Elon Musk predicted that chess "will be fully solved one day," arguing its enormous decision tree overstates how many moves are worth considering, prompting Chess.com to fire back with a two-word rebuttal.
Musk Says Chess Will Eventually Be Solved
"The actual number of moves that are not utterly stupid in chess is tiny and chess will be fully solved one day," Musk wrote Thursday on X, replying to a post illustrating how rapidly possible continuations multiply.
Chess.com quote-posted Musk’s remark with only "skill issue," turning the technical debate into a viral roast.
Chess, however, remains unsolved. University of Basel artificial-intelligence teaching material from 2024 estimates roughly 10^40 reachable positions, while a game tree can reach around 10^154 nodes when assuming about 35 legal moves per position and 50 moves per player.
AI Engines Already Prune Vast Search Trees
Musk’s intuition echoes a principle used by modern chess engines: rather than exhaustively evaluating every possible continuation, they prioritize promising lines.
Alphabet Inc.’s (NASDAQ:GOOG)(NASDAQ:GOOG) DeepMind says AlphaZero used a neural network to guide Monte Carlo Tree Search toward promising branches, examining about 60,000 positions per second compared with roughly 60 million for Stockfish in its 2018 evaluation. AlphaZero defeated the then-strongest version of Stockfish after learning chess through self-play.
That fits Musk’s broader conviction that machines will surpass humans cognitively. He has predicted AI could become smarter than any individual human and later declared that "we have entered the Singularity" as AI capabilities accelerated.
Checkers Precedent Fuels Musk’s Chess Forecast
After Chess.com’s response, Musk doubled down. "Chess is hard for humans, but not for computers. One day, it will be fully solved like checkers," he wrote. "Between now and then, or even after, have fun. People enjoy many games/sports where machines are vastly better."
His checkers comparison has precedent. University of Alberta researchers announced in 2007 that checkers had been solved after an 18-year computing effort, proving that perfect play from both sides produces a draw.
Musk’s forecast, for now, remains a prediction. It is worth noting that modern engines like Stockfish, LCZero and Berserk already play far beyond human strength, but no system thus far has mathematically proved whether perfectly played chess ends in a White win, Black win or draw.
Photo courtesy: Shutterstock
Login to comment