Danny Hillis' genetic algorithm experiments with his Connection Machine led to his formulating the Red Queen Hypothesis. Basically it is a coevolutionary model. GAs traverse a fitness space and find maxima in that space; and the problem, as Hillis observed, was that genetic algorithms/systems have the propensity to get stuck at a local maxima. It makes perfect sense in hindsight, if you think about it. To remedy the situation, he introduced another GA type (a 'parasite' species) to compete with the original type. The introduction of the predators destabilized the system and provided impetus for the original type to actually descend from a local peak and thus (possibly) find a better relative (or possibly absolute) maxima in that space. These flies were isolated. The system was homogeneous. The longer hair and more eggs are a local maxima.
I love these types of experiments... Just the other day, I was wondering if I turned up the temperature on our cell incubator by 0.1 °C each month, and continued to culture the same cell line in it, how high could I get the temperature before the cells gave out? Unfortunately, I don't think my boss will allocate an incubator to satisfy this curiosity. I share too many crazy ideas with him as it is.
- So in short, evolution can appear a lazy process if it can get away with it.
Hm. Perhaps I misunderstand what you are saying, but it is not a matter of taking its time getting somewhere ("lazy"). It has already arrived at what is a relatively optimal genotype, if the ecological context remains stable and a low probability mutation doesn't land a subset on a higher fitness gradient. Unfortunately, I don't think my boss will allocate an incubator to satisfy this curiosity. I share too many crazy ideas with him as it is. I have a feeling you will be successful, so perhaps soon you will be able to buy all the incubators you want! :)Just the other day, I was wondering if I turned up the temperature on our cell incubator by 0.1 °C each month, and continued to culture the same cell line in it, how high could I get the temperature before the cells gave out?
- Hm. Perhaps I misunderstand what you are saying, but it is not a matter of taking its time getting somewhere ("lazy"). It has already arrived at what is a relatively optimal genotype, if the ecological context remains stable and a low probability mutation doesn't land a subset on a higher fitness gradient.
I guess I was being a bit silly with 'lazy'. But what I meant, is that evolution leads to the most easily attained maxima and settles in, even if it isn't the most advantageous or stable one. Evolution to a 'good enough' state. Thanks for the vote of confidence!I have a feeling you will be successful, so perhaps soon you will be able to buy all the incubators you want! :)