There's no actual gobal optimum, as environment changes, and adaptability to those changes would tend to be higher when you're not in a global optimum for the particular situation, but diversified in some way so that as the fitness function landscape changes, at least some of the population would be above the threshold for death. It is simply impossible for multicellular organisms to be able to change as fast as bacteria, partly due to far longer life cycle, partly due to lack of horizontal gene exchange, and mostly due to the fact that a larger portion of possible mutations are detrimental or even fatal.
Quote:
Originally Posted by peelax
just because bacteria are optimised for eating poo
|
They're optimized to survive. And from an evolutionary perspective, which is what I'm discussing as was clear with my initial post, that is all that matters.
Ultimately, of course, even they would disappear. Even if space-faring bacteria riding on chunks of rock struck off the planet by an asteroid impact are able to escape Sol's death, accelerating expansion guarantees that life (whether biological, mechanical, pure energy, whatever) will eventually become impossible in the universe. The expansion guarantees that non-gravitationally bound galaxy clusters will in the end spread apart from each other faster than the speed of light (possible since the space itself is expanding and the limit applies only to matter/energy traveling through space) . Thus, a finite amount of matter/energy will remain in any given Hubble volume, and though globally the expansion prevents entropy from equalizing things everywhere, locally in any Hubble volume there will be no energy gradients left with which to do work (such as living, or anything else). Beyond that, on extremely long time scales baryon decay and in even larger ones a quantum smearing of positions (I'm using the term in the same sense as H. Stapp) essentially blurs out any kind of structure.
Back to optimization: what happens when you have an evolutionary algorithm and you keep diversification (mutation etc.) but remove selection? If the population is elevated towards some (local) minimum, then over time its average fitness will decrease as it randomly diffuses (less paths lead towards higher fitness unless you're closer to a minimum). Humanity is in that situation. The majority of selection pressures have been removed. Clearly, natural selection cannot be practically applied to humans, but artificial selection can, and should, be used.