Garry Kasparov on AI, Chess, and the Future of Creativity

Conversations with Tyler is produced by the Mercatus Center at George Mason University, bridging the gap between academic ideas and real-world problems. Learn more at mercatus.org. And for more conversations, including videos, transcripts, and upcoming dates, visit conversationswithtyler.com. I'm here with Garry Kasparov. Garry, of course, is the world's greatest chess player ever, but more importantly, I think of him as a man who has never ever shied away from a challenge. He is one of the least complacent people I know. Truly brave and courageous. He has a new book out. The title is "Deep Thinking." Gary, of course, himself is an extremely deep thinker in virtually all areas. The subtitle is "Where Machine Intelligence Ends and Human Creativity Begins." Whereas in Gary's previous book, well, winter was coming. This time around, it's a little more cheery. And at least in some regards, well, finally, Spring Has Arrived. Gary, thank you for coming on the podcast.

Thanks for inviting me. And you're right, yeah, this book is more cheery than the previous one, so I thought it would be a good idea to keep a balance. And I think in both cases I went against the dominant trend at the moment. When Winter Is Coming was written, and published, many people here in this country or in the free world, they had doubts about my strong statements about Russian dictator Vladimir Putin. Now, regarding this book, Deep Thinking, when I was writing the book, many people tend to believe Elon Musk, Stephen Hawking, and other pessimists who predicted doomsday scenario when artificial intelligence would inevitably take over our world. And I, this time, I took the opposite side, the optimistic side, and this book was an attempt to write a real story of artificial intelligence and its connection to the game of chess, why the founding fathers of computer science like Alan Turing and Claude Shannon had such an affection for the game of chess, and why they believed that solving the game of chess, or more likely making machines capable of beating strongest players, could help them to solve the problems of artificial intelligence, and why, by the way, they were wrong.

Let me ask you some of my questions. We'll start with topics related to your book. If we think of computer chess as a kind of project, along with AlphaGo, as somewhat of a referendum on human intelligence, that is, for decades, even centuries, humans have been trying to solve chess and go and play it better and understand how the games work. And then artificial intelligence comes along, computer chess, and they teach us how far off we were. So if you think of artificial intelligence as teaching us something about humans, what kind of mistakes we make, how smart we are or how stupid we are. How would you judge the results of these referenda?

Oh, um, I don't think the founding fathers of computer science had the same affection to Go or shogi because they were not familiar or had very little knowledge of these games that were played in other parts of the world. Now, chess was a game that for centuries had kind of an ultimate— sign, a mark of unique intelligence. And it was quite natural for them to look at the game of chess as the milestone for breaking up this mystery of human intelligence. Sure. And what we discovered in 1997 when Deep Blue was successful in the match against the current world champion, and I was world champion at that time, that chess could be crunched by brute force once hardware got fast enough, databases got big enough, algorithms got smart enough. But even while Deep Blue was victorious, it was anything but intelligent. Of course, you can start arguing about the definition of intelligence because by the definition of its output, the grandmaster level chess, Deep Blue was intelligent. But at this incredible speed, at certain points, Deep Blue could reach a phenomenal speed of— even today it's phenomenal, but 20 years ago it was mind-boggling— of 200 million positions per second. It offered us very little dreamt-of insight into the mysteries of human intelligence.

So you think humans are smarter than we might have thought?

No, it's just different because it's— any game, whether it's chess or we can move on with Go or shogi, still offers us a very limited area of human intelligence. Because at the end of the day, the way we try— we're trying to understand who is better in the respective game, it will be a result, win or loss. And machines have a decisive advantage because they, you may call, they have steady hand.

Yes.

Humans are vulnerable because we cannot keep the same level of vigilance required to play with the machines, even if we understand chess as well as machines do, even if we can survive the, the brute force of calculation. But at the end of the day, the pressure on a human player facing the machine is simply unbearable.

So when humans are playing chess, and here I mean professional chess players, what do you think is the major human systematic bias during the game? Is that they're not —patient enough, or they take too many chances, they don't take enough chances? What's the bias in us as human beings across the board?

I don't think humans at the highest level are playing machines anymore, so it's, it's, it's useless because you know the outcome. Oh sure, but Caruana sits down against Nakamura.

What biases do they exhibit in their play? Psychological biases?

No, but it's— are we talking about two humans? Two humans. No, that's another story, because if you have two humans, then you have two individuals, and most likely they know each other. I played, I guess, 182 games in the World Championship matches and many more games, hundreds of games against other top players in, in different competitions. And I knew almost all my opponents. I knew what to expect from them. I knew what to expect from myself. Human chess is a form of psychological warfare. So it includes a psychological element because you should know how to play a game against a very specific opponent. And sometimes, not very often, but sometimes you may look for, for certain moves that may not be, may not be the best purely from chess point of view, but they could create situation at chessboard that might push your opponent off balance. Now with machine, it's totally different. And it's the, it's, uh, the humans are facing an opponent that is not vulnerable to any psychological pressure, and moreover, opponent that doesn't care about what happened one move ago. Uh, in any human, human game, you always have not necessarily blunders or mistakes, but inaccuracies. Because if we are reaching winning position, the complacency is, is hard to avoid.

Uh, and so you think complacency is the main human bias you see in top chess games?

