board games like chess.
And then I bought my first computer when I was eight with winnings from a chess tournament. Ever since then, I felt that computers were this sort of magical device that could extend the power of your mind. Virtual environments in games.
We think they're the perfect platform for developing and testing AI algorithms. Games are very convenient in that a lot of them have scores, so it's very easy to measure incremental progress. So I'm going to show you a few videos of the agent system, the AI.
So let's start off with Breakout. So here you control the bat and ball, and you're trying to break through this rainbow-colored wall. The agent system has to learn everything for itself, just from the raw pixels. It doesn't know what it's controlling, it doesn't even know what the object of the game is.
Now, at the beginning, after 100 games, you can see the agent is not very good. It's missing the ball most of the time. We decided to get the hang of the idea that the bat should go towards the ball.
Now after 300 games, it's about as good as any human can play this, and pretty much gets the ball back every time. We thought, well that's pretty cool, but we left the system playing for another 200 games. And it did this amazing thing. It found the optimal strategy was to dig a tunnel around the side and put the ball around the back of the wall.
The researchers working on this are amazing AI developers, but they're not so good at breakouts. And they didn't know about that strategy, so they learned something from their own system, which is pretty funny and quite instructive, I think, about the potential for general AI. So for us, what's the next step now? Go is the most complex game pretty much ever devised by man.
Beating a professional player at Go is a long-standing... grand challenge of AI research. Whoa. I really want to kill myself.
No, I don't think so, but I'd like to survive. Yes, that's it. I'm a gold professional player. I'm also European champion. You'll never die.
It's just that once you kill yourself, you're so happy and you're not afraid. You're afraid that victory will escape. In general, when you're afraid of that, victory always escapes.
I was born in China. When I was 18, I wanted to change my life. That's why I went to France. I wanted to try to forget the goal.
But it's impossible. Because all the things I learned in my life were with the goal. It looked like a mirror. I see the goal, I also see myself.
For me, the goal is real life. Dear Mr. Pham, my name is Demis Hassabis. I run an artificial intelligence company based in London called DeepMind.
As the strongest Go player in Europe, we would like to invite you to our offices in London, both to meet you in person and to share with you an exciting Go project that we are working on. If you would be interested in coming to visit us, please let us know. Many thanks. Kind regards, Demis. When I see the simile, I don't know if it's true or not.
I will accept about this. Why not? For me, everything is an adventure.
I want to go to visit D-Mind to know what is this Go project. The first visit I think maybe I want me Sit in the special room push many many where in my hand Also my body let me play to scan my brain I don't know, to make some research. We needed to get Fanway to DeepMind to see we were a serious operation and we were serious people doing proper research. And the search time is getting better and better. As we go from, for example, one to eight in any one of these, it always goes...
It goes up. We think of DeepMind as kind of like an Apollo program effort for AI. Our mission is to fundamentally understand intelligence and recreate it artificially. And then once we've done that, we feel that we can use that.
technology to help society solve all sorts of other problems. If you step through the actual game, we can see what AlphaGo thinks, what's the most likely variation that it thinks will happen. We've been working on AlphaGo, our programme to play Go, for just under two years now.
All the little patterns cascade together, layer after layer after layer after layer. I started talking about the Game of Go with Demis more than 20 years ago, and so this has been a really long journey. The Game of Go is the holy grail of artificial intelligence.
For many years, people have looked at this game and they've thought, wow, this is just too hard. Everything we've ever tried in AI, it just falls over when you try the game of Go. And so that's why it feels like a real litmus test of progress.
If we can crack Go, we know we've done something special. So with Fan He, we started talking around what the real purpose of the visit was. It wasn't just a Go project we wanted him to help with, that actually we wanted to play him. And we had a very strong programme. It's okay, just the programme is so easy.
So it will be easy to play. A lot of people have thought that it was decades away. Some people thought it would be never because they felt that to succeed at Go, you needed human intuition.
Go is the world's oldest continuously played board game. And in some sense, it is one of the simplest and also most abstract. There's only one type of piece.
There's only one type of move. You just place that piece on the board. And then your goal is to create a linked group of your stones that surrounds some empty territory.
And when you surround enemy stones, you capture them and remove them from the board. You earn points by surrounding territory. And at the end of the game, the person with the most territory wins. It seems really simple, but then you sit down to play and you realize right away, it's like, well, I technically know what I'm allowed to do, but I have no clue what I should do. It goes 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. So even if you took all the computers in the world and ran for a million years, that wouldn't be enough compute power to calculate all the possible variations. If you ask a great Go player why they played a particular move, sometimes they'll just tell you it felt right. So we have to come up with some kind of clever algorithm to mimic what people do with their intuition.
Nice to meet you. Oh, hello, nice to meet you too. So with Fan Hoi, we agreed a best-of-five match, and we agreed it would be filmed. You know, we would treat it as a serious match. I play with AlphaGo.
Aja Huang pushes the stone for AlphaGo. And of course I think I will win the game with AlphaGo because it's just a program. The first game, I make some mistake and I lose the game.
In that first match I think something clicked for him that this wasn't an ordinary Go program. We weren't just doing the same as everyone else that something new was happening. The second game I tried to change my style but the problem is also I lose the second game and also for third game, fourth game, even the last game. I lose 5-0.
AlphaGo win all the game. After losing, losing, losing, you can feel his pressure is getting heavier and heavier. And several times after the game, he said he wanted to go out for fresh air. I said, oh, I can go with you and have a chat. He said, no, I want to go myself.
I feel something very strange. I lose this program and I don't understand myself anymore. I worried whether he He didn't even come back.
He seemed very low. He spent about an hour away from the office. He came back completely changed.
I told Maddy, this is the first time in history that a human professional goal player loses with a programme. She told me, yes we know. But for you, I said, are you OK?
I told her, not and yes. I am not happy to lose a game, but I will be happy to play in the history. Artificial intelligence researchers have solved the game of Go a decade earlier than expected. The computer named AlphaGo was able to beat the European human champion. Artificial intelligence researchers have made a significant breakthrough.
