We, Robots

by Simon Ings

  And yet they are like us, for all the awesome differences. We know already that a digital machine can never spark with consciousness; regardless of the task to which we harness it, or of the physical processes we simulate in it, it will remain forever apsychic. Since, to simulate man, it is necessary that we reproduce certain of his fundamental contradictions, only a system of mutually gravitating antagonisms—a personoid—will resemble, in the words of Canyon, whom Dobb cites, a “star contracted by the forces of gravity and at the same time expanded by the pressure of radiation.” The gravitational center is, very simply, the personal “I,” but by no means does it constitute a unity in either the logical or the physical sense. That is only our subjective illusion! We find ourselves, at this stage of the exposition, amid a multitude of astounding surprises. One can, to be sure, program a digital machine in such a way as to be able to carry on a conversation with it, as if with an intelligent partner. The machine will employ, as the need arises, the pronoun “I” and all its grammatical inflections. This, however, is a hoax! The machine will still be closer to a billion chattering parrots—howsoever brilliantly trained the parrots be—than to the simplest, most stupid man. It mimics the behavior of a man on the purely linguistic plane and nothing more. Nothing will amuse such a machine, or surprise it, or confuse it, or alarm it, or distress it, because it is psychologically and individually No One. It is a Voice giving utterance to matters, supplying answers to questions; it is a Logic capable of defeating the best chess player; it is—or, rather, it can become—a consummate imitator of everything, an actor, if you will, brought to the pinnacle of perfection, performing any programmed role—but an actor and an imitator that is, within, completely empty. One cannot count on its sympathy, or on its antipathy. It works toward no self-set goal; to a degree eternally beyond the conception of any man it “doesn’t care,” for as a person it simply does not exist… It is a wondrously efficient combinatorial mechanism, nothing more. Now, we are faced with a most remarkable phenomenon. The thought is staggering that from the raw material of so utterly vacant and so perfectly impersonal a machine it is possible, through the feeding into it of a special program—a personetic program—to create authentic sentient beings, and even a great many of them at a time! The latest IBM models have a top capacity of one thousand personoids. (The number is mathematically precise, since the elements and linkages needed to carry one personoid can be expressed in units of centimeters-grams-seconds.)

  Personoids are separated one from another within the machine. They do not ordinarily “overlap,” though it can happen. Upon contact, there occurs what is equivalent to repulsion, which impedes mutual “osmosis.” Nevertheless, they are able to interpenetrate if such is their aim. The processes making up their mental substrates then commence to superimpose upon each other, producing “noise” and interference. When the area of permeation is thin, a certain amount of information becomes the common property of both partially coincident personoids—a phenomenon that is for them peculiar, as for a man it would be peculiar, if not indeed alarming, to hear “strange voices” and “foreign thoughts” in his own head (which does, of course, occur in certain mental illnesses or under the influence of hallucinogenic drugs). It is as though two people were to have not merely the same, but the same memory; as though there had occurred something more than a telepathic transference of thought—namely, a “peripheral merging of the egos.” The phenomenon is ominous in its consequences, however, and ought to be avoided. For, following the transitional state of surface osmosis, the “advancing” personoid can destroy the other and consume it. The latter, in that case, simply undergoes absorption, annihilation—it ceases to exist (this has already been called murder). The annihilated personoid becomes an assimilated, indistinguishable part of the “aggressor.” We have succeeded—says Dobb—in simulating not only psychic life but also its imperilment and obliteration. Thus we have succeeded in simulating death as well. Under normal experimental conditions, however, personoids eschew such acts of aggression. “Psychophagi” (Castler’s term) are hardly ever encountered among them. Feeling the beginnings of osmosis, which may come about as the result of purely accidental approaches and fluctuations—feeling this threat in a manner that is of course nonphysical, much as someone might sense another’s presence or even hear “strange voices” in his own mind—the personoids execute active avoidance maneuvers; they withdraw and go their separate ways. It is on account of this phenomenon that they have come to know the meaning of the concepts of “good” and “evil.” To them it is evident that “evil” lies in the destruction of another, and “good” in another’s deliverance. At the same time, the “evil” of one may be the “good” (i.e., the gain, now in the nonethical sense) of another, who would become a “psychophage.” For such expansion—the appropriation of someone else’s “intellectual territory”—increases one’s initially given mental “acreage.” In a way, this is a counterpart of a practice of ours, for as carnivores we kill and feed on our victims. The personoids, though, are not obliged to behave thus; they are merely able to. Hunger and thirst are unknown to them, since a continuous influx of energy sustains them—an energy whose source they need not concern themselves with (just as we need not go to any particular lengths to have the sun shine down on us). In the personoid world the terms and principles of thermodynamics, in their application to energetics, cannot arise, because that world is subject to mathematical and not thermodynamic laws.

