Epistemic Planning in a Fast and Slow Setting

Abstract

AI applications are by now pervading our everyday life. Nonetheless, most of these systems lack many capabilities that, we humans, naturally consider to be included in a notion of “intelligence”. In this paper we present a multi-agent system, inspired by the cognitive theory known as thinking fast and slow by D. Kahneman, to solve Multi-agent Epistemic Planning (MEP) problems. This is an instance of a general AI architecture, referred to as SOFAI (for Slow and Fast AI). This paradigm exploits multiple solving approaches (referred to as fast and slow solvers) and a metacognition module to arbitrate between them and enhance the reasoning process, that, in this specific case, is concerned with planning in epistemic settings. The behavior of this system is then compared to a state-of-the-art MEP solver, showing that the newly introduced system presents better results in terms of generality, solving a much wider set of problems with an acceptable trade-off between solving times and solution accuracy.