JULY 24, 1959
NEW YORK—Some remarkable things are being done in the International Business Machines research center south of Poughkeepsie, N.Y., but one scientist in particular is delighted because he has taught a machine to play checkers. And the machine is winning, improving its score with each game.
The scientist is Dr. Arthur L. Samuel, and he is pleased because the machine is "learning," whereas other machines have been taught to play checkers and chess with never showing signs of improvement.
So Dr. Samuel believes that his computer has developed an ability which could have a profound meaning in the field called "generalization learning." This might be applied in industry and in the solution of real social and economic problems. It would mean that much of the running of a business could be turned over to computers instead of using human minds on the problems involved.
A type of rote learning requires a machine to remember all past positions, such as board positions in checkers, and the outcome of each. But in generalization learning, the machine forgets specific things but constantly revises the list of factors it uses to analyze each board situation.
Dr. Samuel agrees that this may not be used for 20 to 50 years, but this is automation that goes beyond anything that most of us have thought about.
Perhaps even men with scientific minds may wonder if their gifts someday will not be made useless by machines capable of more accuracy and better reasoning in analyzing problems. This presents the question of whether a machine can be made to "reason" or whether reason will always remain the exclusive capability of the human mind.
The whole field of Dr. Samuel's research must be of great interest to everyone, I think, because there will come a day when the things we consider as God-given gifts to man, which make the difference between man and animal, may no longer be man's exclusive property.
Still, through man's invention but perhaps without man's functioning himself, work requiring reasoning powers may go on. It would be interesting to see how this question will be answered by the coming generations.
* * *
On Monday of this week I went to Bard College at Annandale, N.Y., as I have done for the past few years, to meet with members of the Cooperative Institute.
The group was there to discuss cooperative problems and, since the institute covers most of New England and other Eastern states, it is mainly a consumer cooperative with a small amount of production. There is a dairy cooperative and bakery in Massachusetts; otherwise it is a consumer proposition.
On the whole, these men and women are interested in other parts of the world, but it is an interest developed through the spread of cooperatives over the glove. So I was fortunate to be able to report to them on Israel, which has made great strides in use of cooperatives.