It really is a big leap forward. There's a big difference between the way the IBM computer beat Kasparov, which was programmed by expert chess players, and the way the Go-Playing computer more or less learned itself. The way we start off, training AlphaGo is by showing it 100,000 games that strong amateurs have played that we've downloaded from the internet.
And we first initially get AlphaGo to mimic the human player. And then through self-play and reinforcement learning, it plays against different versions of itself many millions of times and learns from its errors. These specific ideas that are driving AlphaGo are going to drive our future. The technology is at the heart of AlphaGo.
They're what are called deep neural networks, which essentially... mimic the web of neurons in the human brain. It's a very old idea, but recently, due to increases in computing power, these neural networks have become extremely powerful, almost overnight. Big neural networks that operate on big data can achieve surprising things. AlphaGo found a way to learn how to play Go.
Learning is the key thing here. It's machine learning. The whole beauty of these types of algorithms is that because they're learning for themselves, they can go beyond what we as the programmers would like to do.
know how to do and allow us to make new breakthroughs in areas of science and medicine. So AlphaGo is one significant step towards that ultimate goal. My wife called me. She told me, don't see internet.
Don't connect internet. Because people talk terrible things about you much with AlphaGo. Fan Hui is living so long time in Europe.
He's not real professional, he's just an amateur player. This is so hard for me. So hard.
The Go world were skeptical about how strong really was AlphaGo and how much further did it need to get to beat the top professionals. So our program is improving over time and we want to push the AI algorithm to the limit and see how far this kind of self-improving process can go. So we needed to look for an even greater challenge. A match like no other is about to get underway in South Korea.
Lee Seedol, the long-reigning global champ. This guy is a genius. will take on artificial intelligence program AlphaGo in the ultimate human versus machine smackdown.
This is a huge moment for both the world of artificial intelligence and I think the world of Go. So far, AlphaGo has beaten every challenge we've given it, but we won't know its true strength until we play somebody who is at the top of the world like Lee Sedol. We chose Lee Sedol because we wanted a legendary historic player, somebody who has been acknowledged as the greatest player of the last decade. I will not be too arrogant, but I don't think that it will be a very close match.
The level of the player that Africa went against in October is not the same level as me, so given that a couple of months was only past, I don't think that it is enough for it to be able to catch up with me. My hope is that it will be either 5-0 or... zero for me or maybe four to one. So the critical point for me is to make sure I do not lose one.
We don't know how well our system will play against someone as creative as Lee Sedol. Also, he's very famous for very creative fighting play. So this could be... difficult for us, but we'll see.
Maybe in some ways he's the most difficult opponent we can pick. There was still so much of a question about whether or not they could beat someone like Lee Sedol. Fan Wei is a good player, but he's nothing like the very top players. Lee Sedol is a nine down. professional.
Nine-down professional is the highest rank that you can achieve in Go. The ranking system. Fan Hui, who is the European champion, is just a two-down professional.
On the other hand, the very top, Lee Sedol, is nine-down professional. Lee Sedol is to go what Roger Federer is to tennis. He's playing in Wimbledon to win the Grand Slam.
And it's not just this year. He'll be there next year, and the year after that, and the year after that. Of course, in Internet and all the Go community. Everyone. Maybe not 100%, maybe 99.999% think Lee Sedol will win very easily.
Go, hold the stone beautifully. Don't hold the stone like this. Don't hold it like this.
Hold it like this. Go is taken seriously in Korea. It's so much part of the culture, like breathing or, you know, like taking swimming lessons or something. On the surface, it's a game, but inside had a very deep philosophy. Go board reflects the individual who's playing.
The true gonna show itself on the board. You won't be able to hide it. In ancient China, Japan, Korea, Go is one of the four noble things, like four accomplishments for literati, with music, poetry, and painting. So people think the goal players are very smart and very noble.
Master Kwon started this school so that he could produce great players in Korea. Lee Se-dol, when he was age of eight, he came in here attending class 9 a.m. till 9 p.m.
seven days and then he stayed with Master Kwon. His face was very yellow, his eyes were very black. He was a boy from an island, so he looked a little strange.
But his eyes were different from normal kids. His eyes were very bright. Lee Se-dol was a little boy that I remember when I was a student studying Go.
He was very young and rural enough to think that pizza grew on trees. When I was young, I started playing for fun to beat older people. After that, I started to create my own creative work.
I think it's a little different, a little different. Something that's unique, something that others didn't think of. Lee Se-dol plays things that are interesting, where you felt like he's beyond winning and losing.
He wants to do something that's... innovative or takes things to the next level. So every Go player studied his game for sure. Roughly 10 years he dominates the professional Go world. He won 18 world championships.
He is a genius who will come out in 100 years. I think he did a great job for Lee Se-dol. I live with satisfaction. Definitely Lee Se-dol is going to win the game 5-0. Yeah, all game.
It's gonna be an all game. We had our evaluation match last week. We won a game and we lost a game.
And we lost a game in a way that would have made us look extremely foolish if that happened publicly. It means that we still have work to do and we need to take this really seriously. It's just too much risk that actually we could lose overall.
Not only that, but we could lose in a way that makes us look rather silly. So I guess... Yeah, Ajay, did you want to say anything about what you're trying? I'm walking hard.
We're working around the clock at the moment, training our algorithms further, trying to incrementally keep on improving right up to the moment where we have the match. We've collected together people with different skill sets onto the same team, so we have researchers, engineers, evaluation guys. So you mean what happens if it creates something new which wasn't easy to be made?
Then we don't see it. It's the first thing to know. Oh, we're thinking about the perfect solution.
Adjo is the lead programmer and built the original search engine. So Adjo's responsibility is quite a big one. He will be the one sitting opposite Lee Sedol and actually playing the moves AlphaGo makes. I'm feeling excited. Lee Sedol is a great player.
I feel honored to play with him. Many of my friends they are I'm very excited about the match. They keep telling me that the whole world is watching. Just prepare AlphaGo.
Hello. Hello, Dennis. Can you hear us? Yes, I can hear you guys.
Can you hear me OK? Yes, perfect. I was curious. It's a real pleasure to meet you. I love the game of Go.
