Chaos

Home / Chaos

What should be done for those in my circle of care?

I have started and restarted this as least three times, and I am not sure what angle I can tackle it from.  I think that there are three primary problems that I am running into (though I reserve the right to find more before I am done without enumerating them).  First, I haven’t yet defined what ‘good’ is over a universal time horizon. Second, I haven’t yet defined how broad my circle of care should be.  Third, it is difficult to know what the real consequences of my actions will be, even setting aside the issues of induction. I am going to try to wade in and see if I can come out with anything meaningful.

I will try to set aside the first two problems  (good and circle of care)  with a little bit of logical maneuvering.  I read something very interesting this week in Victor Frankl’s Man’s Search for Meaning that I think is instructive here.  He brings up the story of a chess grand master.  What would happen if you asked a chess grand master what the best move in chess is?  He is very likely to say that it very much depends on the arrangement of the pieces on the board.  While the goal is to put your opponent in checkmate, if you were, for instance, in check at the time, then the only move that you can make is to move your king out of the line of fire.  This is the best move at the time, but it is certainly not how to win the game.  Frankl goes on to posit that each of us can have our own meaning, and that, like small actors on chess board, we can all find significance in that.  My purpose here is not to debate Frankl at this time, but to focus on his question and accept the premise.  Here is the logical maneuver that I would like to make to allow me to put these first two problems to one side for the moment.  Let me presume, that I have such a tightly defined circle of care as this king on a chess board, and I only have one move to take him out of check.  Let me presume that this is my starting point, and then play with the consequences of this, not on a black and white grid, but in the real world.

It would be reason able to ask why a post about a what should be done for my circle of care is entitled chaos.  I entitle it this because the farther I move out in time, the more unpredictable the consequences of my actions become in terms of cause and effect.  We don’t live on a chess board.  So even if I control for knowing what to do, and who to do it to, I cannot guarantee the outcome.  This is conventionally known as Chaos theory, or more popularly, the butterfly effect.  The traditional domain of this is weather modeling.  It is very difficult to get accurate forecasts beyond a few days because very small changes in the initial conditions can lead to cascades of changes.  Put poetically, the ‘flap of a butterfly’s wings in Africa may make a hurricane on the other side of the Earth.’ I would like to return momentarily to these non-linear effects in the same medium, but first I would like to make a less conventionally-made point about chaos theory.

Non-Domain Specific Chaos
While it may be true that the flap of a butterfly’s wings can create a storm elsewhere, the changes we make are not restricted to the domain that we make them in.  To carry on the conventional analogy, the hurricane started by the butterfly may have been Hurricane Andrew, one of the most costly hurricanes in American history.  That storm could have bankrupted insurance companies, and that could have sent a cascade through the American financial system.  Perhaps it is better to be more concrete than to keep speaking in the hypothetical.  I have just been reading a book called How we Got to Now. It is a very entertaining read but he draws a very understandable link between the following:

  • Problem: Icebergs and Submarines in WWI stopping Atlantic shipping
  • Solution: Echo-location/Sonar from a well-meaning Canadian engineer
  • Consequence: People’s lives saved as the engineer intended
    • Unintended Problem: Gestational malformation
    • Solution: Sonogram born from Echo-location/Sonar
    • Consequence: Babies saved and sex determined prior to birth
      • Unintended Problem: Chinese one child policy and cultural preference for boys
      • Solution: High resolution sonograms used to determine the sex of a baby
      • Consequence: Sex-selective abortion in China and Male:Female ratios of 110:100

We don’t know where this chain of causality may end, but the point is that the Canadian who originally worked to use the technology to protect lives and cargo inadvertently helped end the lives of babies in China decades later. This is certainly not something that he could have foreseen, but it is likely not something he would have been pleased with.

To be sure, not everything ends negatively, this Norwegian Commando credits a chance purchase of a bolt-cutters in the UK with stopping Hilter from getting the A-bomb.

Let’s circle back to the premise now.  In the non-chess board world that we live in, even the most conventionally noble missions can become horribly distorted and lead to negative consequences in other domains.

