Skip to content
Breaking News Alert House Speaker Kills Effort To Stop The Feds From Spying On Americans Without A Warrant

How Will The Rise Of The Machines Change The Future Of Chess?


The World Chess Championship is being played in London right now, and an American is competing. Magnus Carlsen of Norway, the number one-ranked player in the world, is pitted against the number two player, Fabiano Caruana, who moved to Italy in 2005 but returned to the USA in 2015. He has a fighting chance to be the first American in more than 40 years to become the global chess champion.

It will be an uphill battle. Carlsen has been the champion for five years and is the highest-rated chess player of all time, according to rankings by FIDE, the Fédération Internationale des Échecs, or World Chess Federation. Caruana would be the first American champion since Bobby Fischer crushed Boris Spassky (and the whole Cold War Soviet chess machine) in Reykjavik, Iceland in 1972, in what was dubbed the “Match of the Century.”

Can Caruana Win the Championship?

Carlsen and Caruana are scheduled to play up to 12 games over three weeks, with the first player to earn 6.5 points declared the winner. Each win is worth a point, and the players split a half-point for draws, or tie games. Draws are commonplace at this level, with high-rated, equally matched players in high-risk matches not wishing to risk a blunder of even the loss of a single lowly pawn.

A paranoid Fischer refused to defend his title in the ‘70s, and U.S. chess lost its most incandescent star and perhaps its cultural breakout moment. It’s hard to keep the attention of a nation of so many different kinds of people pursuing a myriad of different interests. Still, the game has lasted for more than 1,000 years and unlike, say, jai alai or greyhound racing, it isn’t going away.

The start of this year’s world championship even made the front of the New York Times sports section, begging the age-old question: is chess a game or actually a sport? It’s certainly become a game for the young: challenger Caruana is 26 and Carlsen is 27.

At the halfway point of the November 16 match, the score stood tied at three points each, with all six games ending in draws. The general sense from observers is that, after some shakiness in the first game, in which he was unable to exploit white’s first-move advantage, the challenger Caruana has found his feet. The first game was a very long draw of 115 moves and seven hours in which Carlsen (playing the black pieces) was the aggressor but made some missteps that cost him advantage.

But game six proved that not all draws are alike, and was the clear highlight of the match so far. Carlsen, with the advantage of the white pieces, moved his knight twice in the early stages, generally a no-no when one’s goal is to “develop” your pieces and give them space to operate. (Of course, the usual rules don’t always apply to world champs). Grandmaster Susan Polgar called it a “rare move.”

In the middle game, Caruana developed a surprising advantage with black, and it took spectacular defending by Carlsen to salvage a draw after nearly seven hours and 80 moves. A computer evaluation showed that Caruana missed a hard-to-find move that would have led to a long and drawn-out win.

The Guardian explained how not even the second-best player in the world understood the computer’s reasoning: “The Stockfish evaluation engine found a forced mate in 30 moves for black….‘I’m not going to disagree with the computer,’ Caruana said when showed the mating move. ‘I just don’t understand it.’”

It’s understandable that a tired player, in a tense game, under time pressure, missed a complex continuation. Still, that brings us to the specter that hangs over this match’s board, and every other chessboard: the rise of the machines.

The Machines Are Taking Over

NPR succinctly summed it up before the last championship match in 2016: “Next month, there’s a world chess championship match in New York City, and the two competitors, the assembled grandmasters, the budding chess prodigies, the older chess fans––everyone paying attention––will know this indisputable fact: A computer could win the match hands down.”

That contest pitted Carlsen against Russian Sergey Karjakin. Carlsen won and this year is defending his title again, as an aside.

Since retired chess champion Garry Kasparov’s two controversial tussles with Deep Blue in 1996 and 1997 (Kasparov won the first match and lost the rematch against a different version), there has been unstoppable improvement in the realm of chess computers, their brute computational force now completely outstripping human strategizing and pattern recognition. Computers have built-in advantages: they are invulnerable to flagging concentration, petty irritations, and romantic distractions.

Carlsen is considered a relatively “human”-style player, not a slave to machine analysis. Kasparov calls Carlsen’s playing style “intuitive positional chess.” But even Carlsen trains with computers and is routinely beaten by them. The chess program on your smartphone could probably do the same.

Despite early machine age predictions, computers can’t yet think or play like humans (although that may be changing with AlphaZero, which teaches itself to play the game through constant iteration). Humans come in with a purpose. They craft strategies, look for patterns, and aim for superior positions. Computers excel in open positions with sharp tactical possibilities and are devilish defenders.

Humans being human, they want to win, which can make them press too hard in weak positions or stress over a bad move. By contrast, as Kasparov ruefully noted while playing then computer-chess champion Deep Junior in 2003, “the computer doesn’t even know it’s playing chess.”

The irony is that it’s easier for computers to locate the perfect move than to actually make it. Robotics has a long way to go to catch up to humans in sports or even basic movements like crossing a street. A chess-programmed robot could beat you in chess, but it would probably knock the pieces over trying to do so. Will a computer ever become unbeatably God-like in its perfect play?

Checkers was “solved” a decade ago when the Chinook computer program analyzed every possible position, requiring perfect counter-play against it just to draw. Chess isn’t there yet, but in a couple of decades, with a speed assist from quantum computing, who knows? Only the ancient board game Go, which has exponentially more possible moves than chess, would seem to be beyond computational mastery, for the time being. You might want to keep those pod bay doors open a crack, just in case.