Friday, September 5, 2014

Unfortunately, The "Formulization" Of Other Disciplines Is Also Proceeding. Institutional Momentum vs Aggregate Adaptive Momentum.

   (Commentary posted by Roger Erickson)

(evolving hitching patterns in different size mule teams)

And you thought that NeoLiberalism in economics was bad!

Rebels in many professions are resisting our current period of systemic decline. Take DoD Personnel Doctrine. Please! Or law enforcement mis-training. Or mis-education in general.

And now we can add sociology. For example, the following paper strikes me as an infatuation with formulae, in this case computer modeling.  The article may make some people think (about SOMETHING), but it is functionally useless as a guide to practitioners of Democracy in the real world.
Conditions for the Emergence of Shared Norms in Populations with Incompatible Preferences
"By means of computer simulations, we study conditions ... "
Is it different this time? Have new tools made distributed logic* obsolete? No.

If there's a computer, there's programmed software. And if there's a software program, 99.999% of the time, there is still a presumed formula for handling a presumed static context.

The insidious, main outcome? Creeping acceptance of formulaic methods reinforces a VERY BAD HABIT. Namely, the naive, underlying belief that systems adapt by predictively reorganizing previously describable degrees of freedom, rather than by completely unpredictable & constantly changing sorting following repetitive trial and error. That sort of institutional momentum delays our endless pursuit of Outcomes Driven adjustments and prolongs wasteful wanderings in Ideologically Driven dead ends.

The simpler, more useful message was already and always known. There are no Cookbook Recipes for adapting and evolving - i.e., no formulae, and no reliable program to use. Our degrees of freedom grow faster than our knowledge, habits, methods and practice base.

Our only recourse is ceaseless expansion of distributed trial-&-error, and frantic selection from our own aggregate feedback. No new tool changes the fundamentals of evolution.

1) We always face new aggregate challenges and capabilities, both demanding and bestowing orders of magnitude more degrees of freedom than we currently perceive. Constant exploration is required.

2) We have zero predictive power, yet seemingly unlimited adaptive power.
We can leverage that adaptive power IF we industriously utilize our full distribution of feedback and analysis. Tempo matters. [None of us is as smart as all of us, or as quick thinking when it comes to aggregate context.]

3) How do we survive every new niche we drag ourselves into? We start trying many things, and then start finding out - ASAP! - what starts to work. In the process, we reshape and tune ongoing, distributed momentum, based on distributed feedback.

4) So please cease, forever, the habit of trying to predict unpredictable adaptive formulas beforehand, and in the process constraining the very distributed activity & feedback which we need to drive massively parallel SELECTION.

5) Preserve fundamentals AND evolve more variables. Constantly retune a GRADIENT involving re-standardizing carefully selected infrastructure, while simultaneously promoting active diversification near the outer edge of your evolving system. [To use a building analogy, consolidate nearer to the foundations, and ceaselessly innovate nearer to the top floors. Our spectrum of innovation must include highly conserved elements as well as increased variance in some elements, just for OUR system to evolve. While not letting innovation fall completely to zero anywhere. To survive, EVERYTHING in our system must change. Just not at the same rates everywhere.]

What we have here - 2014 in the USA - is a lull in an unceasing civil war to evolve.

One combatant is our own urge to rapidly over-adapt to transient context, leading to efficiency traps.

The other combatant is our own requirement and urge to re-orient to changing context, sometimes leading to complacency about transient contexts.

Survival, obviously, requires surviving today as well as endless different tomorrows - but never any of those in isolation.

[Then there is also the constant burden of the untrained zombi or deadweight element, which actively resists any and all change, and actually believes that sitting in the middle of the road isn't suicide. They're not on anyone's side. They're just in the way. If we stop our civil war and pay a bit more attention to education and training requirements, the zombies should soon become statistically insignificant.]

Clearly, we need transient efficiency on demand, plus resiliency on demand, not either in isolation.

Right now, we've hitched up half or more of our own team backwards! Some are demanding too much efficiency, and some are demanding too much resiliency. How do we actually get our swelling ranks of combatants to lubricate all the random frictions, and make recombinant love, not factional war?

Since a formula for solving one, static context can be optimized ... does that mean that EXTENSION OF THE SAME FORMULA can provide optimal solutions bridging multiple, different contexts?

No. Most grandparents learn that the hard way, yet most fail to adequately teach it to their kids. Hence - SO FAR! - most grandchildren must relearn many things that their grandparents learned, but didn't pass on as a permanently incorporated part of our expanding cultural toolkit.

Ok. Does that mean that ANY particular formula will work for the unending stream of different contexts we face? NO!!! We've also learned that too, by trial and error. 

You might say that the only formula that works for any system, is to permanently discount excessive belief in any PERVASIVE formula whatsoever. The only formula that's workable long term is to decline all formula except that of a shifting gradient of distributed innovation.

Slowly change foundations, while rapidly sifting through new experimental variables.

That's how we preserve our aggregate adaptive momentum, while varying it too.

Does this work? Yes. And we actually have a tremendous opportunity to increase our Aggregate Adaptive Rate! 

How? By remarkably subtle and simple tuning of our citizen-development methods. It's a given that all citizens will face new horizons that none can predict or specifically prepare for. The most valuable talent to practice is comfort and familiarity with embracing and extending NEW perspectives on our own aggregate and it's situation.

Practice at audacious innovation - not formulas - is what used to drive American Ingenuity.

There's a difference between Aggregate Adaptive Momentum, and Institutional Momentum, and the former always wins in the end.

There's no reason why most citizens can't learn all this by age 10. Unfortunately, we're not even trying to prepare them!

Summary. New tools and methods are:

a) always prompting us to slowly, recursively improve our highly preserved foundation or "logistic" formulae,

b) while also distracting us from practicing further aggregate innovation.

There's a required duality between foundation and exploration. We survive by coordinating, not consolidating those complementary activities.

Aggregate success means grandparents embracing the inventions of grandchildren PLUS challenging them to explore agile vs formulaic applications of new capabilities.

Youth or beginners start with Institutional Momentum in perceived static contexts, and must grow into perceiving Aggregate Adaptive Momentum in unpredictably evolving contexts.

Without adequate aggregate challenges to generate experience, new generations inevitably get stuck in Static Context outlooks, which reduce their Adaptive Rate.


Our current challenge is to invent AND USE subtly altered education and training methods which permanently capture and preserve a higher adaptive rate in yet a larger population.

Froebel seems to have come close to realizing that, 200+ years ago, and most disciples of Kindergarten promptly forgot it! Go figure.



* Casey Haskins defines logic as "using evidence to draw conclusions".
Evolving, distributed logic is therefore: "using accumulating, distributed evidence to draw NEW conclusions about always changing aggregate situations." Clearly there are multiple, simultaneous levels of logic, and the very concept of logic, like data, is meaningless without context.




No comments: