Monday, April 1, 2019

Consciousness, intelligent systems and Brexit

Last week (as appended far below) , I had some intense discussions with folks in UK and Ireland about Brexit, which is at a critical point of choice right now. here is what I sent them this morning, after some further meditation: 

Before, I noted that there is a fundamental flaw in the decision systems which led to the present debacle, and failed to give a better, creative understanding of the options and of what the public really wants.
This is really serious and fundamental. In Neural networks 2009  (, I noted that the "mass action" learning rules of the human neocortex appear to be based on TWO measures of the quality of output of the cortex: (1) accuracy of the PREDICTIONS of the thalamo-cortico-thalamic System (TCT); and (2) value of the OPTIONS for decision and action which the cortex sends to the basal ganglia, a part of the brain whose higher-level importance has come to light in recent decades. The information system which feeds into Parliament clearly has not been optimizing (2). That is due in part to how the option creators did not have input form a cause-and-effect (predictive) understanding of what the public wants; The INTEGRATION of (1) and (2) is what makes mammals different from their ancestors; the folding together of two hree-level coftices into one six-layer cortex is the essence of the evolution of the mammal brain. Parliament has not yet reached that level.

Before I speculated that worries about immigration are one of the REAL concerns of the public; brexit was supported only as a means to that end, and a better way to serve that end could have been found (a better set of options, certainly to include more dialogue with Italy among others to better address this common problem). 
All of that still stands,  but I feel embarrassed that I failed to mentioned another hugely important  fundamental aspect here. Another reason for the public to vote for brexit is a growing MISTRUST of trade agreements all over the world. Whatever else you can say about Elizabeth Warren, she certainly has articulated that powerful current of public feeling; see the link posted at . The public is not REALLY arguing against economists who say that trade is essential to the level of prosperity in our world. Rather, they feel that the actual AGREEMENTS which implement grade are rigged in a way which hurts people. that's how Trump got elected -- though I wish people had listed more to my Senator, Time Kaine of Virginia, who got a level closer to the truth in HIS speeches for the Vice President debates. There really is a deep problem of trust which is frankly well-justified. The challenge is to modify all trade agreements so that they not ONLY achieve the kind of Pareto optimality which economists study but ALSO meet some kind of standard of transparent fairness -- which also requires some kind of agreement or conscious mutual acquiescence to some concept of fairness, INCLUDING the kind of assurance Kaine talked about, that the theoretical agreements will not be subverted by faulty, biased or corrupt implementation machineries. Absent trust, exit. 

Many in Virginia feel it is a shame that Kaine himself does not seem to be a plausible candidate for 2020! 

But looking ahead: What if, in the future, trade agreements are enforced by IT systems, and if the dialogues which give input to Parliament are also implemented on IT? What if we might inject some intelligence, and even some level of consciousness and soul into such IT systems or to the human/computer aggregate system in which the IT provides the backbone? How do we ensure the fairness of such systems? How we devise transparent fundamental rules to be implemented in the IT backbones?

In discussing this informally with my wife this morning, I explained the core issues as follows. To survive, an organism needs BOTH a brain AND an immune system (and other stuff of course). The same goes for our entire species, and even our entire noosphere. The immune system include some relatively less intelligent rules across the board. In a way, the Ten Commandments could be seen as a Gen 1 attempt to create rules/social-contract to m ake more intelligence and growth possible, on a firm foundation. I view the Constitutoin of the US as kind of Gen 2 of the Ten Commandments (yes, more sacred to me than the Gen 1 version), a little longer, very carefully designed with a certain amount of authentic spiritual input (admittedly, some in an Irish bar in Philadelphia for which my family kept those bar bills, but some in Scottish rite lodges and so on). But now, in the age of IT, we need to move on to some kind of Gen 3, specifically in IT. It is not a simple task, and it worries me that today's IT and trading systems may be eroded away by termites, like an old wooden building,  before we have a more robust steel structure to replace it, a structure designed so that life and soul may thrive inside. 

This is no small matter. But even Brexit is not a small matter. If only the EU had had the kind of dialogue it needs to address these vital concerns more effectively, making brexit itself unnecessary and moot! Too many termites at work already, at risk of turning all of Europe into what termites leave behind. 

(By the way, I do have a couple of nice photos of termite nests in the seven albums of images and explanation which I put onto Facebook this past week. That post on the nature of afterlife was just one of seven, based on the Caribbean. Perhaps someday I will also post photos from Brazil, including the Amazon,  but the issues there were too complex for this kind of list or Facebook this month.)


i do hope that this would warm the heart of my old PhD thesis advisor, Karl Deutsch of Harvard, author of books which helped CREATE the EU, but also author of the Nerves of Government, which argued that neural network ways of thinking should be applied to understand how states and other organizations could actually work. 

The facebook posts I mentioned, which get very deep into actionable esoterica, are at:

Devil's island:
OOps re swimming with sharks (related to a previous HAL cruise):
And reflections on our fate which grew in my mind during the cruise:


Just FYI: here was my earlier post to my friends on this subject:

A few months back, many professional psychiatrists were excited to see how many people, watching what unfolds on CNN these days, validated the importance of their profession in understanding what is really going on in the world. Are policies being warped in damaging ways by narcissistic and psychopathic personalities, fitting blatant textbook models of obsession with their organs or overuse of psychological mechanisms like denial (as described in the beautiful important work by Valliant in his longitudinal study of Harvard graduates). Even as I type this, they tell me that Rand Paul wants to start an investigation of Obama... and I react by saying there is a lot to investigate, not by lawyers, but by... other approaches. 

