A Set of Component Types as a Substrate for Intelligent Systems

Abstract Advances in AI are weakening the ability to distinguish biological from artificial intelligence based on external criteria of which the Turing test is most famous. The development of internal criteria would enable comparison of intelligent systems at the level of their internal organization and substructure. The article proposes such a criterion based on the definition of a substrate as a set of component types. A number of basic concepts are refined such as component subassembly organization. A new concept of complexion is introduced meaning a set of conditions that realize the system in its physical form and operation. The proposed formulation of the definition of substrate is methodologically grounded on set theory and aims to unify the understanding of the concept of substrate in the fields of technical sciences and philosophy.