Many Agents Are Better Than One


Georgeff, M. P. (1987). Many agents are better than one. In The Frame Problem in Artificial Intelligence (pp. 59-75). Morgan Kaufmann.


This paper aims to show how much of the frame problem can be alleviated by using domain models that allow for the simultaneous occurrence of actions and events. First, a generalized situation calculus is constructed for describing and reasoning about events in multiagent settings. Notions of independence and causality are then introduced and it is shown how they can be used to determine the persistence of facts over time. Finally, it is shown how these notions, together with traditional predicate circumscription, make it possible to retain a simple model of action while avoiding most of the difficulties associated with the frame problem.

Read more from SRI