The Revolution Before the Industrial Revolution

Gave this talk at GROK Conf '26 this morning. A few people asked for it, so here it is — audio, slides, and the framework I landed on at the end.

The short version: the Industrial Revolution didn't start with smokestacks. The factory system was invented in 1678 in Piedmont silk mills. The world didn't reorganize around it until 1800. What bridged the 120-year gap is a useful test for anything calling itself a revolution today.

The framework

A revolution needs three ingredients, all at once:

  1. A new machine that changes the unit economics of a task.
  2. A staple-fiber market — a use case so ordinary and universal it can't be sidestepped.
  3. A labor model that lets the machine meet the market.

Silk mills had 1 and 3 for two centuries. No revolution. Silk was a luxury — you can't power a ship with silk sails, wrap wounds in silk bandages, or clothe a soldier in silk. The factory model existed and worked. It just couldn't rearrange the world until cotton gave it a staple-fiber market to rearrange.

That's the test. Apply it to anything.

  • Does the machine exist? (The easy one. Usually yes.)
  • Is there a staple-fiber market? Or is it still silk?
  • Does a labor model exist that lets the machine meet the market?

Pick your favorite revolution of 2026. Which of the three ingredients does it actually have — and which one is missing?


Source: Virginia Postrel, The Fabric of Civilization (Basic Books, 2020). If you liked this, read chapters 1–4.