factual
Toroidal Chess Test Shows No Reasoning
Asked about a chess variant where the board is a torus (opposite edges adjacent), GPT-4 produced sophisticated-sounding but entirely wrong filler instead of seeing the trivially correct answer that white can capture black's king on move one, demonstrating it doesn't envision or reason about a changed board but only echoes familiar chess discourse.
factualpending
Speaker
Sean CarrollEvidence Quote
“GPT is not envisioning a chessboard and seeing how it would change by this new toroidal chessboard kind of thing, and is not analyzing or reasoning about it, because LLMs don't do that. They don't model the world.”
Created: 6/13/2026, 12:12:00 AM
Your notes
Loading notes...