I'm not very good. I'm not very good at it. I'm not very good at it.
I'm not very good at it. I'm not very good at it. I'm not very good at it.
I'm not very good at it. I'm not very good at it. I'm not very good at it.
I'm not very good at it. I'm not very good at it. I'm not very good at it. I'm not very good at it. I'm not very good at it.
I'm not very good at it. I'm not very good at it. I'm not very good at it. I'm not very good at it. I'm not very good at it.
I'm not very good at it. But I don't know if you know, but we actually have something in common. I actually also trained at a game at a young age. I used to be, when I was young, a professional chess player.
I used to play for the England team, and then I stopped playing when I was about 14. When I was 13, I was the second highest rated player in the world. After the match, I would like to propose that I give you one teaching... game of Go and you gave me one teaching game of chess.
How does that sound? That sounds very good. I would be very honoured to do that.
I'm not sure if it's okay to ask you this question, but I saw the games against Fan Hui and I didn't think it was quite at the level to play with me, but I heard that it's getting really stronger. Can I ask how much it got stronger since then? Yeah.
I can't say too much, of course, but it's definitely got significantly stronger. With more time, I think a logical approach would be to follow up on Azure's local search and run that in Fastel to generate a whole new data set and iterate for more time. But we're just out of time for that, so we have to be realistic. We realised that if we wanted to prepare for a mammoth task like taking on Lisa Dole, there would be nothing better than... talking to a professional, and we couldn't have picked a better person than Fan Hui.
We invited him as an advisor, because during the match, we found he is a man of good spirit. Welcome back. Thank you very much. It's very precious that we have Fan Hui with us. He was crushed by AlphaGo, but then he became very positive and a big help to us.
When they tell me, can you come back to help us make AlphaGo more stronger, I feel this respect. Of course I accept about this. I play with AlphaGo to understand where is a strong point about AlphaGo and where is maybe the weakness.
I play in the morning, afternoon, all the time. And I find something. I find big weakness about AlphaGo.
The big one. The superheroes always have a hidden vulnerability, right? And the same is true of AlphaGo. It's unbelievably superhuman in general, but it has some particular weaknesses in some situations. We can think of...
there being this space of all the things it knows about, and it knows about most of it extremely well, but then there'll be these tricky lumps of knowledge that it just understands very poorly. And it's really hard for us to characterize when it's gonna enter into one of these lumps. But if it does, it can be completely delusional, thinking that it's alive on one part of the board when in fact it's dead or vice versa. So there is a real risk that we could lose the match.
Everybody in the team tried to work more and more to fix the problem, but I think it's difficult to fix this very quickly. We've got version 18 in the pipeline, haven't we? Well, we'll only go for version 18 if it makes a significant improvement, which we're having to rerun all the tests because they went wrong.
Subtitles by the Amara.org community unfortunately, so we're sort of basically a day behind in our evaluation because of that. Well, maybe we just need to be realistic that we've tried a bunch of things and we've come to the point where we said we would actually start freezing the code and saying, look, it's not and actually not panning out. These delusions are still a realistic possibility for the match.
We have some weaknesses that we, I don't think we're going to fix fully before the match. So That's causing us a little bit of anxiety. Lee Sedol is getting ready to rumble. On Wednesday, live across the internet, this professional South Korean Go player will take on artificial intelligence program AlphaGo. Tomorrow.
Julian and George, they pack up version 18, they stick it on a laptop and they fly out to Seoul. Everything is all set for the victory against Issezo? Everything is set I think, but when I get to the hotel I'm going to catch up with the team. What we should do in the time remaining is list things that could go wrong in the solidity of the system. We need someone to cook.
I have to go under here. Alright rehearsal, let's go please, max! How do you arrange the baseball?
At least I have to stalk. In a difficult situation, he likes to stalk. Yes, there is a terrace. And we will have security, so he will be able to go up there and be by himself. Are you here to stand on the stage or to talk?
Uh... I'm a little embarrassed. But look at my clothes.
Okay, no problem. I'm the judge. When I got here, I didn't expect the attention on the maps. It was literally front-page news.
About 8 million Koreans play the game of Go. And even those who don't recognize Lee Seedol, he's a national figure. And so there's that, right? There's some national pride involved, but it's more than that.
It's just a... Just the very thought of a machine playing a human as something like this, I think is inherently intriguing to people. We could bring him in. We've got a warrior from the show.
Yes, we can. We've got a warrior from the show. I still have confidence.
I think it's too much for human intuition and senses to follow computer AI. I will do my best to protect humans. All the world gave the prison to Lysidarch.
Before this he played the tournament for country, for himself. But this time he played for the human. I just really hope we win this first game.
If you lose the first game, you literally have to win three out of four. Which is hard work. Where are you going to be?
I'll start in the match room, and then I'm going to come in here and make sure everything's... Yeah, you should probably be in there. I know you're nervous.
Yeah, I'm nervous. I'm nervous. How are you?
You're not nervous? A bit. A little bit, yeah? That's good to have a little bit of nerves.
Be fine, be fine. Hello and welcome to the DeepMind Challenge Game 1, Round 1, live from the Four Seasons here in Seoul, Korea. I'm Chris Garlack of the American Go eJournal. I'm here with Michael Redman, 9DON professional.
Welcome, Michael. I want to give a shout out to all of the folks watching around the world. Well, the excitement's pretty palpable here in the hotel. I've never seen a crush of interest to reporters from around the world. No idea.
Overwhelming. All right, folks, you're here. You're gonna see history made. Stay with us. Five minutes, guys.
Five minutes. I just thought we should take a moment together and just think about what's about to happen. Extremely excited to be there, extremely proud of every one of you and what we've done.
And win or loss, I think it's just amazing that we're here. Isn't it weird fighting with a machine gun? I don't want a machine gun to win a battle. The first time I saw a Chinese man in a Chinese restaurant.
The first time I saw a Chinese man in a Chinese restaurant. Most of them apply Chinese culture and have a long face. Most of them apply Chinese culture and have a long face. The time limit is 60 seconds and 3 days.
