From Lloyd of It Looks Different From Here :
When the Human Genome Project was completed everyone was surprised at how few gene there were. Only about 20-25,000 protein coding genes. It was not that much larger than that for much simpler organisms. So where was the greater complexity of humans and other mammals coming from? Where was the coding for the difference between human brains and those of other mammals?And in the comments:
The talk I went to today was on estimating how many micro-RNAs there were in a given species. It turned out to be a lot. An awful lot. In a mouse they estimated that there were over a million micro-RNAs. Less complicated organisms had an order of magnitude or more less micro-RNAs. And what was really interesting was that humans had over three million different micro-RNAs. Nothing else came close.
Guess what they think most of the extra micro-RNAs in humans are doing? That's right. They are probably a major part of the plan of the brain.
It seems to be the construction plans. Of course the brain modifies itself, there is no sharp distinction between hardware and software. And these micro RNAs will be involved in the modifications.Debugging self-modifying systems is notoriously difficult.