(first published somewhere else 2005/7/28 )

Many people say they don’t believe in evolution. To me, as a computer scientist, that is a ridiculous statement. I have seen evolution at work, and used it in a program. Of course I understand these people mean the evolution of organisms, leading to homo sapiens by coincedence. It is debatable whether there was some intelligence behind that or not. Personally, I believe evolution by itself is intelligent enough to accomplish that, when given a couple of million years. But that’s not what I want to write about today.
Evolution is a process in which generations of individuals in a population change by mutation, often also crossover, and selection of the best/fittest solution. An example:
Let’s say you plan to do a few things today. You could do those in different orders. Some plans would be ridiculous and impossible, like shopping for food after cooking that same food. Others are impractical, and there are a few plans that are practical, of which one is optimal. Your goal is to find a close to optimal planning for today. You might make up a few different ones, compare them (selection), change them (mutation), or combine parts of different plannings (crossover). This is evolution! “Wait”, you say, “I make intelligent choices, not random mutation and crossover”. Well, those intelligent choices might help you get a good planning quickly, but a “dumb” evolutionary algorithm running on a computer will find a better solution in less time. Trust me, I know from programming experience that evolution works. It is actually used to find good solutions for problems that are impossible to find an optimal solution for before the universe will collapse/explode/freeze.
But that’s not the point. What I’m trying to tell you is that not only species evolve, data can evolve too, and even ideas evolve. A large part of the process of science IS evolution. It is also known that languages evolve. Understanding what individuals, population, selection, mutation and crossover mean for science and languages is left as exercise for the reader.

Now, if ideas evolve, that means we can take a evolutionary algorithms / computer science type of view at how innovation works.

For example, take copyrights. Making copies of most copyrighted stuff is not allowed, but you can reuse the ideas. This means more work to produce a mutation, so it’s slowing the mutation speed. With random mutation, that is not always a bad thing, but with our intelligent mutation operator called “thought” it probably almost always means less optimal solutions.
Conclusion: the free/open way of using copyrights is better than the proprietary way.

What about patents? A patent means only one organization can use an idea for a while. Patents reduce the population size. A smaller population means less, possibly good, mutations can be explored. What’s worse, patents make crossover impossible. Crossover is known to make evolution go much faster. (By the way, crossover is the reason we reproduce in couples instead of budding clones like bacteria do. Think about that.)
Conclusion: patents are a very, very, very bad idea.