Optimality Issues of Universal Greedy Agents with Static Priors

摘要

Finding the universal artificial intelligent agent is the old dream of AI scientists. Solomonoff Induction was one big step towards this, giving a universal solution to the general problem of Sequence Prediction, by defining a universal prior distribution. Hutter defined AIXI which extends the latter to the Reinforcement Learning framework, where almost all if not all AI problems can be formulated. However, new difficulties arise, because the agent is now active, whereas it is only passive in the Sequence Prediction case. This makes proving AIXI's optimality difficult. In fact, we prove that the current definition of AIXI can sometimes be only sub-optimal in a certain sense, and we generalize this result to infinite horizon agents and to any static prior distribution.