Back in the seventies, my dad managed the local branch of a national lumber company. The building had a computer that took up an entire room. To communicate with it, you punched little slots into cards which were then manually fed into the computer’s maw. It lived alone in its lair, giving off a constant whirring racket. And when it broke down, the company had to fly in a computer guy from out of state, and this guy was revered like some kind of swami. And when that guy couldn’t fix it, they flew in another guy. I think it ended up being a total of three or four guys before the problem was fixed, and each time a new one appeared the hierarchy shifted and an erstwhile big genius got relegated to getting coffee.
I thought about that computer while watching The Invisible Boy. This film, released in 1957, features a sentient supercomputer that wants to rule the world. The computer takes up not one but two large rooms, with an additional 5300 cubic yards of microtransistors overhead and a wall of glass tubes connected to feedback units. The main console has huge panels filled with flashing lights, switches, levers, and dials. According to Dr. Merrinoe, its operator, this computer represents “the sum total of human knowledge, constantly being revised and brought up to date.”
Well! That probably makes it more powerful than the computer at my dad’s office, let alone any real-life 1950s computers. One 1959 model with the catchy name of PDP-1 stored data on those punched cards and had a memory of 9 kilobytes, which is less than 10% of what might be used today to store one photo of average size. And the 1970s computer at my dad’s office had less processing power than a graphing calculator. (I guess. Actually I kind of pulled that last bit out of the air. When I asked a friend in the tech industry for a valid point of comparison, he replied that in terms of processing, memory, and static storage, modern computers are exponentially larger than 1950s counterparts to such a degree that he couldn’t even think of a metaphor.)
The idea of a piece of tech becoming sentient and turning on its makers is one we’ve all seen explored lots of times in story form—Hal, Skynet, Ultron, replicants, the Omnidroid, the Machines. My introduction to the concept was probably Tron. But here it is in a movie made way back in 1957. Clearly our uneasiness with artificial intelligence has a long history.
And no wonder. Computers have been bound up with national defense right from the get-go. In 1950, after President Truman green-lit plans to develop a hydrogen bomb, atomic scientists needed a way to work mathematical calculations of enormous complexity, beyond the capabilities of mechanical calculators. A new machine based on vacuum tubes was put to work on the job, and later improvements to such machines led to the development of computers. (This strikes me a bit like inventing calculus so you can study planetary motion, as Sir Isaac Newton reportedly did. These math guys are awfully thorough.)
In the 1950s American mind, then, computers were inextricably linked to the hydrogen bomb. And the hydrogen bomb was more destructive than the atomic bomb by an order of magnitude of…a lot. Sorry I can’t be more precise. I read about the difference between fission and fusion but I’m still a little fuzzy on the details.
But although I’m weak on nuclear physics, I’m fairly quick with semiotics, and I did pick up on something noteworthy while researching the H-bomb. Check out these images.
Now take a gander at the piece of hardware crowning the supercomputer in The Invisible Boy.
Did you see them? Those freaky eyeballs drawn onto the glass dome for who knows what deranged and misguided purpose? They look like cross sections of nuclear bombs! (The dome itself looks an awful lot like Robby the Robot’s head, but more of that in a later post.)
The concept of artificial intelligence is far older than computers, reaching back to antiquity. Like the computers of the 1950s, the artificially intelligent beings of mythology were usually built with defense in mind. (Interestingly, one of the non-defense ones, Galatea, started off as an ivory statue that Pygmalion made and fell in love with. So the still-popular idea of artificial intelligence as romantic partner has a long history as well.) Hephaestus, Greek god of blacksmiths, craftsmen, and metallurgy, built Talos, a walking, thinking bronze statue designed to defend the isle of Crete, which Talos did by patrolling the shores and throwing rocks at pirates and invaders. (Hephaestus also built some metal automata to serve him in his workshop, an early example of AI’s industrial applications.)
