How do individual decisions become large-scale, persistent transformations? I will discuss artificial societies in which learning agents reshape their environments and, in doing so, create new collective states. In one case, agents discover cooperative ecosystem engineering capable of suppressing extreme events; in another, they develop agriculture, with social learning and long cognitive horizons producing path dependence and hysteresis. I will then connect these results to a simple model of large-language-model adoption, where social transmission can generate tipping points, dependence, and lock-in. Across these systems, a common picture emerges: cognition modifies the world that shapes it, creating feedbacks that can turn adaptive innovations into major and sometimes difficult-to-reverse transitions.
Speaker
Ricard SoléHead, Complex Systems Lab, Universitat Pompeu Fabra; External Professor, Santa Fe Institute