(Warning: the following text contains loose metaphorical hand-waving arguments.)

Genetic algorithms can be used to solve search, optimization, and control problems. Perhaps weblogs behave like "the island model" of GAs, converging to almost distinct populations containing similar individuals. This could be like solving a search problem for a common question, or a control problem of finding a reponse to a change in the outside world.

The concept of "the island model" means that you evolve several GA populations almost independently, and only occasionally exchange individuals (or genetic material) between the populations. The island model is used to parallelize GA computations, and sometimes also to speed up convergence.

The convergence of weblogs towards similarity can be a good thing. However, I fear this is a bad thing for the sustainability of weblogs. When the outside world once again changes there is not enough diversity in an island of weblogs to generate new solutions which meet the requirements of a changed world.

I hope the weblog toolmakers invent mechanisms for transferring "individuals" or "genetic material" between populations. This may not be easy, because the island populations may have became so different that any outside influence is quickly rejected or extinguished. Or the outside influence wipes out any local originality. Thus, there should be some kind of feedback mechanisms which would encourage and reward those who move between islands, or from one island to another, at the same time protecting the island ecosystems.

Am I on the right track? Or perhaps this is just hand-waving to explain my laziness in selecting only weblogs which contain easily digestible material for me?

M. Mitchell Waldop urges us to manage projects from the bottom up. In [...] The Management Secrets of the Brain he draws parallels to recent understanding of how our brains work to managing organizations. [...] Waldop makes five claims:

1. Never try to micromanage a large, complex organization.
   There's not enough executive attention in the world to ironmonger this level of activity.
2. Don't let bottom-up self-organization go wild.
   Without leadership standard operating procedures are directionless and blind.
3. The best way to control your subordinates is to just point them in the right direction.
   This new model...assumes that [leaders have] just one job, which is to generate a neural map of the [organization's] goals, strategies, and current situation.
4. Be careful listening to the voice of experience -- that voice could be your own.
   Sometimes an organization has to break out of its rut and try a new approach.
5. The organization can't succeed without passion.
   Unless we know what's important, what matters, then all the rationality in the world gets us nowhere.

Waldop makes a great case for managing projects on an agile or lean basis. The brain is ideally suited for project complexity, uncertainty, inevitable learning, and the underlying humanness of the endeavor. Why would we even try a different approach.