A godlike role, yes. But Cunningham resists confronting that notion. He does not believe he is trying to be God; he does not even want to think about God. His family had been on comfortable terms with God—Uncle Tim was a priest, there was an archbishop somewhere back a few generations, his parents and sisters moved cozily within the divine presence as within a warm bath— but he himself, unable to quantify the Godhead, preferred to sidestep any thought of it. There were other, more immediate matters to engage his concern. His mother had wanted him to go into the priesthood, of all things, but Cunningham had averted that by demonstrating so visible and virtuosic a skill at mathematics that even she could see he was destined for science. Then she had prayed for a Nobel Prize in physics for him; but he had preferred computer technology. “Well,” she said, “a Nobel in computers. I ask the Virgin daily.”
“There’s no Nobel in computers, Mom,” he told her. But he suspects she still offers novenas for it.
The angel project had begun as a lark, but had escalated swiftly into an obsession. He was reading Gustav Davidson’s old Dictionary of Angels, and when he came upon the description of the angel Adramelech, who had rebelled with Satan and had been cast from heaven, Cunningham thought it might be amusing to build a computer simulation and talk with him. Davidson said that Adramelech was sometimes shown as a winged and bearded lion, and sometimes as a mule with feathers, and sometimes as a peacock, and that one poet had described him as “the enemy of God, greater in malice, guile, ambition, and mischief than Satan, a fiend more curst, a deeper hypocrite.” That was appealing. Well, why not build him? The graphics were easy—Cunningham chose the winged-lion form—but getting the personality constructed involved a month of intense labor and some consultations with the artificial-intelligence people over at Kestrel Institute. But finally Adramelech was on line, suave and diabolical, talking amiably of his days as an Assyrian god and his conversations with Beelzebub, who had named him Chancellor of the Order of the Fly (Grand Cross).
Next, Cunningham did Asmodeus, another fallen angel, said to be the inventor of dancing, gambling, music, drama, French fashions, and other frivolities. Cunningham made him look like a very dashing Beverly Hills Iranian, with a pair of tiny wings at his collar. It was Asmodeus who suggested that Cunningham continue the project, so he brought Gabriel and Raphael on line to provide some balance between good and evil, and then Forcas, the angel who renders people invisible, restores lost property, and teaches logic and rhetoric in Hell, and by that time Cunningham was hooked.
He surrounded himself with arcane lore: M. R. James’s editions of the Apocrypha, Wake’s Book of Ceremonial Magic and Holy Kabbalah, the Mystical Theology and Celestial Hierarchies of Dionysius the Areopagite, and dozens of related works that he called up from the Stanford data base in a kind of manic fervor. As he codified his systems, he became able to put in five, eight, a dozen angels a night; one June evening, staying up well past his usual time, he managed thirty-seven. As the population grew, it took on weight and substance, for one angel cross-filed another, and they behaved now as though they held long conversations with one another even when Cunningham was occupied elsewhere.
The question of actual belief in angels, like that of belief in God Himself, never arose in him. His project was purely a technical challenge, not a theological exploration. Once, at lunch, he told a co-worker what he was doing, and got a chilly blank stare. “Angels? Angels? Flying around with big flapping wings, passing miracles? You aren’t seriously telling me that you believe in angels, are you, Dan?”
To which Cunningham replied, ‘You don’t have to believe in angels to make use of them. I’m not always sure I believe in electrons and protons. I know I’ve never seen any. But I make use of them.”
“And what use do you make of angels?”
But Cunningham had lost interest in the discussion.
He divides his evenings between calling up his angels for conversations and programming additional ones into his pantheon. That requires continuous intensive research, for the literature of angels is extraordinarily large, and he is thorough in everything he does. The research is time-consuming, for he wants his angels to meet every scholarly test of authenticity. He pores constantly over such works as Ginzberg’s seven-volume Legends of the Jews, Clement of Alexandria’s Prophetic Eclogues, Blavatsky’s The Secret Doctrine.
It is the early part of the evening. He brings up Hagith, ruler of the planet Venus and commander of 4,000 legions of spirits, and asks him details of the transmutation of metals, which is Hagith’s specialty. He summons Hadranel, who in Kabbalistic lore is a porter at the second gate of Heaven, and whose voice, when he proclaims the will of the Lord, penetrates through 200,000 universes; he questions the angel about his meeting with Moses, who uttered the Supreme Name at him and made him tremble. And then Cunningham sends for Israfel the four-winged, whose feet are under the seventh earth and whose head reaches to the pillars of the divine throne. It will be Israfel’s task to blow the trumpet that announces the arrival of the Day of Judgment. Cunningham asks him to take a few trial riffs now—"just for practice,” he says, but Israfel declines, saying he cannot touch his instrument until he receives the signal, and the command sequence for that, says the angel, is nowhere to be found in the software Cunningham has thus far constructed.
When he wearies of talking with the angels, Cunningham begins the evening’s programming. By now the algorithms are second nature and he can enter angels into the computer in a matter of minutes, once he has done the research. This evening he inserts nine more. Then he opens a beer, sits back, and lets the day wind down to its close.
He thinks he understands why he has become so intensely involved with this enterprise. It is because he must contend each day in his daily work with matters of terrifying apocalyptic import: nothing less, indeed, than the impending destruction of the world. Cunningham works routinely with megadeath simulation. For six hours a day he sets up hypothetical situations in which Country A goes into alert mode, expecting an attack from Country B, which thereupon begins to suspect a preemptive strike and commences a defensive response, which leads Country A to escalate its own readiness, and so on until the bombs are in the air. He is aware, as are many thoughtful people both in Country A and Country B, that the possibility of computer-generated misinformation leading to a nuclear holocaust increases each year, as the time-window for correcting a malfunction diminishes. Cunningham also knows something that very few others do, or perhaps no one else at alclass="underline" that it is now possible to send a signal to the giant computers—to Theirs or Ours, it makes no difference—that will be indistinguishable from the impulses that an actual flight of airborne warhead-bearing missiles would generate. If such a signal is permitted to enter the system, a minimum of eleven minutes, at the present time, will be needed to carry out fail-safe determination of its authenticity. That, at the present time, is too long to wait to decide whether the incoming missiles are real, a much swifter response is required.
Cunningham, when he designed his missile-simulating signal, thought at once of erasing his work. But he could not bring himself to do that: the program was too elegant, too perfect. On the other hand, he was afraid to tell anyone about it, for fear that it would be taken beyond his level of classification at once, and sealed away from him. He does not want that, for he dreams of finding an antidote for it, some sort of resonating inquiry mode that will distinguish all true alarms from false. When he has it, if he ever does, he will present both modes, in a single package, to Defense. Meanwhile, he bears the burden of suppressing a concept of overwhelming strategic importance. He had never done anything like that before. And he does not delude himself into thinking his mind is unique: if he could devise something like this, someone else probably could do it also, perhaps someone on the other side. True, it is a useless, suicidal program. But it would not be the first suicidal program to be devised in the interests of military security.