A Framework for evolving complexity in mobile social agents

Walid Nasrallah, American University of Beirut, Libanon

Using iterated prisoners' dilemma (IPD) as en operational metaphor for social interaction is an established convention in agent-based simulation of evolving norms. Many past investigations have found that strategies at a low level of complexity tend to out-compete more complex strategies. The introduction of contingent movement, whereby agents choose to avoid unfriendly interactions by walking away from them, adds a new twist this line of research. Agents that move on purpose can change the nature of the selective pressures that act on their own evolution by shifting their location in space, which changes the mix of neighbors with whom they will interact in future. This paper aims to provide a framework for answering the question: ``Does the selective pressure for less complex strategies remain when contingent movement is a possible choice?" The answer to the question has implications for the origins of complexity in both life and human society.