Night Thoughts #0

I* had weeks of sleeplessness during this time, would lay in bed wide awake, envying my peacefully sleeping wife.  But I couldn’t stop my mind turning things over, examining evidence, books and papers I had been reading. While thinking about the probable results of humanoid robots with better AIs I had another interesting idea that supported my thoughts about how to reorganize society to emphasize human values and also improve our civilization’s technological levels.

I was thinking about the Boston Consulting Group’s ‘Experience Curve’. From an Economist article :

Bruce Henderson, the founder of BCG, put it as follows:

Costs characteristically decline by 20-30% in real terms each time accumulated experience doubles. This means that when inflation is factored out, costs should always decline. The decline is fast if growth is fast and slow if growth is slow.

 There is no fundamental economic law that can predict the existence of the experience curve, even though it has been shown to apply to industries across the board. Its truth has been proven inductively, not deductively. And if it is true in service industries such as investment banking or legal advice, the lower costs are clearly not passed on to customers.
I emphasized that “no fundamental economic law”, because I think it wrong.

First of all, the phenomena is very general. Other studies, which I can’t find now as Google is increasingly useless the further back or more obscure the items you are searching for, found that the costs of products from eggs through semiconductors decline at a predictable rate with every doubling of aggregate production. Kurzweil shows the declining cost curve is consistent from electromechanical through the latest computer chips.

I was thinking how that might interact with general-purpose robots, that is, robots with the capabilities to construct themselves. I could see how that could produce rapidly falling prices for even very sophisticated ‘bots, as labor costs would be very low compared to the cost of humans doing the same work.

Idlely speculating, I somehow connected this to Adrian Bejan’s Constructal Law. That law is a third fundamental law of thermodynamics which predicts how flow systems evolve in time, necessary because the first two laws of thermodynamics do not deal with changes in time.  Flows include fluids, heat,  materials, electricity, people, and systems of those flows are river basins, cooling systems in electronics, transportation, lightning and walkways. Those have structure, e.g. the drops of rain on leaves and rocks coalescing into rivlets which combine into trickles … to the deep and wide rivers joining the ocean.

The Constructal Law says that the evolution of those structures of flow, e.g. the erosion of obstructions to the flow of water from mountains to sea, will be to one that maximizes the flow per unit of energy expended. That increases the rate of producing entropy, because every drop of water hitting a leaf or rock has potential energy that is dissipated by its movement down to lower levels, and the faster it gets there the higher the rate of loss of that energy. Obstructions dissipate energy == increase entropy, but slower than the fast flow of the mass or charge or energy from high energy states to lower.

Gregory West’s scaling laws deal with flows.  For example, the movement of air in and out of the lungs produces entropy directly by resistance to the flow and the energy used by muscles driving the ribs and diaphram and indirectly via the oxygen enabling metabolism. Oxygen use by the total organism is much larger than the energy dissipated in breathing. Minimizing the flow resistance via the 3D structures that evolution has optimized maximizes the rate of metabolic entropy. The Constructal Law defines the evolution of the airway, blood vessel, and bone structures, and West’s scaling laws show how well biological genetic evolution has done.

Another view of all this is von Bertalantffy.  That writeup for some reason does not include the most important aspect of his thinking, that life is a state of disequlibrium that comes into existence wherever there is an energy gradient through a system, and does so in order to maximize the rate of entropy production.  That is, an un-obstructed flow of energy through a system, e.g. the sun’s energy falling on a sterile earth, generates entropy by being absorbed and re-radiating into space, by the mechanical effects of differential heating producing winds and ocean currents, evaporation of water and clouds and rain, and the resulting erosion.

Von Bertalantffy noted that their rate of producing entropy is much lower than an earth with life which absorbs more light via photosynthesis that in turn drives the entire food chain. Those ultimately re-radiate the same frequencies of light, but now as much as the ecosystem has caused to be absorbed, far more total energy. The ecosystem thus produces more entropy per unit time from the same energy flow from the sun striking a sterile earth.

Thus, a chain of reasoning connecting life and the products of life with fundamental physics.  Flows of information have also been related to entropy, and we can expect both scaling laws and energy dissipation analysis to apply to those.  Bejan discusses the structure of the internet and shows that it is also governed by the CL.

