> A reasonable default assumption is that each switch in our "genetic configuration" is set to a quite good value. I'd question that assumption-- each switch

> A reasonable default assumption is that each switch in our "genetic configuration" is set to a quite good value. I'd question that assumption-- each switch in our "genetic configuration" is set to a value, and it's set to one that results in survival often enough for the species to reproduce, but it's not necessarily the optimal value, or even a particularly good one. It's just one that happened to be able to survive under environmental pressures at some particular time. To put this into a computer science context, genetic algorithms have the same problem-- they can find local maxima easily, but can have difficulty (depending on the nature of the search space) locating global maxima because they can't sacrifice short-term fitness to gain long-term fitness.