Many people have argued about whether or not computing machines can think. Can they or can't they? Almost certainly, a machine can do any work in thinking that a person can do-if the correct way for doing the thinking work can be stated exactly. Besides, many programs have been made guiding machines so that they behave in very clever ways. Machines have been taught to play checkers (下棋) and to learn from what happens in the games, so as to make their play better and better. Machines can beat many persons who play checkers. Machines can translate from one language to another. They can prove statements in geometry as taught in school. Machines can recognize printed marks on paper, so that a person is no longer needed to look at the marks. There are many ways in which computers can be used in business, science, industry, engineering, and government. Computers study oil-field surveys made by airplanes; they keep track of oil sales; they study stresses (压力) in pipes; they help in planning and running refineries (炼油厂); and much more. If people cannot state exactly the correct way to do certain work in thinking, it is an open question as to how much of this work a machine can do. Suppose a person is "teaching" the machine and telling the machine "yes" when the machine is right, and "no" when the machine is wrong. Then the machine can keep on searching among possible ways for solving a problem-until the machine can do as well as the person, and perhaps much better. But, what is thinking? People can agree about much that is thinking, and much that is not thinking; but they argue about the rest because the word "think" is not a scientific word with an exact meaning. A lot of the argument is wasted breath because the persons who argue are using different meanings of the word "think." People, however, have only had experience with machines that appear to think since 1944. It would be rash to state that after the next 200 years, machines will not be thinking. And to consider that machines can think gives people a big push to give to machines more and more work in thinking. ( 382 words )