But now: Brexit offers us a truly rich case history today, this past week, combining four elements great for researcher: (1) the database of the past week is public enough and structured enough that comparative analysis is possible; (2)  the bottom line that things are going grossly dysfunctional seems relatively clear as such things go; (3) the dysfunction does involve a variety of important and general issues in systems theory, management, free will, rationality and so on; and (4) if folks don't find a way to get out of the box, fast, the damage could be very great indeed. Unlike some of the more intractable case examples I see in the US and China, the dysfunction seems mainly a matter of well-educated, well-meaning and relatively sane people somehow not being able to make things work. But psychiatry does also have something to contribute, to any realistic integrated holistic understanding.

The indicative votes held yesterday  really imptessed me emotionally, a bit like a kind of Greek tragedy without the clarity of a Sophocles play (but certainly with a few fatal flaws in play). The leadership of Commons really tried hard to work at a systems level, not dictating the outcome (as tyrants try to do), hoping that a workable way forward would emerge. But people could not agree. In discussions of a second referendum, social psychologists were equally baffled, seeing no way to phrase the questions in a way which would lead to clarity; that is a major part of why the idea of a second referendum did not win an indicative vote, even though it came closest of anything.

My naive knee-jerk reaction was "Haven't any of these folks read Ken Arrow, Thomas Jefferson, Locke or Schelling, let alone Raiffa or Von Neumann? Isn't this a case where stronger systems theory and understanding of intelligent systems might have shown them a way to escape from the swamp?"

More precisely.. Thomas Jefferson wrote many things. One was something called "the Jefferson Manual," a book of parliamentary procedures intended to allow an institution like Congress or Parliament to actually work. It was a foundation stone of a major serious intellectual effort to address the kind of value or utility aggregation challenge discussed at lenth by Ken Arrow (Arrow voting paradoxes) in the real world. When people do not agree, how can we build a system which nonetheless avoids the worst, and which improves the chances of attaining some kind of Pareto optimal outcome (an issue which Von Neumann and Raiffa explained far more clearly than older writers who neglected the basic fact of human uncertainty and limited knowledge, a fact which is central to neural network mathematics and statistics). 

Knee-jerk reaction...

I remembered three-way elections in scientific societies, where no one candidate or program could get a majority. There is a reasonable, well-known practical way forward, "n choice" voting. The cleanest version is to ask each identify his or her first choice, second choice, third choice, and so on if there are more choices.  Then the lowest choices are removed and votes reallocated from first choice and so on. It may not be metaphysically or ontologically perfect, but its simplicity is already hugely important, as Schelling's book Strategy of Conflict, should allow people to understand. So I wondered: don't the parliamentarians in the House of Commons know about that? Don't they know how to structure votes designed to get SOMETHING passed? US Congressional leaders in years past have not done exactly that, but they have known at times how to structure procedural motions which then had a similar effect. (But again there are other problems at work leading to dysfunction in the area where I live.) 

But beyond the knee-jerk reaction:

Systems level thinking, looking for workable procedures, trying to fashion institutions and rules of the road which themselves are like intelligent systems, is not a complete silver bullet... or at least, not when it focuses on higher level decision systems without getting "down and dirty" to analyze the systems which give input to the decision systems, and which engage with concrete reality testing. Brexit is a great illustration of that complexity as well. In fact, one could get very down and very dirty in any honest realistic discussion of brexit politics, but for now, just a couple of aspects... after a general comment at the systems level.

People are often confused by the issue of "following the right procedure" versus "doing what is right concretely." This challenge is very similar to the challenge of putting together a higher-level econometric model which does not specific the narrow technical facts on the ground, but is able to capture a wide variety of possibilities. I described in detail how to do that, based on successful first hand experience, in  Werbos, Paul J. "Econometric techniques: Theory versus practice." Energy 15.3-4 (1990): 213-236./EIAeconometrics.pdf. In a way, the job of the parliamentarian is to concoct a procedural bill very much in the way I constructed econometric models, not getting so down and dirty, but AFTER a very intensive effort to scope out the range of possibilities for concrete facts at the down and dirty level. ("The weeds.") This requires parallel efforts to scope out what those folks call "the weeds." This reminds me of how the power of the mammal brain over its immediate ancestors is the circuitry to map out the SPACE OF POSSIBILITIES.

The issue of how to design a second referendum is a very serious intellectual issue, a key part of mapping out the space of possibilities more completely and intelligently.  It was just as much a source of dysfunction here as parliamentary procedures themselves. The approach used by normal professional pollsters suggested a lot of sheer incoherence,  craziness and randomness out there in the public. But perhaps a professional psychiatrist or a decision analyst like Riaffa could have helped in providing a richer mechanism for getting input from the public, through questions intended to explore a variety of choices and uncover what really matters to people in the end.

Here I have my own speculation, my own thought about ONE of the possibilities which should be considered in the map. Could it be that the mainstream people in the middle did not want out of Europe so much as they wanted out of Merkel's glowing but risky policies on immigration? Did they really want Mexit, exit from merkel's immigration policies? Would they have been happy to stay in the EU if changes in policies for immigration from outside the EU (and policies for determining citizenship) were tightened up, in a wide variety of possible ways? Could a deep understanding of the concerns of the British public show some points of commonality with voters in OTHER EU nations, which might allow a balance of forces far better for UK and EU than the crazy trainwreck now in progress? 
But a proper second referendum would include this kind of issue as just ONE of the issues to be probed. And perhaps it could be done as a step towards creating a more real time online dialogue, which Macron has advocated for France but does not know how to create from a technical point of view. In the end, god help us, sustainability might depend on making room for even deeper "weeds," such an unbreakable operating systems to support fair systems of constructive online dialogue, a serious issue in intelligent systems design, way beyond what facebook now even imagines doing. 

Best of luck. We need it, here and now and in the long-term both.

No comments:

Post a Comment