Let us assume that we are able to design a computer that can learn, reflect and create. Would that prove that human beings are merely sophisticated machines?
What if this computer turns out to be better than humans at the very skills on which we base our self-esteem? Would we then have to hand over the management of our lives and societies to these computer parents and assume the role of children?
In the last century, computer enthusiasts had predicted that computers were soon going to perform every conceivable function and also carry on conversations with ordinary people. Such optimism seems to have been unwarranted.
The chess-playing computer Deep Blue which had been developed by Carnegie Mellon in 1984, was later hired by IBM in 1984. A match between Deep Blue and the Russian Grand mater Gary Kasparov was arranged.
Despite initial defeats, Deep Blue went on to become, the first computer system to defeat a reigning world champion in a match under standard chess tournament time controls.
However, it is relevant to mention that Deep Blue was not even” aware” that it had beaten Kasparov!
Some AI experts wondered that If this ability to defeat even the best chess players is not a sign of intelligence then what is the true criterion or yardstick.
The answer is that as yet no computer is intelligent in the human sense. Their greatest handicap or limitation is that computers are still too literal-minded, error-prone and just plain stupid. This is because of the human brain and the computer work in fundamentally different ways. People identify realities by matching patterns, computers by matching numbers. When the human brain recognises a bus it does so by connecting it to an idea, a form, a concept of a bus. A computer, on the other hand, counts the definable external features or the image and then sees if the final numbers resemble those associated with buses. This inability to see underlying continuities and identities beneath slight changes in surface details is the computer’s major limitation.
But, however advanced a chess computer is, the machine’s style is essentially defensive; its “plan” is to wait for its human opponent to make a mistake and then capitalise on it. The human is always setting the pace defining the strategic structure of the game. The machine lets you do what you want, but catches you if you made a mistake while doing it. In other words, it exploits your errors.
This degree of compliance is perhaps a virtue in a commercial product but it is not what one expects in a human or even a so-called intelligent machine. A really intelligent machine would be one that has its own purposes, knows what they are and plans its own advancement; one that could use humans as a means to its ends. Such a machine should also be” aware” of its own actions in the sense we humans define awareness. Only when that such a stage is reached would humans have a cause to worry!
One computer specialist succeeded in designing a computer which gave the impression that it was already conscious—aware of itself and, its inner workings, capable of describing its needs and its nature. However, according to its own creator, this computer mechanically chose words from a sea of nouns and verbs and the like, and merely manipulated them according to the rules of English grammar!
The question as to whether a computer can be made to be conscious or not would no doubt depend to a large extent on our definition of consciousness. Experts in artificial intelligence are of the view that consciousness is overrated and that what we now regard as consciousness is only a “very imperfect summary” in one part of the brain of what the rest is doing. According to these specialists people who ask whether a machine could be conscious, do not have even the vaguest notion as to how ideas are generated. They, therefore, feel that when there is a better theory about how certain parts of the brain summarise what is happening in other parts, then we will understand it and be able to make machines also do it. However, neurobiologists do not share this view.
Leon Cooper of Brown University, who won the Nobel Prize for Physics starting analysing neural networks in the brain to determine how to create something similar in a machine. He also feels that consciousness is regarded by some people as something indescribable and mysterious and held undeserved reverence by them. On the other hand, he believes that consciousness is only a subtle consequence of a large system of interacting neurones and that once we understand what the principles are, the question of whether consciousness can be built into a machine or not, would depend on how complex a phenomenon it is.
In other words, computer specialists feel that the riddle of consciousness cannot be solved within the conventional scientific fabric but only by a proper understanding of how complex a phenomenon it is. But the complexity according to them may not be undecipherable. Marvin Minsky of Massachusetts Institute of Technology compares the human brain to a system of computers and neural networks that are interconnected in a manner that is not understood at present.
Some computer scientists like Nobel Laureate Herbert Simon of Carnegie Mellon do not think that technology is a limiting factor. He concedes that you do need a big parallel computer if you are going to simulate the human eyes. But he states that if for a moment we ignore sensory input and motor output—and just try to simulate a deep thinker sitting in an armchair—the hardware available already is a good deal faster and more powerful than the human brain. However many people do not share his optimism.