The time limit is 60 seconds and 3 days. The time limit is 60 seconds and 3 days. The time limit is 60 seconds and 3 days.
The time limit is 60 seconds and 3 days. The time limit is 60 seconds and 3 days. The time limit is 60 seconds and 3 days. Please cover your eyes.
Please choose a word. Please start the game. We had worked so hard to make sure that this would go technically smoothly. We tested it and tested it and tested it and still there comes that moment when you're live all the TV cameras are broken.
I was a bit nervous. It's the first time that I sit in front of a world-class goal player. I actually can feel the spirit and courtesy of a great goal player like Lee Seo-il. Because I think it is the first time he faces a strange opponent.
I think he's not a human, has no emotion, it's cold, but he stays very calm. And I can feel his mental strength. Oh?
Oh. Did you look into it? It's a surprising number.
Alphago was the first to release a unique move. It was like, oh, he's trying to turn around like a human. I'm Andrew Jackson.
With me here is Myung Won Kim, a nine-time pro from the Korean Baduk Association. We are here live at the Four Seasons Hotel on the 21st floor. What are we looking at?
How's the game going? Oh, it's fighting from the beginning. From the very beginning?
Yeah, yeah. Alphago plays very well. Yeah? Just like a top professional.
Just like a top professional? Yeah, he's... very aggressive. Now the fight is getting really complicated.
This is actually the first time I've seen AlphaGo playing a game that has this difficult a fight. I can shake this a little bit. Of course, I use AlphaGo for 1 minute to 1 minute and 30 seconds. This part is definitely not close to a person.
It's useless to be careful. The opponent always puts it as if he knows everything. I just saw him looking at his opponent's face, and that's just a kind of a habit. It's just an instinct as a player to look at the person across the board. Yeah, it's sort of like something that Lisa Lowe would do when he was wondering how his opponent was feeling.
It's just a habit. So it's not as if Aja Huang is going to do any giveaway, because Aja Huang isn't AlphaGo, right? Oh? He came and went, but now he's smiling.
I think it was a mistake. No, I almost heard something ridiculous. I was here. What are you doing now? Did you lose today?
I was going to turn here, but my hand went down without thinking. Because there are so many people who want it. It's hard now. I was a little embarrassed. This is what we're most afraid of.
Yes. It's shaking. With humans, when you play, you can have an exchange by feeling.
I look at you, I know, okay, maybe you won't talk with me. Maybe you fear about me. I can feel many, many things.
But, I'll have a go. You can feel nothing. So, when you can feel nothing when you play, you have more than one question about yourself. At the beginning, you think, okay, my move is good. And, really good?
Really good? Oh, maybe bad. Oh, terrible. Why do I play here? More than more.
It's hard to know where to be. All the different rooms are exciting in a different way. Isn't it? Quite nice to be here at the heart of the operation.
Yeah, I feel safe here. White's thinking of doing this huge invasion here from its thick wall. Yes, go for it.
He's done it. He's gone in. We'll look at his face. Look at his face. That is not a confident face.
He's pretty horrified by that. I can't believe what I see right now. Oh, really? She thought, you know, she's a little bit behind.
And she made a very aggressive move here. OK. And make it very complicated.
OK. I think if he doesn't respond correctly, then he can collapse. I mean this one is kind of unthinkable as a human role player.
What if she knows everything about what's going to happen next? Here's our search data. We're going to search 50 or 60 moves ahead. That's the maximum. number of moves ahead that AlphaGo is looking from the current game position.
It's typically over 50, it's often over 60. In the games we see often around move 150, AlphaGo goes for the kill. We're at move 115 now, so we're getting to that critical point. We are all astonished just in the middle of the game because AlphaGo seems to be doing a much better job than we all thought it would.
I thought this adult would be leading the game. Comfortably, but it turned out that he's struggling at the moment, but I think eventually he will prevail I hope I was a you know more five to zero But now I'm not sure about that Oh, you can well I'll talk with you soon AlphaGo made a mistake. Oh my god, he made a mistake. That might be the first mistake. Kind of clear mistake that White made.
He made a mistake from the beginning. But if AlphaGo won, then AlphaGo could be the winner. I stopped talking because... I calculated it and it seemed like the alpha was a bit too high.
I didn't believe it. It seems like a mistake, but it's not a mistake. Right. If it's really calculated... It's so scary.
From a person's perspective, it's like playing with a person. Youngwan, it looks like you've got a count. Yeah, right.
What do you think? Yeah. Yeah, but why won a lot at this time?
If it's like this, why is this won by a lot? 2, 3, 4. Oh, I can't believe this. Wow, it's so shocking. I expected AlphaGo to win only one game.
I can feel his pain. He couldn't believe, he couldn't accept it. It takes time for him to accept the outcome. Maybe AlphaGo is very strong now, but he don't want to believe he will lose. It's just a goal-possession player.
We can't believe this. Because for us, it's something very, very far. It's not can be coming now. It's impossible. But the reality is now.
I think he resigned in a very polite way. This is like the piece black, but he put a white stone. I think he resigned.
Oh my gosh. Look at the time it's stuck. Wow.
I feel really good. I feel like I really believe in AlphaGo. Of course, it's natural that humans want humans to win. I mean, I think that's a natural response.
But AlphaGo is human-created, and I think that's the ultimate sign of human ingenuity and cleverness. Everything that AlphaGo does, it does because a human has either created the data that it learns from, created the learning algorithm that learns from that data, created the search algorithm. All of these things have come from humans.
So really this is a human endeavour. The battle between man vs machine, a computer just came out the victor. Deep mind.
It's computer program to the test against one of the brightest minds in the world and won. AlphaGo beat a professional player who has 18 Go World Championships under his belt. The victory is considered a breakthrough in artificial intelligence.
I was so surprised. I didn't think I would lose, but I was so surprised. I think the failure of the first half was continued to the end. I didn't expect to lose in the beginning, but I was so surprised.
I've won several world championships, and I've experienced a lot of things differently. I didn't want to be shaken by the fact that I lost one. I think there's no such thing.