  Before long, the experimenters came to the conclusion that contacts between personoid and man, via the inputs and outputs of the computer, were of little scientific value and, moreover, produced moral dilemmas, which contributed to the labeling of personetics as the cruellest science. There is something unworthy in informing personoids that we have created them in enclosures that only simulate infinity, that they are microscopic “psychocysts,” capsulations in our world. To be sure, they have their own infinity; hence Sharker and other psychoneticians (Falk, Wiegeland) claim that the situation is fully symmetrical: the personoids do not need our world, our “living space,” just as we have no use for their “mathematical earth.” Dobb considers such reasoning sophistry, because as to who created whom, and who confined whom existentially, there can be no argument. Dobb himself belongs to that group which advocates the principle of absolute nonintervention—“noncontact”—with the personoids. They are the behaviorists of personetics. Their desire is to observe synthetic beings of intelligence, to listen in on their speech and thoughts, to record their actions and their pursuits, but never to interfere with these. This method is already developed and has a technology of its own—a set of instruments whose procurement presented difficulties that seemed all but insurmountable only a few years ago. The idea is to hear, to understand—in short, to be a constantly eavesdropping witness—but at the same time to prevent one’s “monitorings” from disturbing in any way the world of the personoids. Now in the planning stage at MIT are programs (APHRON II and EROT) that will enable the personoids—who are currently without gender—to have “erotic contacts,” make possible what corresponds to fertilization, and give them the opportunity to multiply “sexually.” Dobb makes clear that he is no enthusiast of these American projects. His work, as described in Non Serviam, is aimed in an altogether different direction. Not without reason has the English school of personetics been called “the philosophical polygon” and “the theodicy lab.” With these descriptions we come to what is probably the most significant and, certainly, the most intriguing part of the book under discussion—the last part, which justifies and explains its peculiar title.

  Dobb gives an account of his own experiment, in progress now for eight years without interruption. Of the creation itself he makes only brief mention; it was a fairly ordinary duplicating of functions typical of the program JAHVE VI, with slight modifications. He summarizes the results of “tapping” this world, which he himself created and whose development he continues to follow. He considers this tapping to be unethic
al, and even, at times, a shameful practice. Nevertheless, he carries on with his work, professing a belief in the necessity, for science, of conducting such experiments also—experiments that can in no way be justified on moral—or, for that matter, on any other non-knowledge-advancing—grounds. The situation, he says, has come to the point where the old evasions of the scientists will not do. One cannot affect a fine neutrality and conjure away an uneasy conscience by using, for example, the rationalization worked out by vivisectionists—that it is not in creatures of full-dimensional consciousness, not in sovereign beings that one is causing suffering or only discomfort. In the personoid experiments we are accountable twofold, because we create and then enchain the creation in the schema of our laboratory procedures. Whatever we do and however we explain our action, there is no longer an escape from full accountability.

  Many years of experience on the part of Dobb and his coworkers at Oldport went into the making of their eight-dimensional universum, which became the residence of personoids bearing the names ADAN, ADNA, ANAD, DANA, DAAN, and NAAD. The first personoids developed the rudiment of language implanted in them and had “progeny” by means of division. Dobb writes, in the Biblical vein, “And ADAN begat ADNA. ADNA in turn begat DAAN, and DAAN brought forth EDAN, who bore EDNA…” And so it went, until the number of succeeding generations had reached three hundred; because the computer possessed a capacity of only one hundred personoid entities, however, there were periodic eliminations of the “demographic surplus.” In the three-hundredth generation, personoids named ADAN, ADNA, ANAD, DANA, DAAN, and NAAD again make an appearance, endowed with additional numbers designating their order of descent. (For simplicity in our recapitulation, we will omit the numbers.) Dobb tells us that the time that has elapsed inside the computer universum works out to—in a rough conversion to our equivalent units of measurement—from 2 to 2.5 thousand years. Over this period there has come into being, within the personoid population, a whole series of varying explanations of their lot, as well as the formulation by them of varying, and contending, and mutually excluding models of “all that exists.” That is, there have arisen many different philosophies (ontologies and epistemologies), and also “metaphysical experiments” of a type all their own. We do not know whether it is because the “culture” of the personoids is too unlike the human or whether the experiment has been of too short duration, but, in the population studied, no faith of a form completely dogmatized has ever crystallized—a faith that would correspond to Buddhism, say, or to Christianity. On the other hand, one notes, as early as the eighth generation, the appearance of the notion of a Creator, envisioned personally and monotheistically. The experiment consists in alternately raising the rate of computer transformations to the maximum and slowing it down (once a year, more or less) to make direct monitoring possible. These changes in rate are, as Dobb explains, totally imperceptible to the inhabitants of the computer universum, just as similar transformations would be imperceptible to us, because when at a single blow the whole of existence undergoes a change (here, in the dimension of time), those immersed in it cannot be aware of the change, because they have no fixed point, or frame of reference, by which to determine that it is taking place.

