Quote by John86
Does evolution by your means always incorporates growing complexity in the sense that at some moment there must be some starting point, where the evolution or cutoff moment starts.

If we picture observers or arbitrary complexity, then the evolution I envision has to include mechanism for increasing, as well as decreasing complexity. I picture this process closely related to generation of mass.
In the gaming analogies, this can be assocaited to how one player, by beeing more clever and sometimes also having luck, can conquer control over it's environment and "grow" in it's quest for selfpreservation.
It's this mechanis, where two sytems try to conquer each othre that is responsible for the attractive force of gravit in my view. Their constant communication makes the information distance between them shrink, and thus the distance in between them shrinks.
But evolution process isn't just the complexity. When the complexity has reached a steady state for a given observer, there is still an ongoing evolution which can be interpreted as a redisposition and remapping of available resources.
So, no it does not always increase complexity, as in an monotonous increasing function. That would make no sense. The point is that complexity growth or decreasing are very SLOW processes as compared to the redisposition processes. this is why on a short time scale, these "gravitational like" effects can be accounted for my constants.
Edit: In my picture the PART of evolution having to do with scaling complexity, is the thing that I associated closests to "gravity". The part of evolution that is superimposed ontop on, or taking place within the constant complexity domain, are what I associate to the other forces (nongravitational ones). Still I'm quite convinced that even the understanding of the forces, in particular their unification, requires an analysis of also the complexity scaling processes. Most certainly, someting closely related to, or even exactly equivalent to Quantum mechanics as we know it, is the result when complexity is frozen. But this is clearly a special case, and interesting insight await if we understand the general case.
/Fredrik