What should I ask Garry Kasparov?

I will be doing a Conversations with Tyler with him, no public event, podcast only.  Today by the way is his birthday, so send along some good questions as a birthday present to him, and a non-birthday present to me!

Garry’s forthcoming book Deep Thinking: Where Machine Intelligence Ends and Human Creativity Begins is just superb, and the podcast will be released around the time of book publication in early May.



Whose a better chess player: Trump or Putin?

Why does Nassim Taleb beat you badly and so often in Twitter wars? And why do you support jihadis in Syria?

Who's your favorite character in Game of Thrones?

Has he ever tried any of the board games from the resurgence from the mid-90's on, or even better, hex-and-counter wargames?

Would you have beaten Bobby Fischer at his peak?

Of course, Kasparov wouldn't know the answer to that. But it would be very interesting to see what he says. Great question.

met him before, he said he can't know for sure, but likely yes.

I once asked that same question to Karpov. He said that today champion will beat his predecesors, because in chess there is cumulative knowledge. But that no one was so ahead of his competitors as Fischer in 1972.

Take a smart young person, maybe around age 10-12. What are the chances this person learns chess and tries to become good at it now? Is that percentage too low, given the number of other options? What would you do to change that number?

O.M.G.!!! I am going to have a heart attack...

Ask Garry if he thinks ratings inflation exists in the 2700 level and above (of course he'll say yes) based on the fact the same people playing when they were younger had lower Elos than today, which defies Father Time (unless you feel--like marathon runners--older chess players play better with time, which is not true statistically), and, as a followup, what he feels the Elo change is (i.e. +50--my best guess--more Elo points)?

Ask him if he still practices secretly online, like Bobby Fischer did.

I bunch of other questions you could ask, besides the obvious political ones, but I'd like to know the above two.

Of course if you want to be a troll interviewee, which would make your guest a bit uncomfortable, you could ask Kasparov about this:

1/ Russian scientist/professor Anatoly Fomenko's view of history

and as a joke, ask Garry if he's invented this:

2/ Equivariant Kasparov theory of finite groups via Mackey functors https://arxiv.org/abs/1105.3028

But I don't recommend 1/ since it's almost like asking a Baptist fundamentalist--like computer scientist IM Ken Regan--about his creationist religious views. It's not really a question that's friendly, it's like questioning their religion. Best to stick to non-confrontational views.

+ ask: is chess art or sport? IMO: Art!! not Sport?!

+ ask: can gambling save chess, if it's a decentralized prediction market (like GMU professor Hansen is working on, at least from what I saw online a few months ago)? That way, you don't need permission from a state to gamble (traditionally the central stumbling block in chess gambling, btw NY state has ruled since WWII that chess gambling in Washington Square NYC is not gambling since chess is a game of skill not luck, which is somewhat not true). Or, will such decentralized gambling actually hinder the game of chess, by encouraging cheating?

Ask him how a chess sponsor like me can make money! LOL, I want to make money off some of the players I've helped in the Philippines. But, like being an opera sponsor, I don't think you do it for the money, more for the love of the game. However, it would be nice to back a young player and then have him turn into a Wesley So, and make 100x your money back!

There have been players like Botvinnik, Korchnoi and Anand who certainly aged well Elo wise and probably were not at their best when younger.

Most of the great players peaked in their 30s. The current crop of 20-something 2700 players is unprecedented, only 8 of top 45 active are over 40. It doesn't require "inflation" for ratings to rise @ top, today's world-class GMs have many more opportunities to play strong fields, many more tournaments than in '70s, '80s, even '90s. Ratings reflect performance, not strength, so playing overall stronger average opposition can yield a higher rating than was possible most of Kasparov's career - for several years, he & Karpov were the ONLY active players over 2700. today there are 45 - and five over 2800.

