Algorithmic Learning in a Random World