Chess and geopolitics in the supercomputer era

August 10, 2022 10:51
Photo: Reuters

For centuries, chess has been a metaphor for war in myth and in literature. In the next world championship match, which will take place in 2023 between Russia’s Ian Nepomniachtchi and China’s Ding Liren, the comparison may be more apt than ever, with the outcome likely to be decided as much by superiority in multipurpose supercomputing as by individual human ingenuity in chess. And while the Russian military’s dismal early performance in Ukraine hardly suggests an ability to benefit from artificial intelligence in warfare, China is the real deal on that front.

The match between “Nepo” and “Ding,” as the chess world calls them, has arisen because the world number one and defending champion, Magnus Carlsen of Norway, has decided that, having won the world title five times since 2013 (not always easily), he is ready to step down at age 31. (Nepo is 32, and Ding is 29.)

Nepo, who, like many Russian athletes, must forgo the Russian flag to compete internationally, has spoken out clearly against the Ukraine war, and was one of 44 leading Russian chess players who signed an open letter to President Vladimir Putin in early March. “We are against any military action on the territory of Ukraine and call for an immediate ceasefire and a peaceful decision to the conflict through the path of dialogue and diplomatic negotiations,” they wrote. “For us, it is unbearably painful to see the catastrophe that is happening these days with our people.”

Few commentators or top players expected Carlsen to relinquish his title, but his decision is understandable. He has already staked a claim to be considered the best chess player of all time (alongside the Russian Garry Kasparov and the American Bobby Fischer), and has established a successful chess platform. He did not want to contest another world championship match without the necessary steely focus, involving an extraordinary level of memorization, more so than even ten years ago. (Ironically, Nepo’s name in Russian means “forgetful person”; he is anything but.)

Preparations for head-to-head championship chess matches have become increasingly tense, with computers playing an ever-bigger role – more so than for normal multi-player “all-play-all” tournaments. In a normal competition, doing months of expensive computer-supported research to win a single game against a strong opponent is generally not worth it. There are too many games and too many tournaments, and the element of surprise fades once a new idea is employed prominently even once.

But in short world championship matches (the most recent one in 2021 between Nepo and Carlsen was set for up to 14 games), one early win can have a huge effect because the leader can then afford to draw all the remaining games. While a single win does not confer a decisive edge, it means much more than in a normal tournament, where at least a few more wins than losses are usually needed to prevail.

Carlsen won the 2021 world title match decisively in the end, but only after his opponent, playing in his first championship, cracked after losing game six. Over the first five games, Nepo, who had been able to test his new ideas on the Russian Zhores supercomputer, was on the verge of winning twice, but couldn’t quite convert against Carlsen’s brilliant defense.

For next year’s Nepo-Ding match, Ding will most likely receive enormous help from the Chinese tech community. Whether Nepo can still get similar support from Russia is less clear, even if Putin’s chess worldview is that it is Russia’s destiny to be on top once again. That said, Nepo still has a great deal of material in hand from his 2021 title match with Carlsen – material he used to win resoundingly the recent candidates tournament in Madrid to determine who would challenge the champion.

With Carlsen stepping down, the second-place finisher in Madrid, Ding, gained the right to play the title match. Ding, who might well be a better player than Nepo, came to the candidates tournament woefully underprepared because he had been unable to travel freely to compete since the COVID-19 pandemic began.

But the soft-spoken Chinese player just edged out Hikaru Nakamura, which greatly disappointed the American’s devoted fans. Nakamura, who is famously quick-witted both in chess moves and conversation, has become a Twitch superstar, gaining 1.5 million followers by simultaneously playing high-level speed chess and opining intelligently on everything from what kind of car to buy to dating to AI.

Had Nakamura only drawn his last-round game against Ding in Madrid, he would have become the third American since Fischer in 1972 to play in a championship match. Fischer made global headlines by defeating Boris Spassky to win the world title, thereby singlehandedly ending the hegemony of the Russian chess juggernaut. Although Nakamura is very much a self-made player, one imagines that Google’s DeepMind, which has pushed the boundaries of AI, might have decided to lend him help had he faced a Russian opponent in the world championship.

Chess is going gangbusters globally. Online chess exploded during the pandemic and, although the surge has faded somewhat, metrics such as membership in the leading virtual playing sites indicate that interest remains well above the level of three years ago. The success of The Queen’s Gambit, which was the most successful Netflix series of all time when it aired in 2020, and won an Emmy for best limited series, certainly helped. While the miniseries is not exactly for children, it nonetheless attracted a new generation of young girls into the game, as Jennifer Shahade discusses in her insightful new book Chess Queens.

By the way, I predict that Ding will be the next world champion, though it has taken longer than I thought back in 2018. If he loses, it will tell us not only something about Nepo’s character but also that Chinese supercomputing is perhaps not as advanced as we think.

Copyright: Project Syndicate
-- Contact us at [email protected]


Professor of Economics and Public Policy, Harvard University. He was formerly chief economist at the IMF.