And let’s not forget the Golem—or golems, as there are many accounts of them in Jewish folklore. The name “golem” is derived from a Hebrew word for unformed material or uncultivated person. Like Adam, all golems are shaped from clay, but unlike Adam a man-made golem will never be more than a shadow of a living being created by God. The makers of golems are wise and holy people, often rabbis. Interestingly, golems are mute, but written language is the means of bringing them to “life”: any one of the Names of God, called a shem, is written on a slip of paper and inserted into the golem’s mouth, or inscribed directly onto its forehead. To shut the golem down, the operator removes the paper or rubs out one of the letters on the forehead, invalidating the shem.
Golem legends are numerous and fascinating and often associated with actual historical figures. The most famous golem is the one of Prague, supposedly created by Rabbi Judah Loew ben Bezalel. Like Talos, and the supercomputer in The Invisible Boy, the Golem of Prague was made for defense. Rudolf II, Holy Roman Emperor from 1576 to 1612, had threatened to kill or exile all the Jews of the city, so Rabbi Loew made the Golem, whose powers included making himself invisible and summoning the spirits of the dead.
Like all golem stories, this one turns sour at the end. In one version of the story, the Golem desecrates the Sabbath when Rabbi Loew forgets to remove the shem in time; in another, he falls in love, is rejected, and goes on a violent rampage à la Bride of Frankenstein. In other golem accounts, the creatures crush or scar their makers, become surly and uncooperative, or take instructions too literally. In some cases the maker has second thoughts and deactivates the golem before it can do any harm. All the accounts conclude with the golems either disintegrated back to dust, or dismantled and stored in an attic somewhere.
Frankenstein’s monster is a type of artificial intelligence. So is the One Ring, and so is Gurthang, the sword carried by Túrin Turambar in The Children of Húrin. In all these cases, the AI is closely identified with its master (the Elves actually call Túrin and Gurthang by the same name, Mormegil, or Black Sword), and the master is guilty of great hubris in making or wielding the instrument. He knows what he’s doing is problematic at best, but he does it anyway, and the instrument causes a lot of destruction and grief. And in the end, it brings about its master’s death.
Stories in which the sentience of an artificially intelligent being is treated as a positive thing exist, but they’re not as common, and when they do occur there’s still the possibility of trouble. Offhand I can think of Robby the Robot, Data (though Lore kind of cancels him out), and J.A.R.V.I.S., who I understand is going to do some awesome stuff in the next Avengers movie. And WALL·E and EVE. But our prevailing feeling toward artificial intelligence is one of deep unease.
Why? For the same reason people in the Matrix won’t accept an idyllic existence as reality and David Aames keeps messing up his lucid dream: because we know ourselves and our deserts. No matter how hard we try to smother the knowledge, it keeps rising up again, pointing an accusing finger. We are so corrupt that even our best actions have mixed motives.
As it is written, There is none righteous, no, not one: there is none that understandeth, there is none that seeketh after God. They are all gone out of the way, they are together become unprofitable; there is none that doeth good, no, not one. Their throat is an open sepulcher; with their tongues they have used deceit; the poison of asps is under their lips: whose mouth is full of cursing and bitterness. Their feet are swift to shed blood: destruction and misery are in their ways: and the way of peace have they not known. There is no fear of God before their eyes.
We can’t escape ourselves. And when we make a powerful tool to serve us—a tool with potential personhood, no less—we worry. What are we going to do with it? Will we use it wisely? Will we use it to extort and oppress? Or will we just get lazy and let it do all the work for us, leaving ourselves vulnerable to tyrannical takeover? And what will it do with us? Will it foul up its orders by taking instructions too literally, like the Golem, or refusing to stop, like the broom in The Sorcerer’s Apprentice? Will it fall into the wrong hands and be used against us? Will it develop a will of its own and seek to dominate us?
Fear of AI is fear of ourselves, of what our darkest impulses could accomplish with vast power and no accountability. We have begotten it; how can it escape the corruption that is bound up in our every cell? We do well to tremble.