@Boris_Badenoff - IM Ken Regan and NY professor of computer science has said there is probably ratings deflation not inflation. And we all know Gaussian statistics and what happens when more chess players enter a pool (pile of sand gets bigger, so more outliers). However, I am firmly convinced, even though I agree with the above, that above 2700 Elo there is ratings inflation due to the relatively 'closed' nature of the pool. Strong chess players seem to agree with me, I've had masters tell me that the top 2700 players avoid open tournaments where they might lose to an ordinary IM or GM. Don't you think this might cause about 50 points Elo inflation above 2700? I think so. "Further research is required".

I will also add I agree with the studies done based on GM John Nunn's suggestion of the Carlsbad 1911 chess tournament and others that find today's tournament player, on average, makes fewer blunders (and hence is stronger) than yesteryear's. But this does not go to the "top 2700" chess players question, only to the issue that today's average active chess player is stronger than yesterday's, which I agree with.

Has playing a strategic game such as chess made you (more) paranoid than you otherwise would be? In chess, and especially at the level that you played, few events on the board were due to chance. They were the result of a deliberate action by your opponent following some plan. This is perhaps not the case in the world. People make more mistakes and there is more coincidence in the world than there is in high level chess. To what extent do you think that you over-interpret events as being the consequences of some plan put into action and how do you deal with this problem? <- This question is in relation to seeing the events in St. Peterburg as a false flag event. https://twitter.com/sahouraxo/status/848982714212634627

Another question:

Do you think that your talent has been wasted on chess? Clearly you are a very intelligent human being. Wouldn't you have contributed more to "society" / human welfare as say a physicist or a doctor? The marginal benefit of a slightly better chess player in terms of entertainment value are low. Certainly lower than the marginal benefit of say a physicist or a doctor. In general, how much talent do you think is "wasted" in chess? Wouldn't the world be better off with Magnus Carlsen the innovator rather than Magnus Carlsen the chess player?

It's a good question if you assume the skills are transferable, which I'm sure there is no consensus on that. A talented chess player won't necessarily make a good doctor, for example.

Intelligence is not a skill. https://en.wikipedia.org/wiki/G_factor_(psychometrics)

Another question:

1. In other sports we see increasingly more physical training. How important is physical fitness in chess? To what extent and at what level of chess does a lack of peak physical fitness become a hindrance to your play?

2. If you meet a smart kid who is talented at chess would you advise her to continue to practice to become a GM and eventually a contender or would you dissuade this kid from the attempt? Why yes, why no? (This is related to the question regarding the social value of chess).

3. How happy would you have been if you never became a world champion and were instead a top 10 player? How big of a difference does being the best mean to you?

4. Lawyers, so I've heard, tend to make bad spouses because they're trained to spot mistakes and can become overly critical towards their spouses. To what extent do you think playing chess has made you more critical when others make mistakes?

5. What's your best memory?

6. How far has your chess ability declined since you stopped playing competitively and aged? Do you worry that one day you will not be able to play at all?

His playing strength ought to be rusty, but he held his own in a blitz event held around one of the elite events in St Louis last year, without a couple errors pretty clearly from lack of practice, he might have won it outright.

Dictatorships tend to create a set of wrong incentives ending up promoting crooks and cronies while sidelining capable and honorable people and its legacy usually survives its historic leaders, in that sense how realistic is the idea that Putin's removal of power will give us a vibrant and democratic Russia in the foreseeable future?

Good question.

How pessimistic are you, Mr. Kasparov, about Russia's ability to change its longstanding political/institutional trajectory, which seems to involve some kind of powerful warlord or Tsar-like ruler, a strongman in other words?

why are men's and women's chess in separate divisions? One can argue for separating the genders in the 100 yard daxh, but why in chess?

Very interesting question. In poker women and men play along.

In poker, it takes longer for skill differences to become apparent, so differently skilled players can have fun playing together for a while. But they're still present. There are no women in the world top 50.

The top female chess player isn't in the top 100 players currently, right?

Honestly, I'm not sure poker, chess, and the 100 yard dash are all that different in the extent they favor men. Different, yes, because the best female sprinter wouldn't crack the top few hundred male times but would probably make the top 1000. I can't easily figure that out online.

I'm a bit surprised by that.

