Suddenly the world is racing to find ways to create cheap, clean energy. A study of the Internet’s long development since 1946, arguably when modern computing began, offers valuable lessons that can help keep energy innovations on course for a comparable 62 years:
Hardening of the Categories. In the early days of the Internet there were computers, monopolized by IBM, and there was communications, monopolized by AT&T. The U.S. Department of Justice regulated IBM, and the Federal Communications Commission regulated AT&T. Officials in Washington, D.C., actually had years of hearings to decide which agency was in charge of which: computers versus communications. That turned out to be a silly hardening of the categories; the Internet soon became a merger of computing and communications, not one or the other.
Voice, video and data were once separate categories of communications, too, narrowly construed by their respective monopolies to be telephone, television and the Internet. Over the decades voice became more than telephone, video more than television, and both are now integrated into the Internet. For a long time the goal was to integrate voice, video and data by carrying them on the same copper wires. The old telephone and television monopolies were surprised when voice and video became data on the Internet. So don’t be surprised as we learn not to harden the categories, for example, among feed, food and fuel; corn ethanol is teaching us that right now.
Choosing Laws. Few people remember that during the early years of the Internet, computing resources were centralized. Grosch’s Law said that bigger computers were better, so IBM introduced its big IBM 360 mainframes in 1964. But starting the very next year, Moore’s Law came out of Silicon Valley saying many little computers were better than a few big ones. Then, if I may say so myself, Metcalfe’s Law came along to say that the more you networked computers the better. The Internet’s laws changed. Our way of looking at computing and communications changed. We learned too slowly that cheap and clean communications would be distributed. Personal computers and mobile phones surprised us. What do we think energy’s “laws” will be? Will cheap and clean energy come from centralized power stations? Internet history makes me think not.
We also believed, in the early days of the “information explosion,” that we needed to conserve bandwidth. The huge multibillion-dollar infrastructure of copper wires around the world was limited by some version of Shannon’s Law to carrying not much bandwidth. So our first priorities were on bandwidth conservation, on data compression, multiplexing, buffering and smart terminals—the low-hanging fruit. But after 62 years of building the Internet, do we use less bandwidth now? Was building the Internet mostly about conservation? No. Today we are not using less bandwidth, not twice as much, not 10 times as much, but something like a billion times more.
In 62 years from now, in 2070, are we going to use less energy than we do now? No. Today I do walk more, bike more and take trains more, and I’ve recently switched cars, dropping from 12 to four cylinders. I’ve already ordered my next car, which will move me from 30- to 40-plus miles per gallon. I plan to go all electric after that. And I turn my energy-efficient lights out more religiously. But I don’t think for a moment that we’re going to conserve our way out of the energy crisis. Internet history shows that prosperity depends on abundant bandwidth. Prosperity (gross domestic product, per capita) is proportional to energy use. We are not going to lower per capita consumption of energy in the U.S. We are going to enable the rest of the world to be as prosperous by using not less but more energy. We need to make energy cheap, clean and therefore abundant—really abundant, for a really long time.
Going to Washington. I’ve also heard that the green energy movement is going to Washington, probably right after the elections in November. For the Internet we went to Washington to build Al Gore’s Information Superhighway. And the Internet bubble burst. Remember, live by the sword, die by the sword. When innovators go to Washington, it’s a Pro-Am match—professionals versus amateurs—and innovators are the Ams. The status quo can be dangerous and downright mean, and they own Washington. The status quo’s monopolies are very good at lobbying and litigation.
When you go to Washington to get stuff, sometimes you get the wrong stuff, like subsidies for corn ethanol—the wrong feedstock for the wrong fuel. The best thing about corn ethanol is that taxpayer money is being misspent in the Middle West instead of the Middle East. Washington thought that the markets for corn—feed, food, fuel—were separate. Wrong.
When your movement goes to Washington, you also get things like the Department of Energy. The next time we go to Washington, could we start by fixing the DOE? The department developed bad policies, such as not building nuclear power plants for 30 years. The DOE is also malfunctioning by soaking up energy research dollars in its huge and inefficient, congressionally mandated labs—dollars that would be better spent at competing research universities. Be careful when you go to Washington, thinking it’s going to fix things, thinking that’s the way to scale up energy innovations.
Scale. It is indeed a big challenge. The world needs scalable solutions. But be careful about demands for scale early in the process of technological innovation. One of the weapons used by defenders of the status quo is to set high hurdles for innovations. They say, “It’s got to be safe, of course, and it has to scale, or we and our friends in government are not going to let you do it.” In Washington legislators make laws that discourage new technologies when it is not immediately obvious how the technologies are going to scale. The telephone monopolies got Washington to make rules about the Internet that said, “If you can’t serve everybody, then you can’t serve anybody.” Which meant innovators couldn’t start small and grow new technologies, driving them down steep cost curves. You had to serve everyone, all at once, right out of the lab. That’s too high a hurdle for new technologies. So watch out when people demand that we consider only energy solutions that scale, which often turns out to mean only mature technologies already in the hands of the status quo.
Note: This article was originally printed with the title, "Learning from the Internet".