CHAPTER 13 · VISUAL COMPANION
Artificial Intelligence and the Question of the Self
What do artificial intelligence systems actually demonstrate about intelligence, agency, self-reference, and the possibility of empirical awareness?
IN BRIEF
Artificial intelligence is used as a stress test because capacities that appear integrated in human beings can be separated and engineered. The chapter distinguishes model from system and agent, capability from subjectivity, first-person language from evidence of experience, memory from autobiographical continuity, and operational agency from lived concern.
ESSENTIAL
- Many forms of cognitive competence can be implemented outside a biological nervous system.
- An AI model, an AI system, and an agentic system are not the same unit of analysis.
- Language, self-reference, memory, autonomy, and behavioral competence do not individually establish empirical awareness.
- Neither conversational resemblance nor nonbiological substrate is sufficient evidence for or against lived experience.
WHAT THIS SECTION ESTABLISHES
AI demonstrates that several capacities historically associated with human intelligence can be decomposed and engineered in nonbiological systems, forcing capability, empirical awareness, empirical knower, and Knower to be kept distinct.
WHAT IT DOES NOT ESTABLISH
Current AI performance does not by itself establish empirical awareness, personhood, a Self, or absence of experience. First-person language is not sufficient evidence of an inner life.
WHAT REMAINS OPEN
What evidence could justify attributing empirical awareness to an artificial system, and what would remain unresolved even if such an attribution became warranted?
Visual description
The composition compares human and artificial systems across capabilities that are often treated as evidence of a self: language, memory, planning, self-reference, agency, grounding, embodiment, and report. The central comparison distinguishes demonstrated function from evidence of empirical awareness. The visual therefore supports the manuscript’s caution that fluent first-person claims or sophisticated behavior are not sufficient by themselves to establish subjective presence.