[FRIAM] International firm to invest $254M in ABQ hydrogen factory -- ABQ Journal

[David Eric Smith]

> On Mar 11, 2022, at 11:13 PM, Frank Wimberly <wimberly3 at gmail.com> wrote:
> 
> According to Seibert and Rees a small population is the only solution, right?

Yes, this is their more-or-lsss assertion.

This is the drive-by-shooting problem that Marcus so well characterizes.

What S&R can do solidly is show that there is a lot of distortion, and generally quite a lot of dishonesty, in claiming success for various technological systems where a valid accounting shows that not to be the case.

From there they jump to an assertion that “There is no solution” (except population decimation), which is a non-existence assertion.  I won’t bother going into the logic of how absence of evidence isn’t evidence of absence.  

In my mind, I have a metaphorical picture of how I would like to frame this, which comes from Nitat Ay’s way of trying to quantify complexity of joint probability distributions (but we can extend it to measures over lots of kinds of joint events, such as complex processes).  In a simplex of distributions, there are various low-dimensional exponential families that are either simple or constrained in many ways, with only a subset of remaining degrees of freedom.  For any distribution in the simplex, we can measure its “complexity”, a la Nihat, as its distance from the nearest point in the exponential family of the “simple” distributions.  The point of various Information-geometry constructions is that such relations are defined.

There are various uses of this picture that come up in my work where these notions of excess complexity and distance also have implications for dynamics.  There may be fast transients toward the manifold, and then slow flow toward stable points within it, under natural selection or similar mechanisms. 

One doesn’t need the technicalities for the part of the metaphor I am using.  All I have in mind is that, out of the vastly diverse conceivable systems that include people and their artifacts on Earth, there is a family of those systems in which more-or-less everything is recycled, and the timescales all line up.  The biosphere before people, away from times of major transition, lives within this manifold, even though lots of local Darwinian churn is always going on.  Even in the course of major transitions, most of it probably remains there, and all of it remains there if one thinks of very short timescales.  Non-hominids seem to fit fine, and even people before the Younger Dryas mostly fit, though there is some argument that already the Neanderthals were having large-scale impacts on converting forests to parries by impacting mammoth survival.  (I haven’t read up on that latter, but someone shot an article across my bow somewhere.)

I would like to be able to measure “how far from the total-recycling manifold” our current world is — and importantly, getting down into details, in “what directions” — and then go after the question: for what human population is there a reasonable expectation of existence of a point in the recyclable manifold?"  For some claimed number and system model to be “reasonable”, we should know enough of what to do that it isn’t _all_ pie in the sky, and the unknown parts should be things that there is ground to regard as solvable.  Then we ask “what are transients from here to there, which minimize some cost we don’t like?"

Where S&R, and much of that crew, take a step I disagree with (and that I find annoying), is that they conflate the limits of their own imagination with that of the biosphere.  For them, the only paint in the fully-recyclable manifold is the one people occupied before industrialization, and perhaps before the Younger Dryas and the rise of annuals-based agriculture.  So everything modern has to be jettisoned.  I am exaggerating their position somewhat (probably unfairly much, because on much of this they elide details too), but the move is of that type.

If there were a contribution of science to make, to this question, it should be something along the line of “how do we assess the generative capacity of the biosphere, including the fact that it now has imagining, simulating minds in it?  To what extent does the capacity of the biosphere to recycle extend to artifacts and processes that may never have existed before, and that perhaps can only exist under active management by people?”  There are precedents for thinking such solutions exist, at least sector-wise.  Wes Jackson’s paradigm of perennial polyculture requires maintaining species on a genome-evolution instability, from which a natural system would run away to either being an annual with minimal roots, or a perennial with minimal seeds (exceptions being fruit trees, for which the genome-evolution reason is mostly understood, I am told).  The role of the human managers (breeders and farmers) is to stabilize this hybrid point that non-anticipatory selection protocols would remove.

So try to extend that proof of principle however far we can.  It will still clearly be the case that we are vastly far from the recyclable manifold in most of industrial method, meaning that the current method can only exist as a transient, and that wholesale methods have to be replaced.  That will be extreme and difficult, and perhaps impossible.  To assess the likelihood that it really is impossible — that is, to prove a _real_ non-existence theorem -- would also be a valid goal.

The frustration and misanthropy, however, seem to me to get in the way of whatever the best version would be for doing that.

Eric