Elon Musk: “The actual number of moves that are not utterly stupid in chess is tiny”

Elon Musk builds rockets, runs a carmaker, and keeps up a running audit of the limits of human intelligence. This week he pointed that audit at chess — the game that has stood for a century as shorthand for deep thought — and pronounced it, in effect, a solved problem waiting to happen. The wording was pure Musk: absolute, dismissive, and fired off to millions with the certainty of a man who treats every hard problem as a scaling exercise.

“The actual number of moves that are not utterly stupid in chess is tiny and chess will be fully solved one day”

He posted it on X, answering a thread about how a chess game branches into exponentially more positions with every move. An account for Chess.com replied with two words — “skill issue” — and the argument was on, later picked up by Primetimer and others as the spat escalated.

There is a real idea inside the first half of the sentence. Strong engines already discard the vast majority of legal moves as garbage and calculate only the handful worth considering; pruning the tree down to the moves that are not utterly stupid is, roughly, what a chess engine does for a living. On that narrow point Musk is describing something true.

The second half is where the mechanism breaks. “Solved” is not a mood; it is a technical claim — a proof of the outcome from every position under perfect play, the way checkers was solved. Beating humans is not solving. Stockfish and AlphaZero have outclassed the best players alive for years, and neither has come close to solving chess, because the obstacle was never skill. It is size. As Chess.com pointed out, the number of possible games runs to dozens of orders of magnitude beyond the count of atoms in the observable universe. The word “tiny” is carrying a universe on its back.

The tell is what Musk did next. Told he was wrong, he did not engage the math; he reached for the org chart, waving the reply away as coming from “some random intern at a chess website.” Informed it was a full-time employee, he went silent. It is a familiar shape — the same confidence that turns “full self-driving next year” into an annual tradition, and casts the specialists who understand why a problem is hard as the thing standing between him and solving it.

Chess may well be solved one day; checkers was, and nothing in theory forbids it. But “one day” is doing the same quiet labor as “tiny.” Musk lectured a chess company on a word — solved — that has an exact meaning in the field he was lecturing. On X, that was never the point.

Tags: Elon Musk, chess, AI predictions, Chess.com, tech feuds