To be more precise, I think it's 5 to 5 now. I would like to express my deep respect to the two programmers who made AlphaGo and the other programmers. In research, we normally work to produce an academic paper.
It gets published and maybe we get to talk about it in a conference, if we're lucky. This is not normal for research. In fact, I've never experienced any media attention remotely close to this. So it's a special moment for us all and we're just enjoying it while it lasts.
In a matchup between man and machine, who wins? So far, it's the machine. In Seoul, South Korea, the artificially intelligent computer defeated the global champion in the ancient Chinese board game Go.
Lee Seedol. lost the first matchup, but he's got four more chances. These people are the ones who are doing this.
They're the ones who are doing this. We don't have any responsibilities, just the man on the beach. Sorry.
It is a bit strange being the front cover and everything as a computer scientist. Normally you sit there in your corner, you code, nobody really knows about it. Perhaps heard the joke, how can you tell that a computer scientist is an extrovert and not an introvert?
If he's an extrovert, he looks at your shoes when he's talking to you instead of at his own. Like if you look at Aja, he avoids all the cameras like crazy. He's like, the game is finished and he's like out the back and back in his room. And I think a lot of computer scientists would be like that.
We were more about about doing our work than standing in the spotlight. Hello and welcome to game two, round two in the Google DeepMind Challenge, throwdown between man and machine. Game one, the machine takes down man.
Huge shock, headlines around the world. The reactions on the ground here. The folks in Korea were just stunned.
They estimate 60 million people watched the game in China alone. It's probably bringing up to maybe 80 million people watch this game worldwide. It was just incredible.
And if anything, today is probably even more of a madhouse. So Isidore knows today. He knows that this is just a really important game, right?
He's gotta win. Fighting! How do you feel?
Do you look nervous? Yes, I'm very nervous. I think I overslept yesterday.
Isidore on white. I think probably looking for a little payback. My impression is that maybe he underestimated AlphaGo and he's going to change his tactics.
Yeah, but Lee's playing much, much slower today. Yeah, which is kind of a shame, because he's playing at alpha speed, actually. It's hard to say who's better, who's better, or who's better. When I look at the two players, I feel like AlphaGo is very weak. You can't be too careful about that.
I don't think I can handle it. Well, if I say anything about AlphaGo that is not normal, it's maybe the way it handles a game when it thinks it is ahead. Yes, we're actually going to have a visit.
from one of the team, and we'll talk about exactly that point from the inside. Thanks so much for coming by, Thor. I really appreciate it.
Can you sort of share a bit of what is going on in AlphaGo? So AlphaGo has these three main components. There's the policy network, which was trained on high-level games to imitate those players.
And then we have a second component. We call this the value net. And it can evaluate the board position and say, hey, what is the...
probability of winning in this particular position. And the third component is the tree search, where it would look through different variations of the game and try to figure out what will happen in the future. So if we now take a position like this, first the policy network would scan the position and come up with what would be the interesting spots to play.
And it builds up a tree of variations and then employs this value net that tells it how promising is the outcome of this particular variation. So AlphaGo tries to maximize its probability of winning, but it doesn't care at all about the margin by which it wins. Okay, so when you see a slow-looking move, that's maybe an indication that AlphaGo thinks it has a good chance to win.
Yeah, that is a little giveaway. A little tell. We're looking for a tell.
Oh, looks like Lee is taking a little bit of a break. Lee Sedol, go to smoke. And AlphaGo just play, don't think about the open if it will be 0 or not.
So, Ahjap sees AlphaGo plays the move 37, and Ahjap pulls the stone in the board. Wow, really... I really can't believe it.
The value... That's a very surprising move. I thought it was a mistake.
When I see this move, for me it's just a big shock. What? Normally humans would never play this one because it's bad.
It's just bad. We don't know why. It's bad.
It's a little bit high. Yeah? It's fifth line.
Normally you don't make a shoulder here on the fifth line. So coming on top of a fourth line zone is really unusual. Yeah, that's an exciting move. I think we've seen an original move here. That's the kind of move that you play go for.
Hey. Interesting stuff. This fifth line shoulder hit.
I wasn't expecting that. I don't really know if it's a good or bad move at this point. The professional commentators almost unanimously said that not a single human player would have chosen move 37. So I actually had a poke around in AlphaGo to see what AlphaGo thought. And AlphaGo actually agreed with that assessment. AlphaGo said there was a 1 in 10,000 probability that move 37 would have been played by a human player.
So it knew that this was an extremely unlikely move. It went beyond. its human guide and it came up with something new and creative and different.
I am very much watching the game through these commentators. That's the way it works. So when they're confused, I'm certainly confused.
At the same time, I'm latching on to the fact that they are confused, right? That is an interesting moment. When everyone else is confused, who's not confused, right?
Besides the machine. I want to see Lee Sedar when he sees this movie. He's back, Lee is back.
AlphaGo is... I thought that PC was only a machine to calculate probability and win. It's not the moment to look at the tree. AlphaGo is creative enough.
It's a very creative tree that expresses the beauty of the beautiful and the beauty of the wave. Normally, you have to think about one, two minutes, no more. But this time, he thinks more than 12 minutes. The more I see this move, I feel something changed.
Maybe for human we think it's bad, but for AlphaGo, why not? Go is like geopolitics, like something small that happens here that have a ripple effect, you know, hours down the road in a different part of the board. The game kind of turned on its axis at that moment. This move is very special because with this move, all the stones played before is work together, is connect.
It looks like a network, link everywhere. It's very special, very special. I think that's the point, that it's a side of thinking again.
But what is creativity about? This help. I think it's a very meaningful book. This is a tough game for Luis Edel.
AlphaGo is just not letting Luis Edel do what he wants. Right. Black has almost 60 points. Oh, that's a lot. That's not a good sign.
Oh, oh. Luis Edel just slapped himself on the side of the head. Oh, wow. I think black's ahead at this point. It's looking good, isn't it, on that steady, um...
steady path now. I just see Lee Sedol. He lose so much.
Normally we reason before long time ago, but he won't try. He play, he play, he play. He just don't want reason.
