• Choose a category
• All
• Classic-Fiction

By Root 1160 0

compact, self-ambulatory computer—capable on occasion of independent decision making and real control of his or her environment. And, as the old joke goes, these computers can be constructed by unskilled labor. But there are serious limitations to employing human beings in certain environments. Without a great deal deal of protection, human beings would be inconvenienced on the ocean floor, the surface of Venus, the deep interior of Jupiter, or even on long space missions. Perhaps the only interesting results of Skylab that could not have been obtained by machines is that human beings in space for a period of months undergo a serious loss of bone calcium and phosphorus—which seems to imply that human beings may be incapacitated under 0 g for missions of six to nine months or longer. But the minimum interplanetary voyages have characteristic times of a year or two. Because we value human beings highly, we are reluctant to send them on very risky missions. If we do send human beings to exotic environments, we must also send along their food, their air, their water, amenities for entertainment and waste recycling, and companions. By comparison, machines require no elaborate life-support systems, no entertainment, no companionship, and we do not yet feel any strong ethical prohibitions against sending machines on one-way, or suicide, missions.

Certainly, for simple missions, machines have proved themselves many times over. Unmanned vehicles have performed the first photography of the whole Earth and of the far side of the Moon; the first landings on the Moon, Mars and Venus; and the first thorough orbital reconnaissance of another planet, in the Mariner 9 and Viking missions to Mars. Here on Earth it is increasingly common for high-technology manufacturing—for example, chemical and pharmaceutical plants—to be performed largely or entirely under computer control. In all these activities machines are able, to some extent, to sense errors, to correct mistakes, to alert human controllers some great distance away about perceived problems.

The powerful abilities of computing machines to do arithmetic—hundreds of millions of times faster than unaided human beings—are legendary. But what about really difficult matters? Can machines in any sense think through a new problem? Can they make discussions of the branched-contingency tree variety which we think of as characteristically human? (That is, I ask Question 1; if the answer is A, I ask Question 2; but if the answer is B, I ask Question 3; and so on.) Some decades ago the English mathematician A. M. Turing described what would be necessary for him to believe in machine intelligence. The condition was simply that he could be in teletype communication with a machine and be unable to tell that it was not a human being. Turing imagined a conversation between a man and a machine of the following quality:

INTERROGATOR: In the first line of your sonnet which reads “Shall I compare thee to a Summer’s day,” would not “a Spring day” do as well or better?

WITNESS: It wouldn’t scan.

INTERROGATOR: How about “a Winter’s day”? That would scan all right.

WITNESS: Yes, but nobody wants to be compared to a Winter’s day.

INTERROGATOR: Would you say Mr. Pickwick reminded you of Christmas?

WITNESS: In a way.

INTERROGATOR: Yet Christmas is a Winter’s day, and I do not think Mr. Pickwick would mind the comparison.

WITNESS: I don’t think you’re serious. By a Winter’s day one means a typical Winter’s day, rather than a special one like Christmas.

No device of this sophistication has yet been built, although I am not sure how many humans would pass Turing’s human test. But the amount of effort and money put into artificial intelligence has been quite limited, and there are only about a half-dozen major centers of such activity in the world. One of the more striking results obtained in a very limited universe of discourse—that of children’s blocks—has come from the work of Terry Winograd, then at the Massachusetts Institute of Technology. Here is a dialogue between man and machine,