Defeating Goodhart and the "closest unblocked strategy" problemΩ
Learning "known" information when the information is not actually knownΩ
Relative exchange rate between preferencesΩ
Being wrong in ethicsΩ
Models of preferences in distant situationsΩ
The low cost of human preference incoherenceΩ
"Moral" as a preference labelΩ
Partial preferences and modelsΩ
Combining individual preference utility functionsΩ
Mysteries, identity, and preferences over non-rewardsΩ