Elon Musk and Chess.com just turned a technical question into a public argument: can AI ever fully solve chess? Musk’s claim is that the number of sensible moves is small, computers already crush humans, and a future system — including what he calls ASI — will compress the game the way checkers was solved. Chess.com’s reply is the Shannon-number argument: there are vastly more possible games than atoms in the universe, so mapping every line is not the same as “solving” the game. Both sides are talking past each other a bit. Engines already play far above every human. That is not the same as a complete tablebase for the starting position. Checkers is solved. Chess is weakly solved in some endgames and still open at the root. The interesting part is not who won the reply thread. It is that “solved” now means two different things: unbeatable play versus a finished map of every legal position. Would you count chess as solved when no human can beat the machine, or only when every line i...