x
Bayesian Utility: Representing Preference by Probability Measures — LessWrong