  The utilization of “two chronological gears” permitted that which Dobb most wanted—the emergence of a personoid history, a history with a depth of tradition and a vista of time. To summarize all the data of that history recorded by Dobb, often of a sensational nature, is not possible. We will confine ourselves, then, to the passages from which came the idea that is reflected in the book’s title. The language employed by the personoids is a recent transformation of the standard English whose lexicon and syntax were programmed into them in the first generation. Dobb translates it into essentially normal English but leaves intact a few expressions coined by the personoid population. Among these are the terms “godly” and “ungodly,” used to describe believers in God and atheists.

  ADAN discourses with DAAN and ADNA (personoids themselves do not use these names, which are purely a pragmatic contrivance on the part of the observers, to facilitate the recording of the “dialogues”) upon a problem known to us also—a problem that in our history originates with Pascal but in the history of the personoids was the discovery of a certain EDAN 197. Exactly like Pascal, this thinker stated that a belief in God is in any case more profitable than unbelief, because if truth is on the side of the “ungodlies” the believer loses nothing but his life when he leaves the world, whereas if God exists he gains all eternity (glory everlasting). Therefore, one should believe in God, for this is dictated very simply by the existential tactic of weighing one’s chances in the pursuit of optimal success.

  ADAN 300 holds the following view of this directive: EDAN 197, in his line of reasoning, assumes a God that requires reverence, love, and total devotion, and not only and not simply a belief in the fact that He exists and that He created the world. It is not enough to assent to the hypothesis of God the Maker of the World in order to win one’s salvation; one must in addition be grateful to that Maker for the act of creation, and divine His will, and do it. In short, one must serve God. Now, God, if He exists, has the power to prove His own existence in a manner at least as convincing as the manner in which what can be directly perceived testifies to His being. Surely, we cannot doubt that certain objects exist and that our world is composed of them. At the most, one might harbor doubts regarding the question of what it is they do to exist, how they exist, etc. But the fact itself of their existence no one will gainsay. God could with this same force provide evidence of His own existence. Yet He has not done so, condemning us to obtain, on that score, knowledge that is roundabout, indirect, expressed in the form of various conjectures—conjectures sometimes given the name of revelation. If He has acted thus, then He has thereby put the “godlies” and the “ungodlies” on an equal footing; He has not compelled His creatures to an absolute belief in His being but has only offered them that possibility. Granted, the motives that moved the Creator may well be hidden from His creations. Be that as it may, the following proposition arises: God either exists or He does not exist. That there might be a third possibility (God did exist but no longer does, or He exists intermittently, in oscillation, or He exists sometimes “less” and sometimes “more,” etc.) appears exceedingly improbable. It cannot be ruled out, but the introduction of a multivalent logic into a theodicy serves only to muddle it.

  So, then, God either is or He is not. If He Himself accepts our situation, in which each member of the alternative in question has arguments to support it—for the “godlies” prove the existence of the Creator and the “ungodlies” disprove it—then from the point of view of logic we have a game whose partners are, on one side, the full set of the “godlies” and “ungodlies,” and, on the other, God alone. The game necessarily possesses the logical feature that for unbelief in Him God may not punish anyone. If it is definitely unknown whether or not a thing exists—some merely asserting that it does and others, that it does not—and if in general it is possible to advance the hypothesis that the thing never was at all, then no just tribunal can pass judgment against anyone for denying the existence of this thing. For in all worlds it is thus: when there is no full certainty, there is no full accountability. This formulation is by pure logic unassailable, because it sets up a symmetrical function of reward in the context of the theory of games; whoever in the face of uncertainty demands full accountability destroys the mathematical symmetry of the game; we then have the so-called game of the non-zero sum.

  It is therefore thus: either God is perfectly just, in which case He cannot assume the right to punish the “ungodlies” by virtue of the fact that they are “ungodlies” (i.e., that they do not believe in Him); or else He will punish the unbelievers after all, which means that from the logical point of view He is not perfectly just. What follows from this? What follows is that He can do whatever He pleases, for when in a system of logic a single, solitary contradiction is permitted
, then by the principle of ex falso quodlibet one can draw from that system whatever conclusion one will. In other words: a just God may not touch a hair on the head of the “ungodlies,” and if He does, then by that very act He is not the universally perfect and just being that the theodicy posits.