There is an argument that proteins coded by 20 amino acids via the DNA program has so few successful protein formation that a search never can reach the complexity of it via natural selection and genetics. This is a very similar argument as that of the impossibility of the eye. Again the answer is probably that the advanced proteins evolved from other proteins or from proteins that was smaller and now extinct. But still there is another theoretical possibility that I find interesting. As I understand the cell can correct mutations, but if the DNA can encode how to correct we get a self modifying code at least to a degree that can tune and change the mutations to more successful protein mutations. As a fan of Allan Turing and his turing machine, that would be a cool feature of the genetics in the cell. Is there any proofs or indications of mutation cooprated self modifications of the DNA program?
Actually there is an indication. The DNA is the program code that is a blueprint for the cell itself and also the self correction mechanism. Hence there are code points in the DNA that has the possibility to control how the corrections of mutations is done. Hence code points in the DNA affects how changes in the DNA is done. So not really a direct way of modifying, but an indication it is. This is interesting as a change of the correction mechanism can change the speed of mutations and hence also the speed of evolution.