Computation and the Human Predicament
The Limits to Growth and the limits to computer modeling
Forty years ago this spring, a little book titled The Limits to Growth landed with a big thump. The authors reported on an exercise in computer modeling, which they grandly described as “Phase One of the Project on the Predicament of Mankind.” According to the model, the human predicament was bleak, with less than a century to go before civilization would crumple under the burdens of overpopulation, famine, resource depletion and pollution. As a young journalist I was fascinated by this apocalyptic vision. I was also intrigued by the remarkable idea that computation might be a useful tool for understanding the human predicament.
In 1972 I had no way to explore the workings of the Limits model for myself. Twenty years later, though, with a desktop computer and ready-to-run modeling software, I was able to twiddle the model’s various knobs and observe the effects on the outcome. I wrote about that experience in 1993, in the first column published in these pages under the rubric “Computing Science.”
Recently I have turned to the Limits model yet again, this time delving into details of its implementation—the 150 equations that govern the evolution of the simulated world. Closer examination of the model’s structure has not increased my confidence in its predictive power. On the other hand, the hope that computation might have something to tell us about the fate of the planet remains very much alive. We don’t have an abundance of better tools for seeing into the future. An interesting challenge is to clarify what distinguishes the computational methodology of The Limits to Growth from other models that policymakers take more seriously, such as the climate models that now underlie much of the discussion of global warming.