It's, it's complacency if you're on the winning side, but it could be also the desperation if you're on the losing side. Um, one of the greatest problems against, uh, against any machine is how to, to, to, to deliver the final blow, to, to win the game. Even if you lucky and you get a very good position, then again you have to play play with precision that is not required in human chess. So everything that you learn throughout your professional career— yes, it just— it's suddenly working against you. Um, and my experience, and I'm talking about best games played by top players, tells me that inaccuracies are inevitable. And any inaccuracy against the machine, even if you made already 45 good moves and 4 great moves, but the 50th move that is simply inaccurate may deny you victory that you think you deserve by playing brilliantly. You don't have to lose necessarily, but if your position is really good— but machine, let me just emphasize, machine, playing against the machine requires precision and concentration that I don't think it can be achieved by humans in a regular game of chess or another game like Go or shogi. I think it's, if we want to continue this experiment or just to put it in an extreme form, I would say that if we have Magnus Carlsen, the current world champion, offered an opportunity to play against the computer, uh, say 10 games match, uh, on the conditions that he wins the match if he wins one game. Yes. Which means he can play all the games without thinking about losing the game by making a mistake, or just, you know, he can play with— without this terrible psychological pressure. Maybe there's still a chance.

I would bet on the computer though, would you?

I know, but he has to win only one game.

I know, but to win a game against Ribbeka is very difficult. I said if it were draw one game, I might bet. No, no, no, no, no, draw I take.

No, if you're talking about drawing, then, you know, um, uh, and Magnus plays white, for instance. Let's take Magnus. I think it's, it's still 50-50 chance. Magnus is pretty good. So this drawing was— no, I'm talking about winning. So this is not drawing. You'll make few draws, that's for sure. I have no doubt about it. Now winning, if you give him this big handicap that he doesn't have to— he doesn't have to think about losing the games, it's all about concentrating and looking for the best moves, I think there's a chance. I wouldn't bet my bottom dollar. I still think it's below 50%, but there's a chance. There's a decent chance. It's not, uh, it's not slim to none that the best player may win. But again, we're talking about handicap event. So I think that's— the matches I played with Deep Blue and following matches that I played with Deep Junior and Deep Fritz and the matches that Kramnik played with Deep Fritz, they just, they ended the story of human-machine competition on equal terms. Close this chapter of the search of the mysteries of human intelligence in the realm of chess.

Now, a lot of humans don't play chess, but we're looking at a future where AI will make decisions about who gets a monetary loan, who is diagnosed as being schizophrenic or bipolar, how cars drive on the road, increasingly is controlled by software. The fact that the decisions of the software are not so transparent, and you see this also in computer chess, how will ordinary human beings respond to the fact that more and more of their lives will be quote-unquote controlled by these non-transparent processes that are too smart for them to understand? Because in your book you have emotional contact with Deep Blue, right?

Exactly, exactly. And I'm just, you know, I'm just telling you that it's inevitable. There's certain things that are happening and it's called progress. Right. This is the history of human civilization. I think the whole history is a process of a steady process of replacing all forms of labor by machines. It started with machines replacing farm animals and then manual laborers. And then it kept growing and growing and growing. There was a time, I mentioned in the book, people didn't trust elevators without operators. They thought it would be too dangerous. It took a major strike. And in the city of New York, that was equal a major disaster if you had to climb to Empire State Building with paralyzed elevators. I understand that today people are concerned about self-driving cars. Absolutely. Right.

But when do they swerve?

But no, but just imagine that there was a time I'm sure people were really concerned. They were scared, Steve, of autopilots. Now I think, you know, if you tell them that autopilots don't work in the plane, they will not fly because they understand that, again, in the big numbers, these decisions are still more more of the qualitative. And while I understand that also the fear of people who might be losing jobs, and they could see that machines are threatening their traditional livelihood, but at the same time, even these people whose jobs are on chopping block of automation, they also depend on the new wave of technology. To generate economic growth and to create sustainable new jobs. This is a cycle. And the only difference with what we have been seeing throughout human history is that now machines are coming after people with college degree, political influence, and Twitter accounts.

But say I'm at the self-checkout in the supermarket. If something doesn't work, I sometimes get fairly frustrated. If I'm with a human cashier, maybe a new person, they're just training, and they don't do it properly, I don't get frustrated at all. I'm very tolerant because it's a human. But the machine is much harder for me, just like it can be harder to play against Deep Blue than another human. So are we doomed to a future where maybe things work more effectively, but we're just frustrated all the time because there's not the human-to-human interaction to ease—

It's, again, it's this, that's hard for us to imagine, but we have been moving in this direction for a long time. So many things are just being done by machines already. So that now machines will be more actively engaged in human interaction. I can hardly imagine the supermarket with no humans at all. There will be other humans who will be supervising machines. And I think that's—

But you could run a supermarket with no humans. I think there are a few of them in Japan.

Yeah, but I still think at the end of the day, you need some humans Again, different kind of job because you understand exactly what you said, frustration and certain accidents. At the end of the day, you run supermarket to make profit. You don't want your customers to run away if they're unhappy. So you will need maybe some human machine psychologist to make sure that the customer satisfaction will be up to a certain level. Again, let's look at this historical process. So machines that replace manual labor, they have allowed us to focus on developing our minds. Now, more intelligent machines, more intelligent machines, they, I understand that they can, they will take over more menial aspects of cognition and will elevate our lives towards curiosity, creativity, beauty, joy. You can continue this line. We just have to understand that everything that we know how to do, machines will eventually do better than us.

