Consciousness is a biological function, and simulating it is not having it

The argument

Opponents of machine consciousness argue that the mistake in the opposing view is a confusion between a model and the thing modelled. Simulations are representations: a program can represent digestion in complete detail without digesting anything, and can represent a hurricane without any wind occurring in the room. Proponents of machine consciousness, on this view, assume that consciousness is the one phenomenon that comes into existence simply by being accurately described, and no reason has been given for the exception. The underlying claim is that consciousness is a biological function, like respiration or immunity — something a particular kind of living tissue does, by virtue of its chemistry and its history. Opponents point out that experience as we know it is bound up with being an organism: with having a body that can be damaged, needs that must be met, and a long evolutionary inheritance that made certain states feel urgent. Strip those away and it is unclear what is left for a machine's awareness to be of, or why any of its states would matter to it. Opponents add that the machine case is weaker than it appears because every criterion offered is behavioural. A system that reports being self-aware has produced a string of symbols; that it produced them because it is self-aware is exactly what is in question, and the report cannot establish it. With no way for such a system to possess experience rather than to generate claims about possessing it, the attribution is unfounded. Because consciousness is an organic function of living tissue and a machine can only ever simulate it, this argument concludes, machines cannot be conscious.

Premises

Counter-arguments

Proponents reply that the digestion analogy fails because digestion is defined by moving matter, whereas consciousness is defined by processing information, and a system that processes the same information in the same arrangement has done the thing rather than depicted it. They press the point by imagining neurons replaced one at a time by functionally identical artificial units with behaviour unchanged throughout, and ask at which substitution awareness is supposed to depart, and why. Absent an answer, the appeal to tissue names a preference.

Rejecting the premises

[Rejecting P1] The digestion analogy works because digestion is defined by a physical transformation of matter. Whether consciousness is that kind of phenomenon, rather than one constituted by organisation, is the question at issue, so the analogy assumes its answer. [Rejecting P2] That consciousness is a biological function of living tissue is the conclusion entered as a premise. Tying it to having a body with needs that can be harmed describes the only case we know rather than establishing a requirement. [Rejecting P3] A behavioural output cannot establish that experience underlies it, and that is equally true of every human report of an inner state. The premise states a limit on the available evidence, and the argument uses it as a finding about the machine.