← COVER

When Earth Became Computable

DRISHTI · CHAPTER 3

When Earth Became Computable

How the watching became data and the data became a calculation: a slightly wrong metre, a baby in 176 squares, a whale shark found by starlight software, and a forecast that took six weeks to be wrong.

EARTHVISION LAB · ~8 ARTICLES

Four centuries of watching left behind an enormous pile of paper, and paper cannot be calculated.

Turning it into something a machine could use took longer than the watching did. A thermometer reading had to become a number on a scale that strangers had agreed on. A handwritten rainfall total had to become a row in a table. A hectare of living ground became one square in a grid, a whale shark became an identifier with a life history, and the archive stopped being posted to the computers and was parked next to them instead.

Then came the question that made all that effort worthwhile: could the numbers be pushed forward? The idea that tomorrow was a calculation rather than a guess is a little over a century old, and it began with the atmosphere because the atmosphere's equations were already known. Water, forests, animals, crops, power grids and ships followed, each with different mathematics and a different reason for being difficult.

Every step left something out. The metre came out 0.2 millimetres short and nobody minded. The temperature on the news describes a lawn inside a white box. A forecast model leaves out whatever it cannot afford to compute. None of these were mistakes. They were the price of being able to compare one place with another, and of being able to calculate at all, and they were almost never written down beside the answer.

The useful questions are therefore precise ones. What did each record decide to keep, and what did it decide the world could do without? Where does a forecast stop working, why does it stop, and how much of the doubt should survive the journey to the person holding it?

CHAPTER CONTENTS