Let's say there's a mortgage lending algorithm and someone charges that it's discriminatory. It lends too often, say, to white people and not enough to black people. And there's some kind of trial or regulatory inquiry. And the first thing everyone realizes is that no human can even understand the algorithm, much less judge it. What happens then legally? Or is it simply the case that so many new chains of cause and effect are so non-transparent to us we don't even know what to do?

My answer is I don't know. But that's exactly why we should not worry about the future because the problem with the future is if we know what's going to happen. Yes. That's always not good. We don't know. That means that in the situation that you described, we'll need new kind of lawyers, we'll need new kind of experts that will be able to solve these problems. There will be redistribution of jobs. Many jobs today, like drone operators or 3D printer managers or social media managers, they didn't exist 10 years ago, 50 years ago. No doubt that in 10, 15 years there will be many jobs, maybe the best-paid jobs that don't exist today. And we don't even know how these jobs will look like. And I think that's natural. All we have to do is to just to realize that this process is inevitable and we have to prepare us mentally, but also to have some sort of safety cushions to help people that will have great difficulty in adjusting. People that come from manufacturing jobs, manufacturing sector people that are older and just don't have the same knowledge of machines, and how to make sure the society will not be shaken badly by another dramatic shift in technology.

But say more and more jobs are redistributed toward AI. It can be very daunting to work with artificial intelligence. So with Deep Blue, they had Joel Benjamin, who, as you know, is a very smart guy. Is our educational system up to the shift, and won't it just mean the return to skilled labor just keeps on rising and income distribution becomes less and less equal? Or do you think somehow we will adapt? Oh, that—

now, now we're talking. It's like 38— no, you mentioned the word education. Yes, that's what we should concentrate, because I think it's quite obvious today that the educational system that is still dominating this area of knowledge, it's inadequate. When you look at a classroom, it hasn't changed since the University of Bologna in certain 14th century. So we still have the same school infrastructure. Or college infrastructure that normally has— it's like a one-way street. The information always comes from the teacher to the class, while this new generation is operating in two ways. It's always interactive exchange. And the biggest problem, and I've been talking about for quite a while, that we're still teaching very specific knowledge in the schools. And instead of teaching what, we have to teach how, because this knowledge may be redundant 10 years from now. We are preparing kids for, for the world that will change dramatically. By the way, we already know it will look different. So what's the point of trying to teach kids at age 10, 11, 12, uh, um, without recognizing the fact that when they, uh, when they finish college, when they will become adults looking for jobs, the job market will be totally different.

And most don't even finish college, right?

Maybe they won't finish college, but now we have to start thinking already, not even thinking, but even implementing new techniques to teach them how, to teach them certain basics, certain skills that that will be required in the world that we'll know will be quite different. And these new jobs will require definitely knowledge of computers and certain traditional skills that have been on demand for generations simply will vanish.

But won't AI just adjust much more quickly? AI progress in chess exceeded your expectations, the expectations of David Levy, many other people. It seems our system of education at K-12, the high school graduation rate has been almost stuck since the late 1960s. College graduation rate, you know, maybe is approaching 40%. It could take 30, 40 years to turn that around, even if we started doing everything right now. So, aren't, in a sense, some people still left in the dust by these differential rates of change? I mean, should we be so optimistic?

Yes and no. We tend to forget that every dramatic change in human history caused also major problems. Mm-hmm. Whether we like it or not, but the most dramatic technological advance happened at a time of war. Where people thought only about winning and having the best, and, and they didn't think about sacrifices or about comfort. I hope we can avoid repetition of this cycle, but naturally we can expect substantial problems for many people who will have difficulty of adjusting to new realities. So what we can do, just let's be objective, let's talk about it, and let's not terrorize them by the images of the Hollywood movies like The Terminator or The Matrix. It's a challenge that is ahead of us, but it's not the end of the world. There will be downsides, but there will be upsides. And let's see how we can manage our, our move into the future, because what is the alternative? We don't get to choose when and where technological progress stops. It's happening. So what I believe is the right approach is to understand that while machines can do eventually better everything that we know how to do, But there's still so many new things that only we can reveal. I'm always calling for us looking for new frontiers. There's no need to slow down. We just have to speed up. And there's so many new challenges. It's up in the space, down in the oceans. So many things that we're not doing now because they're too risky. And we like our comfortable life. Now, I guess artificial intelligence, though I'm not as optimistic about the development, I don't think it's going to happen overnight. It's happening, but it's not a speed that is threatening us as predicted by the doomsayers. But that will give us a push towards more risky adventures and forcing us to take the tasks that now we are rejecting because they are offering questionable returns with obvious risks.

You've been a pioneer in what's sometimes called advanced chess, freestyle chess, or centaur chess, where you pair a human being with a computer or a set of programs. Today, 2017, do you still think it's the case that a human paired with a set of programs is better than playing against just the single strongest computer program in chess? Oh, there's no doubt about it. But the human will make some mistakes, so the human will ask Stockfish, Komodo, Rybka, what's the best move, collate the different outputs, make some kind of judgment, explore some lines more deeply, put that against Rybka cluster. Is Ripka Cluster really going to lose many games? I think so.

Again, it depends on the qualification of the operator.

Sure, but the best operator in the world, whoever that may be, maybe yourself, maybe Anson Williams.

