Can a computer think? That depends on what you mean by 'think'. If solving a mathematical problem is 'thinking', then a computer can 'think' and do so much faster than a man. Most mathematical problems can be solved quite mechanically by repeating certain straightforward processes over and over again. Even the simple computers of today can be geared for that. It is frequently said that computers solve problems only because they are 'programmed' to do so. They can only do what men have them do. One must remember that human beings also can only do what they are 'programmed' to do. Our genes 'program' us. Our 'program' is so much more enormously complex, though, that we might like to define 'thinking' in terms of the creativity that goes into writing a great play or composing a great symphony, into developing a brilliant scientific theory or a profound moral judgment. In that sense, computers certainly can't think and neither can most humans. Surely, though, if a computer can be made complex enough, it can be as creative as we. If it could be made as complete as a human brain, it could be the equivalent of a human brain and do whatever a human brain can do. But how lung will it take to build a computer complex enough to duplicate(复制) the human brain? Perhaps not as long as some think. Long before we approach a computer as complex as our brain, we will perhaps build a computer that is at least complex enough to design another computer more complex than itself. This more complex computer could design one still more complex and so on. In other words, once we pass a certain critical point, the computers take over and there is a 'complexity explosion'. In a very short time thereafter, computers may exist that not only duplicate the human brain but far go beyond it. What information about computers can we get from the passage?