There is more hand-waving in this than a rigorous physicist can tolerate, of course. Some of that is due to the Constructal Law not being an equation, some from me not being a physicist. But there are clear connections to rates of new entropy in all of these and thus a physical law  basis for BCG’s Experience Curve.  Loosely, every doubling of aggregate production is equivalent to a larger river drainage basis, and requires a wider and deeper main channel from each new tributary on to where it meets the ocean. Alternatively, the same size river basin but with more rainfall, and so wider and deeper at every point from leaf to ocean.

The experience curve lowers costs, maximizing creation of the product per unit time, exactly as smoothly flowing air in and out of the lungs maximizes metabolic output energy. Building a product is the use of formed components and processed materials. Those can be thought of as containing anti-entropy, the result of the entropy produced when they were processed to useful forms. Production is producing anti-entropy ==> entropy, and the rate of production of goods is proportional to the rate of the production of entropy. Smooth, low energy manufacture of a product maximizes the rate of producing entropy, a result of the production of components and their combining into a product.

For the experience curve, the activities of people and organizations in pushing costs down is funded by the larger money flows produced by increasing volume produced by and producing the decreasing price, equivalent to eroding the channels to increase the rate of flow, flow being the dissipation of energy.

If you buy that, and I think careful physicists would consider it at least an interesting hypothesis to be explored, economics, as we outside of that profession have always claimed, is solidly rooted in a reality their analyses and models do not grasp : information. Economics has no independent basis : Dollars are the denomination of everything, and are used to define a universal value. But dollars have no independent value themselves, in fact they are only defined in terms of the things they can buy. Dollars cancel out in every equation if you reduce it to “a market basket of goods”, a dollar holds no independent value, no connection to anything out of that market basket of goods.  Economics is entirely the relationship between market baskets of goods. It’s connections are entirely to human behavior.

Information, on the other hand has an independent physical definition, also in terms of entropy. Entropy is everywhere in thinking about statistics and communications and data.

Increasing the rate of knowledge, the average scientific paper as an example, increases entropy production.  Increasing the world’s steel stock by a ton, ditto.  Both can be measured/estimated.  The relative rates would be a guide to improving civilization in a way that money costs cannot. It could well be that at the beginning of the Industrial Age, producing a ton of steel was the use of human mind and energy that produced the maximum rate of producing entropy, whereas now knowledge and education devoted to research do so.

Even without that connection to economics, human society ‘is’, i.e., can be modeled as, flows of people, materials and information. Those flows must/will-evolve-to conform to all of the above fundamentals of nature. Within that frame, and remember my many warnings** about frames and warnings, designs, whether produced by evolution, guided evolution, or planning, must/will-evolve-to improve flows of information and materials, increasing the rate of production of entropy.

That gave me a way of checking designs such as I was doing. First it justified my ‘gradual, no explosions’ approach. Explosions are equivalent, it seems to me, to a massive obstruction in a flow, say blocking the airway or a blood clot in an artery.  Certainly there is an immediate increase in the rate of entropy production, declining as the body decays. However, that is not sustained. Equally, anything degrading the flows in a society results in, at best, a short-term increase in entropy, and must be, in this frame of analysis, working against evolution, an aspect of nature. Nature’s intent, or at least a minimum-resistance path inviting us not to fight nature, is positive-sum games.

Second, as generating opposition tends to produce un-sustainable obstructions through explosions, metaphorical explosions of chaos as well as the real thing, it supported the non-coercive, maximum cooperation, positive-sum basis for such designs.

Cross-checking, there is a great deal of wisdom stating that actions that make a situation worse are rarely good in the long-run, and experience with contrary assertions such as “you have to break eggs to make an omlette”.  Reformations asserting those have produced social catastrophes, uniformly.

Nature favors positive-sum, no-opposition strategies I concluded. I am merely the latest in a long line of reformers of scientific bent who have reached as far toward this reasoning as their knowledge would allow, the US’s founding fathers were notable examples.

Next thought.

In psyops, the message is the op.

I conclude that scholars aren’t going to run out of things to research anytime soon.

*Generalissimo Grand Strategy, Intelligence Analysis and Psyops, First Volunteer Panzer Psyops Corp.  Cleverly Gently Martial In Spirit

**My collaborator, owner of the blog where we publish together, refuses to write another warning.  He thinks you read them.

Added later, after the fact cherry-picked evidence, blatant confirmation bias : Even phages demonstrate the power of cooperation and information flow to optimize the energy usage of a sub-system.

4 thoughts on “Night Thoughts #0

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s