Because then why, oh, he resigned. It looks like Lee Sedol has just resigned. There was this heavy sadness over that whole floor.
You could feel it during the game. I felt it during the game. And I'm leaving the commentary room to go to the press conference, and I was stopped by someone, another technology reporter.
At first, all he wanted to talk about was the technology and how great this was, but then even he kind of slipped into this moment of melancholy, where he was upset as well. I am quite speechless. I admit that it was a very clear loss on my part.
From the very beginning of the game, that was not a moment in time. that I felt that I was leading the game. You feel elated and you feel a little bit scared. There is something, I think, frightening to people about a machine that learns on its own.
For us, AlphaGo is obviously just some computer program. But looking at the commentary on the internet, I already saw the commentators call AlphaGo like he and she during the games, completely unconsciously. Which, AlphaGo is really a very, very simple program.
It's not anywhere close to full AI, and we already see that happening. So I found that very interesting. The tendency to anthropomorphize AI systems is one of the big obstacles in the way of actually trying to understand how AI might impact the world in the future. For example, the conversation is about what could go wrong, like what the risks are, then invariably you see this Terminator picture.
Every single time there are these red glowing eyes, right? We're really closer to a smart washing machine than Terminator. If you look at today's AI, we are really very nascent.
I'm extremely excited and passionate about AI's potential, but AI is still very limited in its power. I think that people are right to think that there is a danger that as we continue to improve these systems, that we might miss that threshold where we do cross over into danger. But the good news is that there are already people thinking about those dangers.
You know, there's a lot of talk now, and we're leading the discussion on this, that maybe there should be a kind of... cross-industry best practices working group or something where the leaders of the research teams in those organizations that you know the big ones that are working on ai ibm microsoft so on come together and make sure that ai is used ethically and responsibly i think what is important is that there is this community of people who are leading the cutting edge of ai who are interacting with academics and already are thinking about the the long term and how we can ensure that innovation is responsible as the power of these machines gets even greater. This is it folks.
Day three, game three, Lee Sedol, Go Master, back to the wall. He's down 2-0. He's got to win today to keep hope alive.
I heard that 4pros went to visit him to console him and to review Avago's... Console him? You think he's upset? I mean Isidore's upset. Yeah.
In the beginning, there was a fierce fight and then Avago played very well. So he secured a very early lead. From move 50, the win rate was very high. already and was climbing toward 100%. Oh, that's a great move.
What's your probability right now? I'm telling you what it means. Six. On 17? The sixth move for White is to win.
White wants to win, but the pressure is getting stronger and stronger. I think White will be the most disappointed today. The game is a game where you can lose if you try to win. He tried to fight directly in the game, but it's not his style.
When we change our style to play with opponent, normally it's very very bad. So it's a game more easy for AlphaGo. It's looking good for us.
Our black group is huge and it's got nowhere to go and it's going to be running around. I don't know how to describe the situation. If I were black, I would resign. We should admit that we are facing the strongest existence ever, ever in the Go history. There's no point in playing out the end game, and you're gonna lose, right?
Even if black can live here... Oh, he resigned. Okay.
Wow. Wow. You saw history made here tonight.
AlphaGo has won again. Three straight wins. Three straight wins.
He's won the match. When Lester is in the game, he looks like he's unhappy. It's not just that he loses the tournament.
It's just especially about this game. Because he doesn't play his game. I'm very hurt about this.
Very. But I can do nothing. You know, certainly I feel a bit ambivalent about it, given I'm a games player and, you know, Go is the pinnacle of board games. But I really like the statement one of the top Chinese professionals said of, you know, if AlphaGo wins, maybe we'll really start to get to see what this game's about. I couldn't celebrate.
It was fantastic that we'd won, but there was such a big part of me that saw this man trying so hard and... being so disappointed. I see internet many many people are talk about the LisaDoll maybe I don't play the best you know we are Go players okay sometimes in China, in Korea, in Japan we see Go we're like ah we are artists you know we play our He was best for goal. So, this gentle with Lizard.
He's very, very good player, great player. I'm in the room, I see Lizard, he won't win. He try everything. We just, we can't.
Just this, so. I want to say sorry to the audience. I think you guys were expecting a lot from the performance and the performance itself. I'm sorry I couldn't show you my best performance. I've never felt this much pressure or pressure.
I think I'm not good enough to overcome that. This was unbelievable. No matter what the situation is, I lost 5-0, not 3-0. I lost 5-0. This is a huge scratch to my pride.
I'm sorry to those who cheered for me. I know that Lee Se-dol has the strongest heart among the people I know. Nevertheless, I can see how hard it is to fight an invisible opponent.
There are still four or five countries left. If we can show the same level of fighting as the Isedol team, I think we can win. You could see that he was more relaxed after he had lost three games in a row. That said, the stakes were still high.
You know, in the end, it isn't about pride. I don't feel like confidence, but I feel light. Can Lee Sedol find AlphaGo's weakness? Is there, in fact, a weakness?
I was thinking I just pulled a plug. You just pulled a plug. Everyone's still cheering for Lee Sedol.
And then... Yeah, it's not going very well. It feels pretty good for Black at this point. It feels pretty good for Black, yes. I want Alistair to play his game.
Because for the moment, he tried many, many things to play with AlphaGo, to understand AlphaGo, but he never tried to play himself. I don't know. I told many many times AlphaGo looks like the real mirror.
When you play with AlphaGo, you feel very strange. Look like you and all the time naked. So first time you see this, you want to see because oh, it's me, real me? And more than more, you need accept. Oh, it's real me.
So how now, how I can do? It's developing into a very, very dangerous fight. This is really Lee Sedol's type of game.
He likes this kind of fight. White has to find something inside Black's territory. I think he is already planning on trying something. So you believe in Isidore's ability to live in the very small area, right?
Yes, yes. There is a little potential there. And I think maybe. Maybe he's gonna try to do something.
He said all magic. Oh, what would be the magic move? Lee said he's running short on time, but he's gonna have to use up all his time.
He's just burned like seven or eight minutes just on this move already. I don't see anything. We can't find anything.
