Idealized Bayesians don’t have to be logically omniscient—they can have a prior which assigns probability to logically impossible worlds.
Idealized Bayesians don’t have to be logically omniscient—they can have a prior which assigns probability to logically impossible worlds.