Alfred Hubler on Stabilizing CAS

With his permission, I am posting an email thread between myself and Alfred Hubler.  I had contacted him on the recommendation of John Miller when Kevin and I were posting on the possibility of dampening boom-bust cycles in the financial markets through policy or other mechanisms.  Here’s what Hubler had to say:

I would compare large scale boom-bust cycles to catastrophic forest fires.

Two thoughts:

  • Policy impacts only small forest fires: When small forest fires are suppressed, large forest become possible and, more importantly, the untorched forest changes the local climate and therefore the forest may grow faster and faster. At some point forest fires are potentially so devastating that policy makers have no choice but to suppress them.  Eventually the amount combustible reaches a threshold where the forest fire can not be prevented, the catastrophic forest fire takes place, and the cycle starts over.
  • Self-adjusting systems suppress catastrophic boom-bust cycles – therefore catastrophic wildfires are rare in un-managed forests: self-adjusting Systems avoid chaos [1,2]. However such adaptation to the edge of chaos occurs only if the system parameters change slowly, compared to the dynamical variables, i.e. if we change policy faster than the period of the boom-bust cycle, then self-adjustment will not suppress them. The good news is, that almost any type of self-adjustment suppresses chaos[3].

[1] P. Melby, J. Kaidel, N. Weber, A. Hübler, Adaptation to the Edge of Chaos in the Self-Adjusting Logistic Map, Phys.Rev.Lett 84, 5991-5993 (2000):

[2] P. Melby, N. Weber, A. Hübler, Robustness of Adaptation in Controlled Self-Adjusting Chaotic Systems, Phys. Fluctuation and Noise Lett. 2, L285-L292 (2002):

[3] T. Wotherspoon and A. Hubler,”Adaption to the Edge of Chaos with Random-Wavelet Feedback” is accepted by the Journal of Physical Chemistry:

I then asked him if he had any specific thoughts on either Sumner’s proposal or synchronized chaos.  Here was Hubler’s response:

In desert oasis, larger, more specialized ants often dominate. Smaller non-specialised ants have smaller populations and competition pushes them to boundary of the oasis. However the small, non-specialised ants species are around for very long periods of time whereas the larger, specialised ant species are often wiped out by a change in climate and catastrophic events. The smaller ants may initially suffer from a catastrophe too, but have a higher likelihood of survival, and even thrive temporarily after the catastrophe, since the competition from the larger ants is gone. The group around Jim Brown, at UNM, did work on ants and we just started work on bacteria colonies in changing environments.

By analogy, I conclude that boom-and-bust cycles are more harmful for larger and more specialised organizations which flourish in a certain environment, and disappear if this environment changes suddenly. A forest fire helps smaller plants and increases diversity.

Yes, chaotic synchronization sounds attractive and I have published many papers on this topic over the past 20 years. It work wells with simple nonlinear oscillators but it does not seem to work with living organisms, such as yeast cells. We do not understand that. Maybe there are adaptive processes or evolutionary advantages suppressing synchronization. For years we tried to entrain the life cycle of cancer cells and yeast cells with heat pulses, microwave pulses and other stimulus, but the cells always escape synchronization.

By analogy, I conclude that it might be difficult to synchronize economic systems.

I’d like to hear your opinions on what this means for policy, regulation, incentive plans, etc.

  • Paul Phillips

    Viewed from a thousand miles, the financial system has a incalculably large incentive to fail catastrophically as frequently as it can do so without killing the goose that lays the golden eggs.

    As long as there is such a thing as “too big to fail” and trillions of dollars are available for siphoning, according to what logic can this cycle be dampened? Nobody has to explicitly pursue this outcome (although there are many who will) for it to be inevitable; the system obeys its own logic above all else.

  • Alex Golubev

    I think it comes down to the tradeoff between growth and survival, but it is not in a vacuum. We CAN have a more stable system, but what’s the optimum point on the utility curve? Absolute level of growth doesn’t matter. it only matters what your relative size and growth is compared to others (individuals, companies, countries). So now we have to introduce yet another variable of competition. Ultimately, the ant analogy breaks down, cause the specialized ants cant resort to a highly superior military to get them out of trouble. There are too many assymetric payoffs out there. A system of a huge number of selfish agents can lead to growth of the least long-term adaptive agents and subject the whole system to extinction. I think reality is incredibly unfair. So my opinion is that the system will not improve. But the 2nd questions is can we protect ourselves as single agents. I think that’s possible to a large extent. Be very careful with leverage particularly if you’re using it along with the majority. conditional probability of failure will increase. Ultimately the gov’t has shown that when the forest fire does get out of control, noone will be spared (taxes, devaluation, confiscation, etc…)

  • @Alex, you make some good points. Since you argue convincingly for mechanisms of instability, doesn’t that undermine your claim somewhat about having a more stable system? You don’t really give evidence as to how that might be accomplished.

  • Pingback: Stability Through Instability « The Emergent Fool()

  • Pingback: Best Reader Comment Award « The Emergent Fool()

  • Pingback: Cancer as a Complex Adaptive System « The Emergent Fool()

  • If you widen the scope of your encapsulation of the question to include metabolism, and then, you extend the the encapsulation you may find a path.
    The RMCM asserts that metabolisms can be simple, complex, modern, or primitive. A primitive metabolism would be similar to what we call an echo system; its mechanisms are widely distributed and loosely bound. Catabolism and anabolism reside independently of the Rational Paradigm (life form). The Primitive Rational Paradigm is little more than serendipitously synchronized behaviors. The all-in-one metabolisms of modern Rational Paradigms are indeed there except they are metaphors abstractions or trajectories. These are contexts that can be flipped as the surroundings change. A Catabolic mechanism for one becomes an anabolic mechanism or a few, and an anabolic mechanism for one becomes a catabolic mechanism for a few.

    • Rafe Furst

      You realize of course that since this is cryptic even to me that it’s just gobbledegook to everyone else :-) Why don’t you unpack the jargon and relate it to the post and we can see if your model and claims have merit?

  • The RMCM asserts that Complex Mechanisms are metabolisms.
    The RMCM defines complexity as Polymorphic Multibehavioralisms.