I just looked up high school 100m rankings. Last years women's Olympic champ ran a 10.71 which would have been 364 among u.s high schoolers.


Yeah, there's a bigger gap, but not astoundingly bigger. The best woman in chess is around #150. In the 100m she's around #1500. And I'd venture a guess that more people record 100m times than play tournament chess, so the difference is less than it appears.

I'm not saying women are great at track, I'm saying they're almost as far behind in chess and that's surprising.

This is one of those things where if you have to ask...

They aren't. There are women's chess tournaments, but no men's chess tournaments. Lat I checked (it's been a long time), the "women's chess champion" was not the best female chess player, but the bets (and maybe second best also) female player refused to play in women only tournaments.

At one point, the backgammon engine, Snowie, was able to comfortably beat professionals in standard positions, and had good evaluations of its winning chances, so did well in the 'doubling dice' wagering element of game. However, human players discovered one position that it totally misevaluated and were able to win money against it by steering it into this situation and wagering heavily that they would win [1]. Are there analogies for other fields adopting narrow machine intelligence? I am watching a lot of just eat takeaway delivery drones being tested in my borough of London, and can see them being terrorised by kids leaving school as soon as the observers aren't following 100 yards behind. I've also seen plenty of colleagues gleefully screwing with transparent productivity measures. I also observe commuters happily scamming ticket barriers and self-checkouts when they would be noticed/feel shame when dealing with people.

How can issues this be addressed? Increased randomness, or opaqueness of engines? Human supervision? Different procedures/governance rules?


Also, does he think there could be a role for computers to generate sharp and human-like middle games for professionals to finish, negating opening preparation and making professional chess more interesting?

My proposed question: How does your (Kasparov's) unpublished 'The Blueprint' collaboration with Peter Thiel & Max Levchin intersect with Tyler's themes in 'The Great Stagnation', 'Average is Over' & 'The Complacent Class'?

Is the human mind best thought of as algorithmic? Is creativity algorithmic? Does he have thoughts on David Cope's algorithmic music? I ask as I find results from algorithmic music to be informative of the creative process.

Has he participated in building AI or algorithms for creative tasks? Any insights from the process?

Does he ever regret his commitment to chess given the opportunity cost of his intelligence?

Or is the opposite true -- did his intelligence come from playing chess?

"Surely, you're joking Mr. Thorson"

Does he see a realistic path to democracy for Russia? Putin's bloodless exit, rise of serious (not Navalny-like) opposition, etc.? Or should we expect a christian (unless Chechnya, where s/christian/muslim/g) fundamentalist dictatorship taking hold there for the foreseeable future?

Several years ago I had dinner with Garry and he argued that if the price of oil dropped below $100 for a sustained period of time, Putin would not be able to hold on to power. So, what happened?

How much does he feel Chess is based intuitive skills and not learned efforts ? It would certainly be one field where 10,000 hours wouldn't guarantee you will get a GM norm.
I tend to feel it is 33% inspiration ; 67% perspiration. And without both elements you don't reach the 99% level.
I always wonder . I am a significantly below average player ; when I was sixteen I drew a game against a player in a low-level college tournament. A year later that guy was in the top 20 in the country , could it have been by effort alone or did he have some innate skills that I didn't realize he had ?

Regarding AI:

"A couple of people talked about how the quest for “optimal Go” wasn’t just about one game, but about grading human communities. Here we have this group of brilliant people who have been competing against each other for centuries, gradually refining their techniques. Did they come pretty close to doing as well as merely human minds could manage? Or did non-intellectual factors – politics, conformity, getting trapped at local maxima – cause them to ignore big parts of possibility-space? Right now it’s very preliminarily looking like the latter, which would be a really interesting result – especially if it gets replicated once AIs take over other human fields."

Can Kasparov grade the chess community here? What high-level lessons have been learned from Computer Chess? How well did humans do at solving chess? Are there any glaring and consistent inaccuracies that have only been corrected by computer analysis?


I endorse this question and would love to hear his thoughts. I am curious to see how this turned out for chess.

