(November 5, 2017 at 1:31 pm)Abaddon_ire Wrote:(November 5, 2017 at 8:31 am)Mathilda Wrote: That is actually a valid criticism of how the theory of evolution is taught. There is this tendency to describe how a single beneficial mutation will start to propagate throughout a population. But there are so many other ways that a single agent can die early that a beneficial mutation may not confer enough of a survival advantage for it to start propagating and therefore is lost.
What's not often described though is how a single mutation will confer a survival advantage on average. That is if it does start to propagate throughout a population, then there will be many agents with that mutation, with some still dying off early for other reasons, and others living for longer. And in the same way that there are animal viruses constantly spilling over into the human population with most not be able to survive in the human body and just occasionally one being able to survive and spread, the same mutation may happen independently multiple times and if it has the potential to confer a survival advantage will on average be more likely to propagate.
This was the sole reason I stopped using genetic algorithms and took my understanding of how they worked to develop my own evolutionary algorithm that kept track of how many times a mutation had been tested. It is for this reason that the evolutionary process takes so long, because even a beneficial mutation needs to be tested over the course of many lifetimes as it spreads throughout a population.
Or conversely a deleterious mutation may survive and spread because it has occurred in an agent that has many other survival traits. But eventually that mutation will die out because it does not confer the survival advantage itself, it just struck it lucky. This also suggests that there are also many other beneficial mutations that may have occurred but due to sheer bad luck never actually spread. But then the evolutionary process is not directed, there is an element of chance of which evolutionary strategy will be adopted.
So when we talk about a beneficial mutation propagating, or a deleterious mutation dying off, then we always need to remember that this is a process happening on average over the course of many, many different lifetimes.
That really puts the onion in the religious ointment. Evolution happens to populations over time, not to individuals in an instant.
It's kind of like global warming deniers who confuses temperature ("only a 0.5 rise") with heat ("what a 0.5 increase in the oceans amounts to").
