Saturday, December 21, 2019

Cosmic technological selection

An important person who wants to stay anonymous sent me an amusing, one-week-old physics.pop-ph preprint by Jeffrey M. Shainline, a photonics-chip guy
Does Cosmological Evolution Select for Technology? (PDF)
that offers the readers an even bolder modification of the Cosmic Natural Selection, a meme by Lee Smolin, a self-described seer who will forever be deeply grateful to Jeffrey Epstein for the unexpected and generous funding of the ludicrous anti-string-theory crusade.

The anthropic principle is a meme saying the parameters describing the low-energy effective laws of a universe are heavily constrained by the condition that the values allow the birth of stars or life or intelligent life. The correct laws surely have this property because stars, life, and (occasionally and mildly) intelligent life are observed, experimental facts. So the minimal statement of the anthropic principle is a tautology of a sort. The real point of the anthropic principle only emerges in a stronger version of it: the anthropic principle really says that the existence of stars, life, or intelligent life should be considered a sufficient justification of the values of parameters and other properties – so that further explanations "why the Universe is what it is" are rendered unnecessary.

Clearly, this strengthened anthropic principle is highly problematic because the implication cannot be proven. Intelligent life could probably emerge within a vastly more diverse collection of the laws of physics. The strengthened anthropic principle could have been used at any moment of the history of science to stop further progress because the strengthened anthropic meme is basically equivalent to the statement "God created whatever He wanted and you must shut up and stop asking annoying questions if you don't want to be burned at stake as a heretic".

On top of that, the anthropic principle is often strengthened in another way. The minimal anthropic selection assumed above is "negative" in the sense that it just eliminates the Universes that are "hopeless" as cradles of stars, life, or intelligent life. As long as life... has some reasonable probability to emerge, even a small amount of life, the Universe "passes the test" and is on par with every other Universe that has passed the test.

However, in practice, many anthropic authors believe in an even much more questionable "positive" version of the anthropic principle. A promising Universe should not only allow life. It should allow a lot of it. The more stars, life, intelligent life is predicted for a set of values of the parameters, the more likely the Universe becomes. Needless to say, this "positive" version of the anthropic principle is much more problematic because we need to "count" the stars, life, or conscious intelligent observers. Is it possible at all? Isn't the number really infinite? Should we count the maximum number at some moment (and in that case, how can we define the "moment" in a highly curved connection of cosmic patches in general relativity?) or some total number of observers in the whole spacetime? Does the longevity matter? And so on (such questions are often labeled the "measure problem" but I think it is a misleading label, too). None of these questions may be quantitatively answered and none of these questions has been "rationally argued" to be relevant for the likelihood of a given Universe. All these claims are mere guesses and there is no objective protocol to decide which of them are better than others. On top of that, their proponents often do childish mistakes in their application of the probability calculus.

In the anthropic scheme, the competing Universes start at the beginning and some of them are supposed to be selected by the fishy anthropic criteria. But otherwise all the candidate Universes participate in a single race with the same start line and the same finish line. Lee Smolin has brought the readers of popular books a modified, more dynamical version of the anthropic principle, the Cosmic Natural Selection, in which the competing Universes don't start to compete "at the same moment", at the very beginning. Instead, the selection of the right Universe and the right values of the parameters is a long-term process that is supposed to be analogous to Darwin's natural selection in biology.

In biology, animals want to reproduce themselves and those that can do it well will be logically widespread. So humans are descendants of various previous life forms that tended to be more viable or capable of reproduction than their competitors. In Smolin's Cosmic Natural Selection, it's the Universes that are reproducing. Their values of parameters are analogous to the DNA of a species and the offspring are supposed to have "similar" values of those parameters as the parents. Microscopically, the birth of new Universes is proposed to be the birth of singularities in the black holes.

Instead of having sex, Smolin says, Universes are producing black hole singularities in stellar collapses and those grow into offspring Universes. The values of parameters that can maximize this production of the new Universes "win" and we must live in some of those.

That's a playful analogy of Darwin's mechanisms in another context. The real problem is that nothing works. According to our knowledge, it is not really true that the newly formed black hole singularities produce new Universes. In quantum gravity, there is really no qualitative difference between "having a singularity" and "not having a singularity", either (elementary particles are continuously "becoming" black holes if they become heavier than the Planck mass or so). And it is not true that the parameters of our Universe are particularly optimized for the production of black holes. Black holes are rather rare. It would be trivial to adjust the parameters so that the number or density of the black holes would be far higher than it is in the Cosmos around us (just reduce the QCD-Planck hierarchy, you will make stars smaller, and they will collapse to black holes earlier: clearly, the Cosmic Natural Selection seems to predict that hierarchies shouldn't exist, even more strongly than naturalness says the same). There is zero encouraging empirical evidence that those ideas are right and even at a purely theoretical level, the popular "demo" of these ideas can't really be clarified or elaborated upon. The "demo" of the theory from the popular books is everything that exists. The readers who buy these books are being deceived into thinking that the theory really works and someone could check those things in more detail if he read the technical papers but this suggestion is completely untrue. It is just stuff to impress non-experts, it can't impress the experts.

Shainline adopts most of Smolin's ideology but he proposes a conceptually new replacement for the "sex". Instead of producing new Universes from the black hole singularities resulting from the natural collapse of stars, the Universes on our branch of the "tree of cosmic life" are producing the new offspring Universes from the black hole singularities resulting from.... (so far it is the same) some technological collapse of matter that occurred because of the advances of the civilization.

So Universes have parameters that are optimized to allow civilizations that are good at many things, including deliberate collapse of matter that produces singularities, hopefully from black holes (although he vaguely discusses some strange experiments in condensed matter physics, too). Well, I originally thought that Shainline would suggest that the "technological Kurzweil's singularity" may be considered an equivalent replacement of the general relativistic singularities but he doesn't go this far.

