OMNI Vol. 16 No. 2  November 1993 Souls in Silicon by Frederik Pohl & Hans Moravec * Painting by Michel Henricot

Souls in Silicon

by Frederik Pohl & Hans Moravec

OMNI   November 1993   pp. 66-76

It is the year 20-something--we don't know the exact date yet, but figure 20 to 50 years from today--and your doctor has just given you some really bad news. That nasty little pain in your lower abdomen turns out be serious. The doctor explains to you with great tact and kindness that, although medicine can now fix almost everything that can go wrong with the human body, there remain one or two really ferocious ailments that cannot be cured. You won't be in pain, he says. You won't even be bedridden, except at the very end--but the fact is that you have just six months left to live.

Naturally, you don't enjoy hearing that. It comes at a particularly bad time, you think, because now that you're approaching 90, you've just begun thinking seriously about how you're going to enjoy your retirement years. Then the doctor clears his throat and says, "Of course, there is an alternative."

That gets your attention right away. "Alternative?" you say. "You mean I don't have to die in six months after all?"

The doctor purses his lips, professionally precise. "That isn't exactly what I mean, he says. "Your body is certainly going to die. There's nothing we can do about that; but that death doesn't have to be, well, fatal. You're a possible candidate for a mind transplant."

So a couple of weeks later you're undergoing tests in the best surgical hospital in your area. You're surprised to find out that most of the tests aren't medical. They're psychological, and they test things like your memory retention, your reflex speed, even your IQ. The CEO of your company comes by your hospital room with a dozen roses. There is a small flask of her best 1990 Scotch hidden among the flowers, and when the nurses aren't looking, she shares a last drink with you. "You'll be back in the office in two weeks, " she predicts. And then one morning they wake you up, give you a tranquilizer, slide you onto a gurney, and wheel you down to the operating room.

They don't put you to sleep, only make you a little drowsy so you can drift off to sleep if you want to. They do anesthetize your entire scalp, because the skin of your head is sensitive to pain, but the real cutting-and-splicing operation is going to be done on your brain, which has no nerve endings to feel pain. They've fixed it so that you can watch the whole thing on a TV monitor if you want to. (You're not at all sure you want to, but every once in a while you sneak a look.) You already have a good idea of what they're going to do, because the surgeons and the computer people have gone over it with you, with a model of the brain. You're a little impressed with the number of people in that operating room, all concentrated on you--two brain surgeons, an anesthesiologist, four nurses . . . and five computer experts. That's a lot of highly trained specialists to be working on you, you think . . . but you can't help feeling a little lonely in that crowd.

What they do to you doesn't hurt. You feel a kind of gently pushing this way and that as they lift a flap of scalp to expose the skull, and you definitely feel the vibration as they cut the bone and lift it away. You stop looking into the monitor at that point. You close your eyes and try to concentrate on thinking about your wife and kids, all waiting in one of the hospital's lounges, along with a couple of people from your company. Maybe you even do drift off to sleep. . .

And while you're doing that you know the surgeons are exposing that tough lump of meat in the midsection of the brain that is called the "corpus callosum. " You even know what the corpus callosum is, because they've told you that its half a billion fibers operate as conduits, passing information back and forth between the two halves of your brain.

Then the surgeons step back, and the computer people take over. They don't touch any scalpels themselves. They operate micromanipulators which gently slide a very dense and fine comb into the corpus callosum. The comb has some hundred thousand tines, and each tine has a hundred thousand connections. The tines slip elastically into the space between the fibers, until each fiber has made a contact with one or more of the connections. It's a mammoth job, but it is done nondestructively. The whole thing takes only about eight hours.

And then they bandage you up and wheel you back to your own room, and you really do go to sleep.

When you wake, you don't even have a headache--you're full of selective analgesics--and the doctor's there grinning at you. "Congratulations," he says. "Welcome to immortality."

Of course, that's science fiction--today. (But nuclear power, spaceships, tele vision, and robots were also science fic tion--once and now they're all over the place.)

The idea of storing human intelligence in some kind of machine is pretty old stuff in science fiction, almost as old as science-fiction magazines themselves. One of the earliest writers to use the notion was Neil P. Jones, who published his short story "The Jameson Satellite" in 1931. It told of a college professor named Jameson who, learning that he was soon to die, decided he would like something better than the usual funeral and burial. Like all college professors, Jameson was of course very rich. (When college professors read this story this is generally the point at which they start laughing uncontrollably.) So he took some of his money and built a spaceship in his backyard. When the professor did at last die, his executor loaded his corpse into the spaceship and fired it off into low Earth orbit. There Jameson remained, frozen solid, for a long time--40 million years until some wandering aliens called Zoromes discovered it and decided to recruit the professor into their band. So they surgically removed the brain from the frozen corpse, thawed it out and implanted it in a robot machine that resembled a breadbox with tentacular metal arms and legs. Then, renamed 21MM392 by his new Zorome friends, the professor went on to have endless adventures in space.