I, I'm not— by the way, I exclude myself from this category because I'm not a very good operator. I'm, I'm a very good chess player, but a great operator does not have to be necessarily a very strong player. What makes for a great operator? Someone who can work out sort of the most effective combination, bringing together human and machine skills. I reached formulation that a weak human player plus machine plus a better process is superior not only to a very powerful machine, but most remarkably, to a strong human player plus machine plus an inferior process. At the end of the day, it's about interface. Creating an interface that will help us to coach machine towards more useful intelligence will be the right step forward. And I'm a great believer that if we put— if we put together a good operator, still, it's a decent chess player, not necessarily a very strong chess player, running 2-3 machines and finding the best way to get this knowledge, to translate this knowledge into the quality moves against Rybka cluster, I would probably bet on a human plus machine.

A few questions about another one of your interests, Russia, formerly the Soviet Union. If we look back on centuries of Russian history, do you think there's something in Russian geography or demographics or geopolitics? What has it been that has led to such unfree outcomes fairly systematically? Where do you find the roots of tyranny in the history of Russia? Is it like a mix of the size of the country, its openness to invasion, its vulnerability, something about being next to dynamic Europe, on the other side China, or what is it?

Look, it's a long, if not endless, theoretical debate based on our interpretation of certain historical events. I'm not convinced with this— with these arguments about some nations being predetermined in their development and alien to the concept of democracy and the rule of law. And it's the reason I'm quite comfortable with this denial, that we can move from theory to practice. And while we can talk about history and certain influence of historical events to modernity. But we can look at the places like Korea, Korean Peninsula, right? The same nation, not even cousins but brothers and sisters, divided in 1950. So that's by historical standards yesterday, and we could see the difference. So one country is the demonstration of the worst Orwell's nightmare. It's a gulag run by a paranoid dictator, a family of dictators, 3 generations. And on the other side, the very same nation built one of the most sophisticated market economies and a very resilient democracy, democracy that people in Europe can envy. Just recently, they had a president impeached and tried and the CEO of the biggest corporation also being charged with crimes.

The two most important people in the country, right?

Exactly. So that's— and the same nation. Greece. And then obviously we can talk about two Germanys, uh, but I would rather look now to, um, China and Taiwan. Again, the same people, and we could see the difference. Obviously China is, is, is growing, but comparing to that dynamism and, and, and, and, and political stability of Taiwan, it's— and, and Taiwan is just, it's a small rocky island with 20+ million people versus 1 point whatever billion Chinese in continental China. And we could see that Taiwan is far more advanced, more dynamic. And the separation happened at the same time. Now, we can even look, because you just mentioned it's Asia and Russia, Asia and Europe. We can even look at former Soviet Union and, and, and try to analyze from our knowledge of 2017 what's happened after collapse of the Soviet Union and why certain, um, uh, some nations that are close to each other now moved in opposite directions. Let's look at Russia and Ukraine, and let's look not at the whole Ukraine, but just at eastern Ukraine. Eastern Ukraine is populated mostly by ethnic Russians. In the former Soviet Union, the borders between republics, they were very nominal, so people could move around. It was not a big deal. Even after the collapse of the Soviet Union, the official state border between Russia and Ukraine was respected, but people still could move around, and they didn't need special visas. So when we look at ethnic Russians born and raised in Kursk and Belgorod on the Russian side, and across the border, say, in Kharkov and Dnipropetrovsk on the Ukrainian side, there were people that could be hardly separated anything. They read the same newspaper Pravda, watched the same television, spoke the very same language, not even accents. But somehow in 2014, after Putin's annexation of Crimea and invasion of eastern Ukraine, we saw a huge difference. Most of ethnic Russians in Ukraine signed for Ukrainian army, fighting against Putin's invasion. Against the same Russians that came from other side? Again, it could be a long debate, but I would say that one of the main reasons is that Ukraine experienced in 1994 a gradual transition of power from one president to another after sitting president Leonid Kravchuk lost elections and walked away. So Ukrainians somehow got an idea that power is not sacred and the government can come and go and they can remove it by voting. And even despite the fact that Ukraine never experienced higher living standards than Russia, um, people realize that keeping this, this freedom, keeping this ability to influence their bureaucrats and, and government through peaceful process of voting and, if necessary, striking, far more effective than Russia's, quote-unquote, stability, where the same leader could be in charge of the country with his corrupt clique for a long, long time. How vulnerable is Putin right now?

And to the extent he is, is there a path toward democracy at the end of that process?

