[Nebullis]

Changes in the comment field have a probability that looks like P1*P2*P3...PN.

In evolution, any change looks like that. I wonder why you differentiate between neutral changes and other changes.

[maro]

No, a non-neutral change, one that affects function, will (or is likely to) be selected for or against to one degree or another. For example, sickle shaped hemoglobin has been selected FOR in Africa, apparently because having this trait helps with malaria. Sickle shaped hemoglobin is selected AGAINST outside the malaria belt. Traits that are selected for or against cannot be said to continue to have the original probability of occurring, PI. This is the core of evolutionary theory. It is a powerful idea, and true--within its narrow domain of changes that can be reached in a small number of bitflips from a prior state of DNA. My argument throughout this thread has been that this theory CANNOT apply (or seems implausible given our current state of knowledge) to changes that require a large number of bitflip changes. In THAT case, it is difficult to see how a path of intermediates could exist and be selected for, culminating in an individual having the trait under discussion.