Issues with the solution you mentioned
Troll bridge
Counterfactual mugging
Agent simulates predictor
Coordination games (I discussed these in this post)
Game theory and bargaining in general
So most of the problems seem to be related to making UDT work with logical uncertainty?
Some of them specifically require updateless reasoning, some don’t or might not. All of them involve logical uncertainty.
Issues with the solution you mentioned
Troll bridge
Counterfactual mugging
Agent simulates predictor
Coordination games (I discussed these in this post)
Game theory and bargaining in general
So most of the problems seem to be related to making UDT work with logical uncertainty?
Some of them specifically require updateless reasoning, some don’t or might not. All of them involve logical uncertainty.