Lex Fridman With Magnus Carlsen: An Interview You Don't Want To Miss

In the latest episode of his popular podcast, the American computer scientist and artificial intelligence researcher Lex Fridman had a fascinating, two-and-a-half-hour conversation with GM Magnus Carlsen. The two discussed a wide variety of topics, such as who is the greatest football player of all time, Chess960, reaching 2900, who is the greatest chess player of all time, a new world championship format, and even love. Below are some of the highlights of this truly outstanding interview.

Greatest Football (Soccer) Player Of All Time

"I think it's pretty hard to make a case for anybody else than Messi for his all-round game. And frankly, like my Real Madrid fandom sort of predates the Ronaldo era, the second Ronaldo, not the first one. So I always liked Ronaldo, but I always kind of thought that Messi was better and I went to quite a number of Madrid games and they've always been super helpful to me down there. The only thing is that, like they asked me, they were going to do an interview and they were gonna ask me who my favorite player was and I said somebody else, I think I said Isco [Francisco Roman Alarcon Suarez - PD] at that point. It was like, okay, take two. Now you say Ronaldo. So for them it was very important, but it wasn't, it wasn't that huge to me."

"@MagnusCarlsen spoke to @vgnett about meeting the @realmadrid players: "It was super cool!" pic.twitter.com/8PylQc3GIo

— Tarjei J. Svensen (@TarjeiJS) 30. November 2013

Carlsen's Approach To Chess

"I feel like I've had two peaks in my career, in 2013-2014 and also in 2019. And in those years, I was very different in terms of my strengths. Specifically in 2019, I benefited a lot from opening preparation while in 2013-2014 I mostly tried to avoid my opponent's preparation rather than that being a strength. I think my intuitive understanding of chess has over those years always been a little bit better than the others even though it has evolved as well. Certainly there are things that I understand now that I didn't understand back then, but that's not only for me, that's for others as well. 

"I was younger back then, so I played with more energy, which meant that I could play better in long, drawn-out games, which was also a necessity for me because I couldn't beat people in the openings. But in terms of calculation, that's always been a weird issue for me. I've always been really really bad at solving exercises in chess, like, that's been like a blind spot for me. First of all, I found it hard to concentrate on them and to look deep enough. Usually you have to look deep and then when I get these lines during the game I very often find the right solution even though it's not still the best part of my game to calculate very very deeply."

I've always been really really bad at solving exercises in chess.

"For me it's more like I'm at the board trying to find the solution and I understand the training at home is trying a little bit too to replicate that. Like, you give somebody half an hour in a position like, for instance, you might have thought for half an hour if you play the game, but I just cannot do it. One thing I know that I am good at though is calculating short lines because I calculate them well. I'm good at seeing little details and I'm also much better than most at evaluating, which I think is something that sets me apart from others."

I'm also much better than most at evaluating, which I think is something that sets me apart from others.

Opening Preparation

"These things have changed a lot over time. Back in Kasparov's days, for instance, he very often got huge advantages from the opening as White. There were several reasons for that. First of all, he worked harder, he was more creative in finding ideas. He was able to look at places others didn't and, also, he had a very strong team of people who had specific strengths in openings that he could use. And he would also very often come up with them himself.

"Also at the start, he had some of the first computer engines to work for him to find his ideas, to look deeper, to verify his ideas. He was better at using them than a lot of others. Now, I feel like the playing field is a lot more level. There are both computer engines, neural networks, and hybrid engines available to practically anybody. So it's much harder to find ideas now that actually give you an advantage with the white pieces. I mean, people don't expect to find those ideas anymore. Now it's all about finding ideas that are missed by the engines, either they're missed entirely or they're missed at low depth. And using them to gain some advantage in the sense that you have more knowledge. It's also good to know that usually these are not complete bluffs, these are like semi-bluffs so that even if your opponent makes all the right moves, you can still make a draw.