I don't know whether there's a path to democracy. So what I know is that it's the regimes like Putin's, they're most aggressive and most dangerous, because it's not even Soviet Union or modern China where you had, as in the USSR, or we have today as in China, a Politburo, Central Committee of the Communist Party. So a group of people, very powerful people, calling all the shots. Putin's Russia is a one-man dictatorship, and dictator doesn't care about the future of the country or about transition of power. There's nothing of interest for him if he's gone, because normally it's not the retirement. It's, uh, he loses power, he loses everything, including very often his own life. So he will be, he will be desperate in, in protecting this power and, and throwing all resources in his disposal to stay on the top. So that's why he is less sensitive for new wars and new very aggressive moves, international moves, because his power is not limited. Every group of people, even the worst of people, the mafia families, they always come to more balanced decisions because there's always a search for some kind of compromise. As in Iran, yes. Exactly. They have to look for compromise because they know that they're all part of the same system. And if the system goes down, they all, you know, could lose everything. Putin doesn't have to listen to it. Okay, he may listen, but he doesn't have to listen to these advices. And his powers are limited. By other elite groups. All he has to do—and he proved to be a very capable dictator—to make sure that these groups are fighting each other. And he could be kind of a judge on top of that. I always call it the spine of the system, which means that if he goes down, you don't expect just a simple transition of power. Will it be a democracy? The answer is no. It's not going to be democratic next day. But I'm strongly resisting the idea, popular idea in Russia and now in the free world as well, that Putin's collapse could lead to even something even worse. They say open fascist dictatorship, as if today in Russia we don't have a fascist dictatorship. No one, or no other group that will replace Putin in power, will have the same, the same ability to cause damage inside the country and outside the country. And most likely, these groups will try to communicate with the West, to find some kind of an understanding and to reduce the tension. So while I don't expect an immediate democratic reconstruction in Russia after Putin's collapse, I think that Putin's fall, Putin's demise will serve Russia well as well, and also will help to reduce the pressure with our neighbors.

What if the provisional government of Kerensky had made peace with the Germans in World War I before Lenin reenters the country? Could you imagine a path for Russia where today would be more or less like a normal European country?

Oh, I think that if we try to make some projections into the future, you can hardly imagine Russia surviving as a state in its current borders, though I'm not sure it's feasible at all with all the pressure from China and radical Islam from the, from east and from the south. But for Russia to find its spot in the family of civilized nations in the 21st century, it has no other choice but to work with Europe and the United States. And also, going back to this Russian-Ukrainian conflict, and now open war in eastern Ukraine, we can revisit early, early days of Russian history, going back nearly 1,000 years. And to remind people that there was another Russia. Actually, the cradle of Russian civilization was, was Kyiv's Russia, and the trade republics in the northwest of Russia, like Novgorod-Tomskov, that were— those entities were fully integrated into European political and trade space. And maybe what what is happening now in Ukraine and what will happen eventually in Russia, it will be the final showdown between Kiev's Russia and the pro-European Russia and the dominant tradition that was inherited by Moscow from the Golden Horde and Byzantine Empire. And Russia will reinvent itself and will go back to its original roots.

Now, we have a segment in all of these podcasts called Underrated or Overrated. And I name something and you give me your sense of how important it is to you, or if it's underrated or overrated. And I'll start with the Bulgakov novel The Master and Margarita.

Overrated or underrated? It's hard to say, uh, because it's my— it was my favorite book. It's still my favorite. Then it's underrated, right?

Uh, yeah. Okay, why is the novel important to you?

It's hard to explain for people who did not grow up in the Soviet Union, because it has so many unique aspects that were close to the hearts of people who experienced communism and the life under these draconian laws. But it's a unique combination of revisiting history and going back to the days of Jesus and telling us in a very original form the classical biblical story and projecting it to the life in Moscow in the '30s. And for those of us who read it in the '70s and '80s, we still could actually feel that it was— because it was about human nature. It was just— it reminded us that human nature doesn't change. With all the progress, people still have the same vulnerabilities, the same weaknesses, the same temptations. And this connection of times from the ancient days to modern days, for me, it still stands.

What from American fiction has almost the same power for you? Faulkner, Melville, Mark Twain? Oh.

You've become an American in some ways. Talk about the nature Human Nature, uh, All Kings Men. Yeah, Robert Pennell Ward. Yes, so that's the, uh, uh, Ides of March. Yes, uh, uh, Thornton Wilder. Again, this is— I'm always looking for very— yeah, the books that are just, you know, they, they are about something else, about days that, that, that were just in, in distant past. But again, there's this something that is so This is so close because we're humans and certain images from these books, you know, we could see in today's political arena.

The films of Andrei Tarkovsky, overrated or underrated?

Stalker, yes. It's a mixture. And I also don't like when you say films. So with the, uh, um, overall, uh, I, I, um, I'm not the biggest fan. So I picked up one or two movies, they are just— but as I mentioned, Stalker. But otherwise, I'm not, uh, I'm not the, the greatest fan.

The chess opening, the King's Gambit, 2.f4, underrated or overrated?

Um, It's a great piece of history, but it has no relevance today. So that's why, again, it doesn't pass my test of connecting the past and the present and the future. There's certain things that stay in the past because they were great moments in the game of chess. These gambits helped to create beautiful, beautiful games that we can still enjoy.

But, but Spassky beat Fischer with it once. That was a long time ago.

You don't think that could happen? Having, having totally lost position after move 15.

So that's the databases, you think? No way, ever again.

No, but it's just, it's the, it's this— so when I looked at these openings and I just came to the conclusion that they just simply don't stand modern technique of defense. And counterattack. It's just, they're not feasible. If you are looking for good game, professional game, they're not feasible. By the way, I was the one who reinvented and brought back the Scotch, the Evans Gambit. But I just have to realize that, okay, I don't play anymore, but the way that the chess has developed made most of these openings, not all of them, but mainly gambits, simply not feasible weapons at the top level.

If you took a top 10 player from the '70s, someone like Timman or Ulf Andersson, a very good player—

are you sure Timman is top player of the '70s?

No, top 10 player at times.

I don't think he was top 10. Well, the '70s, okay, it's just he was someone who was in the candidates, never champion.

