Evolution likes local maxima. Getting out of them is difficult. That’s what the OOP meant with “evolution was powerless to correct it”.
Getting out of local maxima means you first have to go with a worse setup until you get to a new, better local maxima. That’s why evolution doesn’t really do that all that often and instead prefers small optimizations.
(I use “like” and “prefer” not to say that evolution has goals or emotions, but to say that that’s what the “algorithm” of evolution leads to.)
Evolution likes local maxima. Getting out of them is difficult. That’s what the OOP meant with “evolution was powerless to correct it”.
Getting out of local maxima means you first have to go with a worse setup until you get to a new, better local maxima. That’s why evolution doesn’t really do that all that often and instead prefers small optimizations.
(I use “like” and “prefer” not to say that evolution has goals or emotions, but to say that that’s what the “algorithm” of evolution leads to.)