There were plenty of other such stories, but almost all of them assumed you would have to store the physical, organic human brain in some kind of machine. That seems pretty unlikely as a really long-range solution to the problem of immortality, since, sadly, the human brain is by its organic nature subject to rather rapid decay (as well as being afflicted with the steady deterioration that costs each one of us a few thousand brain cells dead or decrepit each day). When computers came along, they offered a more hopeful place to store intelligence.

We don't have any computers today that can come anywhere close to the capacities of the human brain--a typical late twentieth-century computer has roughly the "brainpower" of a housefly--but the things keep getting better, and they do it very fast. Between the early days of this century, when the first mechanical adding machines began to be useful, and the arrival of electric calculators during the World War II era, machine computation increased a thousandfold in speed and capacity. Then electronic machines came along, so that between 1940 and 1980 there was an additional millionfold improvement as vacuum tubes and then transistors took over. Since then, with accelerating miniaturization and the use of advanced integrated circuits, the curve continues to steepen, while future computers--using such techniques as quantum de vices, diamond semiconductors, increasing miniaturization down to the atomic scale--suggest that computing power will continue to grow at its historical rate or better for an indefinite time into the future.

The human brain, with its 100 billion neurons and roughly 100 trillion connections, requires a lot of computing power, to be sure . . . but not more than computers early in the next century should provide.

Given the probable existence of such hypercomputers within the lifetimes of many of us now alive, how do we go about getting all the memory, speed, and flexibility of a human mind into the machine?

That's where the corpus callosum comes in. Suppose that a neural comb like the one we have described is slipped into it and connected to an ex ternal computer. At first that computer does nothing but pass the brain's traffic from one hemisphere to the other and eavesdrop on it. It retains what it learns. Over time, it constructs a model of what goes on in your brain. More than that, the computer can put enough signal on each connection point to overwhelm the normal traffic if it needs to so that, when the model is nearly complete, the computer begins to insert its own messages into the flow. The computer becomes an auxiliary brain; then, when the original, organic brain begins to deteriorate, the computer smoothly assumes its functions.

And when the brain at last dies, as all organic things must, your mind is complete and functioning--in the computer. Optical, auditory, chemotactic, and other sensors let you know what is happening in your environment; speech synthesizers and graphics programs let you communicate with others in the "real" world . . . and you live on, though your body has died.

So here you are, a couple of months after you've had your operation.

It's now the Labor Day weekend. You've spent most of the summer "convalescing"--not really convalescing in the usual sense, because the operation didn't leave you particularly damaged, but getting used to this new companion in your mind.

You have to wear this portable computer all the time, of course. By 20something the thing is made with quantum-effect devices, a hundredth the size of the microchips of the 1990s, so it isn't particularly large or heavy. Still, it's got to be able to hold a lot of infor mation in its data file, so it's as big as, say, a 1990s' laptop.

That doesn't mean it has to look like a laptop. The engineers have built the whole thing into a sort of helmet, which covers your entire head. Although your children are now in their fifties and sixties, they still remember the old movies they saw in their kindergarten days; their affectionate name for you is " Darth Vader." The helmet does cramp your style a little. You can't swim while you're wearing the helmet, and it's not a good idea to ski or play football--but at your age you've pretty nearly decided to give up the more violent sports anyway. Apart from that, you can do anything you ever did.

No, that's not true. Actually, you can do a great deal more than you ever did--with your mind, at least--because the companion in the helmet is actually helping out your memory. Baseball? When you and your great-granddaughter watch the Tokyo Mets playing the Vladivostok Dodgers you astonish her by remembering the batting averages of every Mets catcher since 1960. It's all in the data file. Cooking? If your wife thinks of making the chicken-and-wine dish you once had in Paris for dinner, she doesn't have to look up the recipe. You could recite it for her, if you chose--actually, you probably just go ahead and make the dish, though eating any of it when it's done is a little trickier. Business? When you go back to work, your entire corporate financial dossier for the last 50 years is right there in your memory, and you can tell, off the top of your head, which divisions pulled their weight and which have generally underperformed . . . and even why.

