In a genetic algorithm you will often see two different local optima being generated, of which the combination operator always creates worse solutions, unless one of these solutions manages to climb back to one of these local optima.
So imagine a fitness landscape with a few huge peaks, and a lot of valleys around it. Combining two of the peaks will lead to a point in the valleys that will either move up back to one of the peaks, or miss out in the next generation.
What if this applies to humans. What if there are genetically good solutions to environmental problems, with huge masses of failure in-between.
"Fitness" is subjective. Fit to do what? To make money? - Jew wins. To play basketball? - Nigger wins. To breed? - Muslim wins. To advance computer science? - Anal Tuning wins (because you have to be gay)
If Turing wasn't gay, how come he was messing with computation, instead of cuddling with girls? Right! He wasn't interested in girls, leading to the only possible explanation: he was gay.
In a genetic algorithm you will often see two different local optima being generated, of which the combination operator always creates worse solutions, unless one of these solutions manages to climb back to one of these local optima.
So imagine a fitness landscape with a few huge peaks, and a lot of valleys around it. Combining two of the peaks will lead to a point in the valleys that will either move up back to one of the peaks, or miss out in the next generation.
What if this applies to programming languages. What if there are linguistically good solutions to programming problems, with huge masses of failure in-between.