x
Open Problems in AIXI Agent Foundations — LessWrong