News – 01-AUG-99

The success of The Matrix this summer suggests that despite our love affair with computers, we are simmering with anxiety about their growing power. Young people flocked to the film, riveted by its concept that artificial intelligence had taken over the world and was keeping human beings captive and immobile in pods as a kind of neural battery pack. The humans’ brains were wired into a giant computer which fooled them into thinking they were living in the late 20th century world. It took Keanu Reeves, as the fallible but courageous hero, to slip out of the virtual domain and save humanity’s soul.

The conventional answer to computer phobia is that these machines are nothing more than a tool wielded at the whim of their human masters. Like any technology, so the argument goes, computers’ purpose is simply to make our lives easier and better. Certainly most engineering would be unthinkable these days without them. “Intelligent buildings,” like the BBA office featured on page 40, for example, are no longer unusual, with their automated systems that respond as the occupants move around the interior. Engineers in the resource sector are using even more advanced systems. In the petroleum industry they can predict a well’s output with up to 15 per cent accuracy using “artificial neural networks.” The prospectors bombard the computer with data, and from this it “learns” to recognize patterns and relationships in a way similar to how the brain works by establishing neural pathways.

If computers can be taught to act like a human brain, then should we be wary of their power after all? After reading a book by Ray Kurzweil, I’m certainly afraid. Kurzweil, who invented the first text-to-speech synthesizer, wrote The Age of Spiritual Machines: When Computers Exceed Human Intelligence (New York: Viking-Penguin, 1999). He reckons that computing power will grow so fast that by 2029 a standard model will have the calculating and memory capacity of not just one human brain but a thousand. First he argues in his “Law of Accelerating Returns” that the evolution of technology exponentially speeds up as time passes. Then he points out that California companies are already developing the hardware that will vastly increase computers’ processing capacity. Contending materials in the circuitry stakes are light photons, crystals (a trillion bits of storage for each cubic centimeter), DNA molecules, “nanotubes” (carbon based molecules so small that 50,000 of them side by side are as thick as one human hair), and finally and most formidably, quantum computing. When processes operate at this infinitesimally small scale it is hard to see a difference between mechanical structures and biological ones like the human brain.

Indeed, by the time we reach the age of Kurzweil’s “spiritual machine” during the next century, he says we will simply copy our brain activity onto a computer. The programmed version(s) will then take over and we can cast off our frail flesh. Immortality by computer hardware. The prospect raises serious questions about the nature of consciousness. Can a computer that is so advanced it apparently thinks become self-aware? And will it have a conscience; will it need one? Even Kurzweil doesn’t have all the answers, but he suggests it is fruitless to resist.