As we debate energy independence, fossil fuels and renewable energy, it's a complex challenge to foretell the winners and losers. As Yogi Berra put it, "Prediction is very hard, especially about the future." But that sure doesn't keep us from trying.

Perhaps there are lessons to be learned from the world of computer technology.

Three decades ago, a data-processing director would have predicted a bright future for centralized mainframe computing, with dumb terminals for the end users--a hub-and-spoke configuration. We know how things turned out, but it wasn't so clear back then, even though Intel co-founder Gordon Moore had predicted back in 1965 that the number of transistors on a chip would double roughly every two years.

This column appears in our monthly Infrastructure email newsletter. Click to subscribe.

What came to be known as Moore's Law essentially proved to be true: As chipmakers continued to shrink hardware, desktop computers and distributed computing gave control to end users. Now, notes Intel's website, "the evidence of Moore's Law is everywhere, embedded in devices millions of people use every day, such as personal computers and laptops, mobile phones, and common household appliances and consumer electronics."

Is there a parallel in the energy field? Will there be a Moore's Law for energy technology, one that predicts the impact of local, "generated-on-site" energy-conversion technology?

We're certainly seeing plenty of innovation in the field of locally generated power, from small wind turbines to solar panels and fuel cells. For the most part, though, in the world of energy we're still about where we were with information technology in the 1980s. Large centralized power plants convert energy (principally fossil fuels) from one form to another (coal to electricity, for example) and distribute power to end users: our homes and businesses and transportation system.

Roger Duncan should know. Duncan, a former general manager for Austin Energy, the Texas capital city's municipally owned electric utility, points out that almost all of the energy-conversion technologies we use today were invented in the 19th century. Those technologies continue to define the physical parameters of our hub-and-spoke energy system.

These days, Duncan looks at energy in a different way as a research associate at the University of Texas' Center for International Energy and Environmental Policy. "Energy comes in a multitude of forms, and those forms are constantly changing," he says. "Sometimes the change is very slow, like the erosion of a mountain, sometimes very fast, like the explosion of fuel in an engine."

It remains to be seen whether the development of new energy-conversion technologies will have the impact that what Gordon Moore predicted had on information technology. In 1965, the Internet, personal computers, mobile phones and the World Wide Web were unknowns. Think for a minute where'd we be right now if our economy still relied primarily on centralized mainframe computers and dumb terminals.

Yogi Berra was right about the difficulty of predicting the future, but in the world of energy, some facts are clear. Fossil fuels are limited resources, and as supplies diminish their extraction costs increase. Sun and wind are essentially limitless resources, and their energy conversion costs are technology-based and decreasing.

Predicting that renewables will at some point become the planet's predominant energy source is easy. Predicting when that will happen is next to impossible. Will there be a new energy-technology paradigm guided by the equivalent of the computing world's Moore's Law? I sure wouldn't bet against it.