The Globe and Mail

Saturday, February 27, 1999
We, robots?

Hans Moravec predicts that computers will match human
cognitive capability within 50 years, and as robotic
work-forces rise, capitalism will die within a century.

ROBOT: Mere Machine to Transcendent Mind
By Hans Moravec
Oxford University Press, 1999, 227 pages

I've often wondered where Hans Moravec's obsession with robots came from. Was there a seminal piece of science fiction that kindled his fancy? Or did it arise during his teens, as a sort of intellectual refuge from the intense social demands of a North American adolescence? Whatever the case, Moravec has kept the faith in an anti-robot decade: a decade when talk of robots brings snickers and blank stares, when cheap, solar-powered, lawn-mowing robots languish on warehouse shelves thoughout North America due to lack of public interest.

But Moravec is not alone in his speculations about the future of humankind. He is a member of a trio of visionary scientists that also includes nanotechnologist K. Eric Drexler and Freeman Dyson, the cosmic engineer. All of these science popularizers, and others, have published books that confidently predict the dawn of the superman. But this superman will appear in a form that would have been inconceivable to Friedrich Nietszche, for it will arise out of the fertile conjunction of augmented humans and intelligent machines.

Hans Moravec was one of the original founders of the prestigious robotics program at Carnegie Mellon University, and he has decades of hands-on experience in robotics and artificial intelligence under his belt. He rode out two decades of under-funding that ended in the early eighties. Unfazed, in 1988 he published an astonishing book, Mind Children, which was not only intensely stimulating but well written and carefully thought out. As a result, I looked forward to the publication of Robot, wondering if he could repeat the performance of a decade earlier. I was not disappointed.

Robot turns out to be every bit as visionary as Mind Children, though Moravec builds his case more slowly this time. In the first few chapters Moravec acquaints the reader with the evolution of human beings, then of computers and robotics. These chapters are informative and sometimes technical. They are also replete with some extraordinary facts. It was shocking, for example, to learn that a robot car called NAVLAB drove from Washington, D.C., to San Diego, Calif., at an average speed of more than 100 km/h, with a robot in control 98.2 per cent of the time.

About mid-way through the book, Moravec changes gears and begins his survey of things to come. He predicts computers will match human cognitive capability within 50 years. With the rise of largely robotic work-forces, he claims that capitalism itself will end within a century, as -- under pressure from corporate competitors -- companies that squander resources by paying their owners become uncompetitive. There will still be a capital economy, but it will be one that is driven by radically evolving competition. In the short run, Moravec calls for strong government intervention in the new capital economy, particularly by taxing the extraordinarily efficient robot corporations. This, he reasons, will provide generous pensions for most humans from birth onward. Moravec sees robot technology as spawning tremendous wealth and leisure for humans.

This is a more bureaucratic Moravec than we encountered in Mind Children, and there are times that Moravec the libertarian seems to have been replaced by Moravec the authoritarian. But my worries were premature. All his regulatory strategies are methods to get superintelligences off the earth and into space where they can interact, compete, evolve and absorb each other freely. When any human decides to go "ex" then he or she, too, would have to leave the earth.

By chapter six, The Age of Mind, we realize that the Moravec who wrote Mind Children is back in town. In it he envisages "Exes," as he calls posthumans (consisting of both "up-loaded" biological humans and artificial intelligences), quickly evolving into an inconceivable future. He sees an explosion of various physical entities housing superintelligences, but over time these will transform into pure intellects pulsing as digital patterns through connected entities in a sort of realized cyberspace. These entities, as they master quantum effects and gain the ability to distort space/time to their own ends, then evolve into unrecognizable energy forms organized into sub-molecular, constantly expanding wave-fronts.

These are indeed heady speculations, but Moravec's scientific underpinnings always inform his sometimes poetic prose. He is a gung-ho supporter of artificial intelligence and the eventual robotic transcendence of human beings. Rather than being alarmed or dismayed at the possibility of our intelligent machines (or "mind children," as he refers to them) superseding us, he says, "It behooves us to give them every advantage and to bow out when we can no longer contribute."

Moravec is not naive, either. He realizes that there will be some resistance to the existence of conscious machines, and that Luddites will deny, at first, that these machines could indeed be conscious. Parodoxically, this misperception will occur particularly in those who make the first artificial intelligences. Just as the makers of Deep Blue, the computer that beat Gary Kasparov in chess in 1997, were very careful to state that it was a case of mere quantity (massive computing) prevailing over human "quality." However, Kasparov himself felt that he was dealing with an unsettling form of consciousness that he called an "alien intelligence."

Certainly, Moravec has developed some real humility in the past decade, marked by passages in which he claims that his own predictions for the next century may be as wrong-headed as Jules Verne's were for ours. But given that caveat, Moravec goes on to map out an extraordinary scenario for human evolution as it extends through the new supermen that we will create. In the last chapter, Mind Fire, Moravec sketches the extreme end-point of all evolution. He describes the Omega Point, brainchild of John D. Barrow and Frank J. Tipler (co-writers of The Anthropic Cosmological Principle), as the logical outcome. The Omega Point is the point at which consciousness includes and controls not only every particle in the known universe, but every particle in all possible universes.

Christopher Dewdney is the author of Last Flesh: Life in the Transhuman Era.

Recent Related Reading

How We Became Posthuman: Virtual Bodies in Cybernetics, Literature, and Informatics, by N. Katherine Hayles (Chicago, 350 pages, unpriced). Hayles, a UCLA English professor and science writer, ranges through the history of technology to examine the fate of embodiment in the information age.

Copyright © 1999 The Globe and Mail