Remember that, all this time, the computer that sits on your head isn't only teaching you, it's also learning you. It learns who your friends are, and what experiences you shared together, and what they mean to you. It learns what music you like to hear, and what sorts of books you like to read, and what plays and films you enjoy. It learns everything you know about your own life, from the first three-year-old birthday party (when you didn't, after all, get the Super Nintendo you had your heart set on) to the last disagreement you had with your wife . . . and how pleasingly you made up afterward. It remembers everything you remember, likes everything you like, worries about everything that worries you. . .

It is you. And when that demon in your belly at last makes the body you have occupied all these years useless, and the couple of pounds of wetware in your skull has to die . . . you live on in the machine.

Do you really call that living, you ask? Well, what do you call living? Is Stephen Hawking alive, for instance?

Hawking is generally acknowledged to be the world's greatest living theoretical physicist, but his body has been all but dead for many years. He is a victim of the disease called ALS--Lou Gehrig's disease--and he cannot even feed himself; worse, a complication a few years ago cost him his voice. But that does not keep Hawking from being a great scientist and a loved human being. It does not prevent him from traveling, or even from lecturing in public, though to be sure he must use a speech synthesizer for the purpose. Since the synthesizer is American made and Hawking is very English, he apologizes for its American accent--but it is still Stephen Hawking speaking. And you, with your advanced hardware and sophisticated software of the next century, can certainly do better than that. You will be able to speak in your own voice--or to sing--sing as well as you ever did, and, if you like, much better than ever, with the voice of any opera star who ever lived.

But at least Hawking does have a body, you argue, although admittedly one in bad repair. When he is speaking through his voice synthesizer peopIe can see him, anyway. You won't have even that much, right? Wrong! The mere lack of a body won't keep your friends from seeing you--just as you were, or as much handsomer as you wish. (Push back that receding hairline, smooth out those wrinkles--why not?) All you need for the purpose is a TV monitor. You can be the one who controls the image it shows, and that image can be you, made up out of the data bits your computer mind will generate for you. By then, the image will probably be in 3-D as well. Possibly it could be even physically present as a sort of puppet operated by your computer mind so that it can be touched and embraced.

Well, that's all well enough for your friends, you say, but what about yourself? Can you feel? you ask. Can you hear and see and smell? Can you perceive heat and cold? Can you feel the sensation of pain, or the touch of a byer's gentle caress?

Of course you can. It is not your brain that feels or sees any of those things, you know. Your brain can't. It doesn't have the necessary equipment. The brain is blind, deaf, and without sensation. All the brain knows is what the sensory organs of the body tell it, and your machine-stored mind can have all the sensory organs you like: video eyes, microphone ears, transducer sensory to convey the physical sensation of touching. Indeed, that could be only the beginning for you. The machine brain can be equipped with far better sensors than the standard accessory package that comes with the human body, for there are better designs on the market. The human eye, for exampIe, cannot see infrared or ultraviolet (but video cameras can); the ear misses the bat's shrill squeak and the low frequency sounds of nature (but microphones do not); there is no human sense that can pick up radio waves direct, but machines do it all the time--why not be in yourself your own TV set, pocket radio, or even radar?

It's possible, though, that adding new senses might not be a good idea. The brain you were born with had to work hard in order to learn how to interpret all those sensory inputs. There is some evidence that, after a certain point, new kinds of sensory inputs can be emotionally damaging. Young people who have their sight restored at maturity, after having been blind from birth, find the experience disorienting; Dr. Jerome Lettvin of MIT has found that many such people commit suicide.

Of course, your machine brain isn't bound by the same rules as your old organic one. Very possibly a few extra programs could be written in, or a little extra hardware added to your system, and you could then easily enough deal with senses that would allow you to "see" and "feel" anything that any instrument can detect.

All right, you say at last, but that's still not living. What about eating? What about the taste of a fine wine? For that matter, what about the buzz you get from a six-pack? And then you get right down to the question that's really on your mind: What about sex ?

The answer: Don't worry. No problem at all.

Well, no theoretical problem, anyway. Remember the main point: Everything you experience is experienced in the brain. It is the brain that interprets all those sensory inputs, including the pleasures of love-making. Once the eavesdropper in your corpus callosum learns how your brain works, it is only a step or two to reach the point where it can create for you any array of sensual inputs you like, not just sex. Not even just very good sex. Incredible sex, without such penalties as AIDS or unwanted pregnancy or even the wrath of a jealous lover, since all of it takes place in your mind.