And you give them a 2017 openings database that they can use and no one else ever has access to, how much stronger a player do they become?

It could be quite significant progress.

100 rating points? 200 rating points?

I don't know, but it will definitely add probably 100 rating points. So it's— the reason I'm thinking about your question, and I have some difficulty answering this question, is that for how long will it stay secret? Because people will start— Of course. It's the moment people see new ideas, they will start moving into this territory. So you can hardly imagine that you can sit on this database for too long and others just falling victims of your deep knowledge. But it will be quite a significant improvement. Look, in the match, if you're talking about the database given to a world championship, one of the participants of the match, say Korchnoi-Karpov, that could be— It would decide the match. That could be the decisive factor, yes.

Tolstoy, overrated or underrated?

Do I have the— You can answer whoever you want. Again, very uneven. So there's certain novels or just, you know, the short fiction, right?

Haji Murad is phenomenal.

There are many of them phenomenal. The problem was that sometimes there's so much philosophy, like in in War and Peace. That is just— I think it's outdated. Outdated, like The King's Gambit. I would— so Dostoyevsky, underrated. That's the— because it's all about human nature. That doesn't matter whether you live in Russia in the second quarter of the— the third quarter of the 19th century, or in the first quarter of the 21st century. It is the, the human emotions, the human temptations, the human reactions, they are still very much the same.

Back to chess. Now we both know Ken Rogoff, highly successful as an economist and a very good chess player. If we take most top chess players at the very top and somehow they could not have played chess, in other endeavors, how successful do you think they would have been?

No idea. No idea. No idea. But they're smart, right? But let's see though, how specialized is talent? But you talk, you talk to the very general audience. When you say a very good chess player, you have to expect top 5. You know, no, no, but Ken Rogoff, that— I mean, he's not top 5. No, no, no, no, but he was a very strong grandmaster. I mean, this is the man who played professional chess. Because the moment you say very good chess player, some people think, oh, maybe a good club player.

Very good by your standards.

No, no, he was very good, he was very good by any standards, so that's the— and that makes him unique, so whether he was top 100 or not, but somewhere there in the world at his peak in the mid-'70s. I don't think that the aptitude of playing chess is anything but aptitude for playing chess, so it's— there's no universal rule. Some people could move to other areas. I was very, very happy I was very lucky that I could, you know, I grew up with my mother who helped me to sort of get education outside of chess, and I always had various interests and looked for life in its entirety, and that's why after I stopped playing professionally in 2005, Then already it was the help of my wife, so I could move from chess to other areas, and now I have a very full life. Sure, you've done— Without playing chess professionally, but that's, again, that's not the rule. That's not happening with other players, though some of them I'm sure could do extremely well. By the way, some of them were very good business people, some of them made good investments. If you look at some of the British grandmasters or other, you know, players, in Europe. No rule, but I think that's the— chess helps to some of them to have so very— to find the right approach to their new endeavor. It's not— it may not be specific. For instance, I'm more of humanitarian, so I prefer history, literature. I I'm, I guess I'm a good speaker, but I'm not as good in number crunching. So it's very important to understand what you're good at. That's, that's what chess helps you. So to recognize, uh, what is your strengths and also to be very open about your own weaknesses.

Perhaps you may recall your 1999 game against Topalov, right? So a key moment in that game, you play Rook takes d4. Now, when you played Rook takes d4 and made that sacrifice, did you see at the time the later brilliant moves coming up— your c3, your Bishop f1, your Rook d7— or did you just feel there's going to be something that comes up? How much was it one or the other?

Not sure exactly at what point I saw the final combination, but, uh, but, uh, first of all, after playing Rook takes d4, I mean, I knew I was, you know, it's— if he took— if he would have taken the rook, which was a mistake, right? You should have played King b6. Exactly. But it's— and he got very good position. But I knew that this would already be good for me, though I was not sure about the outcome immediately. But within a move or two, actually, I saw the whole line. And it's the— I was quite lucky, but at the same time, you know, it prevented me from finding the forced win in the middle of the combination. Because at one point I made a second-rate move. Yes. Just because I already saw the whole line and it was so aesthetically pleasing. It was so easily— and I didn't just— and I just looked for obvious moves for Black. By the way, both me and Topalov missed a much better defense for Black that could prevent this combination, though of course the endgame—

you mean 30, Rook h to e8?

Yes, yes. Does Black hold with that? Yeah, you just— yeah, I think the endgame was still, uh, it's, it's much better for white though it's not winning. So the, the move, uh, pointed by Kavalec, Rook a7 immediately, so it would be, would be, would be winning by force. So it would be, it would be mate, uh, a mating net that black couldn't, uh, couldn't escape from, where black king couldn't escape. Um, but because I saw this, it's all the way up to the Rook d7, so I just, I just And you look at the time I spent, I played almost blitz because I was so anxious to reach the position and to demonstrate how far I could see. And also I was amazed by the beauty of this geometry.

So I didn't prepare you for this question, you simply remember, correct? The game against Topalov. I didn't come to you before this session and tell you to to go back and look at the game. You simply remember—

I have to confess that, you know, I, I don't remember the exact number of the moves. Uh, I can, I can reconstruct the game. Of course, looks like the game, but I'm not— when I said Rook h8, I was actually to figure out Queen c4. And of course, exactly, that then, then it would end up with the endgame when I, I would be, uh, I would be changed down, but I had rook and bishop and, and, and several pawns for his two rooks. And, uh, it didn't If I remember correctly, it's probably— it's winning, but it's not forced.