Let's try to drink. I don't think I can. What is Lisa doll up to here? Yeah, he's really concentrating. He really is.
Look at that. Lisa is very patient. He waits, he waits, he waits this moment. I feel something look like the wolf. Wait in the forest in the winter.
It's cold. It feels very, very cold. But I need patience.
But the moment coming, it go out to attack. This is the edge of the game. Oh, look at that move. That's an exciting move.
Oh, he found the wedge. Whoa. It's gonna change the equation because now black cannot escape. That would be so cool if that works. Ooh!
AlphaGo has just played something maybe unusual. You know, I'm not actually sure what AlphaGo is trying to do here. What's that about?
I don't really understand it. Well, well... That was a...
sharp drop in the lead? That's the sharpest drop. in win rate received.
It's up to 8%. That deals the jewelry, I get that. Wow.
This could be that it actually can't find a way through. I think this is, it's looked far enough ahead to see that it doesn't work, and now maybe it's on tilt. I don't know. There comes the... There comes the...
It looks like it's fallen off the cliff. Yeah, it's made a mistake. Did anything strange happen in the... No, it all looks normal. Well, there we go.
We can definitely say, well, there's weakness. We can definitely say it's a mistake. I felt this mixture of this sinking feeling in my stomach where I was wondering if AlphaGo was becoming delusional in this situation, where I could see that it was starting to play strangely. And at the same time, relief of Lucidal, that he was actually in with a chance now. I just like trying not to look horrified.
I knew after move 17, after like 10 or 20 moves, I saw AlphaGo strange moves. I know AlphaGo somehow became crazy, but I didn't realize why. Whoa. We searched 95 ahead at that point? At the point where we made a mistake?
I think there's something went wrong. That's the longest reception time I've ever seen. Yeah.
I think it's like it's searched so deeply it's... It's got tile. I do get the impression that AlphaGo has sort of gone off on a tangent.
What is it doing there? Well, maybe it has a master plan. No.
It doesn't even think it has, does it? So it knows it's made a mistake. It starts evaluating it the other way. Look, look, look.
Lisa's confused. What is it doing? That's not a I'm scared confused.
That's a what is it doing? What is going on? You know, you asked if it was a bug.
I've said before that if... Yeah, right. If DeepMind has figured out how to write code that doesn't have bugs, that is a bigger news story than AlphaGo. Are you kidding me? Hmm?
Literally this next movie we're going to play. I think they're gonna laugh. I think Lee's gonna laugh.
Oh? Oh? Oh? What is this? What is this?
I think something's wrong. I don't really know what AlphaGo is trying to do here. That's the understatement of the app. Is it mouse misclick? Nope, nope.
That's the move. Aja Hong makes no misclicks. Isidore is very confused.
These are not human moves. This move also, sort of inexplicable. I mean, those, you can clearly call those mistakes. Yes, of course.
This is the first time in the four matches that we've seen moves like that. Right. Oh, the value dropped even more.
That's weird. We're like 45%. White. White.
White. Sweetie. Come on, lose it all.
This won't be the world. I'm right. I was so confident that this Black Moya would just be consolidated by Black and that there was nothing there and somehow he's just a... Erased it all, like it's gone. So he's found his weakness, that wedge move probably surprised him.
The wedge? Yeah. AlphaGo seemed to have such a good game and then this whole sequence in the center sort of changed that. He pulls off a miracle.
He managed to make it so complicated that the artificial intelligence doesn't evaluate it correctly. I think it looks like Ajahn Huang, maybe sitting there, also knows that the game is over. I can't, I can't, you know, stop smiling.
Did he smile? No, I haven't seen him smile yet. I think he is still checking. He's so serious.
He wants to be careful. He cuts off a lot. He really does his best.
I think he fixed something like, you know, this possibility of burden. How far go with the percentage? What's the percentage? Percentage is 18.2.
What does it make? Wow, she's very present. Thank you.
AlphaGo resigned. Looks like AlphaGo has resigned. Wow.
The most amazing game, I'm almost going to tear up, was game four, where he comes back and wins, right? Yeah! He's so good! I heard that people cheered for me when I lost to AlphaGo. The reason is clear.
A little bit of helplessness, fear. I was wondering if I was such a weak human being. As a winner, I think I've still kept my promise.
It's going to be really hard to win the AI in the future. But... You've won that one game?
Ah, that's enough. This is enough. I heard from so many people saying, you know, they were running out in the street.
They were so happy. You know, they were chanting, they were celebrating. Especially after the deal.
At the time, it seems to be hopeless that the end of the world is coming, but we see the light. Usually, I'm happy when I win, not like my colleague wins. But this time, it felt like my win.
I never thought I would win this much. I just couldn't believe it. I just couldn't believe it. Thank you. I think it's my first time to be congratulated after winning a game.
I can't be this happy after winning the first game after 3 games. My question is about number 78 move. Chinese taco player Gu Li said it was a God's play.
It does leave me a little bit in awe of the human brain's power, in particular Lee's amazing ability to cause AlphaGo problems and find something seemingly out of nothing. And so we really want to understand what had happened. So it really was Lee's move.
The key is the centre. If the centre goes on, there's no chance. We were winning before this.
If AlphaGo thought it was winning and it couldn't convert that win, that means it wasn't actually winning. Otherwise, it's a win. Otherwise it would have had the right responses.
Would we have played it? What probability does it give it as white? 0.007% of our interest.
Is that in the position that he played? That's when he played it. I see. So we thought this was the one that's impossible.
Yeah. The value's really low. So the God move was literally a God move because we believe that only one in ten thousands humans would have found that move. It doesn't come to mind as the top five, in the top five morphs that people would consider.
Unless you're Lucy Dole, he said it was the only one. And he thought it was the only one, yeah. Who might be the one that's now. Welcome back to the Four Seasons. World Champion Lee Sedol went looking for AlphaGo's weakness in Game 4, and he found it today, last round.
He's looking to see if he can repeat that. We'll see what happens. The place is a madhouse downstairs. Maybe as many, if not more journalists here today than there were for game one. We all were really excited for the fifth match to see what would happen.
