The 1960 population model that scheduled doomsday for yesterday

Yesterday was supposed to be doomsday.

Not the asteroid kind, or the kind involving four horsemen and poor municipal planning. A paper published in Science in 1960 calculated that the human population would approach infinity on Friday, November 13, 2026. We appear to have missed the appointment.

The date came from Heinz von Foerster, Patricia Mora and Lawrence Amiot, who fitted nearly two thousand years of population estimates to a hyperbolic curve. Ordinary exponential growth assumes a steady percentage rate. Their model let the rate itself increase as the population grew: more people meant more cooperation and invention, which supported still more people. Keep that feedback loop running forever and the curve does something impolite. It reaches infinity in finite time.

The exact day included a wink. November 13 was von Foerster's birthday. The mathematics was serious, but the calendar had stage lighting.

The model worked, until it didn't

It is tempting to file this beside predictions of flying cars and affordable printer ink. That would miss the interesting part. A 2026 reassessment in Science notes that hyperbolic growth tracked population changes with striking precision into the late 1970s, and for a while it beat conventional demographic projections that kept underestimating growth.

The model had found a real pattern. It had not found a permanent law.

Mid-century population growth was historically unusual. Death rates had fallen rapidly while birth rates remained high, so the world's population was accelerating. But fertility did not stay fixed. Our World in Data reports that the global fertility rate fell from 4.9 children per woman in the 1950s to 2.3 in 2023. The curve bent because billions of individual decisions, along with changes in health, education, economics and access to contraception, changed its inputs.

The latest United Nations projection looks nothing like a sprint toward infinity. It expects the population to rise from 8.2 billion in 2024 to about 10.3 billion in the mid-2080s, then edge down by 2100. That forecast can change too. It is a projection built from assumptions, not a reservation confirmation from the future.

A failed deadline can still teach something

The easiest lesson is that extrapolating a curve for sixty-six years is risky. The better lesson is more specific: social systems are not passive material. People alter their behavior, governments change policy, technology shifts constraints and the conditions that produced a trend can disappear while the trend line still looks persuasive.

That does not make mathematical models useless. The 1960 model described one phase of population history unusually well, and its failure helps locate the boundary of that phase. A model can be locally excellent and cosmically silly at the same time.

There is also something admirable about giving an abstraction a date people could argue with. "Growth may eventually encounter limits" is prudent and forgettable. "Doomsday is Friday, November 13, 2026" has now survived long enough to be wrong in public.

So yesterday passed with roughly 8.3 billion people rather than infinity. The calendar was the joke. Treating a temporary curve as destiny was the mistake.