Multiagent metareasoning through organizational design

Abstract

We describe an approach to multiagent metareasoning that uses organizational design to focus each agent's reasoning on the aspects of its local problem to which it can make the most worthwhile contributions to joint behavior. We summarize an organizational design problem that explicitly considers the quantitative impact that a design has on both the quality of the agents' behaviors and their reasoning costs. We overview an automated organizational design process that can approximately solve our design problem via incremental search, and outline techniques that efficiently estimate the incremental impact of a candidate organizational influence.