I prefer function to equation, because function shows the dynamics. How to prevent reductionism that models and metrics may suggest? I'm wondering how Monica Anderson's work on Model Free Methods for artificial intuition could apply here.

Here are videos of Monica's work: http://videos.syntience.com/index.html

I have run across Monica many moons ago, will check it out - to your point on 'function' vs 'equation' you are absolutely right that what I am really talking about is a function (value function, value equation) where I map a number of dimensions on to a scalar representing an abstract 'stock' - I have found using the word 'equation' is more evocative in most folks minds, and the mathematicians and mathematically inclined immediately correct the language - I will use them interchangeably with you - Also regarding reductionism, or I think more pertinent the achievement of a stable state in the operation of the model, there are three possibilities when applying a generative set of rules that may be configured in use, one is total chaos (divergence, likely with attractors coming and going) another is oscillation between strong attractors, another is convergence (this is the ossification I think you fear) - I am more worried about total chaos with limited useful patterns generated than with convergence on a single pattern, though my wife gave me a jolt by asking whether my platform for an ecology of business forms might coalesce to the existing system, then what.. LOL

Your wife is right... I would say the 'coalescence' to the existing form is the momentum of exiting forces keeping the system in its status quo 'direction'. That's what we want to 'break' with new code...

what is sought sounds alot like 'robustness' in complexity/chaos theory. the story of robot cockroaches is illustrative - consider a system of a dark room with a spot light and a boot stomping in the spot light, half the cockroaches run from dark to light, half run from light to dark, if you run this system for long enough it will become optimized (no cockroaches that run from dark to light), however if you start the boot stomping now instead in the dark, an extinction happens (not robust) - in order to introduce robustness let's say the cockroaches randomly change their mind - the solution is now sub-optimal but more robust as even if you change the conditions, there are always some (formerly stupid) cockroaches that will run to the light - long winded way to say we need to build robustness into the system, and that the idea of 'mutations' needs to be considered - currently I am solving this with sufficiently large numbers of permutations, but that may not be sufficient as the system may optimize to one or a small number of such permutations

Yes, more macro. I refer here to the powers that derive from the operations of the system, from the position in the system, from the 'mechanic' accumulation and distorsions they may create. Power laws and network effects are not the result of governance, for example.

I think we are talking governance perhaps a bit differently - if we look at the macro 'money is power' of todays system the accumulation of money is very relevant to power without reference to any macro forces - My take on this is that money is valueless or abstract, and so should not be a basis for generalized power, but that we need to name what things we feel are potentially valid contributors to power (dimensions of governance) and directly measure these, while agreeing in a given context which of these factors should be weighted more heavily to get the results we seek - In the operation of such a transparent power function, one could differentiate power by its source and context, and make it less portable, which in my mind will create a more efficient application of resources, or at least a better understood one - So my short point is that power is an accumulation of governance rights, and has the characterstic of portability - 'power laws' or scale free network effects, in my mind are to be moderated actively by a system that aims for stability and sustainability - this can be approached through the use of decay functions on factors of governance rights and more generally by radical inclusiveness