Skip to content

Recombinant Innovation

Definition

Against economist Robert Gordon's "innovation-as-fruit" view — that big inventions get picked and their benefits gradually exhausted, so growth must eventually slow — the book argues innovation is mainly recombination: "building blocks" (technologies, techniques, pieces of knowledge) get combined into new configurations, and each new combination becomes an additional building block for future combinations. Because combinatorial possibilities explode, building blocks are never used up; the binding constraint becomes the capacity to search the space of combinations for the valuable ones, not the supply of raw ideas.

In the Book

The chapter grounds the idea in Kary Mullis's Nobel Prize-winning invention of PCR, which Mullis himself described as pure recombination of existing biochemistry techniques ("there was not a single unknown in the scheme"), and in economist Paul Romer's "new growth theory," including his concept of "meta-ideas" — ideas about how to support the production of other ideas. The book then argues digital technology is recombinant innovation in its purest form: Waze recombines a location sensor, a phone, GPS, and a social network, inventing none of them; the Web recombines TCP/IP, HTML, and the browser; Instagram recombined mobile photo-sharing with filters. Economist Martin Weitzman's mathematical model of "recombinant growth" is used to show that as the number of "seed ideas" grows, the number of possible combinations explodes even faster — so the limiting factor shifts from generating ideas to processing them, which the book argues digital tools (and more connected human "eyeballs") directly address.

Why It Matters

This reframes stagnation debates: if progress in a field feels like it's running out of "big ideas," the recombinant view says look instead at whether the number of combinable building blocks and the capacity to search among them are still growing — because if they are, the supply of valuable new combinations is not close to exhausted. It's a domain-general lens for any system where value comes from combining existing components (research, product design, business models, even culture) rather than from de novo invention.