A lay-readable book on the mathematics of optimal intelligence (Bayes, Solomonoff induction, AIXI), the specification problem, and what they imply about modern AI and alignment.
machine-learninglatexreinforcement-learningbookagiartificial-intelligencebayesian-inferenceai-safetykolmogorov-complexityalgorithmic-information-theoryai-alignmentlarge-language-modelsaixisolomonoff-inductionspecification-problem
-
Updated
Jun 25, 2026 - TeX