Carlsen: AI shows humans are not that good

I rewound the footage of Magnus Carlsen's interview with AS on October 21, and stopped at a sentence that Vietnamese media barely mentioned: "AI makes chess more interesting, but it also makes it increasingly difficult to create an advantage over opponents." It is a double sentence. One half for the audience. One half for himself.

For 15 consecutive years, Carlsen has held the world No. 1 ranking. He has five classical world championship titles. Since 2026, he has not defended the crown. And at 35, he says his physical condition has declined, that he no longer plays those 5-6 hour games as before, that he tries to win a little faster. Those are the facts. The rest is how we read them.

Context: a player who left the championship cycle

The World All-Around Championship in Budapest, from November 10 to November 21, gathers 24 of the strongest players, including Le Quang Liem. The format combines classical, rapid, and blitz chess. This is not an event in FIDE's world championship cycle. It is an independent commercial product, and Carlsen's participation is a signal.

In the interview, Carlsen says he is interested in many other aspects of chess more than in a world championship match. But he does not oppose the idea of returning, if someone pays around 20 million USD. That figure sits next to the actual number: the winner of the classical world championship match receives just over 1 million USD. The ratio is about 20 times. That is not an offer. That is a benchmark.

Analysis: AI flattens preparation advantages

In the engine era, every top player has access to the same strongest machines. When opening knowledge becomes universal, the marginal return on deep opponent-specific preparation falls. Carlsen says this explicitly: he has shifted from studying specific lines to "more general ideas." That is a rational cost-benefit adjustment, not a sign of declining capability. But it is also a consequence of no longer committing full time to classical chess.

The mechanism is clear. When everyone has the same tools, the decisive differentiators migrate to over-the-board calculation, time management, and psychological resilience. There, age is a non-negotiable variable. Carlsen himself admits his fitness has declined. He says he used to score well in 5-6 hour games thanks to sustained concentration, and now he tries to win faster. This is a genuine technical-tactical shift with a clear mechanism: if the stamina premium in the late game erodes, compress the game to convert remaining strengths into chances before stamina becomes the binding constraint.

But there is a paradox. Carlsen says the average quality of his moves has improved. This is the most technically interesting and least evidenced claim in the piece. It implies an internal assessment of accuracy, but no metric is disclosed. No ACPL, no engine match rate, no score. I trust footage more than testimony, because footage does not lie. Here, we only have testimony.

Contrarian angle: the 20 million USD figure is not an invitation

Media will read the sentence "does not oppose a return if there is 20 million USD" as a signal of homecoming. But that sentence is surrounded by two layers of hedging. First, Carlsen makes clear he is interested in other aspects of chess more than in a title match. Second, the 20-fold gap from the current prize is a deliberate opening position, not a reservation price. A transfer contract is a final stretched over weeks off the pitch. Here, the 20 million USD is a final stretched over years in the meeting room.

More noteworthy is the structure of the Budapest event. A 24-player, three-format, 11-day event outside the FIDE cycle is a product competing directly with the classical world championship. If it succeeds commercially, it will create a new standard for measuring a champion's value. And if Carlsen wins, he will have another title that does not need FIDE recognition.

There is a blind spot in the media narrative. They focus on the sentence "humans are not that good" as a philosophical statement. But that sentence, read carefully, is a statement of epistemic humility: humans understand chess less well than machines. It is not "chess has been solved." These two readings are different, and the second is the one that diminishes the sport's appeal.

Takeaway: a new benchmark for professional chess

What I take from this interview is not the sentence about AI. It is the 20 million USD figure placed next to the 1 million USD figure. When the world No. 1 publicly sets such a measure, he is not talking about himself. He is talking about the commercial position of the world championship. And that is a debate that will last longer than any chess game.

I have run hurdles all my life, only I was never timed. But I know one thing: when the best leaves the track, people start re-measuring the length of the track. Carlsen left the championship cycle in 2026. The question now is: who will re-measure that length?