The Future of the Mind (44 page)

CAN CONSCIOUSNESS LEAVE YOUR BODY?

Some scientists who have investigated near-death and out-of-body experiences are convinced that they are by-products of the brain itself when it is placed under stressful conditions and its wiring gets confused. However,
there are other scientists who believe that one day, when our technology is sufficiently advanced decades from now, one’s consciousness may truly be able to leave the body. Several controversial methods have been suggested.

One method has been pioneered by futurist and inventor Dr. Ray Kurzweil, who believes that consciousness may one day be uploaded into a supercomputer.
We once spoke at a conference together, and he told me his fascination with computers and artificial intelligence began when he was five years old and his parents bought him all sorts of mechanical devices and toys. He loved to tinker with these devices, and even as a child he knew he was destined to become an inventor. At MIT, he received his doctorate under Dr. Marvin Minsky, one of the founders of AI. Afterward, he cut his teeth applying pattern-recognition technology to musical instruments and text-to-sound machines. He was able to translate AI research in these areas into a string of companies. (He sold his first company when he was only twenty.) His optical reader, which could recognize text and convert it into sound, was heralded as an aid for the blind, and was even mentioned by Walter Cronkite on the evening news.

In order to be a successful inventor, he said to me, you always have to be ahead of the curve, to anticipate change, not react to it. Indeed, Dr. Kurzweil loves to make predictions, and many of them have mirrored the remarkable exponential growth of digital technology. He made the following predictions:

•  By 2019, a $1,000 PC will have the computing power of the human brain—twenty million billion calculations per second. (This number is obtained by taking the one hundred billion neurons of the brain, multiplying one thousand connections per neuron, and two hundred calculations per second per connection.)

•  By 2029, a $1,000 PC will be a thousand times more powerful than the human brain; the human brain itself will be successfully reversed engineered.

•  
By 2055, $1,000 of computing power will equal the processing power of all the humans on the planet. (He adds modestly, “I may be off by a year or two.”)

In particular, the year 2045 looms as an important one for Dr. Kurzweil, since that is when he believes the “singularity” will take hold. By then, he
claims, machines will have surpassed humans in intelligence and in fact will have created next-generation robots even smarter than themselves. Since this process can continue indefinitely, it means, according to Dr. Kurzweil, a never-ending acceleration of the power of machines. In this scenario, we should either merge with our creations or step out of their way. (Although these dates are in the far future, he told me that he wants to live long enough to see the day when humans finally become immortal; that is, he wants to live long enough to live forever.)

As we know from Moore’s law, at a certain point computer power can no longer advance by creating smaller and smaller transistors. In Kurzweil’s opinion, the only way to expand computing power further would be to increase overall size, which would leave robots scavenging for more computer power by devouring the minerals of the Earth. Once the planet has become a gigantic computer, robots may be forced to go into outer space, searching for more sources of computer power. Eventually, they may consume the power of entire stars.

I once asked him if this cosmic growth of computers could alter the cosmos itself. Yes, he replied. He told me that he sometimes looks at the night sky, wondering if on some distant planet intelligent beings have already attained the singularity. If so, then perhaps they should leave some mark on the stars themselves that might be visible to the naked eye.

One limitation he told me, is the speed of light. Unless these machines can break the light barrier, this exponential rise in power may hit a ceiling. When that happens, says Kurzweil, perhaps they will alter the laws of physics themselves.

Anyone who makes predictions with such precision and scope naturally invites criticism like a lightning rod, but it doesn’t seem to faze him. People can quibble about this or that prediction, since Kurzweil has missed some of his deadlines, but he is mainly concerned about the thrust of his ideas, which predict the exponential growth of technology. To be fair, most people working in the field of AI whom I have interviewed agree that some form of a singularity will happen, but they disagree sharply on when it might occur and how it will unfold.
For example, Bill Gates, cofounder of Microsoft, believes that no one alive today will live to see the day when computers are smart enough to pass for a human. Kevin Kelly, an editor for
Wired
magazine, has said, “
People who predict a very utopian future always predict that it is going to happen before they die.”

Indeed, one of Kurzweil’s many goals is to bring his father back to life. Or rather, he wants to create a realistic simulation. There are several possibilities, but all are still highly speculative.

Kurzweil proposes that perhaps DNA can be extracted from his father (from his grave site, relatives, or organic materials he left behind). Contained within roughly twenty-three thousand genes would be a complete blueprint to re-create the body of that individual. Then a clone could be grown from the DNA.

This is certainly a possibility. I once asked Dr. Robert Lanza of the company Advanced Cell Technology how he was able to bring a long-dead creature “back to life,” making history in the process.
He told me that the San Diego Zoo asked him to create a clone of a banteng, an oxlike creature that had died out about twenty-five years earlier. The hard part was extracting a usable cell for the purpose of cloning. However, he was successful, and then he FedExed the cell to a farm, where it was implanted into a female cow, which then gave birth to this animal. Although no primate has ever been cloned, let alone a human, Lanza feels it’s a technical problem, and that it’s only a matter of time before someone clones a human.

This would be the easy part, though. The clone would be genetically equivalent to the original, but without its memories. Artificial memories might be uploaded to the brain using the pioneering methods described in
Chapter 5
, such as inserting probes into the hippocampus or creating an artificial hippocampus, but Kurzweil’s father has long passed, so it’s impossible to make the recording in the first place. The best one can do is to assemble piecemeal all historical data about that person, such as by interviewing others who possess relevant memories, or accessing their credit card transactions, etc., and then inputting them into the program.