"And also at the start of 2019, neural networks had just started to be a thing in chess and I'm not entirely sure, but there were at least some players even in the top events who did not use them or did not use them in the right way and then you can gain a huge advantage because a lot of positions they were being evaluated differently by the neural networks than traditional chess engines because they simply think about chess in a very, very different way. So, [the] short answer is: these days it's all about surprising your opponent and taking it into a position where you have more knowledge.

"My team does but personally, I try not to use [engines] too much on my own because I know that when I play you obviously cannot have help from engines and I feel like often having imperfect knowledge about a position or some engine knowledge can be a lot worse than having no knowledge. So I try to look at engines as little as possible.

Having imperfect knowledge about a position or some engine knowledge can be a lot worse than having no knowledge.

"AlphaZero would sacrifice a knight or sometimes two pawns, three pawns, you can see that it's looking for some sort of positional domination, but it's hard to understand. It was really fascinating to see. Yeah, in 2019 I was sacrificing a lot of pawns especially and it was a great joy. Unfortunately, it's not so easy to continue to do that. People have found more solid opening lines since that don't allow me to do that as often. I'm still trying both to get those positions and still trying to learn the art of sacrificing pieces."

I'm still trying both to get those positions and still trying to learn the art of sacrificing pieces.

Chess960 (Fischer Random)

"Some of the positions obviously are a lot more interesting than others. In some of them, it appears that if you don't play symmetrically at the start, then you're probably going to be in a pretty bad position. That's the thing about chess though. So let's say White opens with 1.e4, which has always been the most played move. There are many ways to meet that, but the most solid way of playing has always been the symmetrical response with the 1…e5. Then there's the Ruy Lopez, there's the Petroff opening and so on and if you just banned symmetry on the first move in chess, you would get more interesting games or you’d get more decisive games. So that's the good thing about chess: we've played it so long that we've actually devised non-symmetrical openings that are also fairly equal.

"Symmetry is a good default and it's a problem that by playing symmetrical, armed with good preparation in regular chess, it's just a little bit too easy, it’s a little bit too drawish. I guess if you analyzed a lot in Chess960 a lot of the positions would end up being pretty drawish as well."

Reaching 2900

"Right now my rating is 2861, which is decent, like I think that pretty much corresponds to the level I have at the moment, which means in order to reach 2900, I would have to either get better at chess, which I think is fairly hard to do, at least considerably better. So what I would need to do is try and optimize even more in terms of preparations, everything. Not necessarily like selecting tournaments and so on, but like just optimizing in terms of preparation, like making sure I never have any bad days.

"Basically I can't f*** up ever if I want to reach that goal and so I think reaching 2900 is pretty unlikely. The reason I have set the goal is to have something to play for, to have a motivation to actually try and be at my best when I play because otherwise I'm playing to some extent mostly for fun these days.

I think reaching 2900 is pretty unlikely. The reason I have set the goal is to have something to play for.

"I love to play, I love to try and win, but I don't have like a lot to prove or anything, but that gives me at least the motivation to try and be at my best all the time, which I think is something to aim for. So at the moment I'm quite enjoying that process of trying to, yeah, trying to optimize."

Not Playing The World Championship

The world championship became a topic when Fridman asked: "What would you say motivates you in this, now and in the years leading up to now, the love of winning or the fear of losing?"

"So for the world championships, it's been the fear of losing for sure. Other tournaments, the love of winning is a great factor and that's why I also get more joy from winning most tournaments than I do for winning the world championship because then it's mostly been a relief. I also think I enjoy winning more now than I did before because I feel like I'm a little bit more relaxed now and I also know that it's, you know, it's not gonna last forever, so every little win I appreciate a lot more now. And yeah, in terms of fear of losing, that's a huge reason why I'm not going to play the world championship because it really didn't give me a lot of joy. It really was all about avoiding losing.

