An Agent-Based Design for Distributed Artificial Intelligence

摘要

Recent EEG brain mapping studies, along with studies utilizing Functional Magnetic Resonance Imaging (FMRI) and Positronic Emission Tomography (PET) scans, suggest that the brain operates as a Distributed Intelligence Processing System (DIPS) [I]. This process, within Artificial Intelligence research, is known as Distributed Artificial Intelligence (DAI) and is intended to mimic human processes for complex learning, planning, and decision making [2]. Here, we present an Intelligent Software Agent (ISA) architecture and design to facilitate distributed learning, processing, and communication within an Artificially Intelligent System (AIS), as well as, spatially separated/ distributed extensions to AI systems.