Frances Arnold: So it’s easy to say, “I want to be an engineer of the biological world,” but no one had any clue how to do it. These are complicated things that we know very little about. How does DNA encode a human being? How does DNA encode even a single protein molecule? These are answers that we don’t have yet and we won’t in my scientific lifetime. But there’s a process by which all of that was invented, and that process is called evolution. No human being designed any of this. It came out through a simple algorithm of mutation and natural selection.

So I didn’t know how to design new DNA, but I was aware of this engineering process. And I said, “Okay, if I don’t know how to design, why not breed molecules? Why not breed them like you breed cats and dogs? We don’t know how this DNA encodes a hairless cat, but you don’t have to know. You can breed it by choosing the parents and deciding who goes with whom and which progeny from the next generation go on to reproduce. You can do the same thing with molecules in the test tube.” So I rejected the way that the engineering world was approaching this problem. A biochemist would say, “We have to get the structure. You have to understand everything, then you can engineer it.” I said, “No. I’ll be old and dead before that happens.”

Why not use the process that nature uses and breed these mutant molecules in the test tube? Make mutations in the DNA; recombine DNA from 33 different sources. You don’t have to have two parents in the test tube. And then let the system search through those new products and see which ones are starting to acquire the traits that I’m interested in. And that way, we would completely circumvent this complete ignorance of how DNA encodes function. We would just evolve it. And of course, people thought that was a nutty idea and it wasn’t “scientific.”