【怒加。科幻。評論。科學。Artificial Intelligence

Artificial Intelligence - Can Computer Replace Human Being? ﹝09-10-1997﹞

        I think that most of you remember a few months ago, IBM supercomputer, Deep Blue defeated the Chess Champaign. Do you think that computer can replace human being in the future? Let me tell you more detail, ancient Chinese philosophers had the discussion about the difference between animals and human, but nowadays, scientists and philosophers concern more about the difference between machine, especially computer and human. Can computer think like human? Can computer as intelligent as human, or even more superior than human?
        Some of you may immediately say that computer never be superior than or replace human, because computer cannot do cooking, singing, writing poem...etc. However, in the past, some tasks that people think computer cannot do so, for example playing chess, accept oral command, reading...etc, all these tasks, computer can do now. Then people may not give up, they set up another tasks for the computer to do. When the computer can do that, then they set another tasks again. This is not reasonably discussing on this topic. For such topic, we call that "weak artificial intelligence", or simply call that "weak AI"
        Firstly, I hope that most of you have some idea about the development trend of computer:

Artificial Intelligence	Computers to emulate such aspects of intelligence as speech recognition, deduction, inference, creative response, the ability to learn from past experience, and the ability to make reasonable inferences from incomplete information
Expert System	Computer program that uses artificial intelligence techniques to make decisions or recommendations or predict outcomes based on an analysis of data.
Neural Nets	Neural networks imitate the brain's ability to sort out patterns and learn from trial and error, discerning and extracting the relationships that underlie the data with which it is presented.

        Computer experts hope to build a computer has self-learning, creative idea ability. In 1949 Professor G. Jefferson said that, "a computer is equal to human unless it can think and it has its own feeling, it can write poem or compose song, more than that, it must know what it has written or composed." In 1950 English mathematician Alan M. Turing said that, “Let us assume there is a man and a computer in a closed room, no matter how the people outside the room, ask how difficult questions, the people outside still cannot distinguish who is the man and who is the computer inside, then I will agree that computer can think like human. "For this topic, we call that "strong artificial intelligence", or simply call that "strong AI" Obviously, we are talking about strong AI today.
        Under this topic, strong AI, people divide into two opposite groups, "for AI" and "against AI".
        Since I have just mention two examples how the people who against AI challenge the computer, so let me talk about the opinion about them.
        In 1980, American philosopher John R. Searle suggested a famous experiment, "Chinese room experience" An experiment which cannot carry out but caused a lot of argument.
        The experiment is like that: We assume in a closed room, there are a man who know English only and do not know any Chinese, many many cards that printed a Chinese character on it, a handbook printed in English. People outside the room send some cards through the gap between the door to him. The cards represent some questions about Chinese culture, e.g. poem, eastern philosophy. The man inside just to check the cards he received form the handbook, then send back some cards to response following the instruction stated in the handbook. That is any question cards, the handbook can tell him how to send back some relative answer cards. So people outside the room may think that the man inside must be an expert in Chinese culture. In fact he even do not know Chinese. Searle thought that the room with man is like the computer, the handbook is the computer program. Despite the possibility of writing such program, Searle think that if one day, we can write such program, do it represent that computer can think? The answer is NO, like the man do not know Chinese. Computer has only syntax but no semantics. Searle emphasized more detail that, computer never have self-consciousness and free will. We can see that, his experiment is a reply towards Alan M. Turing, that if computer can act like human, it is still a computer, because it does not have self-consciousness and free will.
        However, For AI people argue that self-consciousness is so abstract, when the computer said that I know I am exist, how people prove its statement is right or wrong? Because it is abstract, untouchable. Then why we agree that every people have his own self-conscious mind, but computer have not?
        Concern about the problem of syntax (The rules governing construction of a machine language) and semantics (relationships between signs and symbols and what they represent), for AI people also pointed out that in a low operating level, we can only notice syntax in computer program. But in high operating level like Chinese room experiment, we can not avoid the semantics involved.
        Computer does not have feeling? For AI people disagree, because feeling is also a abstract thing, even a person tell you that he feel happy at this time, how we judge that he is tell lie or not, even computer can do better, because it do not feel nervous when it "tell lie".
        For AI people has a strong faith that AI is possible to come true. Because how much self-consciousness we have, this is just the result of biological evolution. Simple cell do not have self-consciousness, when it evolve to more complicated organism, self-consciousness will appear after certain evolution. Biological evolution has 10 billions more history, but AI has just more than 10 years history, why it cannot be come true in the future?
        In my opinion, whether computer can think or not, I do not have a sure answer, but I am open minded towards this subject, I also do not worry about that computer may replace human, or even take the control of the world like some science fiction films or stories. I think that somebody or something can solve a problem, answer a question, he/it is intelligence enough, but he/it requires more intelligence to discover a question, man can progress fast, the main reason that is some intelligent people discover new question, then someone else solve it. To raise question or discover difficulty, I think that computer cannot replace man, no matter how the computer intelligent, man are still the master and computer is just a tool to help us.