To add to Halting Thoughts' Go anecdote: AlphaGo plays the game to maximize the likelihood of a win, while the majority of professionals are playing to maximize their score. AlphaGo is content to win by 0.5 points and will begin to simplify the board, even at the loss of points, in order to allow for no chance of a loss. It makes for an (apparently?) unbeatable opponent and amazing beginning game but a less useful/entertaining endgame.

I am immensely curious if something similar happened in chess.

When did chess die? Closer to 1996 or 2005?

Sorry guys, but it's dead.

U mad bro?

How has chess changed since computer evaluation? How can it draw in more spectators? What's the best time control?

Could chess be a realistic way to improve academic achievement in previously academically under-performing areas and countries?


Ask him why he considers himself relevant when no other Russian opposition leaders do. Ask him why he has no skin in the game when it comes to his activism. Ask him how it feels to be part of the complacent class.

Is he complacent if he is out there raising hell and using his bully pulpit to do it.

What does he think of the chance of former Soviet republics and satellite states in the EU staying democracies for the next 30 years?

Who are his favourite contemporary Russian novelists?

When you study how computers play, do you think "Oh, that's so obvious, I should have thought of that!" or do you have to forcefully apply your mind to see the logic? And are there ever moves that you just can't appreciate even with the benefit of study?

How do Go players compared to Chess players? What about Poker?

Does or will machine intelligence have the capacity or inclination for desire?

When is his best tip for a beginner at chess.

What car do you drive?

Fernand Gobet has been publishing research for many decades on the possible educational benefits to playing chess. Along with some co-authors he has a paper out in 2017 on the pervasive and extensive methodological issues that trouble the research in this area. Given that traditionally the research has been looking for benefits in areas such as better grades in mathematics classes, are there any areas where you think chess provides benefits that should be looked at more carefully, like say perhaps epistemic humility, or possibly de-biasing benefits such as learning to play against the opponent's rating instead of their physical appearance?

Ask about Kasparov versus the World (a fascinating story for those who are unfamiliar).

TC: in the 12 April Atlantic piece in which you're quoted, James Somers notes that forty years ago only two players in the world had ratings above 2700 while in 2017 the number of players with ratings above 2700 is forty-four.

Is this a demonstration of how well we learn from machines, or should we be surprised properly that only forty-four players worldwide have attained such ratings, considering how widespread and powerful chess programs must have become by now?

The Flynn Effect at work.

Has he ever tried other strategy games like go, bridge or poker? If not, how well does he think he would perform? More generally, how far does he think his remarkable natural gift extends beyond chess - is it just confined to chesslike strategy games, or does it extend to non-geometric strategy games and other areas like science and mathematics?

Magnus Carlsen has said his father is smarter than Magnus himself. Der Spiegel paid for testing that said Kasparov's IQ was 135 (I assume with s.d. 15). What do you think of this?

Why did John Nunn never become world champion?


Good questions - the one on bridge was one I was going to ask (Bill Gates, who did really well as a freshman at competitive math at Harvard, seems to enjoy playing bridge with really good competitors). Second - How would Magnus know? I did not read the source material, so maybe that is not a worthwhile question (in that maybe it has been answered before)

Happy birthday, Garry! I hope Tyler doesn't startle you when he jumps out of that giant cake to sing you Happy Birthday. Anyway I have only a rudimentary understanding of chess and machine learning (and most things really) so hope these questions aren't too frivolous:

A while back Tyler mentioned a version of chess in which a human-computer team plays against another human-computer team. If you've followed this, is there anything unique or surprising that arises in these games compared to human v. human, human v. computer, or computer v. computer games? In what ways does the computer improve the human's gameplay, and vice versa?

