Consider the account of Danny Hillis and the local maxima. Hillis investigated whether he could generate a number-sorting program through imitating the mechanism of natural selection. He programmed his computer to generate a large variety of mini-programs by random. They were all hopeless for the task he had in mind, but a handful were less inadequate than the rest. Hillis introduced certain parameters. The ones that were most able to sort numbers survived, crossbred and mutated. The rest were deleted. After a few thousand cycles, the computer created a number-sorting program that was actually quite effective. The program had evolved.

But alas, the program which resulted was only quite good. It was inferior to code that could be produced by a competent programmer. Hillis repeated the experiment several times and always came up with similar results.

His solution was ingenious. Hillis programmed a predator into the system which could destroy code that had stopped evolving. This forced the code to evolve in more radical ways, and ultimately led to a much superior number-sorting program.

In other words, the predator forced the evolution process to go down paths it might otherwise have rejected, and some of these alternative routes threw up an even better adaptation down the line.

In short, sometimes we have to make something worse before we can improve it. Evolution is not good at allowing for this.

As a result, catastrophes can prove to be essential tools for creating new adaptations. Extinctions can be essential building blocks in evolutionary history.
Frequently a major shock is the catalyst for change.

It may follow, then, that economic evolution is powered by two forces. A slow, dead slow and stop kind of force and then sudden change that may occur.

Return to Why Black Swan may explain a lot, but not all