Would a top chess engine have played Rook takes d4 today?

I didn't check. Uh, um, it depends on the evaluation of King b6, because they will, they will, they will see immediately, even without calculating the whole combination, they will see immediately that White could force a draw. And by the way, that's what Topalov thought, that I wanted to make a draw, right? White would force a draw, while after King b6, every engine shows immediately that Black is better. I mean, slightly better, but the evaluation is, it's clearly in Black's favor. So that's why they wouldn't take the rook. Now, I don't remember whether rook takes the— I think at that time White had no choice but to take on d4, because otherwise, you know, the whole position—

then Black has much more space. Yeah, exactly, exactly. Now, given this, does this mean, uh, even very, very good chess engines will never play games as beautiful as human beings because you were so excited by this combination. In some ways it may have led you down inexact paths, but it nonetheless led you down those paths and we got the beautiful game that we did.

But now, now you moved back to these chess computers and there's certain things that people should realize because it's, I hate, you know, talking about these things. It's, Hussein Rashid is using a bird language because it's the, you're asking me questions and I'm not sure that 99% of our listeners, they understand understand exactly what we're talking about. Now, there's one thing for people to understand is that chess is, you may call, a mathematically infinite game. So the number of legal moves is more than the number of atoms in the solar system. So machines cannot solve the game. You cannot expect a machine to play e2 e4 at move 1 and announce mate in 16,455 moves. But machines could work, uh, the game of chess from, from the, from the end. So now we all, we know that machines solved, mathematically solved, all positions with 4 pieces, like king and queen versus king and rook. All positions with 5 pieces, all positions with 6 pieces, right? And now, um, 7 pieces. 7 pieces, it's the, it's, it's It's on the way. I don't think it's— I'm not sure it's all solved, but it's— that's— it's, it's, it's, uh, we're talking about, um, 100 terabytes. Yes. Now obviously 8 pieces will be already just insane number, and, and the game of chess is ultimate endgame with 32 pieces. So that's why it's— maybe machines will get to 8 or 9 moves, but that will probably end even for the immense computing power that you can expect in the next 5, 10, 20 years.

Will AI write beautiful music or is there something about—

No, but I want to finish this because this is what we discovered in this process. I mean, I wouldn't overweight our listeners with all these details. I don't want just to throwing them this, the mass information. It's amazing what people will enjoy that.

You'll be surprised.

But in some of these positions, like there are certain 7 pieces positions.

Yes.

When the win, and we're talking about a forced win, can be reached within 500 moves. Yes. Now, 500 moves. I remember I looked at some of the positions. Even at 6 pieces position, it's not intelligible. No, no, no, there's no— it's no intelligence at all. It's just, you know, pieces moving around, you know. This— I saw the position with, you know, with king, 2 rooks, a knight on one side, and king, 2 rooks, and, and on the other side. And, uh, it said mate in 490 moves, forced mate. Yes. Now I can tell you that even being a very decent player For the first 400 moves, I could hardly understand why these pieces moved around like a dance. It's endless dance around the board. You don't see any, any pattern. Trust me, no pattern, because they move from one side to another. And at certain points I saw, oh, but White even, you know, White's position even deteriorated. It was better 50 moves before. Now the question is, and this is a big question, So if there are certain positions in these endgames, like 7-piece endgames, that take, by the best play of both sides, 500 moves to win the game, what does it tell us about the quality of the game that we play, which is an average 50 moves?

It means we're clueless in the entire universe.

Exactly. So that's— it's, it's an interesting philosophical question, and I have to confess I don't know the answer.

Imagine how bad our politics would look.

But I don't know the— now, this brings us to another point. So maybe with machines we can actually move our knowledge much further and we can understand, you know, how to play decent games at much greater lengths. Now, going back to computer chess, and I don't know how much time you spend looking at the computer matches.

Mm-hmm, quite a bit. Quite a bit.

Now, I think you can confirm my observations that, uh, it is very— there's something strange in these games. First of all, they're longer, of course. That's— they're much longer because machines don't make the same mistakes. So they could play 70, 80 moves, 100 moves. So 70, 80 moves is a normal length, right? Still, by the way, it's still 70, 80 moves. It's still not fine. It's, it's, it's way, way below what we expect from perfect chess. Yes, that's the that tells us that machines are not perfect. But most, most of these games are decided by one of the machines suddenly— can I call it losing patience? Because you have a position that is roughly even, it maneuvers and they go around the pieces from one side to another. And then suddenly one machine makes a, you may call, human mistake. Suddenly it loses patience and it tries to break up. Without a good reason behind it.

My driverless car will do this someday. But that's the—

no, that also tells us that these machines, they— that's an interesting observation from also my experience, that machines also have, you may call psychology, the pattern in decision-making. And if we understand this pattern, so we can make certain predictions. So it's not all predetermined. There's so many things that can go one way or another. And that means that we will not be run by this mystical AI. It's not perfect. It's far from perfect. It's still— there's so many areas it doesn't know how to cover. Even setting aside a simple question whether the brains can function effectively separately from the body. So what does it mean that you have brains that are not part of our moving body. So can you separate it and still have the same effect? My answer is I don't know. And that's what I believe is good because that tells us that there's so much to learn in this process. And for those who are predicting that AI is just around the corner and is going to wipe us out, They don't know what they're talking about.

