Physics and mathematics can tell us how the Universe began, but as the cosmologist Stephen Hawking noted: “They are not much use in predicting human behaviour because there are far too many equations to solve.”

The motives, needs and desires that drive human action have long resisted rational analysis. From the volatility of the stock market to fads and fashions that flare brightly and then vanish, the ability of individuals to act unpredictably has undermined attempts to model their behaviour with any level of precision.

The science-fiction writer Isaac Asimov had the right idea. If one considers a sufficiently large population of people, he wrote, then just as the mass movement of a gas can be inferred through simple calculations — whatever the individual molecules might do — so too can the future actions of a large population.

Asimov called his fictional science of predicting people’s behaviour psycho­history. He used it as a central plank of his classic Foundation series of books. The predictions of psychohistory were more than a model, they were a set of instructions for how future societies must respond to a predictable crisis they helped to create.

In a Comment on page 365, Paul I. Palmer and Matthew J. Smith call for human adaptation to climate change to be modelled to help avert a real-life predictable crisis. Existing models of the planet’s changing climate are insufficient, they argue, because they leave out the people. Omitting human behaviour from these mathematical studies, they write, is like “designing a bridge without accounting for traffic”.

Societies will be different in a warmer world, they point out, and we should understand how this will unfold. It is, in essence, another feedback in the climate system, and one that should be quantified and accounted for. Perhaps another seven billion equations will need to be added to the mix.