 |
A K Peters, August 2005, |
Consider this: Robots will one day be able to write poetry and prose so touching that it will make men weep; compose dozens or even hundreds of symphonies that will rival the work of Mozart; judge a court case with absolute impartiality and fairness; or even converse with the natural ease of your best friend. Robots will one day be so life-like that a human could fall in love and marry one.
You think this is thought provoking and controversial, maybe even far-fetched? Well, IM David Levy, expert on computer chess, has just brought out a history of Artificial Intelligence which elaborated on these predictions.
In the 50 years since the inception of Artificial Intelligence, computer scientists have made remarkable achievements that can be seen in computer games, children's toys, your home PC, and nearly every facet of human life. In this popular approach to understanding AI, Levy captures the essence, excitement, and potential of Artificial Intelligence. He lays the factual foundations for his intriguing speculations by presenting the history of AI from its earliest conception to the present day. He then considers the most recent advances and makes predictions about the future of this burgeoning science.
ChessBase: You have just written a book entitled “Robots Unlimited”. What exactly does it mean?
Levy: It reflects my belief that within about fifty years there will be almost no limit to the intellectual and creative powers of robots, nor to the sophistication of their electronics and their electromechanical design.
David Levy at a press conference in New York
ChessBase: You are an International Master and and expert in computer chess. This is a new field for you, isn’t it?
Levy: Not really. Some of my earlier books were on computer chess, but also on the programming of other thinking games, which are topics within one branch of the science of Artificial Intelligence. And in 1997 I won the Loebner Prize, a kind of world championship competition for conversational programs, which is an important topic within another branch of A.I. So this book is something of a culmination of my interest in A.I., which started in 1967.
ChessBase: Why robots? Why did you decide to write this book?
Levy: I have felt for quite some time that although A.I. is a fascinating science, and one that steadily increases its impact on our lives, most people know absolutely nothing about how any intelligent programs work. Of course the fine details of artificially intelligent programs are very much the stuff of academic conferences and learned journals, but in many branches of A.I. it is possible to explain the basic principles of A.I. to a lay reader, although I have never come across a book that does so. My book is intended to fill this void.
ChessBase: In your book the early history of A.I. goes back a long way, a few hundred years. We thought that the subject was not around before the invention of the computer.
Levy: A.I. means exactly what it says – artificial intelligence. Although conventionally, when A.I. is discussed, it is in the context of computer programs, there is nothing sacrosanct about the means of achieving an artificial form of intelligence. What is important is the result, and if that result consists of a music composition, which is the earliest example I give in the book, dating from 1650, or a machine for proving theorems in logic, for which I describe an example built in 1775, then why, just because of the means by which the result is obtained, should we refrain from describing the process as an artificial form of intelligence?
Conference on computer chess: with former world champions Rustam Kasimdzhanov
and Alexander Khalifman (left and right), David Levy and Spanish journalist
Leontxo Garcia, in Bilbao 2005
ChessBase: You concede that the technologies making A.I. possible are quite advanced, the sort of thing we find in academic textbooks and at specialist conferences. Do you believe your book succeeds in explaining these technologies to a general readership, to people who don’t understand how computers work or how to write a computer program?
Levy: Most definitely. Nothing in the book requires of the reader any scientific or technical knowledge. And the basic principles of most of the A.I. technologies I describe are extremely easy to understand.
ChessBase: Your subtitle, “Life in a Virtual Age”, obviously refers to your predictions in the last five chapters. How confident are you in making these predictions, or are they pure speculation?
Levy: I am very confident, partly because of the amazingly fast progress in technology during the past half century, and the dramatic increase in the rate of this progress. The more we know about a science the faster it is to discover even more about that science. But there are other factors as well, for example the inevitability of dramatic increases in computer speeds and memory sizes, both of which will facilitate the development of A.I.
ChessBase: Are you really so sure of your predictions about love, marriage, sex and reproduction with robots? Isn’t this all rather science fiction?
Levy: No, it isn’t science fiction. Do you remember the movie ”2001 A Space Odyssey”, in which the computer on board the space ship defeated David at chess? When, four years before 2001, Garry Kasparov lost to Deep Blue, Arthur C. Clarke’s science fiction suddenly became science fact. My predictions should really be no more difficult to believe than Clarke’s were when they were made. I think that my book provides ample explanation as to the concepts of love, sex and reproduction with robots. And as for marriage, I accept that the concept of marriage to a robot is outside most people’s frame of reference, but it is certainly not fiction. If someone had said, one hundred years ago, that within a century same-sex marriages would become legally and socially acceptable unions, how many people would have believed them? Social mores change, and along with so many other aspects of modern life their rate of change is much faster nowadays than a century ago, so I think it is quite reasonable to predict that people will be marrying robots fifty years from now. By then robots will be just like people in so many ways, just different on the inside.
ChessBase: Do you seriously believe that what you write about robot ethics, robot religion, and so on, is a realistic perception of robot behaviour and how people will treat robots?
Levy: Why not? Given that robots will be very much like people, surely we must accept that they will have legal and civil rights, and that some of them will have religious beliefs. I assume that most people will treat robots with the same decency with which most pet owners treat their cats and dogs.
ChessBase: Your timeline will probably be regarded by many of your readers as optimistic (or pessimistic, depending on how one views the prospect of your predictions). Isn’t fifty years or less too soon for your predictions to become reality?
Levy: No, I don’t think so. A few months ago, John Nunn predicted on your site that his son’s Lego brick would be playing Elo 2800 chess within ten years. Although I believe that John’s prediction about this one microcosm of A.I. is somewhat optimistic, one of the world’s leading technology gurus, Ray Kurzweil, has written that in the next 100 years we will not experience only a century of progress at current rates of development, but 10,000 years of progress. This is the point that the sceptics overlook – the massive rate at which technological progress is increasing.
ChessBase: Do your predictions frighten you at all?
Levy: A little, because mankind seems to have an unfortunate knack for being able to find abominable uses for so many technologies.
ChessBase: Are you happy with the thought that your children will be living in the “virtual age” you are predicting?
Levy: I think so, because I believe that the benefits of A.I. technology will outweigh whatever disadvantages come from its abuse, but of course we cannot be certain.
