The robots go marching - but not quite yet
`THE Revolution of the Robots'' is a script still restricted to science-fiction movies. The futurists who confidently predicted that robots would be as commonplace as toy soldiers by the end of the '80s are now muttering, ``Wait until the '90s!'' The assembly line in Detroit was supposed to be the stage where the robot would emerge as the once-and-future star. But the automobile industry is talking cautiously about the robot as less than the expected ``panacea.'' For now, mechanical hands are merely adding a little ``appropriate flexibility'' to ``conventional'' methods of production.
At research centers like the Massachusetts Institute of Technology, where the robot has been almost an engineer's religion - an electronic idol - the rhetoric has been scaled down to less worshipful terms. One MIT professor has gone so far as to say, ``The robot is only a symbol - just one of many computer-controlled machines on the factory floor.''
Nobody doubts that the robots are coming. But they are coming more slowly - and less spectacularly - than promised, and this gives the robot-rooters and other bystanders a valuable interval in which to reappraise the original heady assumptions. These fantastic and often naive hopes were usefully summarized by Grant Fjermedal - a wary chronicler - in his somewhat neglected book, ``The Tomorrow Makers.'' Traveling from Cambridge, Mass., to Tokyo, and talking to theologians as well as computer scientists, Fjermedal struggled sympathetically but fearfully with the vision of the Brave New World (as he called it) of ``artificial intelligence.''
It was the true believers who scared him more than the ultimate machines. He found a simplistic faith that total automation would solve all the problems humans had wrestled with since the beginning of time. How to achieve peace? In the Utopian electronic future, he was told, a benign supertechnology would ``garner all the weapons of mass destruction into a machine-controlled system, in the same way that you have to take matches from children.''
How to achieve freedom? ``Computers taking over might make us free'' in every other sense. (This speculation, not from a scientist but a divinity professor.)
How to achieve the truth? ``Machine sociology, anthropology, and psychology'' would set new standards of near-infallibility.
How to achieve happiness? Androids would put an end to loneliness, with empathy and loyalty engineered into your mechanical roommate as a matter of quality control. A number of visionaries even foresaw the possibility of marriage between humans and androids, thus making Ann Landers obsolete as well as the divorce court.
For a voice of calm dissent to all this microchip euphoria, one must resort, as Fjermedal did, to Joseph Weizenbaum, author of ``Computer Power and Human Reason.'' Weizenbaum has declared flatly that the robot is ``an intelligence alien to genuine human problems and concerns.'' Indeed, he doubts that a robot can be programmed to know what a human being is - to distinguish your aunt from a hatrack. Unlike many of his colleagues, he is certain of a robot's inability ever to feel love or pity, no matter how sophisticated the strings of binary code.
As a data bank without a sense of joy or tragedy, will the robot restrict our rich and passionate concept of knowledge to less than human terms? Weizenbaum fears so. A child who escaped the Holocaust in Nazi Germany, and also a computer scientist, he sees the computer possessing a sinister potential to become ``an instrument for the destruction of history'' - as collective memory, fully felt.
It is a measure of how extreme and pervasive the faith in computer science has become that Weizenbaum may be speaking for a minority when he says the obvious: ``Man is not a machine.'' And: ``There are some things beyond the power of science to fully comprehend.''
Here is where we are, fighting for this much skepticism in the presence of the sciences - which once prided themselves on being the embodiment of skepticism.
When T.S. Eliot wrote of the dream of a system so perfect no one would have to be good, he was deploring the shallow optimism of humanistic theories that had replaced religion with the hope of heaven on earth. Now if he were using the word ``system,'' he would mean system as in computer system - The Perfect Pushbutton. And what does that tell us about the evolution of intelligence, ``artificial'' and otherwise? A Wednesday and Friday column