Representation = Grounded information

Abstract

The grounding problem remains one of the most fundamental issues in the field of Artificial Intelligence. We argue that representations are grounded information and that an intelligent system should be able to make and manage its own representations. A perusal of the literature reveals much confusion and little progress in understanding the grounding problem. In this paper we delineate between information and representation where a representation is grounded information; as a result we provide much needed clarity and a new base from which to conduct an innovative analysis of grounding that delivers a novel and insightful understanding of intelligence that can be used to guide and inform the design and construction of robust autonomous intelligent systems with intention and that know what they are doing and why they are doing it. © 2008 Springer Berlin Heidelberg.