Machine super intelligence

Abstract

This doctoral dissertation investigates the optimal behavior of agents in unknown computable environments, also known as universal artificial intelligence. The theoretical agents in question are able to learn to perform optimally in a broad range of settings and can be mathematically proven to upper bound the performance of general-purpose computable agents. The dissertation examines the circumstances under which the behavior of these universal agents converges to optimal, the relationship between universal artificial intelligence theory and the concept and definition of intelligence, the limits that computable agents face when trying to approximate theoretical super intelligent agents, and finally the big picture implications of super intelligent machines. – AI-generated abstract.