"When I sit down at the board then it's mostly been fine because then I'm focused on the game and then I know that I can play the game. It's the time like in between. I feel like losing is not an option because it's the world championship and because in the world championship there are two players, there's a winner and a loser. If I don't win a random tournament that I play, then, you know, I'm usually, it depends on the tournament, I might be disappointed for sure, might even be pretty pissed, but ultimately, you know, you go onto the next one. With the world championship, you don't go on to the next one. It's like, it's years and it also has been like, it's been a core part of my identity for a while now that I am world champion and so there is not an option of losing that.

"The only world championship after [2013] that I really enjoyed was the one in 2018 against the American Fabiano Caruana. What made that different is that I'd been kind of slumping for a bit and he had been on the rise. So our ratings were very, very similar. They were so close that if at any point during the match I had lost a game, he would have been ranked as number one in the world. Our ratings were so close that for each draw they didn't move.

The only world championship after [2013] that I really enjoyed was the one in 2018 against the American Fabiano Caruana. 

"And the games themselves were very close. I had a winning position in the first game, then I couldn't really get anywhere for a lot of games, then he had a couple of games where he could potentially have won, then in the last game, I was a little bit better and eventually there they were all drawn. But I felt like all the way that this was an interesting match against an opponent who is, at this position, at this point, equal to me. And so losing, that would not have been a disaster because in all the other matches, I would know that I would have lost against somebody who I know I'm much better than, and that would be a lot harder for me to take.

"That's why it's also been incredibly frustrating in other matches, like when I know when we play draw after draw and I can just, I know that I'm better, I can sense during the game that I understand it better than them, but I cannot, you know, I cannot get over the hump.

"After the last match, I did an interview right after where I talked about the fact that I was unlikely to play the next one. I'd spoken privately to both family, friends, and of course also my chess team that this was likely going to be the last match. What happened was that right before the world championship match there was this young player, Alireza Firouzja, he had a dramatic rise. He rose to second in the world rankings. He was 18, then, he’s 19 now, he qualified for the Candidates and it felt like there was at least a half-realistic possibility that he could be the challenger for the next world championship, and that sort of lit a fire under me. I liked that a lot, I loved the idea of playing him in the next world championship. Originally I was sure that I wanted to announce right after the match that this was it, I'm done, I'm not playing the next one, but this lit a fire under me.

"So that made me think, you know, this actually motivates me and I just wanted to get it out there for several reasons to create more hype about the Candidates, to sort of motivate myself a little bit, maybe motivate him. Also, obviously, I wanted to give people a heads-up for the Candidates that you might be playing for more than first place. Like, normally the Candidates is first place or bust. It's like the world championship.

"And then, Nakamura was one of many people who just didn't believe me, which is fair because I've talked before about not necessarily wanting to defend again, but I never talked as concretely, as seriously as this time. So he simply didn't believe me and he was very vocal about that, and he said, nobody believed me, no other players, which may or may not have been true. And then yeah, he lost the last game and he didn't qualify. But to answer the question, no, I already at that point decided that I wouldn't play, I would have liked it less if he had, if he had not lost the last round, but the decision was already made.

"At least over time, I'm a lot more proud of my streak of being rated number one in the world, which is now since I think the summer of 2011. I'm a lot more proud of that than the world championships."

I'm a lot more proud of my streak of being rated number one in the world ... than the world championships.

Different World Championship Format

"So I think 12 games, or now 14 games, that there is for the world championship is a fairly low sample size. If you want to determine who the best player is, or at least the best player in that particular matchup, you need more games. And I think, to some extent, if you're going to have a world champion and call him the best player, you've got to make sure that the format increases the chance of finding the best player. So I think having more games, and if you're gonna have a lot more games, then you need to decrease the time control of it, which in turn I think is also a good thing because in very long time controls with deep preparation you can sort of mask a lot of your deficiencies as a chess player, because you have a lot of time to think and to defend and also, yeah, you have the preparation. So I think for me to play, those would be the main things: more games and less time.