Domain-Specific Chaos
Let’s return to the notion of chaos in the conventional sense.  Let’s imagine an Nitrogen molecule traveling and several thousand km/hr, as they typically do, and it careens into a hapless water molecule in the middle of the Pacific.  Lets presume that the Nitrogen molecule is primarily travelling laterally, with limited velocity in the vertical/altitude direction.  The initial states of the temperature of the water, the evaporation of the water, the sinking of more saline water, and the turbulence from a diving whale are all so perfectly balanced at this point that there is a small upwelling of water with no lateral direction at all, that is, an upward wave with horizontal motion.  Let’s set aside some of the quantum mechanical weirdness, and presume that the nitrogen imparts most of its lateral energy to this water molecule. We will set off something like the snowball effect, except on a small scale.  The lateral energy imparted by that nitrogen molecule onto one water molecule will cause it to move sideways, The adhesion to other water molecules will mean that many will follow it, as they collectively move they will find that there are less water molecules to support them in the direction that they are traveling, and they will be pulled, by gravity in the direction that this one molecule pushed them.  The chain reaction does not stop there, as they bounce back to the top the next time, they will have a wider surface, and may be struck by even more nitrogen, oxygen, and carbon dioxide molecules heading in the same direction as the first.  If so they will gradually pick up speed and more energy, and crash at the feet of someone on a beach somewhere.

These type of non-linear events are not everywhere.  We believe with a great degree of empirical precedent that, for instance, friction with the air slows down pendulums until their motion is imperceptible.  Though there are at least 6.02X10^23 air molecules that it will come in contact with, some impeding its progress, and some bolstering it.  In the end they net out.  In a relatively still room, pendulums can keep time reasonably well, and are not subject to the feedback loops I mentioned above for water, though they have a parallel wave motion.

Even with water, at levels of remove more microscopic, and more macroscopic than I just mentioned it has very different properties.  At the more microscopic level, water molecules adhere to one another so tightly, and create such wonderfully powerful surface tension, that insects like water-striders sit comfortably on top, even in quite choppy water.  At a more macroscopic level, such as traveling in an airplane or spaceship, individual waves become invisible, and the sun is reflected almost as a perfect mirror.  It is only at our conventional level of resolution that things like body-surfing are possible.  Many people deny the presence of things like rogue-waves where a large number of peaks or troughs will align simultaneously and damage a large ship.

What areas of are the domains of non-linear events and what areas are the domains of linear events?  I would like to draw a parallel here to water.  It is not that a single substance is the domain of a linear or non-linear system, but rather that it is a function of the level of ‘zoom’ you have on it both in terms of time and in terms of its own properties.

First, let me state what I know to be empirically true of non-linear systems.  It is also empirically true that we find non-linear events in areas where the energy in a system is close to balanced at whatever level of remove you are talking about.  To take and example of a ball perfectly balanced on the top of a pyramid. It will happily stay there until nudged by a bit of air in any direction, and then rapidly pick up horizontal speed in the direction of the face it moves along.  Predicting which way it will go is difficult because it is dependent on nearly imperceptible gusts of wind, but once it is moving, it is fairly easy to determine its trajectory.  Thus, what starts out as a non-linear system goes back to a linear system.

Back to my level of ‘zoom’ issue.  If we were to take 1000 balls (or the same ball 1000 times) and try to balance it on the apex of a pyramid.  All other things being equal, the ball would probably roll down each face ~25% of the time or 250 times per side.  At this level of remove too, we see a probabilistic linear function despite a chaotic initial condition.

Much of human existence is in a finely-tuned balance.  There is not a human in the world who could build a conventional computer from scratch.  They may be able to assemble the components, even write in binary enough to make a compiler and have it complete simple calculations, but even they would not be able to know exactly the balance of the right elements and the right temperatures for their chip foundries.  We are, all-of-us, increasingly dependent on one another, balancing our inputs and our outputs and optimizing for some time horizon.  Sometimes, a clever human will lite on an idea that makes the world a bit better in some conventionally defined way, and push it in that direction.  Henry David Thoreau, for instance, writes his essay on Civil Disobedience which is picked up by Gandhi, and used in South Africa (twice), India, and back in America by Martin Luther King Jr. Sometimes.  People like Larry Page will find an idea whose power is not initially recognized, and then makes them fabulously wealthy.

