6.2 Simulating Human Intelligence

Many researchers today believe that the totality of what it means to be human will be known when they can fully describe the activity of the brain. Toward this end a great deal of work has been devoted. So far, it is known that nerve cells called neurons respond to electrochemical signals in the brain in a complex switching operation or neurotransmission taking place at a junction known as a synapse. The patterns for transmissions through synapses change through time, and these changes no doubt have something to do with learning and memory.

Some things are also known about the speed at which a synapse operates and the rate at which signals move through the brain. These turn out to be substantially slower than in electronic switches, by several orders of magnitude. Even though the mechanism by which all these take place is not clearly understood, the information that is available makes some researchers confident that the functions of the human brain can be duplicated in a smaller and faster electronic device. Note that it is not necessary to duplicate the human brain itself, only to understand how it works well enough to build a functional equivalent--that is, a machine that calculates and stores in a way capable of producing the same results.

If a functional equivalent of the human brain could be constructed, two things could be done with such a device. The first possibility is to program it or "teach" it so that it can perform a few simple tasks. Then it can act as an "intelligent" controller or designer capable of making decisions and acting upon them in a way that is the electronic equivalent of the fashion in which the human brain works.

It might eventually be possible to build an ambulatory body for this thinking machine and thus create the mobile robot of science fiction. This perfect servant/slave could be given instructions such as "take out the cat," "bathe the kids," and "go to the grocery store."

It would presumably be able to carry out such tasks without any further human intervention. Indeed, it should be possible to let the machine decide when the house or office routine dictates that something needs doing, and go ahead without asking, or being told. Whether it will ever be possible to discuss child rearing, philosophy, or one's emotions with such a machine is quite another matter.

The motivation to spend the enormous sums of money that would be required to develop such machines would have to be powerful indeed. While building the ideal butler, maid, secretary, lover, or factory worker might be interesting, it is not clear that machines are necessary for such tasks, or that they need either a human shape or the equivalent to a human brain. Perhaps such devices are needed in very hostile environments where humans could not go,--for instance, the ocean floor, space, the moon, underground, or a nuclear reactor core. To do crucial jobs that cannot be done otherwise, machines will be built. They need not look or act anything like a human being for such purposes, and it is uncertain that many robots ever will. Moreover, such devices will not be used at all in the home unless there are substantial benefits to cover the enormous cost.

There is a second, and perhaps even more ambitious potential use for a functional analog of the human brain, however. The most optimistic of AI researchers are confident that not only can such a machine be built, but that a human brain could eventually be scanned on a molecule-by-molecule basis and its activity duplicated in the artificial version. Thus, they believe the totality of the human's thinking would have been downloaded into the mechanical construct. Give the electronic brain a mechanical body to match and the result is hoped to be not just an intelligent robot, but a mechanical copy of the human being.

Since the duplicated human would now reside in a more easily repairable body, or in no body at all, and since backup copies could be made at any time, the body of flesh could supposedly be discarded when the downloading was complete. The net result: immortality would have been achieved in a mechanical form. A human would cease to be blood and bone and becomes a cyborg (part machine) or a fully machine intelligence--with electronic capabilities projected to be many times as great as those of the bodies they currently inhabit.

The end result of this line of research is supposed to be nothing less than the ultimate in man-made salvations from death--eternal life in a manufactured body and brain--not in heaven, but here on Earth. Not only that, but the ability to make backups means that a person really could be in two places at one time, and merge the memories afterward into a single copy. If backups were frequently made, even the fatal destruction of a single unit would cost no more than a few days out of one's life and experience.

Quite apart from any other ethical and moral problems that may come to mind, this goal raises an old conundrum, the answer to which may in part determine whether this is possible.

Is the mind more than the brain?

If on the one hand, the human mind and soul can be expressed unambiguously as the sum of the brain's electrical parts, then downloading its activities to a functional equivalent would transfer a copy of the personality from a "machine made of meat" to one made of electronic parts. This would be regarded by many as offering a final proof that the empirically verifiable material world is the sum total of all existence, that the spiritual and supernatural are fantasies, and that logical positivism is permanently triumphant. On the other hand, if such things as emotions, friendship, anger, fear, intuition, poetic appreciation, conscience, intentionality, self-awareness, and the ability to enquire about the existence of God cannot be expressed as sets of electro chemical impulses, this endeavour may well fail, for the mind is then more than the brain.

These issues touch on the essence of what it means to be alive and what it means to be human. Thus, attempts to achieve practical immortality by such means are sure to touch many raw nerves. Those who oppose such research may say that there are some things that ought never to be tried. To those who support it, the potential prize is great enough to pursue at all cost. Furthermore, there is no stopping such work now that it has begun. To forbid such research and make the prohibition stick is impossible, as long as qualified researchers are not yet satisfied that the question has been answered one way or the other.

Even if such a transfer succeeded, questions would still remain. Would the downloaded person's thinking and memories really constitute the person, or is this a simulated person, and not a full duplicate of the original? Assuming there is a soul, perhaps it would depart when the flesh and blood body was discarded, and would not transfer along with the contents of the brain. The question of whether such a copy is fully human would still remain.

There is also the question of timing. As Grant Fjermedal (The Tomorrow Makers) notes (p. 5):

In the weeks and months that followed my stay at Carnegie-Mellon in February of 1985, I would be surprised and intrigued by how many researchers seemed to believe downloading would come to pass. The only point of disagreement was when -- certainly a big consideration to those still knocking around in mortal bodies. Although some of the researchers I spoke with at Carnegie-Melon, and later at MIT, Stanford, and in Japan thought that downloading was still generations away, there were others who believed we were actually so close to achieving robotic immortality that some of the researchers seemed to be driven by private passions never to die. And perhaps this explained the eagerness of Hans's young research assistants to work through the nights and weekends to further this quest for the life ever after.

Some regarded these developments as imminent two decades ago. Others believe it could be many years, if ever, before such questions need to be answered seriously, for research on the activity of the brain is moving very slowly, and it seems unlikely that it can be functionally duplicated soon, if ever. Indeed, the small progress made in the intervening years seems to argue this work may be at a dead end.

Biological enhancements to the human brain achieved by genetic manipulation or chemical means might obviate the necessity to go the mechanical route altogether. In the meantime, computer and communications research may take other turns, the products of which could also render the production of artificially intelligent brains unnecessary. On the other hand, startling new techniques have a way of appearing on the scene almost full-blown and with great rapidity. Predictions about this area of research and its potential applications have as much likelihood of being proven too conservative as too extravagant.

The Fourth Civilization Table of Contents
Copyright © 1988-2002 by Rick Sutcliffe
Published by Arjay Books division of Arjay Enterprises