So it's more like a future of inscrutability and periodic surprises.

Yeah, but it's the, you know, but we also know that the doomsaying has been always a very popular pastime when it comes to technology. It's because it's easy, you know, it's us against them, the race against the machines, the war against the machines. You can sell it. And I think it's important that people will stop looking at this either from utopian point of view, oh, we'll just, you know, get together, we'll fly to Alpha Centauri with machines, you know, they will serve us, you know, in this endless spaceflight, or to the contrary, it's a Skynet that is going to kill all of us. It's a process. It's a process and we should be objective. This is not too optimistic, not dark pessimism, but something that is more human, which is, look, let's get objective. Let's assess our chances and let's realize this is a process of of developing for the human civilization. And compared to the fears that people had about machines in the 19th and 20th century, we're not pioneers.

Do we do our best thinking under pressure?

Yes, we do much better. I'm not so best, but we definitely perform much, much better. Some people collapse, but as a— Aronian. I know you don't like to say that. No, no, no, no, no. As a humanity. Yes. As a humanity, of course we operate much more effective under pressure.

Look at wars. A lot of mistakes, though, are made in wars. No, but exactly.

But this is mistakes always made. But we're talking about the performance. That's why we can look at the productivity. Look at what people can do, you know, by inventing new things. So this is the moment they're under pressure, you can expect that part of the group will perform way above expectations.

Last question: What has been your favorite city to play chess in and why?

Now it's simple because I don't play chess anywhere, anywhere else but in St. Louis, Missouri. So that's the—

but looking back, is it Moscow? Is it Linares?

Is it some city? I played more, more chess in Linares than anywhere else in the world, I guess.

And you've done very well there.

Yeah, I did well. Now, in terms of— perfect results, I would say Paris. I played a few events, but most remarkably, 5 rapid chess tournaments. And the city never distracts you? And I won 4 of them, and I was in the final in all 5. So that's the— in terms of, again, rating my performance, I don't think any city will come close to Paris.

And the city of Paris never distracts you when you're there? If you're there to play chess, you play chess.

No, look, Paris is always Paris, but again, the results were perfect, nearly perfect. So, and I'm just thinking about other places now, but I don't think that's— Is there a home disadvantage?

So Anand playing in Chennai, his whole entourage is there, he's a big hero in India, his family is around, he has to juggle competing demands.

He lost to Magnus and So. As well.

So sure, that's the, um, so no home court disadvantage in chess?

Uh, it could be advantage, could be disadvantage, because you know, it's, uh, it's a pressure. You play at home. Anyway, I never, I never played at home because my home was— except, okay, 1980 I played in Baku. It was a tournament when I made my second GM Grandmaster norm. Okay, but that was not such a big challenge. Of course, people wanted me to win, but I was much stronger than the rest of the field. It was all about me competing with Beliavsky for the first prize. Everyone knew I would be a grandmaster there. Big matches, when I played Moscow with Karpov or just other events, it was not considered to be home. So it's the— though I was not very successful in Moscow, when you look at, at, at overall performance. Um, uh, if you play at home and just think about Anand and Magnus, or you're a guest, so it's, it's a pressure on both players. It's just an extra pressure, and it depends who can handle it better. But since Magnus beat Anand both in Chennai and in Sochi next year, I don't think that the factor of the home factor, uh, or guest factor played a major role.

Oh, and one final question I forgot: who is the most likely challenger to Magnus Carlsen this next time around?

Considering the, the logic of, of the World Championship history, uh, Magnus should face opposition from a younger player or the same age. So, um, I would say there are 3 players that I guess could challenge him. Uh, uh, Maxime Vachier-Lagrave, I would rate him as the least probable among these three. Second, Caruana, and the most likely, Wesley So. The reason I put Wesley ahead of two others is that he has a phenomenal— he has phenomenal concentration, absolutely phenomenal. It's just— and it's very important these days, concentration trumps everything. Uh, it's, it, uh, I would put it on top of other things, though he's a very, very good player. And, um, I think that he might be among these three, he might be the most dangerous for Magnus because he has nothing left, nothing to lose.

He feels he's early.

It's just, it's because, because his concentration, um, uh, could, um, uh, overboard Magnus. Magnus, Magnus' concentration is what's all his strength. He He was always very concentrated, he could mobilize all his resources for the game. Now we saw Magnus could feel, again, it's by educated guess, Soh hasn't even qualified for the Candidates, he still has to win it, but against Wesley So, Magnus could have some serious problems in reading his opponent and also just massing up his energy, uh, Soh could reflect. Um, though I think it's match with Caruana could be also uncertain. And if Vashela Graf makes some improvements, more psychological improvements, he could be also dangerous. It's— he's the same age. Caruana is younger. Soh is even, even younger than Caruana. So I would, I would bet on these three as, as, as, as, um, one of the Magnus challengers, ranking them in the following order. So, Caruana, MVL.

Gary, thank you for the chat. It's been a pleasure. Thank you. Thanks for listening to Conversations with Tyler. You can subscribe to the podcast in iTunes, Stitcher, or your favorite podcast app. And if you like this podcast, please consider rating it on iTunes and leaving a review. Review. This helps other people find the show.