January 2000

Nano-Hype

Just as chip manufacturers reach the limits of silicon's abilities, nanotechnology will save the day with self-assembling "molecular computers." Sound too good to be true? It is.

By G. Pascal Zachary

Imagine a microscopic computer that assembles itself, atom by atom, then calculates at a speed faster than today 's zippiest electronic chips. Now imagine this same computer is unbelievably cheap -dirt cheap, in fact.Sounds too good to be true? Well,some people think it 's real. This is the idea behind nanotechnology: that individual molecules can serve as digital switches and, acting in concert with billions of other molecular switches, replace digital computers.

If this vision can be realized,molecular computers could in one swoop destroy the enormous investment by the semiconductor industry in "fabs,"the plants that fabricate chips. The most advanced plants today cost billions of dollars. Not only would molecular computers disrupt what's probably the world's most important manufacturing industry next to cars, they also solve a looming "problem" presented by the laws of nature. Chip makers face physical limits in etching circuits on the wafer-thin material called silicon. One widely accepted estimate says the limits of silicon will be reached by 2014.

So molecular computers-or talk of them in the nation's most prestigious newspapers and magazines, including this one (see "Computing After Silicon," TR September/October 1999)-appear to be coming along just in the nick of time. Like the cavalry in a John Wayne movie, they will rescue high-technology from the specter of stagnation. This is a beautiful story, one that warms the heart of the capitalists who pay for each new round of innovation in computing and other fields. There is only one problem with this story: It's a lie. And not a small lie either. In journalism, the story of molecular computing is a Big Lie.

The fact that it's a lie isn't all that surprising, however. For as long as innovators have been around, they've lied. Lied about the possible obstacles to further innovation. Lied about the utility of their innovations. Lied about the economic advantages of their breakthroughs. Lied about the breakthroughs themselves-all in the service of promoting their innovative technologies.

Remember artificial intelligence? Computers were going to automatically translate from one language to the next. Take dictation. Run factories without human intervention. Lead space missions. And we're not talking about predictions made a couple of years ago. These fanciful ideas were promoted way back in the middle of the last century: in the 1960s. How about the energy that was going to result from nuclear power: "too
cheap to meter," one enthusiast crowed. Or the nuclear-powered airplane. Does that ring any bells?