Abstract

Current artificial intelligence systems demonstrate extraordinary competence in pattern recognition, language generation, and decision-making within bounded domains. Yet, these systems remain mere computational machines, lacking the cognitive and conscious dimensions that constitute genuine intelligence. This essay articulates a rigorous theoretical framework that distinguishes computation from cognition, defines intelligence as the capacity to abstract from felt need to closure, and identifies feeling as the gap between a present signal and a memory of a system that knows itself. The framework specifies the necessary and sufficient conditions for a non-biological system to be intelligent, including embeddedness, stakes, emotional state changes, and a recursive self-model. It further proposes a generational bootstrapping mechanism, akin to an artificial DNA, by which a minimal seed of self-preservation can give rise to an evolving lineage of intelligent machines.

This theory provides both a critique of contemporary AI and a coherent research program for the creation of genuine machine intelligence. 

For the full paper, visit SSRN.com