Research

How does agent intelligence affect collective behavior? What forms of emergence can arise from simple local rules, and when do more sophisticated decision-making capabilities become necessary?

I investigate these questions through controlled multi-agent simulation, comparing agents with different levels of complexity under shared environments and interaction protocols.

Collective agreement can emerge even when the shared decision is incorrect. I study how local information, asymmetric influence, communication, and decision rules contribute to incorrect consensus, polarization, and unstable collective dynamics.

I explore how decentralized agents can reorganize their interaction networks using limited local information, and how such self-organization affects connectivity, efficiency, robustness, and system-level performance.

I develop simulation and experimental platforms for studying interactions among rule-based agents, AI agents, and human participants.

These platforms support controlled comparisons of individual decisions and collective outcomes in shared environments.