In a process akin to the summoning of Beetlejuice, if irreducible complexity and constructive neutral evolution get mentioned, I am bound to show up and talk about X-Men.
I explored that simulation further over a series of blog posts to demonstrate that irreducibly complex systems can evolve and they can even do so when complexity provides no selective advantage. Not only is it possible, it is the most common outcome in these experiments.
I’ve been wondering if this could be demonstrated theoretically with a formalism similar to representing entropy in terms of the size of the state space where different microstates have the same macrostate, such that systems tend towards high entropy because those are the largest portions of state space. Likewise, there are simply more solutions that involve two parts than one, and more that involve three parts than two, etc*. I’m not sure if such a result would be interesting or is already known or would be considered trivial.
Caveat: This pattern cannot be extrapolated indefinitely. In this particular case, adding additional parts is not always neutral or beneficial, so there is likely an inflection point past which the number of solutions decreases with additional parts.