Stephen Pratt, Professor, Arizona State University

Social insects are paradigms of decentralized organization. Complex colony traits emerge from the interactions of many leaderless workers, each applying appropriate decision rules to limited local information. In this talk, I will describe efforts to understand this process through a combination of experimentation and computational analysis. This work uses nest site choice by colonies of rock ants acting as a model system for collective intelligence. Sharing the burden of information processing contributes to enhanced colony performance, or “wisdom of crowds”, but I will show that it can sometimes instead lead to paradoxically worse performance. Finally, I will describe how information theory can shed new light on the distinctive communication behavior that allows colony members to share information about nest sites and other important resources.