Cognitive Primitives for Automated Learning

Artificial Intelligence deals with the automated simulation of human intelligent behavior. Various aspects of human faculties are tackled using computational models. It is clear that brain as a cognitive machine is significantly different from a computer. A general intelligent machine needs to incorporate primitives that are identical or similar to those intrinsic to human. We argue for the identification, verification, and development of technologies that will deliver core primitives that are fundamental to human cognition that are applicable to various domains. Mnemonics for aural and for visual cognition are presented in [2, 3]. We propose a generalized set of cognitive primitives that can specialized to various applications viz. a) Existential, b) Structured Recurrence, c) Relative Attachment, d) Harmonized Counterparts, e) Cordial Counterparts and e) Discriminate Counterparts.