If you ever wanted confirmation that patents aren't there to improve innovation take note of the Evolutionary Computation Conference which showcases (among other things) some new technology that evolved via evolutionary algorithms. One of the notes is a "high performance Wi-Fi antennae (evolved to avoid patent fees)" -- Which isn't a hard thing to note, you just put in fitness blocks that say if you use the important features here here or here you're less fit. You simply evolve a good design while avoiding certain designs. You hamstring your design process.
Which brings up an interesting problem with patents. Obviousness. Who is this criteria suppose to apply to? If it's obvious to an evolutionary algorithm shouldn't that count? I mean, isn't that the entire point of obvious criteria to determine if some idea is simply the next obvious step in the making of something and compare that to what else could be done. The point is that the obvious next step shouldn't be grabbed and patented but rather people should know that their hard work won't be for nothing and they will get paid for their innovation. This, sadly, is nothing like the patent system works.
You could in theory patent all of these adaptations and run the program again to avoid these and again and again. Sometimes there are certain bits which are needed. The Good Tricks to designing things that are obvious and are the way things are built.