Functional equivalence is the thesis that what matters about a cognitive system is its functional organization — the patterns of input, processing, and output it instantiates — and not the physical substrate that realizes the organization. On this view, if a silicon system and a biological system implement the same functional relations, they are cognitively equivalent. The thesis underwrites computational theories of mind and the Turing test. Chalmers's framework accepts that functional equivalence captures cognitive equivalence but insists it does not settle phenomenal equivalence: two systems can be functionally identical and phenomenally different.
The thesis has a distinguished pedigree running from Turing through Hilary Putnam's functionalism to contemporary computational cognitive science. Its force is that it explains how beings of radically different physical composition — humans, aliens, hypothetical AI — could share mental properties. If what matters is the pattern, not the material, then minds are multiply realizable, and the question of machine consciousness becomes tractable in principle.
Chalmers's engagement with functional equivalence is nuanced. He accepts that functional equivalence captures a great deal — plausibly everything about cognitive function, and perhaps even everything about psychological consciousness. What he rejects is the further claim that functional equivalence captures phenomenal consciousness. The zombie argument is precisely an argument that functional equivalence and phenomenal equivalence can come apart.
For the AI context, the thesis matters because large language models increasingly approach functional equivalence with human cognitive performance on many tasks. If functional equivalence were sufficient for consciousness, this convergence would settle the question. Chalmers's framework says it does not — that the question of what, if anything, it is like to be a functionally-equivalent system remains open.
The thesis is associated with Alan Turing's 1950 Computing Machinery and Intelligence and was given its modern formulation by Hilary Putnam in the 1960s. It became central to computational theories of mind and to the computational theory of mind that Noë and others have contested.
Functional equivalence captures cognitive equivalence. Same pattern, same cognition.
It does not capture phenomenal equivalence. Same pattern, possibly different experience.
It underwrites the Turing test's logic. And shows what the test does and does not establish.
AI convergence on human performance is functional equivalence. Whatever that is worth.