There is some difference between Smolin's stellar and Shainline's technological meme. The difference is analogous to the difference between Weinberg's anthropic principle where the "birth of stars" is the only thing we really calculate and the more anthropomorphic i.e. more radical (in a different direction of being radical) versions of the anthropic principle where the "birth of human-like creatures" is really the "purpose" of a successful Universe. Shainline's version of the cosmic selection is more "anthropomorphic" than Smolin's. In this sense, it's even... "bolder".

On the other hand, nothing changes about the general features that make both memes hopeless. Just like Smolin's ideas, Shainline's ideas don't work at all. As far as I know, our civilization hasn't created any black-hole-like singularity yet, at least not one that would survive for much more than a Planck time (which is probably needed to have cosmic grandsons and granddaughters), and there are no promising indications that we will achieve such things in a foreseeable future. So the claim that our Universe is optimized for the singularity-making technology of this kind clearly seems to be wrong. It doesn't matter how beautiful it is, what Shainline's name is, and so forth. ;-)

Again, even at the theoretical level, these theories or mechanisms just don't make any sense. Because we're not mass-producing singularities i.e. new Universes, Shainline would have to argue that this mass production will arrive in the future. However, we want to explain why the Universe is what it is already now and by basic causality, it just can't be determined by the future because such a reverse logical effect would be acausal. The past affects – determines the unambiguously calculable probabilities for – the future, not the other way around.

To make the theory causal and scientific, Shainline would actually have to propose and study some pre-history of the parent and grandparent of our Universe where the technologically skillful human-like creatures were creating general relativistic singularities. Just like there is zero evidence that this technological production of singularities is taking place in our Universe (there seems to be negative evidence here), there is no positive evidence that such a mass production of singularities was taking place in our parent or grandparent Universe.

Moreover, all such technological activities would be pretty much totally detached from physics and cosmology. The whole cosmological evolution of our Universe would proceed just like before. The birth from the technologically assisted singularity would be just another Genesis story. As far as I can say, it is exactly on par with the Genesis story in the Bible – and with the pop science Simulation hypotheses, too. The technological singularity Genesis story may look more scientific (because it is technological) but it is a complete illusion. It is perhaps more "impersonal" but if you actually try to convert it to some equations that have sense from the viewpoint of the laws of physics, you will fail pretty much equally as if you try to convert the Old Testament to equations and path integrals. All these kinds of faith about the Genesis story are equally unsupported by science, equally separated from the type of the laws of physics that have been actually connected with the observations, and they're equally religious in character. The Bible is just a more traditional religion and "progressives" may prefer a newly created religion (with simulations or civilizations that mass-produce singularities from asteroids) but the irrational, religious character of all these Genesis stories is exactly the same.

The laymen mostly misunderstand these things and they're ready to believe that if something sounds more either wildlife-like or technological, or more impersonal, it's more compatible with science (just like a few centuries ago, people tended to believe the opposite: whoever said something that was in disharmony with the priest had to be a witch or a Devil or an unscientific crackpot). But this rule-of-thumb doesn't really work. This rule-of-thumb is really just an ideological preconception. The preference for Universes where civilizations technologically produce singularities is as far from the type of laws of physics that seem relevant as a system where someone creates species in 7 days etc. because of some human-like feelings and instincts (of the Lord). Both technologies to hugely compress matter as well as human-like feelings are emergent, complex process. In principle, both of them could matter for some history but both of them are very far from the fundamental equations of physics as we know them such as the Standard Model or string theory.

Smolin's and Shainline's proposals are clearly pop science which is why it's so appropriate that Shainline's preprint appears on the "pop science" section of arXiv.org (which is where all the papers by the other man should have been posted, too). It's "pop science" because lots of people like this cheap stuff. But it is not good physics and the authors of such ideas are really amateur physicists, not credible professional ones. I am afraid that e.g. Trump's planned liquidation of the professional commercial journals will contribute to the liquidation of professional science in general and the shallow, misleading, would-be competing theories created by amateurs – and designed for amateurs as the consumers – will become even more important and will suppress the non-trivial, professional, expert science "that just works, often at the quantitative or precision level".

The suppression is already bad enough today when hundreds of millions of people are drowning in Greta-Thunberg-level "science" but it may become even worse in the future. That's really why I am worried about the liquidation of the paid journals in most scientific disciplines. It's also a liquidation of the jobs of curators, filters, serious and knowledgeable reviewers. There won't be well-defined financial resources for them so their work will be disappearing, too. That will almost unavoidably lead to a further deterioration of the quality of science or "science" that typical people in the public – and even typical people with some degrees – will be imagining under the word "science".

Credible cosmologists or particle physicists obviously don't need a reviewer or curator to see that Shainline's paper isn't promising because it's self-evident that this is a paper optimized to excite the emotions of a science-fiction fan, not a scientist. But there are lots of situations in which the childish character of a paper isn't so obvious and when lots of work – which should be paid – is needed to separate the wheat from the chaff. They become particularly important in fields where too much stuff is being published. In string theory, the refereed journals are unimportant because almost no one can even write a fake string theory paper – it is just trivial for a professional to immediately see that the writer is a layman. But the boundary is much fuzzier elsewhere. Curators, reviewers, and other professionals who do some work and say "No" to many people (for their work and statements) are becoming politically incorrect. Everyone should always be told "Yes", the political correctness of our times says, and in this sense, Trump's proposal to liquidate the paid journals is a part of this PC culture.

But that's just a devastating approach to everything because "No" is extremely important. "No" makes us what we are, it is extremely important for the preservation of quality of virtually anything, and it is essential for the preservation of the technological civilization, too.

And that's the memo.