A more practical way of inserting a person’s personality and memory would be to create a large data file containing all known information about a person’s habits and life. For example, today it is possible to store all your e-mail, credit card transactions, records, schedules, electronic diaries, and life history onto a single file, which can create a remarkably accurate picture of who you are. This file would represent your entire “digital signature,” representing everything that is known about you. It would be remarkably accurate and intimate, detailing what wines you like, how you spend vacations, what kind of soap you use, your favorite singer, and so on.

Also, with a questionnaire, it would be possible to create a rough approximation of Kurzweil’s father’s personality. His friends, relatives, and associates would fill out a questionnaire containing scores of questions about his personality, such as whether he was shy, curious, honest, hardworking, etc. Then they would assign a number to each trait (e.g., a “10” would mean that you are very honest). This would create a string of hundreds of numbers, each one ranking a specific personality trait. Once this vast set of numbers was compiled, a computer program would take these data and approximate how he would behave in hypothetical situations. Let’s say that you are giving a speech and are confronted with an especially obnoxious heckler. The computer program would then scan the numbers and then predict one of several possible outcomes (e.g., ignore the heckler, heckle back, or get into a brawl with the heckler). In other words, his basic personality would be reduced to a long string of numbers, each from 1 to 10, which can be used by a computer to predict how he would react to new situations.

The result would be a vast computer program that would respond to new situations roughly the way the original person would have, using the same verbal expressions and having the same quirks, all tempered with the memories of that person.

Another possibility would be to forgo the whole cloning process and simply create a robot resembling the original person. It would then be straightforward to insert this program into a mechanical device that looks like you, talks with the same accent and mannerisms, and moves its arms and limbs the same way that you do. Adding your favorite expressions (e.g., “you know …”) would also be easy.

Of course, today it would be easy to detect that this robot is a fake. However, in the coming decades, it may be possible to get closer and closer to the original, so it might be good enough to fool some people.

But this raises a philosophical question. Is this “person” really the same as the original? The original is still dead, so the clone or robot is, strictly speaking, still an impostor. A tape recorder, for example, might reproduce a conversation we have with perfect fidelity, but that tape recorder is certainly not the original. Can a clone or robot that behaves just like the original be a valid substitute?

IMMORTALITY

These methods have been criticized because this process does not realistically input your true personality and memories. A more faithful way of putting a mind into a machine is via the Connectome Project, which we discussed in the last chapter and which seeks to duplicate, neuron for neuron, all the cellular pathways of your brain. All your memories and personality quirks are already embedded in the connectome.

The Connectome Project’s Dr. Sebastian Seung notes that some people pay $100,000 or more to have their brains frozen in liquid nitrogen. Certain animals, like fish and frogs, can be frozen solid in a block of ice in winter yet be perfectly healthy after thawing out in spring. This is because they use glucose as an antifreeze to alter the freezing point of water in their blood. Thus their blood remains liquid, even though they are encased in solid ice. This high concentration of glucose in the human body, however, would probably be fatal, so freezing the human brain in liquid nitrogen is a dubious pursuit because expanding ice crystals would rupture the cell wall from the inside (and also, as brain cells die, calcium ions rush in, causing the brain cells to expand until they finally rupture). In either case, brain cells would most likely not survive the freezing process.

Rather than freezing the body and having the cells rupture, a more reliable process to attain immortality might be to have your connectome completed. In this scenario, your doctor would have all your neural connections on a hard drive. Basically, your soul would now be on a disk, reduced to information. Then at a future point, someone would be able to resurrect your connectome and, in principle, use either a clone or a tangle of transistors to bring you back to life.

The Connectome Project, as we mentioned, is still far from being able to record a human’s neural connections. But as Dr. Seung says, “
Should we ridicule the modern seekers of immortality, calling them fools? Or will they someday chuckle over our graves?”

MENTAL ILLNESS AND IMMORTALITY

Immortality may have its drawbacks, however. The electronic brains being built so far contain only the connections between the cortex and the thalamus.
The reverse-engineered brain, lacking a body, might begin to suffer from sensory isolation and even manifest signs of mental illness, as prisoners do when they are placed into solitary confinement. Perhaps the price of creating an immortal, reverse-engineered brain is madness.

Subjects who are placed in isolation chambers, where they are deprived of any contact with the outside world, eventually hallucinate.
In 2008, BBC-TV aired a science program titled
Total Isolation
, in which they followed six volunteers as they were placed inside a nuclear bunker, alone and in complete darkness. After just two days, three of the volunteers began to see and hear things—snakes, cars, zebras, and oysters. After they were released, doctors found that all of them suffered from mental deterioration. One subject’s memory suffered a 36 percent drop. One can imagine that, after a few weeks or months of this, most of them might go insane.

To maintain the sanity of a reverse-engineered brain, it might be essential to connect it to sensors that receive signals from the environment so it would be able to see and feel sensations from the outside world. But then another problem arises: it might feel that it is a grotesque freak, an unwieldy scientific guinea pig living at the mercy of a science experiment. Because this brain has the same memory and personality as the original human, it would crave human contact. And yet, lurking inside the memory of some supercomputer, with a macabre jungle of electrodes dangling outside, the reverse-engineered brain would be repulsive to any human. Bonding with it would be impossible. Its friends would turn away.

THE CAVEMAN PRINCIPLE

At this point, what I call the Caveman Principle starts to kick in. Why do so many reasonable predictions fail? And why would someone
not
want to live forever inside a computer?

The Caveman Principle is this: given a choice between high-tech or high-touch, we opt for high-touch every time. For example, if we are given a choice between tickets to see our favorite musician live or a CD of the same musician in concert, which would we choose? Or if we are given a choice between tickets to visit the Taj Mahal or just seeing a beautiful picture of it, which would we prefer? More than likely the live concert and the airplane tickets.