“The Nature of Mental States,” by the philosopher Hilary Putnam, was first published in 1967 (originally under the title “Psychological Predicates”). It is one of the founding statements of functionalism, the view that what makes something a mental state - a pain, a belief, a desire - is the functional role it plays in a system, not the physical stuff the system is made of. On this view, a mental state is defined by its causal relations: what tends to cause it, what it tends to cause, and how it interacts with other mental states.
Putnam paired functionalism with the idea now called multiple realizability. He argued it was overwhelmingly likely that the same psychological state, such as pain, is realized by very different physical mechanisms in humans, octopuses, and hypothetical alien creatures. If pain can be a particular brain state in one creature and a completely different physical configuration in another, then pain cannot simply be identical to any one physical state. This was a direct challenge to the mind-brain identity theory, which held that each mental state just is a specific type of physical brain state.
The argument matters enormously for artificial intelligence. If mental states are defined by function rather than by being made of neurons, then in principle a system built from silicon could instantiate the same mental states as a biological brain, provided it realized the same functional organization. Functionalism is, in effect, the philosophical underpinning for taking the idea of machine minds seriously, which is also why critics such as John Searle aimed the Chinese Room argument squarely at it.
Why a general reader should care: the question of whether software can ever truly think, rather than merely simulate thinking, traces back to this debate. Putnam’s paper is where the case that substrate may not matter was first made carefully.