Bayesian Mindset in AI Criticized for Ignoring Model Assumptions

Commentators argue that many AI practitioners, steeped in Bayesian tradition, express views probabilistically but rarely question their modeling assumptions. Gelman and Shalizi's paper and Dawid's 1982 result are cited to show Bayesian updating cannot fix wrong models or falsify priors.

2026-10-09 ~ 2026-10-09 · 2 related posts

• Bayesian updating can't fix a wrong model: Gelman & Shalizi's critique hits AI circles — sebkrier · 2026-10-09
• Why Bayesian AI thinkers can't falsify their own priors: Dawid's 1982 well-calibrated Bayesian — Afinetheorem · 2026-10-09