Elon Musk says chess will be solved in 10 years: A quarrel of 10^120 moves

One summer afternoon, the man with more than 200 million followers on X wrote a short line: "Chess will be solved within 10 years." To outsiders, it was just a remark about artificial intelligence. To me, a man who has spent nearly half a century beside chessboards and stadium corridors, the statement tasted familiar. People once said man would never reach the Moon; then they said computers would never beat a world chess champion; then Deep Blue beat Kasparov in 2026. The only difference this time is that the man making the claim does not merely own a tech company—he owns an AI company whose model had just lost to another AI model over a chessboard. Chess.com, the world's largest online chess platform, responded quickly enough to remind me of a legendary arena flash: the player who cannot think, only react. The truth behind the shocking claim lies not in the future but in a number written by Claude Shannon in 2026: 10^120. That is the maximum number of possible chess games, vastly larger than the 10^80 atoms in the observable universe. Musk is not entirely wrong about the direction of AI; he is wrong about the nature of the problem. To "solve" chess absolutely, a machine must determine the optimal outcome from every legal position assuming perfect play. That requires examining roughly 10^44 legal positions—a number beyond any current or near-future computational power. People often confuse 10^120 as the number of positions, but that is the number of games, of paths. The number of legal positions is about 10^44. It is a tiny fraction of 10^120 but still an endless abyss. Checkers—a far simpler game—took more than two decades of computation before being completely solved in 2026. Whenever I hear someone say "chess is about to be solved," I think of a saying from our craft: a game is solved when the machine understands every path; a game is killed when humans believe they no longer need to play it. Before 2026, Musk repeatedly dismissed chess as "too simple compared to real life." He argued chess has no fog of war, no technology tree, and fully symmetrical starting positions. For a man building empires in an uncertain environment, chess might indeed seem too clean. But he forgot that the symmetry of the board is exactly what reveals human asymmetry: given the same position, one person sees attack, another sees defense, a third sees the clock bleeding. What made this affair a real sports story was not the exchange of tweets but the language Chess.com chose. Alongside the usual joke inviting Musk to play a game, the official Chess.com account suddenly wrote in Vietnamese: "Do kỹ năng thôi"—roughly meaning "It's just skill." A small detail, but for Vietnamese people who treat chess as family memory—those afternoons when grandfather taught grandchildren the opening—it was a subtle nod. Perhaps they were just joking; perhaps they know the Vietnamese chess community is growing too fast to be ignored. In the endgame, the most interesting development appeared: instead of debating mathematics, Musk did not answer about 10^120 or about positions; he dismissed the opponent's account as run by an "intern." I call that the move of someone losing the battle. A true player never calls a rival an intern right after taking a sharp counterattack. I often tell young writers: machines can count all 10^120 games, but no machine knows the smell of coffee at a 3 a.m. match. The human game is not in the count. Those who cannot see the historical analogy may want to remember one thing: checkers was solved, and people still play checkers. Chess can be solved theoretically one day, just as checkers was solved. But solving a game is different from understanding why humans play it. Amateurs still play endgames in public parks, still regret a wrong move made 30 years ago, still call old friends to analyze a position nobody remembers. Perhaps the best answer to a shocking statement is not to find a mathematical flaw but to look back at your own arena. More than 40 years ago, I sat beside a chessboard in Quy Nhon and learned that the greatest games usually end with a handshake, not a declaration. Chess may be solved, but the handshake after the game cannot be solved. A record is never old; it is just waiting for someone who knows how to listen. The argument today will cool down, the tweets will be buried under the timeline, but a generation of Vietnamese children opening chess apps every night will still be there. They do not care in what year chess could be solved. They are learning to sit still before a difficult position, to lose gracefully, to win without despising their opponent. They play for reasons no algorithm can model. And perhaps that is precisely why chess will never be solved in the way outsiders imagine. When machines play better than humans, we begin to understand why we love the game. That love does not come from the mystery of algorithms; it comes from sleepless nights, unfinished games, moves we wish we could take back. Solving chess cannot kill that love. The only thing that can kill it is the way we tell its story—as something soon outdated. The story of Elon Musk and Chess.com may be just social-media gossip. But to me, it is a signal. A world where tech heroes declare our games soon obsolete is a world gradually forgetting that the value of sport is not measured by the shelf life of its rules. Chess has lived more than 1,500 years. As I write, somewhere in a coffee shop in Hanoi or Saigon, someone is still flipping the clock, tapping a pawn, and saying: "Check." Machines may know every path, but machines do not know why a move makes the human heart beat faster. That is the one thing that cannot be solved.