Don’t kid yourself: Thinking is hard. You can see this in grandmaster chess players, whose heart rates triple to cantering under their shirts. What separates champion Magnus Carlsen is his lethal stillness, a supremely thoughtful chill under pressure, which makes his recent behavior more startling. Basically, what Carlsen has done to chess is the equivalent of upending the board and scattering the pieces. Carlsen never gets upset — so he must be pretty upset.
What’s upset him is the possibility that Hans Niemann, a 19-year-old American, may have infiltrated the undefended world of table chess to beat him with a machine. Unless he’s just upset that Niemann beat him. Carlsen’s explicit accusation Monday that Niemann cheated in a match by relying on artificial intelligence to help select his next moves — an allegation Niemann denies — has plunged chess into smirky speculations of devices hidden in cavities. The ability to receive such computer-generated advice through hidden signals is also an “existential threat,” as Carlsen says, to an old-world culture in which competitors have played on trust without checking what’s up each other’s shirt sleeves or pants legs.
“The chess world has been pretty blasé and pretty relaxed in terms of taking the possibility of cheating seriously,” says American grandmaster and Twitch star Hikaru Nakamura, who for years played in tournaments where competitors simply hung up their jackets before sitting at a table, or were barely wanded. “ … Magnus has said something along the lines of, he’s not doing this for himself. It’s part of a bigger question, a bigger situation.”
Chess makes for strange bedfellow-fanatics. Amari Cooper of the Cleveland Browns is a chess addict and so was the filmmaker-auteur Stanley Kubrick. Asked once why he found the game so entrancing, Kubrick answered, “It trains you to think before grabbing.” Cooper loves chess for the same reason. All athletic actions are essentially micro-decisions, and even the fleetest and most impulsive seeming NFL wide receiver must make feints, counter-feints, and judgments.
Anyone who questions whether strategic thought requires an athletic-like stamina should consider the physical toll on chess players, who can drop 10 pounds or more in a tournament with their metabolic burn rates. In 1984, according to an ESPN story, Anatoly Karpov lost 22 pounds during his world championship siege with Garry Kasparov. A pair of American physiology researchers, Leroy DuBeck and Charlotte Leedy, were the first to wire tournament chess players with a variety of sensors to verify the relationship between thought and action. The sensors showed that breath rates soared. Adrenaline surged. Pulses galloped; muscles contracted. All while the players sat virtually unmoving.
As Bobby Fischer once remarked. “Your chess deteriorates as your body does. You can’t separate body from mind.”
In recent years, the proliferation of livestreams, fit trackers and other tools have created almost a game within a game at modern chess tournaments. Rubbernecking audiences watch for clues of mental cracking and physical distress in the quirky, contemplative figures bowed over the boards. At the 2018 Isle of Man International tournament, fitness metrics projected on a large screen revealed that grandmaster Mikhail Antipov torched 560 calories sitting stock still for two hours. By way of comparison, the average person will burn just 100 calories running a mile on a treadmill.
The champion at this game within a game has long been Carlsen, who is such a hard-training player that he famously visited the Norwegian Olympic center in 2017 to develop a physical regimen that would help him in the final stretches of five-hour matches. He does high intensity interval stints for 30 to 60 minutes on treadmills, hot yoga, and soccer, tennis, and basketball workouts.
All of which brings us to Carlsen’s quarrel with Niemann, and why he is apparently so suspicious of him. Earlier this month, Niemann, a patently inferior player, beat Carlsen without breaking much of a sweat. Somehow, Niemann anticipated and swiftly blocked a tremendously obscure opening strategy by Carlsen. “I had the impression that he wasn’t tense or even fully concentrating on the game in critical positions,” Carlsen said in a statement released Monday.
This provoked Carlsen to a rare histrionic: In a rematch with Niemann last week, he resigned after just one move and stalked away from the board — a gasp-inducing gesture of protest that earned a reprimand from the international chess governing body. But it also achieved Carlsen’s main goal, which was to subject Niemann’s playing patterns to close examination. The scrutiny forced Niemann to acknowledge he used computer assistance in online matches at chess.com when he was 12 and 16, for which he was banned. Niemann insists his recent rise in board chess has nevertheless been legitimate. Asked at the Baer Cup to explain some of his match play that seemed less than explicable, he replied: “I’m a very intuitive player.” That wasn’t good enough for Carlsen.
“I believe that Niemann has cheated more — and more recently — than he has publicly admitted,” Carlsen charged in his statement on Twitter.
Carlsen appears to have seized the banner on behalf of a group of grandmasters who believe that machine intelligence is outstripping those who play purely with their heads — and that it’s not being captured by current analytics or tournament organizers. Grandmaster Srinath Narayanan of India tweeted, “We all knew cheating was a serious problem. We all knew it was rampant. We all kept quiet, not knowing exactly how to go about it. Magnus spoke about it and in a way that the world had no option but to take notice.”
The rare instances where someone has been caught indicate the possibilities: In 2015 chess officials discovered Arcangelo Ricciardi receiving signals in Morse code in his armpit.
Why should we much care whether a 19-year-old chess prodigy used artificial intelligence or a signal to solve a board challenge? Because the Carlsen-Niemann confrontation raises the important matter of “techno-solutionism.” Too much machine intelligence in problem-solving, as it happens, can be more confusing — and weakening — than helpful. The long-term cost of techno-solutionism can be a fatal slackness, both mental and physical. You don’t want to lose your conditioning for decisive human judgment.
Performing while sitting for five and six hours at a stretch is the most common office experience. We’re all familiar with the peculiar exhaustion that can come from that posture. It’s a tiredness that feels different from any other kind. That’s not your imagination. Clinical researchers have found that “decision fatigue” is a distinct form of expenditure, separable from the other physical or cognitive loads. It affects our behavior, and unaddressed it can result in reduced capacity for problem-solving, as the social psychologist Roy Baumeister and a team of fellow researchers demonstrated in a series of studies. In one, a group of collegians were asked to make product choices, on everything from whether they preferred pens or pencils. After making these small and not particularly important decisions, the choosers showed less physical-mental endurance than a peer group and less inclination to study for a test.
“Recommendation algorithms” can solve some problems, but they don’t always make us smarter or stronger. Not every probability deserves credence. Asked whether tech has been good or bad for chess, Nakamura said, “Wow, that’s a good question. It depends on who you ask. For me, I’d definitely say that I really enjoyed learning the game without having this sort of second opinion, or superior opinion or, like, perfect opinion. I definitely liked not having computer programs which just knew the answer to everything. I think I’m in between. I think it’s been very good for pushing boundaries of our knowledge forward, but at the same time when you have these computers that are so much better than humans, and it’s possible to, in one move, gain an advantage and win a game it also is a problem.”
The limits of algorithmic predictions are abundantly clear in the Niemann mess. Chess observers have tried to use them to evaluate Niemann’s play, only to fall into a morass of argument. One analysis finds that his play falls within an unsuspicious range while another finds his performances improbable.
“At the end of the day when we’re talking about looking at the games, there are probably only a handful of people in the world who can say whether these moves look like they’re human, or not human,” Nakamura said. “There’s a limited pool of people who can have opinions that are legitimate. That also makes it very difficult. There’s really no agreement.”
Carlsen has called for better methods of detection and added, “I hope that the truth on this matter comes out, whatever it may be.” But the chess world may discover that machine intelligence or tech engines don’t solve its new problems any more efficiently than an age-old human practice: the honor code, the development of conscience, which solves problems before they begin. As the Russian chess grandmaster Alexander Grischuck once remarked about the explosion of chess online and the proliferation of tools with which to cheat, ultimately, “Everything rests on decency.”