You don't even have to give up your present mate, either. The technical problems of love-making between some collection of data bits stored in a Cray 100 (or whatever) and your flesh-and-blood nearest and dearest are daunting, but that's only temporary. If you are determined to be monogamous you can arrange with your nearest and dearest to join you in machine storage when the times comes. That may not be quick. Your devotion may require a good deal of patience . . . but then you've got all of eternity.

At least, you have eternity as long as you go on paying your utility bills.

Well, how much "science" is in this particular piece of science fiction?

Quite a lot, actually. That isn't to say that this is something you can count on by the year 2001. It may take longer. It may take much longer, because some pretty daunting technical problems are involved. Brain anatomists will tell you that there are important sections of the brain--for instance, the brainstem and the cerebellum--not directly reached by the corpus callosum; perhaps more connections must be made than we have outlined. Then there is the tricky question of hooking nerves to wires. Nerve impulses are at least partly electrical in nature, but they are also at least partly chemical. It isn't just a matter of taking a soldering iron to the nerve endings in the brain and joining them to an equal number of copper wires. Some sort of interface will be needed, and no one can now say what form it will have to take.

So the mind-transplant procedure has a long way to go to become a mature technology. How long, exactly? Perhaps about as long, say, as computers themselves had to wait in 1945, at the time of the huge, clumsy vacuum-tube things like ENIAC; or as nuclear energy had in 1938, when Hahn and Strassmann first split the uranium atom. But science goes faster these days--largely because of the computer itself, which makes scientists effectively a good deal smarter than their unaided native brains would allow. If the mind-transplant procedure can be done at all, as seems at least theoretically plausible, it is at least a good gambling bet that something like it will be real within the next few computer generations.

By the time you've been back on the job for a few years, you've become fully accustomed to your new existence. You find it's pretty neat; you even wonder why your flesh-and-blood friends put off joining you.

For one thing, you will have a lot more spare time than you ever had before. Your mental life won't be held to the 55-mph speed limit of an organic brain any more. Computer functions go far faster than organic synapses; you can do in seconds what takes your meat friends hours to accomplish.

Fortunately, you're not alone in machine storage. You have machine-stored colleagues and friends to talk with, and relate to, and do things with; they move as fast as you do, and actually you find your "living" colleagues just a little slow and dull.

And the things you do with peers are really a lot of fun. Travel? Why, you can enjoy a simulated Campari on the synthesized Champs Elyses or experience the thrills of skin diving on what your senses tell you is the Great Barner Reef whenever you like.

You know how this works in advance. When you were a child, you remember, you saw Hollywood films filled with such spectacles as the great spaceships of the Empire and the collapse of cities in earthquakes and nuclear wars. You were aware even then that things had never really happened, but were computer-generated images put together by special-effects firms like George Lucas's Industrial Light & Magic. The same techniques, now brought to perfection, can provide you with any "surround" you like for your adventures, as real to you as any weekend on the Jersey shore was when you were still in your body of flesh. For that matter, you're not limited to dull reality. You can choose to invent your own fantasy world (Barsoom, or Middle Earth, or the Arabian Nights', or the Heechee Universe), and the computer will build it around you, complete with food, drink, and companionship. And you have plenty of time for all this sort of fun. Not only do you do things fast, but you never have to waste any time in sleep.

And, of course, you are better at your job than even your best ever was--better than any flesh-and-blood person ever could be.

For that reason, you're not really surprised when your CEO calls you in just before your one hundredth birthday. She tells you that the compulsory retirement rule has been repealed for machine-stored intelligences.

You knew that was coming. When you look at her with your video eyes you feel a little compassion. She's definitely beginning to show her age, and you wish you had been able to persuade her to take the next step to join you.

But, although you like her and sympathize with her, you turn her down.

You've had other offers, you explain. The most interesting has come from NASA. They have a great need for someone like you--several someones like you--for some of the exciting, new long-range space missions they're planning.

After all, living human beings make a lot of trouble for spaceship designers: Flesh-and-blood people need food and air and water; they need to be kept warm (but not too hot!) and shielded from the radiation of solar flares; worst of all, they sometimes get sick, and it just isn't feasible to include physicians and dentists on a normal space mission. On the other hand, it is certainly worthwhile to try to have the presence of a human being to make the on-the-spot decisions, take care of the unexpected glitches, interpret what is discovered. And in you and your kind they have the perfect astronauts.

So, with regret, you tell your CEO that you'll be leaving the company to start training for your new job, which will be investigating the frozen surface of the planet Pluto. And there you are, with a whole new career, and a whole new life . . . and you're still a youth hardly out of your first century!