As deep learning machines become more complex they seem also to become more unaccountable, making decisions or drawing conclusions that their creators can't explain. Is that a problem? Are we likely to develop techniques that will allow us to monitor and understand why a machine is making certain decisions as it's making them, or at least reverse-engineer its decisions after-the-fact? If not, how do you see our trust in these machines evolving as their abilities and responsibilities become more complex but our understanding of them diminishes? Will individuals and societies trust machines they can't understand? These questions were prompted by this article: https://www.technologyreview.com/s/604087/the-dark-secret-at-the-heart-of-ai/?set=604130

That's all I have for chess and A.I. But I've immensely enjoyed your commentary on American and European politics, even (especially?) when I disagree, so wanted to ask about that. In a recent interview with the Columbia Journalism Review you said: "While Putin clearly preferred Trump to Hillary as the next president, his other goal is chaos and weakness in his rivals, and his greatest potential rival is the most powerful country in the world, the United States. Discrediting Americans’ faith in their elections, in their country, leaders, and institutions, and fomenting divisiveness, anger, and hatred, are all key elements of Putin’s propaganda and hacking efforts." This seems to quite aptly describe the current state of American politics, but is Putin responsible or did we do this on our own? Much of it seems to have been building for years, but since the election the Left has pointed to Russian meddling as a way to delegitimize Trump's presidency, our electoral system, and now even the Supreme Court on account of the Garland/Gorsuch debacle. How much tangible effect did Putin have, and how much is just our own irascible nature?

When a Navy Admiral is interviewed for chair of joint chiefs, it is common for him express concerns about and amity for the Army. What lessons does Mr. Kasparov draw from the game of backgammon, and its use of computers as coaches/opponents, handling of gamblers and reputational consequences, regional pressures and ideas arising in the game (middle Eastern in many ways)?

What would he think of a doubling cube in tournament chess?

What is the role of chess in society now that computers are consistently better than humans? Does that imply anything for a future in which computers and robots become better than humans at many other tasks (and jobs)?

Yep, I am curious about his thoughts on computer-human teams now that computers trounce their carbon cousins.

I would like to know what he has to say about early childhood education -- pre-chess even if the child and parents have lofty aspirations, as many do.

Does he acknowledge Jung Jong Hyun as a peer?

Tolstoy- overrated or underrated?

Does the en passant rule simply represent an example of mood affiliation?

Ask him about food in Russia.

In "Violent Entrepreneurs", Vadim Volkov talks about the role of violence and cooperation between organized crime and government agencies (i.e. the local fire department or building inspector) combined to take the place of trust or enforceable contracts in post-Soviet Russia. As levels of trust in institutions of all types has continued to fall in the US, we have also seen lower economic growth and entrepreneurialism. Strong enforceability of contract is mitigated by the expense and delay of accessing the court system.

Do you see risks to the American socioeconomic system coming from this decline in trust? Do you see parallels between the two countries paths?
Do you think Russia is still in this sort of system, where state corruption is a necessary for engaging in large complicated transactions?

1. How can politicians (and the person on the street) best effect positive political and cultural change in Russia.

2. How can U.S. National Intelligence agencies best address Russian cyber spying.

3. Should the U.S. National Intelligence agencies go on "offense" with respect to international cyber threats, or does the U.S. have too much to lose in this respect (and/or the agencies are just not sufficiently competent), and thus should stay on defense. Or is some combination of the 2 in the U.S. best interest.

4. What is the state of the U.S. counterespionage effort/capabilities?

1. What must happen for a democracy to become a dictatorship. Which factor is most important?

2. Have countries trended toward populism for similar reasons? If so, what can be done to stop it?

Tyler, if you have the time, check out the (fairly short but rather deep) 2016 comment thread on "Magnus and the Turkey Grinder" from IM Ken Regan's blog - as an accomplished conversationalist, I guess you could come up with an interesting question based on the theories spun out on that amazing comment thread (amazing to me, anyway - maybe it is basic stuff that Kasparov will cover in his next book - in brief, it is about the next stage in the chess world's (taken in the broadest sense) reply to the question of how many interesting and deeply competitive chess games are possible and accessible, given the current capabilities of AI) (Ken Regan's April 1, 2017, post was interesting, too, the Platonic ideal of an April Fool's Post that has a small element of unbelievability but an overwhelmingly larger element of believability. I almost always am bored by the science 'humor' of scientist's April Fool's Day "jokes" but this one was really good).
If you ask about books or works of art (and I am not sure I care if you do, the times are long gone when I wanted to read a book simply because someone I admired thought the book is good), it would not be uninteresting to know which novels or paintings or movies, or whatever, a former chess champion has found to be intellectually (or emotionally, or nostalgically, or just humorously ) satisfying in a way that reminds him of championship days. Not that any interviewee is not interesting simply as a person, without reference to past accomplishments, of course.

