Carlsen: 'AI shows humans are not that good'

The home pitch has no stands, but every time the ball rolls, a sky of memories rushes back. With chess, there are empty stands too: a new tournament in Budapest, 24 of the strongest players, three formats, and a classical world champion who no longer participates in the official cycle. Magnus Carlsen, at 35, is sitting at a larger board than any game: the board of media, money, and human limits. In an interview with AS, cited by Marca, Carlsen said: "AI makes chess more interesting in the short term, both for fans and for professional players." But immediately after, he added: "Technological development also makes it increasingly difficult to create an advantage over opponents." These two sentences, placed side by side, are the core statement of the article. Not a technical discovery, but a way of looking at the engine era. I have followed Carlsen playing live at many major events. What I remember is not the perfect moves, but the times he sat still for a long while before a seemingly simple move. That is the sign of a player considering more than calculating. When everyone has access to the same strongest engine, the advantage no longer lies in opening knowledge. It lies in decision-making under time pressure, in calmness, and in the ability to read the opponent. Carlsen says he has shifted from studying specific opening lines to "more general ideas." This is a deliberate change, not a sign of declining ability. In a preparation market saturated with engines, studying a single line too deeply yields low marginal returns and high risk. A more general approach suits a player who no longer devotes full time to classical chess. He also admits: "I no longer play such long games. So I try to win games a little faster." This is the most notable technical point. At 35, Carlsen acknowledges physical decline. In classical chess, games lasting 5-6 hours are where fitness decides. If he no longer has the stamina to maintain high concentration into the endgame, shortening the game is a reasonable adaptation. He shifts his advantage forward: opening, practical decision-making, and psychological pressure on the opponent. But there is another statement that needs careful reading: "The average quality of my moves has improved." This is a claim without supporting data. There is no ACPL, no engine match rate, no accuracy metric in the article. It may be true, but it is a hypothesis, not a finding. And if true, it may reflect a smaller sample: shorter games, where he has more time per move. The most notable thing in the article is not the statements about AI, but the number. Carlsen said he would not oppose returning to defend the world championship if the winner's prize were around $20 million. Currently, the classical world championship pays the winner just over $1 million. This 20-fold gap is a statement about the commercial position of the tournament, not about Carlsen's intentions. He relinquished the title defense in 2026. He has held the world No. 1 ranking for 15 consecutive years. This is a record of durability, not of peak performance. And it is the strongest argument against any decline narrative. But it also raises a question: is a No. 1 position defended by selective participation still the best measure of chess strength? The new tournament in Budapest, from 10/11 to 21/11, with 24 players and three formats, is another answer. It is not part of the FIDE official cycle. It is called the "World All-Around Championship." The presence of Le Quang Liem in the field explains why this article reached Vietnamese readers. But more importantly, it shows organizers are seeking a broader market, not just Europe. I once wrote about silent stadiums during the pandemic. What does not happen sometimes has more power than what does. Here, what does not happen is a world championship with the participation of the No. 1 player. That means the world championship title is no longer the only way to identify the strongest player. And when the strongest player does not participate, the title loses part of its legitimacy. The statement "humans are not that good" is a philosophical assertion, not a technical finding. It should be read as epistemic humility: humans do not understand chess as perfectly as engines. But it can also be misread as a shocking claim: the world's best player says chess is beyond human comprehension. These two readings are very different. What interests me is the impact of AI on watchability. If AI makes preparation more equal, draw rates will rise. If draw rates rise, drama falls. And if drama falls, the core product of professional chess is diluted. The three-format event in Budapest is a reasonable countermeasure: it compresses time, increases decisiveness, and creates more winning chances. Carlsen is adapting to age by changing the format, not by fighting the age curve. This is a new archetype. Not like Anand, who selects classical events to extend his career. Not like Nakamura, who relies on streaming to maintain competitiveness. Carlsen chooses a third path: moving to mixed events, where fitness matters less and experience remains an advantage. Is this model sustainable? The answer depends on whether the Budapest event succeeds commercially. The prize fund has not been disclosed. That is a worrying sign for a new tournament. If the prize fund is low, a second edition will be difficult. If it is high, it will pressure FIDE to increase the world championship prize. I once thought chess was a sport of silence. But what is happening now is louder than ever: interviews, numbers, conditional threats. And amid all that noise, a 35-year-old player is saying he no longer has the stamina to play long games. That is a statement about human limits, not about the limits of chess. The question left: if the world No. 1 no longer wants to defend the championship title, what meaning does that title still have? And if AI makes all players more similar, what will keep us watching chess games?