I think for me to play, those would be the main things: more games and less time.

"To their credit, this was suggested by FIDE as well: for a start, to have two games per day, and let's say you have 45 minutes a game plus 15 or 30 seconds per move. That means that each session will probably be about or a little less than two hours. That would be a start. Also, what we're playing in the tournament that I'm playing here in Miami [the FTX Crypto Cup - PD] is four games a day with 15 minutes plus 10 seconds per move. Those would be more interesting than the one there is now. And I understand that there are a lot of traditions, people don't want to change the world championship. That's all fine. I just think that the world championship should do a better job of trying to reflect who's the best overall chess player."

Ding Or Nepomniachtchi, Who Is The Favorite?

"Generally, I would consider that Ding has a slightly better overall chess strength. Nepo is even better at calculating short lines than I am. But he can sometimes lack a little bit of depth, like in short lines, he's an absolute calculation monster. He's extremely quick, but he can sometimes lack a bit of depth. Also recently he's improved his openings quite a bit. So now he has a lot of good ideas and it's very, very solid. Ding is not quite as well prepared, but he has an excellent understanding of dynamics and imbalances in chess I would say. He's very, very good at that. And understanding the, you know, the dynamic factors, as we call them, like material versus time especially."

On Garry Kasparov

On playing Kasparov in 2004 in Reykjavik:

"So I remember I was so tired after the blitz tournament, like I slept for 12 hours or something. Then I woke up, like, okay, I'll turn on my computer, I'll search Chessbase for Kasparov and we'll go from there.

Okay, I'll turn on my computer, I'll search Chessbase for Kasparov and we'll go from there.

"Before that, I hadn't spent a lot of time specifically studying his games. It was super intimidating because a lot of these openings, I knew I was like, oh, he was the first one to play that, oh, that was his idea, I actually didn't know that. So I was a bit intimidated before we played. Then of course, the first game, he arrived a bit late because they changed the time from the first day to the other, which is a bit strange, but everybody else had noticed it but him. Then he tried to surprise me in the opening. I think psychologically the situation was not so easy for him, like clearly it would be embarrassing for him if he didn't win both games against me, and then, I was spending way too much time on my moves because I was playing Kasparov, I was double checking everything too much. Normally I would be playing pretty fast in those days and then at some point I calculated better than him. He missed a crucial detail and I had a much better position; I couldn't convert it though.

"I knew what line I had to go for in order to have a chance to win, but I thought I'll play a bit more carefully, maybe I can win still. I couldn't, and then I lost the second game pretty badly, which, it wasn't majorly upsetting, but I felt that I had two black games against Kasparov, both in the blitz and rapid, and I lost both of them without any fight whatsoever. I wasn't happy about that at all. That was like less than I thought I could be able to do. So to me, yeah, I was proud of that, but it was a gimmick. I was like a very strong IM but had GM strength, I was like, it can happen that a player of that strength makes a draw against Garry once in a while. I understand I'm 13, but like still it felt a bit more gimmicky than anything. I mean I guess it's a good thing that it made me notice."

On working with Kasparov:

"We did work together in 2009 quite a lot and that cooperation ended in early 2010, but we did play a lot of training games in 2009, which was interesting because he was still very, very strong and at that time it was fairly equal, like he was outplaying me quite a bit, but I was, I was fighting well, so it was, it was pretty even then. So, I mean, I appreciate those games a lot more than some random game from when I was 13. And I maybe I just don't know what I'm talking about, but I've always found it at least based on that game, you couldn't tell that I was going to take his spot. Like, I made a horrible blunder and lost to a certain kid in the World Rapid Championship in 2018. And I mean, granted, he was part of the team that won gold in the Chess Olympiad [Uzbekistan's board five, GM Shamsiddin Vokhidov - PD], but he wasn't a crucial part. He barely played any games. It wasn't like I would think that he would become world champion because he beat me. I'm always skeptical of those who said that they knew that I was going to be world champion after that game or at all at that time. I mean, it was easy to see that I would become a very, very strong player. Everybody could see that, but to be the best in the world or one of the best ever, that’s hard to say."

