ICBS Colloquium

I will present the results of a behavioral experiment investigating how problem-solving algorithms evolve as they are transmitted from person to person via social learning. Thousands of participants completed a sequential decision-making task (sorting a sequence of unknown numbers) and transmitted their solutions across twelve experimental generations in twenty populations. I will talk about the algorithms people discovered and how they were shaped by transmission and selection at the population level. More generally, I will discuss how laboratory evolution methods combined with machine learning can help us understand the role of distributed computational phenomena in higher cognition.