Musk vs. Chess.com: The Boardless Match and the Battle over Whether Chess Can Be Solved

No pawn was promoted, no clock was pressed. Yet the war of words between billionaire Elon Musk and Chess.com kept the chess community and the tech world buzzing for hours. It all started with a brief declaration by Musk on X: artificial intelligence will "soon completely solve chess." Chess.com's account replied with three mocking words: "Just skill." Musk refused to back down, suggesting that the other side was merely "an intern online." Before fans could pick a winner, Chess.com produced a mathematical barrier: 10^120, the total number of possible chess games — a figure larger than the number of atoms in the observable universe.\n\nOn the surface, this was pure online entertainment, but beneath it lay one of the most common misunderstandings about the relationship between AI and this ancient intellectual game. To understand the substance, one must look at the backdrop on both sides. Musk, the head of xAI, was not chatting casually. He was reinforcing his commercial narrative: a future in which AI models can compress information "far beyond human imagination." Chess.com, meanwhile, was defending its brand authority. It has become the most influential voice in online chess, so letting a billionaire dismiss the intellectual value of the game with a claim that "AI will soon solve it" was not acceptable. Notably, Chess.com's clever replies were deliberate: they echoed the phrase "intern" in Vietnamese, a wink to the large Vietnamese-speaking chess community.\n\nLooking further back, this is not the first time Musk has clashed with chess. In 2026, he canceled a chess event at Tesla. In 2026, he said chess is "too simple for real life" because it has no fog of war, no technology tree, and symmetric starting positions. The refrain "chess is too simple" has followed Musk for more than a decade, and now it is dressed up in AI language. Could he be preparing the market for a larger story?\n\nTechnically, Musk's claim is not entirely wrong, but it ignores a crucial detail: the difference between the number of positions and the number of games. Chess has about 10^44 legal positions — the possible board states. But the total number of games, meaning the sequence of moves from opening to endgame, reaches roughly 10^120. The complexity of the "game tree" — not the number of positions — is the real barrier to solving a game. Chess.com used 10^120 to compare with 10^80 atoms in the observable universe, arguing that chess lies beyond the reach of any machine. In essence, that comparison is correct.\n\nThe core issue lies in the definition of "solve," which the two sides understand differently. A game is considered "solved" when one can determine, from every legal position, the optimal result if both sides play perfectly. Checkers was solved after years of computation because its number of positions is only about 10^31, a domain small enough for modern machines. Chess's 10^44 is not merely larger in a philosophical sense; it is 10^13 times larger in computational scale. That is the gap between a project that might run for a decade and a problem that could not be completed in a thousand years using all current hardware. Musk's argument — that future AI will find compression methods "beyond human imagination" to crack this barrier — therefore belongs to philosophy, not computer science. It relies on a technology that does not yet exist to negate a proven complexity.\n\nHistory shows raw computing power has made leaps before. But "solving chess" is a different matter from "beating the world champion." Machines have been stronger than humans since the late 20th century, starting with Deep Blue's defeat of Kasparov in 2026. Yet being stronger than humans does not mean finding the optimal answer from every position. Popular culture still conflates these two ideas, and that conflation is fertile ground for statements like "chess will soon be solved."\n\nThe irony is that in a recent AI tournament covered by the media, xAI's Grok 4 — founded by Musk himself — lost to OpenAI's o3 model. This context places Musk's claim in an awkward position: when his own model cannot surpass a rival within a fixed-rule chess framework, his claim about compressing the entire game sounds more like reframing a competitive failure than making a serious technical prediction.\n\nLook closely, and this war of words is not really about chess. It is a branding stage. For Chess.com, every fast, witty, multilingual reply is a content campaign riding on a celebrity. Musk's X account has more than 200 million followers — an audience no advertising budget can buy. For Musk, the "AI will solve chess" story serves xAI's commercial agenda perfectly. Both sides benefit. But there is a third party that may lose: chess itself, as a symbol of human intellect.\n\nCasual viewers who do not follow tournaments may swallow the message that "chess is about to become a dead game." When a tech billionaire says so, the prejudice spreads faster than any chess publicity campaign. The effect is not immediate, but in the long run, sponsors and mainstream media may gradually shift attention to sports "not yet touched by AI." The irony is that no single high-level game changes because of this spat. The players' strength remains intact, their competitive value remains intact. But the perceived value in the public mind can shift — and in the attention economy, perceived value often matters more than actual value.\n\nFrom another angle, however, the rise of AI tournaments opens a new role for chess: a benchmark for AI capability. When tech companies measure their models on a 64-square board, they turn chess into a standardized test. This is a double-edged sword. On one hand, it creates fresh content comparing AI systems, attracting tech audiences. On the other, it risks reducing chess to a dry technical standard rather than a human sport with stories, emotions, and limits.\n\nProfessional chess players probably need not worry too much about Musk's claim. The chess elite has heard this refrain many times. What they must do is proactively explain to the public why 10^120 matters, and why AI strength does not strip the game of meaning. The answer lies in why millions of people still sit down at the board every day: they are not searching for a final answer; they are seeking the experience of battle between two human minds.\n\nThe real question is not "when will AI solve chess," but "when will people stop measuring a game by a machine's ability to conquer it." Chess has survived more than 1,500 years, witnessing countless technological incursions — from mechanical devices to supercomputers, then to self-learning algorithms. Through every era, it has endured, because its core value does not reside in mathematical complexity but in its power to mirror human thought. A new generation of machines may change how players prepare openings, but it cannot solve the reason why humans love this game. And that is the final match no algorithm can win.