Greatest Chess Player Of All Time

"I think I can make a case for myself, for Garry, and for Fischer. So I'll start with Fischer. For him, it's very, very simple. He was ahead of his time. But that's like intangible. You can say that about a lot of people. But he had a peak from 1972 when he was so much better than the others. He won 20 games in a row. Also the way that he played was so powerful and with so few mistakes that he just had no opposition there. So he had a peak that's been better than anybody like him. The gap between him and others was greater than it's ever been in history at any other time and that would be the argument for him.

The gap between [Fischer] and others was greater than it's ever been in history at any other time.

"For Garry, he's played in a very competitive era and he's beaten several generations. He was the best, well, he was the consensus best player, I would say for almost 20 years, which nobody else has done in at least in recent times. The longevity. Also at his peak, he was not quite the level of Fischer in terms of the gap, but it was similar to, or I think even a little bit better than mine. 

"As for me, I'm of course unbeaten as a world champion in five tries. I've been world number one for 11 years straight in an even more competitive era than Garry. I have the highest chess rating of all time, I have the longest streak ever without losing a game. I think for me, the main argument would be about the era where the engines have leveled the playing field so much that it's harder to dominate. And still, I haven't always been clear number one, but I've always, I've been number one for 11 years and for a lot of the time the gap has been pretty big. So I think there are decent arguments for all of them. I've said before and I haven't changed my mind that Garry generally edges it because of the longevity and in the competitive era, but there are arguments."

Garry generally edges it because of the longevity and in the competitive era.

In a way of answering a question about the most interesting chess games ever played, Carlsen came back to Kasparov as well:

"There are several games of young Kasparov, like before he became world champion. If you're gonna ask for my favorite player or favorite style, that's probably young Kasparov. It was just an overflow of energy in his play. Extremely aggressive, dynamic chess. It probably appeals to me a lot because these are the things that I cannot do as well. That it just feels very special to me."

The Beauty Of Chess

"What is beautiful to me is when your opponent can predict every single one of your moves and they still lose. It means that at some point early you're planning, your evaluation has been better so that you play just very simply, very clearly, it looks like you did nothing special, and your opponent lost without a chance.

What is beautiful to me is when your opponent can predict every single one of your moves and they still lose.

"In general, I try and create harmony on the board, like what I would usually find harmonious is that the pieces work together, that they protect each other, and that there are no pieces that are sub-optimally placed or if they are sub-optimally placed, they can be improved pretty easily. Like I hate when I have one piece that I know is badly placed and I cannot improve it."

About Love

"I'm not necessarily trying to find love. Sometimes I am, sometimes I'm not, I'm just trying to find my way. And my love for the game, obviously it comes and goes a little bit, but there's like, there's always at least some level of love. So that doesn't go away. But I think in other parts of life, I think it's just about doing things that make you happy, that give you joy. That also makes you more receptive to love in general. So that has been my approach to love now for quite a while, that I'm just trying to live my best life and then the love will come when it comes. And in terms of romantic love, it has come and gone in my life. It's not there now, but I'm not worried about that. I'm more worried about, you know, not worried, but more like trying to just be a good version of myself. I cannot always be the best version of myself, but at least try to be good."

At the time of writing, the Lex Fridman Podcast has over two million subscribers on Youtube. Starting in 2018, Fridman has created 315 episodes so far, about the most diverse subjects such as AI, science, technology, history, philosophy and the nature of intelligence, consciousness, love, and power. Guests have included Elon Musk, Joe Rogan, Jack Dorsey, and Mark Zuckerberg. You can find the podcast also on Apple Podcasts and Spotify.