A reply to “parallel Computation and the Mind‐Body Problem”

(1) Thagard argues that increased speed has metaphysical importance because only intelligent creatures (or machines) quick enough to adequately deal with the demands of their environment will survive. Thus, the best possible serial simulation of a parallel algorithm, while strictly possible, may be hopelessly slow. Lest we consider this conceptually irrelevant, Thagard admonishes us not to ignore such real-world limitations, observing that our theories relating matter and intelligence need no more account for the merely conceivable than Newtonian mechanics need explain worlds with negative gravitation. But this still shows only that the hardware must be able to run the software fast enough to be useful/survive. A program running on a parallel machine that produced some sort of intelligence will also run on a serial machine, and this is enough to show the hardware irrelevant for explaining the nature, if not the evolution, of that particular intelligence. If the program runs too slowly on the serial machine to be useful we would not say that it no longer demonstrates intelligence but only that it is too slow, or that the particular approach, though successful, is impractical. (Some kind of practicality test is relevant to determining whether we have produced an intelligence that works in the same way as, as opposed to being functionally equivalent to, some aspect of human intelligence, but Thagard does not SO constrain his position). The analogy to negative gravitation is to instruct us not to muddy our thinking or burden any non-functionalist position with a purely theoretical multiple instantiation hypothesis and speculation about exotic serial