Essay The Uncomputed World #3

Why Simulate the World

The post-AGI economy arrives in two waves. First, thinking becomes cheap: a design that once took an engineering team a month costs a kilowatt-hour. Then building becomes cheap: robots build robots, factories build factories, and machines supply their own ore and power.

Between the thinking and the building, one question remains: will it work? In the end, proof comes from one laboratory: the world itself. And the world runs at its own pace: a year of wear takes a year, and a storm arrives on its own schedule. Once thinking is instant, the wait for proof becomes the slowest step in all of progress.

A second world

The answer is not to speed up the old world, because no one can. It is to build a second world, from the laws of physics. In this simulated world, time is no longer tied to reality: a year of wear can finish in an hour. In the real world, a test shows only what its sensors happen to touch. In the simulated world, everything is visible: every field at every point, even inside a failure as it happens. The world can be copied: a thousand variants of a design, each tested in its own world, all at once. And nothing is too rare or too dangerous to test: the crash of a real bridge, the storm of the century, the meltdown that must never happen.

Proving becomes fast

The economy rebuilds itself around the second world. Machines will propose new designs, prove them in the simulated world where proving is fast, and build only the winners in the real world. And a design proven in simulation does not need to be proven again. The real world still has the final word, but its role changes: it is no longer the laboratory where ideas are tested; it is the destination where proven designs arrive.

Life evolved in the real world, by trial and error, over billions of years. Technology will evolve in the simulated one, at the speed of compute.

That is the third wave of the post-AGI economy: proving becomes fast. When thinking is cheap, building is cheap, and proving is fast, better engines, better drugs, better materials arrive as fast as the second world can prove them.

Zongfu Yu

Works on the physics, the computation, and the company-building behind first-principles simulation at scale.