Raphael Malyankar; Findler; Nicholas V; Heck, Peter S
We are currently engaged in investigating intelligent agent societies from a macro-level view, by defining certain emergent variables that describe the behavior of such systems and investigating relationships between these variables. Part of this research investigates the role played by satisficing models in coordination within intelligent agent societies. This paper discusses the tradeoffs between modeling of other agents and other coordination-related activities and phenomena, in particular system coherence and optimality of coordination. We describe our past work in this field and our current activities in the area of relating satisficing to coordination. Work in two main areas is described: first, investigation of the effects of satisficing on system coherence, and second, the relationship between satisficing and agent autonomy.