Technology ❯Artificial Intelligence ❯Applications ❯Gaming
By underestimating his 2839 FIDE rating at 1800–2000, the exchange highlights language models’ struggle with precise game state tracking.