On the Other Hand - Medical Report: Infectious Symbolophobia

Strains Many different strains have been identified, each with its characteristic symptoms. Here we will briefly sketch only a few of the most common. Formalist vision arises from imagining oneself to be trapped in a box into which people shove little slips of paper with Chinese scribbling on them. There is perpetually nothing else to do, so the victim matches these symbols with similar symbols found in a large book, follows the rules contained therein, and then passes some other symbols out under the door. The victim often feels as if he were a computer. Our sufferer, under the spell of the formalist vision, claims that he understands nothing—so neither can computers—because he is like a computer and he understands nothing. The Gödelian itch is a belief that Gödel’s incompleteness theorem doesn’t apply to humans, who can therefore correctly decide whether or not they will terminate on every input. In one sense this is trivial, since all humans will indeed eventually terminate no matter what their input, but the Gödelian itch is characterized by the delusion that humans can compute noncomputable functions. This strain of symbolophobia, thought to have been eradicated over a decade ago, has recently reappeared in a new and particularly virulent form. Epidemiological studies suggest that a single human vector may be the source of this recent outbreak. Double mysterianism is the belief that since intelligence (consciousness, etc.) is mysterious, we should find another mystery, perhaps called antimystery, which will cancel it out in an explosion of pure illumination. The commonest variety at present is quantum mysterianism, where the second mystery is the problem of making sense of quantum theory. Patients with quantum mysteriComputers are the only machines which both manipulate symbols and are manipulated by them; they are “physical symbol systems.” As we know, AI uses this insight to implement systems capable of thought. However, some people find this idea of a symbolic machine very unsettling. This tension can be a clinical precursor of a disorder called symbolophobia. Although we thought we had independently discovered and named this disorder, we subsequently found that it was already known to medical science by the same name. Doctor and Kahn (1989) have this to say when discussing symbolophobia: “Individuals may fear the symbols themselves with or without understanding their unconscious representation.” Those suffering from symbolophobia take the word symbol to be symbolic of all that is wrong with AI. Symbols are blamed for rigidity, ungroundedness, disembodiment, formality, seriality, logicality, etc. Victims of symbolophobia often find reassurance by imagining computers that are subsymbolic, analog, biological, socially-situated, or quantum. Most forms of symbolophobia seem to be contagious. The etiology and epidemiology of this worrisome affliction are not fully understood, but some early treatment protocols seem effective and early innoculation is advised.