The Future Of Chess

We live in unprecedented times when AI and the Digital Revolution are reshaping how we live. Recently, Elon Musk even suggested that, in 15 to 20 years, work will be optional and money irrelevant. 

It is difficult to argue with one of the smartest people on Earth, but I heard the exact same phrase as a little kid when the Soviet leaders promised us communism in 20 years. Well, it only proves the words of brilliant Yogi Berra, "It's tough to make predictions, especially about the future."

Since we have the world's largest chess community here on Chess.com, I would like to discuss how recent technological developments might affect our favorite game. I am not an expert in computers in general—and in computer chess in particular, I'm far from it.

However, when you have lived in the chess world for 50 years, you can see some patterns. As the saying commonly attributed to Mark Twain goes, "History never repeats itself, but it rhymes." So, without further ado, let's address some hot topics of modern chess that scare many people.

1) The game of chess will be solved.

I hear this as the main concern for many people. What if one day computers prove that with the best play, a chess game will end as a draw in 80 moves or White will win in 110 moves? To be honest, this should bother us the least. First of all, it is not going to happen in the next 15-20 years, but even if it does, it won't make any difference to us, human chess players.

As of today, all positions with seven or fewer pieces on the board are already solved, thanks to the endgame tablebases. Tablebases have found some remarkable things. This is one of them:

White to play and mate in 549 moves. I saw the solution, and practically all the moves there made little sense to me. But even if I could understand the point of those moves, how is it possible to memorize them?

So, even though I saw the solution, it wouldn't help me one bit if I had this position in my games. The same with the initial position. Even if one day engines will prove that after 1.e4 White wins by force, no one will be able to remember all the variations and sub-variations. 

2) Opening theory becomes too deep and "solved-like."

If you followed the recently concluded XTX Markets London Chess Classic Elite tournament, you couldn't miss the following game:

GM Nodirbek Abdusattorov introduced a novelty on move 29, and after 34 moves, the game still followed his home preparation. When GM Abhimanyu Mishra resigned, the Uzbek grandmaster had more time on his clock than at the beginning of the game. Even super GM Alireza Firouzja couldn't contain his amazement: “It’s insane. I don’t know how these guys do it."

So, has the opening theory gotten too deep, thanks to computers, and must you memorize engine-approved lines just to survive? I remember the exact same sentiment over 40 years ago when GM Igor Novikov introduced a novelty on move 36 and basically won the game thanks to his superior home preparation:

Yet, somehow, chess survived, and we were able to play it for the last 40 years just fine. My prediction is that while opening preparation for elite players will become increasingly complex, for 99.9% of chess players around the globe, it is a moot point.

3) Draw death / Risk of chess becoming less interesting

To me, this is the most superficial danger. Also, it is not new and has existed for the past 100 years, well before chess engines appeared. Here is what Wikipedia says: "Capablanca was concerned with how high-level chess was overly reliant on extensive memorization and study of chess openings and their variations. He feared that eventually most high-level games would end in draws because of this. This threat of 'draw death' for chess was his main motivation for developing and promoting a new chess variant."

Ironically, all these long opening variations that allegedly led to inevitable draws didn't help Jose Raul Capablanca himself to defend his title against Alexander Alekhine. Somehow, Mike Tyson's famous quote comes to mind: "Everyone has a plan until they get punched in the mouth."

Only when they convincingly demonstrate that they are able to stop GM Magnus Carlsen, who is the unquestionable number one in any form of chess, will I start taking all this "draw death" nonsense seriously. 

4) Moving online chess loses its identity.

This is one of the biggest concerns for people who grew up during the "golden age" of chess, when it was treated as a royal game. I have to admit, the following "performance" of the world's top player looked seriously cringy to me:

This episode reminded me of a game I played some 30 years ago. After unsuccessfully trying to win for 70+ moves, the game came to this position:

I played White and saw no more winning chances in this dead-drawn position, so I offered a draw. My opponent, a grandmaster, shocked me by refusing the offer. Yes, I had only about two minutes on my clock, and we used an analog clock, so no increment or time delay was possible.

But it never occurred to me that a self-respecting grandmaster would try to flag an opponent in such a position. I immediately remembered the late GM Eduard Gufeld, who witnessed two young players trying to flag each other in a dead-draw position. They smashed the chess clock as if it were a boxing contest.

After the show was over, Gufeld asked them, "guys, why do you need a chessboard? Set one minute each on your chess clock and just punch the buttons until one of you loses on time!" So, to make sure that I understood my opponent correctly, I clarified: "so, you are going to play this position for a win, right?" My opponent suddenly shook his head, as if he had just woken up from a nightmare, smiled embarrassedly, and said, "Oh no, of course it is a draw."

You can say that these two cases are different, and Magnus tried to flag his opponent in blitz, where time is a major part of the game. Fair enough. However, I grew up with stories like the following one:

GM Mikhail Tal was participating in a blitz tournament, where he played GM Nona Gaprindashvili. Toward the end of the game, the legendary Women's World Champion had practically no time on her clock. So after making his moves, Tal "forgot" to press the clock a couple of times. When Gaprindashvili noticed that, she whispered angrily: "You do it one more time, and I'll immediately resign!"

To be objective, we shouldn't compare Carlsen to Tal or say something like "GM Mikhail Botvinnik would never throw his king to spectators, a la Hikaru Nakamura." The great champions of the past had a totally different mission. They had to push chess horizons, discover new concepts, and amaze the chess world with the strength of their play.

These days, we have chess engines that can do all these tasks much better than any human player. Moreover, today everyone possesses these little monsters, and therefore, the mystery of chess is gone to some extent. For example, Tal's famous 21...Nf4!! from his world championship match vs. Botvinnik was analyzed for decades!

Today, all you have to do is offer this position to a chess engine for an overnight analysis, and in the morning, you'll know everything you need to know about this fascinating sacrifice.

Magnus Carlsen, Hikaru Nakamura, and other elite players are smart. They realized that if they cannot act as professors of chess science, they need to be showmen who promote chess and make it more popular. Hence, meme openings and other stuff that would make old school players cringe. To be fair, chess is more popular than ever, partially due to Carlsen's and Nakamura's efforts. 

5) The risk of cheating will increase

I vividly remember the Y2K scare. While people were expecting Armageddon, Y2K turned out to be a blessing in disguise, helping to modernize outdated computer systems and improve cybersecurity. Instead of disaster, it upgraded the world’s digital infrastructure.

Call me naive, but I expect the same positive effect of AI in chess, and the problem of cheating, instead of increasing, will recede. It could be old detection methods significantly upgraded by AI. Something like a combination of move matching to major chess engines, time usage models, and behavioral patterns (eyes and mouse movement). Or it could be something completely new, like every player having a digital profile (preferred openings or pawn structures, typical calculation lengths and blunder frequency, time-use patterns, etc.). Any significant deviation from this profile would automatically trigger a review.

I will repeat: I am not an expert in computer chess, but the entire history of chess, combined with general life experience, suggests that AI will solve more problems than create.