Yemen Methodology: The Shadow Knows

The lack of detail about the Yemen conflict, as portrayed by open source media, presents an initial choice for the open source analyst.  The IARPA crowdsourcing experiments, out of mathematical necessity, couched questions as “dial-a-pie-chart” probabilities of an outcome, as reported by recognized open-source media. Occasionally, a question was retracted, when it was decided that, based upon the actual outcome, the original terms of the question were undecidable. Ironically, one retracted question was  whether then-president of Yemen, Ali Abdullah Saleh, would leave power by a certain date. Anticipating the difficulty of determining whether Saleh was in power or not, the question was elaborately worded to envision all possible scenarios. Saleh, who was famous for leaving-but-not-leaving, managed to evade all of them.

But unless you are a participant in a crowdsourcing program, you have the luxury of defining your goals in a more expansive, if less scoreable fashion. You might want to know more about Yemen for other reasons, such as the price of oil. Referring to the previous post, you might want to determine when Yemen has been “securely bought”, meaning that Iranian influence, through Houthi proxies, has been securely contained.

A number of analytic techniques have been mentioned in this blog:

and many others.  But in the confusion of Yemen, there is the absence of the focal point that gets one started with tool selection.

Deduction is absent, as is induction, and the rest of the vocabulary of formal logic. Jean Piaget identified an age, and a stage of mental development, called the “age of formal operations”. But formal reasoning is an accessory to a more ancient mechanism. Whether formal reasoning exists discretely, or as some kind of “condensate” hovering around the basic neural processes is not known, but it is of keen interest to IARPA.

That formal logic is not the key to intelligence has  been proven; it was sadly found out in the 80’s, in the last big surge of AI research. In those days, researchers thought of the AI problem as “how-to-stretch-the power of a supercomputer as much as possible.” Since simulating a neural network was inconceivably wasteful, the thrust was  to implement AI through fancy, structured logical systems. Very fancy Block Worlds were constructed, and a few autonomous vacuum cleaners, and that was the end of it.

Although formal reasoning is not the base layer of mentality, you may  gravitate towards it, and so be  frustrated by the vagueness of Yemen. One of the most ancient traditions is that of the wise judge, who carefully ascertains the facts of a case, and renders a decision based upon traditions of jurisprudence. Another tradition is that of the mediator, who attempts to understand the concerns of both parties, so as to serve as an instrument of compromise.

These traditions push the nagging sense  that to be responsible, one has to understand the details of the problem. When one serves as judge, it is appropriate. But generating intelligence always begins with vagueness. Sometimes it evolves to judgment, and sometimes it stays vague. That’s the nature of the game. But while errors of thought, and of mental habit, are of infinite variety, consider: formal education does not offer much for problems swirling in vagueness. The educators have left it to scratch and sniff.

Perhaps it would help with the confusion of how to approach a Yemen-type situation to name it. Perhaps the how-to of approaching a Yemen-type situation has already been covered somewhere in blizzard of academic publication, but we’ll start from scratch. But first, let’s consider the inner, primal You. At birth, your brain was not a tabula rasa, but pre-equipped with helpful gestalt images that jump-start the infant’s understanding of the environment.

A behavior-based theory about the brain has the defects of all theories about black boxes: the theory itself contains input from a gestalt, so it is self-referential. Nevertheless, the Gestalt Principles of Grouping are an important proto-theory. It helps us understand how animals like dog and cows, which cannot do the NY Times crossword, seem to enjoy solving puzzles.

The Gestalt organized abilities of the brain are layered on top of something even more primitive: the pattern matching machinery itself. One of the favorite words on the lips of IARPA scientists is “Lyapunov Function.” If you are mystified at how the brain can create patterns out of nothingness, the Lyapunov is the answer. While actual techniques of learning-network construction have advanced since discovery of the Lyapunov magic bullet, the existence of a Lyapunov is proof that a thinking machine can self-organize.

Since at least some parts of your brain (and notice that, in accordance with modern writing guidelines, it’s now “your brain” as opposed to “the brain”), run according to Lyapunov’s magic, here’s a sketchy explanation. Since things in general,  wristwatches,  storm-water runoff, people in Barca-loungers, and soufflés , tend to lose energy and go downhill, the Lyapunov mimics energy. If a bunch of neurons and a problem-to-be-solved, such as matching a pattern, collectively have a Lyapunov, they want to “cool off”. So if we put these neurons inside a bag with the problem, such as the organization of Yemen, and shake the bag in a special way, the stuff in the bag will “relax”, and the outcome will be a state. The state is a readout of one or more of the neurons, and it could be a pattern-match.

This is the most primitive form of reasoning, and this is how you should approach the Yemen problem. Simply activate your collection of gestalts, and your personal Lyapunov will hand you the answer. Go have a cup of coffee while you wait.

It could be that simple, or impossible. You won’t know until you reach down inside yourself. Who knows what lurks in the hearts of men? To find out, you must study your own shadow, unfettered by the formalisms of academe, the conventions of polite society, or best wishes for mankind.