Tuesday, December 28, 2021

social contracts, game theory, sociobiology and extinction threats here and now e.g.Russia

 Whenever we try to think really deeply, out of the box, to make sense of big challenges right in front of us... it can point to a need for new paradigms. THIS WEEK, the crisis in US-Russian relations has woken me up that way, pointing to very basic issues in fundamental science, BUILDING on but EXTENDING the frameworks given in http://www.werbos.com/mind_brain_soul.htm and in my draft paper on approximation of entropy operators. It comes down to the theory of dynamical games....


Sudden shifts in Russian policy this week have strongly reduced the probability of World War III over Ukraine this year. But even as Western policy leaders all breathed VERY loud sighs of relief, going back to other issues on their plate, a strong warning came to me: "DON'T think this is over, or that 
you folks are home free. Nowhere near that." As I see this playing out, it is a warning that the
"AGI/IOT existential threat" is EVEN more urgent, difficult and compelling than I was already saying months ago. That, yes, plus the nuclear and bio threats entangled with US-Russian relations. 

In the past, I have generally just agreed with Yeshua that these existential threats (with the partial exception of climate changes) can be seen as examples of existential level CONFLICTS, demanding peace as a solution, peace as in some kind of new social contract or Pareto optimal bargain. (Xi Jinping even seems to understand these words, more than other major world leaders, whatever the actions of other Chinese may or may not be.) We have always thought back to the Twelve Tablets of Rome, the ten Commandments, Locke and the US Constitution. (I have even visited Washington's meditation study, and been locked by accident  into a large room containing his private papers.) I view this as an example of Schelling's analysis of nonzero sum games, in Strategy of Conflict.

HOWEVER: WHAT DOES THIS SAY ABOUT THE GAME OF HUMAN LIFE INCLUDING THE POSSIBLE OUTCOME OF EXTINCTION?

Naive judgmental people often say: "Of course this species will not go extinct. Species (like this) never go extinct. The biosphere has a very powerful natural stability, resulting from the mechanisms Darwin told us about." 

It is SO SAD how many "mainstream" academic cultures never learn what other cultures down the hall from them, in the same universities, know much better. To discuss the evolution of life, people really should know the foundational (if incomplete) work by E.O. Wilson, in Sociobiology, the classic work by George Gaylord Simpson before that, and the seminal book by Robert May Stability and Complexity in Model Ecosystems. (For a more complete up to date factual account see Ward and Kirschvink.) 

Back in the 1970's, May used solid mathematical modeling to completely disprove the old fantasies about stability in ecosystems. Species go extinct all the time, both at the bottom and the top of the food chain. What we see as stability is usually just HEAVILY FILTERED by millions of years of selection, BUT MASSIVE CHANGES in the web of ecological relations typically DO lead to massive new extinctions. 

The key point is that HUMANS (and all species affected by us) are at the BEGINNING of even more massive changes in the ecology of the earth, most notably through the AGI/IOT complex but also through many others. 

Here is a fundamental key question: in what kinds of  nonlinear dynamical games is the long-term outcome more or less guaranteed? What degrees of freedom do we really have? 

My common sense, informed by deep knowledge of the entropy functions predicted by the most solid modern physics, tells me that I am asking about the existence of widely separated BASINS OF ATTRACTION in the space of the human cultures which could lead to a sustainable social contract or to entropy and extinction. That in turn depends on the variety and scale of noise and connection in that cultural system. That does not sound much like separability. Nor does it LOOK that way, as I think of examples like the entropic processes guiding Putin and ALL other large human societies and organizations on earth today. (I pray that IEEE will be an exception, and there are reasons why it MIGHT be enough for us tALL o survive, but I will learn more this coming week or two.) Furthermore, the examples cited by May all feel like the kind of nonlinear dynamical process we are involved in; I do not see signs of collective intelligence giving collective free will there, or in any of the history reviewed by Ward and Kirschvink. 

IMPLICATION: ONLY A CHANGE OF GAME gives us much hope. And that comes back to issues
INTERNET design **IS** a kind of game design. Approached that way, with the right kind of immune system included, it may be possible to create enough of the right kind of separability of "QAGI workspace of consciousness" to give us a chance to survive. The odds may be against our species surviving in the face of so many things which we have to get right, but in such a large cosmos SOME species probably get through, so it is only natural that we should try our best... as part of our noosphere, we naturally do respond to the most basic feedback we receive from it, feedback which gently guides us all to help as much as we can.. 

1 comment:

  1. Hi Paul

    I’ve just finished the draft of a paper that describes how to use ML to build generative NN models as evolutionary dynamic systems, that use limit cycle and strange attractors to build morphological fields and augment creative processes. Let me know if you are interested in seeing it. Hope you are doing well.
    guy.lukes@gmail.com

    ReplyDelete