It took three centuries to evolve from Newton’s classical mechanics to Richard Feynman’s Quantum Electro Dynamics. It may, according to some scientists, take a similarly long period for machines to compete with humans in many respects.
In the meantime, all current efforts are geared towards designing expert systems, specialised computers that can control a missile or sculpt an aeroplane’s wing; machines that can read aloud from different typefaces; music synthesisers that can imitate many instruments and advanced word processors you can even talk.
Going back to our question as to whether computers are alive, one has to only understand how the human brain is different from the computer, To illustrate this I would like to invite attention to a viewpoint expressed by Nobel Laureate Francis Crick in his book “ The Astonishing Hypothesis”
Taking up the question of visual perception an important function of the brain, he refers to the work of the world-renowned neuroscientist V.S.Ramachandran as follows.
“What style of explanation shall we need to understand the brain. My own agrees more closely with V.S Ramachandran’s ‘Utilitarian Theory of Perception’. He argues that visual perception does not involve intelligent deduction of exactly the type we use in constructing an argument, nor does it involve the vague idea that the Brain simply resonates to the visual input. Neither does it require the solving of elaborate equations as often implied by A. I by researchers. Instead, he believes perception uses ‘rules of thumb’ sleight-of-hand tricks that are acquired by trial and error through millions of years of natural selection. This is a familiar strategy in biology but for some reason, it seems to have escaped the notice of psychologists who seem to forget that Brain is a biological organ. I also agree with Ramachandran when he states’ The best to resolve some of these issues may be actually to open, the black box in order to story the responses of nerve cells but psychologists and computer scientists are often very suspicious of this approach’. In Ramachandran’s view, the job of the visual psychologist is not this stage to construct elaborate mathematical theories to explain his results but instead to sketch what might be called the ‘natural history’ of vision, especially the earliest stages of vision. When the visual task has been dissected into its many component parts and especially can be shown that certain interactions are weak or absent we shall know just what needs to be explained in neural terms. The explanation may or may not involve elaborate mathematics. They will certainly involve the properties of interacting neurones and the details of their interconnections, Thus because of the complexity of its visual world, one expects to find many rapid, rough and ready processes interacting dynamically in many different ways.”
In other words, the human brain being a biological organ does not depend on mathematical equations for carrying out visual perception as a computer would do. Over millions of years, it has evolved biological mechanisms that have no necessity of employing the kind of functions computers are good at. And what is true of the visual system is true of all other biological organs. Nature had its own way in evolving organs best suited to perform a particular function and did not use sophisticated mathematical models. Taking as a whole the human brain with the advantage and sanction of millions of years of biological evolution employed various tricks to develop organs most suited for its purposes, I.e integrating the information obtained by them and achieving its results. That being so how can a computer however powerful and sophisticated be conscious or get the better of the human brain. True it might perform a million permutations and combinations a minute and come out with a play of Shakespeare. But is it AWARE that it is typing or can UNDERSTAND what it has typed? Does a computer have human emotions like love, passion, joy, hate, ambition revenge, greed, jealousy, and greed, to name a few? Most definitely not. Then how can one believe that a computer is conscious?
To sum up we must admit that as yet we do not have what we call a “conscious” computer. We cannot call the pocket size chess playing machines which play an advanced game an example of Artificial Intelligence. Nor can we perhaps hope to gain insight into the mysteries of the human brain with the help of computers.
Ultimately our knowledge of the working of the brain must come, not from a computer scientist but only from a neurophysiologist, someone like Roger Sperry or David Hubei. And only after a complete understanding of the brain has been achieved can we even venture to think in terms of designing machines that can match the brain in complexity as well as capability.
To end the discussion on a lighter vein I would like to draw attention to an apocryphal anecdote about a man who applied to a computerised marriage bureau listing out the features he wanted in his life partner:
‘Cute, dignified, of good stature and gait, very friendly, a good listener and someone willing to be cuddled and fondled all the time without complaining of a headache.’
The bureau suggested a penguin!