Someone else may have already suggested this, but does Garry think that Russia will ever have anything approaching a liberal democracy? Or is there something in the Russian soul/history/geography/culture/etc that pushes the country towards one form of authoritarianism (Czarism/communism/Putinism) or another? Related question: can corruption in Russia be cleaned up or is it so endemic that there are literally not enough honest people for the job?

Another set of questions: does Garry think that Russia's military and foreign policy posture is fundamentally defensive or offensive? One school holds that, by expanding NATO and the EU ever eastward, we have threatened Russia, which has responded with things like Crimea and Georgia. Another school holds that Putin dedicated to re-creating much of the lost Soviet influence, in part to maintain his domestic popularity, and is fundamentally the aggressor. Which narrative does Garry think is truer?

By being a mixture of open to and closed off from the world, Russia has given birth to some highly productive diverse mono-cultures in the past, e.g. 19th century novelists or post-revolutionary artists. What's the next one likely to be?

Kasparov simply says that Assad should be removed because he is evil without saying what the alternative is. So why does he think that his argument is different from:

Step 1: We should do something.
Step 2: This is something.
Step 3: Therefore let us do it.

What moves did he learn from a computer chess game that he did not see on his own?

Could you ask him about the studies of chess players done by Adriaan de Groot and Herbert Simon? What successful learning strategies did he use in the the study of chess that could be applied to other domains?

Peter Thiel - underrated or overrated, or a question on his thoughts about Peter Thiel in general

I hope this will be a great podcast. As for questions, I think all of my suggestions are obvious. Still, if this helps:

1. Straussian expression. Normally, I view the value of a Straussian message as providing a way for people who lack power to communicate w/o punishment from larger agencies, often governments, which is why TC can often point to Straussian readings of art, esp. films. But in recent podcasts, TC has begun to suggest that the government not only misleads but actually should consciously mislead the public in its policies in a Straussian way. To some extent, I find TC in these moments advocating for a sort of conscious self-deception and deception of others. But Kasparov has actually lived under autocratic rule. To what extent is he comfortable with the idea of Straussian communication made by artists, governments, and/ or economists?

2. Rationality / intuition. In a lot of these CWTs, there has been discussion on the extent to which people are rational/ irrational, often linking back to Hume's thoughts on the matter. Given his background in chess, I'd be interested in hearing Kasparov's thoughts.

3. Is chess overrated seems an obvious question, but I'd still be interested to hear his answer. FWIW, my answer would be yes in the sense that we associate any interest in chess with brilliance and no in the sense that we underestimate how hard it is to become competitive.

4. Ends/ means. TC has suggested that it might be worth it for America to cultivate and maintain high levels of poverty by the standards of the developed world because it encourages innovation that benefits the world. Let's imagine that Russia, as it exists today in global politics, also has an unappreciated purpose. Is Kasparov as (perhaps this isn't the best word, but) comfortable as TC seems to be at ignoring the suffering of others in order to justify larger ends?

5. The most obvious things I could think to ask about that are in keeping with the content of these podcasts were: translation, reading Russian literature and reading in translation, does he feel he has influence or can make a difference, and, of course, food.

Behavioral economics question: ask him what material pieces and what material boards are has favorite, favor his game. What lighting does he prefer. What clocks? How did he study performance with different behaviorial factors that are not visible in the notation and "on paper."

If a general strategy game tournament was created where masters had to compete on 3 different games. One was Chess, One was Go. What would his choice for the third game be? (A long winded way of asking what other closed information strategy games he likes/admires)

Phantom Time Hypothesis, Anatoly Fomenko, and Garry's "Mathematics of the Past"

Comments for this post are closed