Was this gonna be a 3-2 thing, or was it gonna be a 4-1? thing and you know there's a different message with both of those for me i don't think alpha go will make the mistake again but who know maybe it will come in and the listener have now have more confidence about this look like maybe lucid find the key to open the alpha go So, it could be that AlphaGo made a mistake. You can look forward to it.
from game four yet. Yeah. Still to do this?
So White has to do something about this idea. Oh, no. I hope he doesn't play one in.
No, he needs to play that. That's a pass to lose. That's a pass to lose. That's a pass to lose. So it's a pass to lose.
Him play some move is bad move. I feel something, oh, maybe it's weakness come back again. Are we seeing another short circuit or is there something? What's I think it could be a kind of a misreading.
And we're you're pretty comfortable with saying that it is good for at least at all good for at least no. There's no reason for white to be playing that move. It's a bad move and in some cases it's going to lose a point too. It is exactly the same thing.
91% certain now. Someone was telling me that maybe it had more to do with the draw. The whole game we thought that AlphaGo was wrong about the board position. We were super worried that, oh it's going to play garbage, it's going to be like losing in a very embarrassing way. And this continued for the whole game.
But this is a bit weird. When did they play that? Just now. So I don't know. I mean, it doesn't...
Well, that we don't know, but we're not good enough goers to know that, right? Yeah. As it turned out, none of us know Go well enough to accurately judge what AlphaGo is doing.
Why is he winning, though? This one looks like White is winning. We all say some of AlphaGo moves are so weird and strange and maybe mistakes.
But after a game is finished, we have to doubt ourselves. our judgment. AlphaGo making another kind of nonsensical throw-in.
I'm not really sure what that's about. This is what 10 or maybe 11.play looks like. It looks weird, and we don't quite understand it.
I think it is important to study more about Alpago's mistake-like moves. Then maybe we can adjust our knowledge about Go. To me, the most amazing thing to come out of my understanding of Go as a result of watching AlphaGo play are the information that I get.
infamous slack moves. Well, there's something strange about the way it's playing, because it's playing some moves that are not really necessary. Right.
A slack move is a move that looks lazy. You can see these other better moves, and AlphaGo is rejecting them. But what I think AlphaGo is teaching us is that we've been using score as a proxy for chance of winning.
So the bigger my margin of territory, the more confident I am that I'm going to win. And AlphaGo is saying, no, no, no, it shouldn't matter how much you win by. You only need to win by a single point.
Why should I be seizing all this extra territory when I don't need it? The lessons that AlphaGo is teaching us are going to influence how Go is played for the next thousand years. Am I counting? Why won this game? How, by how much?
Are we talking two points? One and a half. One and a half points. One and a half, maybe half, one and a half.
Of course, Go is just a game. could learn important lessons from our computer be so successful at Go. Machines will have the capability not only to crunch through a huge amount of data but also to analyse it intelligently. Just as in the case of the Go games, the machine made moves that surprised even the experts and eventually the machine...
...will gain our confidence because we will see that very very often they make a better guess than we could have made as humans. What's amazing is that the creativity of the The creativity of the The creativity of the The creativity of the The creativity of the The creativity of the The creativity of the The creativity of the The creativity of the The creativity of the The creativity of the Unfortunately for Luis Adela, I think white might have a slight advantage here. How do you know the Korean ones called it? Someone...
Who told us? I don't know. Korean too? The Korean room has called it for AlphaGo.
We'd just like to know. The Korean room thinks AlphaGo is winning. Winning, OK. They're not one. That's what she said.
They're called AlphaGo. I have to go say it's gonna resign. I'm not only joking. Jesus, that was a part of the game. That was ridiculous.
That was very good. Red plate. No, not really.
I couldn't do it in too long. I saw all of your faces, I just couldn't make that. I've been waiting two years to do that.
I've been waiting years to do that. We don't know the score. We have no idea what's going on.
Two and a half. Two and a half. I feel like I'm giving up on this stone.
Yes, I touched the stone. I'm touching it now. After this, I'm fine. Except for that thing.
I'm fine. I'm fine. I'm sorry.
I'm looking at the camera. I'm fine. So wait that out.
I'm sorry. I'm sorry. I'm sorry.
I'm sorry. I mean, I think it's just really is a once in a lifetime thing. I would say it's the most amazing thing I've experienced. You know, for us, it's the culmination of a 20-year dream. It started as a pure...
research endeavor we just wanted to understand can neural networks play the game of go and from there it went on to a level that I never expected and I'm unbelievably proud of the team When I started doing artificial intelligence, it was four or five years ago, and I was really interested in that, but many people would discourage me. They would say, OK, there is no future. So actually seeing that within five years we are at... this stage right now, it's amazing by itself. Everybody say the name of the one.
Panda! M.G.! There are so many possible application domains where creativity in a different dimension to what humans could do. could be immensely valuable to us.
And I'd just love to have more of those moments where we look back and say, yeah, that was just like Move 37. Something beautiful occurred there. At least in a broad sense, Move 37 begat Move 78, begat a new attitude in Lee Cito, a new way of seeing the game. He improved through this machine. His humanness was expanded after playing this inanimate creation.
And the hope is that... that machine and in particular the technologies behind it can have the same effect with all of us. I remember hearing a talk by Kasparov who says that a good human plus a machine is the best combination. This is a unique experience. Nobody can have this experience playing with AlphaGo 5G like this.
So... I hope Lee Sedol can find something in this five game. Maybe some change in his game.
I see today he fight, fight, fight. It's very good. Big master.
Real big master. I felt like I found the reason why I was playing the bard. When people played the bard for me, I felt like that. I thought, if I could go further, I would be able to say, I really did a good job playing the bard. I felt that way.
I really can't forget that moment. It was really that kind of experience. Is this Ava's time?
But for all the story, maybe it's just a big game. We don't know. It's just when I play with AlphaGo, he show me something.
I feel beautiful. Just it. I see the world different before everything begin.
What is real thing inside the Go game? With this thing, I will change something with my game. Maybe he just can show humans something we never discovered. Maybe it's beautiful. Thanks for watching!