There is certainly macroscopic variation in human existence, our equivalents of waves or a ball half-way down the pyramid on one face.  Just as the fates of these lone people are not well understood at the outset, by now, their personal fates, and the fates of their movements are not in question, at least over time horizons that we are accustomed to.  It would seem exceptionally unlikely that apartheid South Africa would pop back into existence now or that Larry and Sergei would go broke tomorrow.  That said, the fates of their children, are by no means secure.  The Rockefeller’s are not the magnates that they once were.

Melding
How should we then live?  Let me try to merge these ideas back together.  Non-domain specific chaos and domain specific chaos.

In the linear component of non-linear systems, the destination is essentially set, and our actions don’t have an outsized impact. If we are the air molecule colliding with a ball that is half-way down one slope of a pyramid, then our actions are inconsequential.  What we do in this system doesn’t matter.  Unfortunately, there are some people who are already on a collision course with this.  They are working in industries that are declining, in a company without a defensible competitive advantage, and with domain-specific skills.  These are people like the wheelwrights, or chandlers or yore who had venerable professions for thousands of years, until they just didn’t exist anymore.  They exist in the non-economic sphere as well, the religion of Baal would be difficult to resurrect unless the child sacrifice component of it were reduced.  Likewise, few cry for the Island Republic of Minerva. These people worked passionately in areas that we can rarely even understand anymore, let alone empathize with.  Unless you are prepared to overcome the inertia of the entire system, your efforts in these types of arenas will not fundamentally change the direction of the system

In the chaotic component of a non-linear system, our action does matter, but the outcome is often uncertain.  Should we be starting in a low-energy chaos-ridden state, then we can start something novel.  We can combine our physical and mental inputs in ways never seen before, and solve for the problems in our immediate circle of care, but be unsure of the effects over longer time horizons.  The first question is thus: Should we act at all?

It is true that most of these effects are more noticeable as the time horizon is shortened, and as the multi-causal nature of the problem is reduced, but that only provides solace if you are content to optimize for that shorter threshold.  There are some, like Nassim Taleb, who advocate something like this.  Minimal dependency, avoiding areas where your personal income is correlated with the vicissitudes of the market, but we can be certain that autarky in every case where it has been tried, leads to poverty.  In a moral sense as well, restricting your circle of care only to those immediately around you could seem like the citizens of Auschwitz saying they had no idea what is going on over the fence, and where all of that smoke was coming from.

I think this last example is instructive, because it reveals the true quandary here. Even inaction is an action.  Going along with the crowd, seeking herd absolution for our errors and our collective foibles is as much an action as choosing to push against them.  There is an old psychological experiment about three people tied to a railroad track, two on one siding and one on another.  A train is hurtling towards you. You are at the switch, and don’t have time to untie them or stop the train.  You can leave the switch, which is currently pointing the train towards the two people, or flip it to one.  You can choose to walk away from the switch, but are you any less culpable for this non-choice?

This is a happy outcome, or sorts, because it points to what type of systems we may want to invest ourselves in.

Action Matrix
I will not rehash the text in the box but merely focus on why the gradient is not flat.  I have some green (meaning it might not be a bad place to be sometimes) in the lower right quadrant because there are indeed times when through our actions we can overcome the inertia energy of a system and through sheer exertion, change its trajectory.  This expends many of our resources, be they physical, mental, moral or financial, but it is possible.  Generally speaking; however, in linear systems, you want to get out of the way of the rolling snowball…
This puts a general preference toward the non-linear, where your nudges are multiplied by the extant but directionless energy in the system.  Why then is the upper-right quadrant more green than the upper left.  This is for two reasons, first, because the action cost in non-linear systems is low, but the information cost is high.  To understand that you are truly looking at a non-linear system, and to determine the flux points is most of the cost.  If you believe that the system has already tipped, you may well be wrong.  In this scenario, it makes sense to at least give it another nudge in the direction that you believe it should go.  A bit of extra energy at the start may well be exponentially multiplied.
I hope this is not the Marine in me merely resonating with this being something like ‘putting yourself at the point of friction and having a bias for action.’
What about the unforeseen consequences?  This I have not solved, and perhaps I can’t given my limited knowledge and processing capacity.

Please, tell me what you thought before reading this, and let me know if this changed your mind.





Leave a Reply

Your email address will not be published. Required fields are marked *