Algorithmic Learning Dynamics

Brian and Sean discuss how algorithms learn better through experimentation rather than teaching. Brian shares insights from setting up a computer stock market in 1988 with AI investors, showing a phase transition in outcomes based on the rate of exploration.