I WENT to Eastern Europe looking for a new generation of computer users. I found Andreas.
A chemical engineer, he and his wife, Petra, are raising two daughters in what used to be East Berlin. Their second-story apartment is a study in contrasts. Petra (also a chemical engineer) shovels brown coal into a stove to supplement the gas heat. The family considers itself fortunate to have a phone. Yet, at the end of one hall, sits a 486-50 MHz computer.
Under the Communists, these people had limited access to technology. Today they have more computing power at their fingertips than I do.
If anyone doubts that developing nations are rapidly clambering up the technology curve, let them come to Eastern Europe. In offices and student quarters, 486-class machines are beginning to pop up. Color monitors display software that any American computer user would immediately recognize. Intel chips and Microsoft Windows are sweeping the world.
Later I found a 486-66 MHz machine in a student-organization's office in downtown Warsaw. A few days after that, I saw another 486 parked in the office of a mid-sized Polish town.
Until a few months ago, these models were the world's most powerful personal computers. Now they're here. Increasingly around the world, computer power is available and cheap - a commodity that perhaps will be as necessary for the next generation as oil and electricity are today.
It's a positive process, but it raises some troubling questions.
Many in the industry reject this commodity analogy. The pace of innovation is far faster in computers than anything in agriculture or mining, they point out. Parts of the industry are hard for newcomers to crack. In 100 years, corn will still be corn. Who knows what computers will be?
But these arguments miss the point. Today's computers are a different kind of commodity - sophisticated but standardized and increasingly interchangeable goods that just about everyone wants. And gets. That's a dramatic shift for Eastern Europe.
Before 1989, when East Berlin was still Communist, Andreas worked on an old 8-bit CPM machine. Not only was it slow, researchers also had to share it, booking in advance to get time on the machine.
What happened? Like many Eastern European scientists who had little access to physical resources, Andreas became very good at the theoretical aspects of his job. If the machine was too slow to calculate a problem, he used better math to simplify the equation. He and his colleagues published at least three articles in scientific journals using that antiquated computer. Their work has been recognized in the West as well as the East.
In his new work finding ways to purify drugs, Andreas has access to much more computing power. He writes programs in a computer language called Turbo Pascal. His access to software is virtually as good as that of any Western scientist. But ask him about his biggest breakthrough in recent years, and he points to the math, not the machine.
Some of his old CPM tricks still come in handy. When a colleague complained recently that an equation was taking hours on an advanced computer, Andreas showed him a simpler algorithm that did it in seconds. An algorithm is a set of rules for solving a problem in a finite number of steps.
Andreas worries that the math skills of the next generation will deteriorate because of computers. ``I see many of these young students who focus too much on the machines, and they play with all these nice programs and at the end of the day, they don't have anything accomplished in their work,'' Andreas says.
Does this mean he'll unplug his 486? No. You don't throw away a superior tool. But you do learn when to use it.
Twenty years ago the United States woke up to the costs of over-reliance on a commodity called oil. Will people one day realize the costs of overreliance on computers? The Worldwatch Institute reports that computers already consume as much electricity a year as does the entire country of Brazil. Their effect on our mental environment remains, so far, unseen.
* Send your comments on this column to CompuServe (70541,3654) or Prodigy (BXGN44A).