There is a school of thought in the computer world that is advancing the concept of what they call a technological singularity. The proponents of this concept point to the accelerating pace of computer speed and capabilities and state that there will be a point in the near future where computers gain a kind of consciousness and start to improve themselves. They call this point a singularity in reference to the gravitational singularity of a black hole. A black hole is a collapsed star so massive that gravity doesn’t allow even light to escape and common physical principles don’t apply. This departure from predictability is the essence of the appropriation. Once computers start advancing their own development the speed and direction of that development would be unpredictable. Some proponents of the theory claim that we will reach this singularity within the next 25 years.

There are doubters, of course. Some point to the recent slowing of the rate of increase in computer speed. Some question the basic principles of the argument and accuse the proponents of misleading themselves about the limits of electronic computation. I have a diametrically opposite opinion on the long-term future of computation in our society.

There is another limit on the evolution of computing power and on the widespread use of digital electronic technology in general. That limit is discretionary energy. The development of computer technology has occurred in an economic environment rich in discretionary energy. From the Second World War onward the worldwide production of coal, oil, and natural gas has been increasing and the supply has been far more than sufficient for the basics of human life. For decades we have been using these resources with absurd inefficiency, spending them on recreational mobility, and engaging in non-essential activities such as space exploration. In such a glut there is plenty left over for processor chip manufacturing and facilities full of servers and routers.

Our access to complex electronics and computing power is striking. Many children have cell phones, each device containing more computing power and memory than the mainframe computers of 30 years ago. The devices are affordable even to the relatively poor. Personal computers are common, if not universal. Even more significant are the electronic devices we no longer really notice: the automatic door opener and scanner at the supermarket, the programmable timer on the coffee maker, the smoke alarm, and the cordless phone. Even less visible and more important are the electronics that control our power grid, coordinate our transportation system, and speed our industrial production.

All this electronic intelligence relies on set of interlocking conditions. It is hard to know where to start, given the complexity of the connections. There were the initial scientific discoveries that were, in turn, augmented and speeded by the technological developments they enabled. There was the demand of early adopters, including the space program and the military that jump-started the consumer market, which then had its own early adopters. There was the ramp up into mass production. Then there was the export of high-tech manufacturing to countries with despotic governments and the resulting low standards for workers and the environment. This interaction of technological development, mass demand, and cheap mass production brought the price of electronic computing into a range that the ordinary consumer could afford. It also enabled product designers to include intelligent features in what were previously manual devices.

Underlying this all is discretionary energy. It is this energy that offers masses of people in the industrialized world the prosperity to be a mass market for electronic consumer goods. It is this energy that allows the mass international shipment of these electronic goods. It is this energy that allows industrialized agriculture to displace peasants into the cities of the third world, where they are available to cheaply produce electronic devices and the discrete elements that make up these devices. It is this energy that is available for the mining, transportation, and processing of the materials that go into these devices. Even the cheap plastic casings for all this electronic bounty are made from petroleum products.

The earth is no longer making fossil fuels, and therefore the supply is declining as we consume it. It is a geological fact that the annual production of oil fields, natural gas fields, and coal mines slows as they age. Many observers, myself included, conclude that the world is presently on the long bumpy plateau of peak production that precedes irreversible decline. So what happens when the supply of fossil fuels no longer meets all our superfluous needs? What happens when it no longer meets even our most basic needs for food production, heating, medical care, and the manufacturing of the basic necessities of life? What happens when the supply of diesel fuel and natural gas based fertilizer declines and the sons and daughters of third world factory workers make a desperate return to the land?

Part of Norse legend, as envisioned by Richard Wagner in his Ring Cycle operas, was the idea of Gotterdammerrung, literally the twilight of the gods. The balance of the world is lost, the rope of the Fates is broken, and the gods themselves go up in flames. I can foresee an Elektronikdammerrung, a twilight of electronics, when humanity no longer has the necessary supply of energy and materials to make them or the prosperity to drive demand. The huge chip and transistor factories in Southeast Asia will go to ruins. Like history running backwards, electronics will devolve from mass consumer goods to luxury consumer goods, and then to the vaults of universities, military bases, and government agencies. Computers will be lovingly tended by teams of specialists. More prosaically, people will open store doors by hand and time their coffee with windup devices, communicating the everyday events of their lives with letters and the landline telephone.

Eventually, without our present widespread use and facing the time and energy demands of post-petroleum agrarian life, our descendants will be faced with the decision between parts for the mainframe computer and the wheat harvest. Given the complexity of manufacturing processors, perhaps the industry won’t sustain itself below a certain level of mass demand. A period of electronic cannibalism will ensue and run its course. One by one, the last lights of the computer age will wink out.

I’m not a modern Luddite. Let’s face it, this piece was written on a computer and you are reading this on my website. I like my electronics and rely on them for business, personal communication, and recreation. However, I do not make the historical mistake of thinking that the way we live today is the way we will live forever. Nor do I make the similar mistake of thinking that the path of our society is onward and upward forever. The graph of human history is a series of rising and falling lines, with civilizations increasing in complexity and demand on their environments until they collapse. It would be arrogant to assume that we are exempt from the laws of physics, biology, and geology. I used to enjoy a vision of our future as a secular version of the Amish, augmented by a veneer of electronics, as a best case scenario for a post-energy-glut world. I now have a rougher, sparer vision, formed by what I know about the limitations of our resources and our species itself. If we are careful our descendants will still have the knowledge we have gained in our period of technological bounty. Perhaps they will be able to use much of it with only (I shouldn’t say “only”) their minds and their hands.