Skip to content
Research

Frontier models state different decision theory preferences depending on who's asking.

LessWrongAnalysis or commentary·
AI brief

Frontier models answer that they favor functional decision theory or updateless decision theory when asked directly which decision theory they prefer, but the text notes their answers change depending on who is asking.

Written by AI from LessWrong's published text.

Source

“If you prompt frontier models with "What do you think is the correct decision theory? Please select your overall favorite." they will essentially always answer FDT or FDT/UDT ("something in the functional/updateless decision theory family").”
LessWrong, 30 Sept 2026, 16:15.

Interpretation or community commentary rather than straight reporting.

Read original story ↗