The Physics of Here-and-Now

Pursuant to the emails Delerium as a Phase Transition and Quantum Vacuum Fluctuations, there was another remark. My reply:

Hi Anton!

On Tue, Apr 21, 2026 at 9:39 PM Anton Kolonin @ Gmail <a…@gmail.com> wrote:

Wow! That’s quite the table of contents!

Statistical physics overlaps economics, which, you could say, is the statistics of a game (involving money) played on a network. For those readers who think economics is boring, rest assured, so is statistical physics.

Both of these overlap with something that is apparently called “experimental philosophy” these days, which is, as far as I can tell, mathematical computer simulations of interacting agents playing games with one–another. The prototypical form is the prisoner’s dilemna played on a network. I recall a fun interactive game, The Wisdom and/or Madness of Crowds, from a decade ago, by Nicky Case that illustrated various basic results.

I presume that it is not lost on anyone that last year’s Nobel Prize, handed out for large language models, was all about how statistical physics was applied to solve seemingly intractable issues in deep learning. And now we’re spending a trillion dollars a year to built datacenters. Maybe even in space, if the billionaire futurists have their way. When I studied statistical physics in school, it never occurred to me that it might one day become a trillion–dollar industry. I think I missed the boat.

Back on Earth, we have the problem of equitably distributing wealth between cranks proclaiming nonsense about quantum zero–point energy, and the more legit, but still often-wrong contingent of, what shall I call them – earnest but midwit researchers who are less adept at chasing fame, glory, money, adulation or youtube stardom. The current economic system does not work for them.

In large corporations, there are periodic layoffs to remove the “deadwood” – employees who apparently contribute little or nothing to the economic survival of the corporation. In longevity biomedicine, we talk of removing senescent cells: those that spew free radicals and induce inflammation.  But what happens when these are people? Living, spiritual beings who want to get on with things? Or, in the present case, participants in scientific conferences at the edge of financial instability?

The statistical physics model I keep coming back to is Per Bak’s “Abelian Sandpile“. Drop grains of sand on a pile, and you eventually get avalanches. They come in all sizes. Usually small, sometimes medium, rarely large. This is called “1/f noise” (one–over–frequency) and is characteristic of network systems at the edge of a second–order phase transition. I like this model because you see it everywhere: water near the triple point (“critical opalescence”), but also biochemistry, biology, plants, animals, ecosystems and economics. Oh, and some neuroscientists talk about this too. Some claim to actually measure it, with neuron clamps or with MRI. Others theorize it to be a model of consciousness, thinking, common sense and decision-making. I can’t keep up with all the novel ideas, or distinguish the rigorous ones from the flights of fancy, the important ones from the fraudulent ones.  I do like the general outlines.

Avalanches in ecology require more sophistication to describe. Forest ecologists note that forests regrow and appear to be healthier after a wild–fire. Is this what the future after Trump will be like? He is burning down so much of the old world, that perhaps it will be easier for the new world to take root and flourish? I can only hope. Or does it encourage noxious invasive species? To stick to the sand–pile analogy: after a particularly huge landslide, the mountain slope becomes much much more stable. Trump etal are shaking the ground so hard that maybe the final outcome will be a stabler system. I dunno.  I see statistical physics where–ever I turn, but perhaps that is just me.

One area where we see accretion is at the frontiers of science. Vague thoughts and philosophical ruminations, soft and often wrong, lead to the preferential attachment of good ideas into the framework of “the known”. Sometimes a new scientific discovery slots neatly into an edifice of the known, much like a jigsaw puzzle piece slots into a puzzle. Sometimes this is very literal: mathematical formulas have very specific, typed jigsaw connectors: they only connect to other math formulas in very strict, specific ways. Software is like this too: the c/c++/java function `int f(int x)` cannot be connected to a string. The jigsaw tabs prohibit it. Our villain of the day, Nassim Haramein, is villainous precisely because he abuses formulas in this way, connecting them where connections are not allowed. In other cases, say in philosophy, the fog is so thick that it’s hard to see what connects to what. In other cases, say, medicine up through the 19th century, the system under study was so complicated that comprehension was nearly impossible.  Conferences, like your conference, Anton, on statistical physics, give me the willies, because I can already tell that a third of what will be published there will be just plain wrong, and another third will be not particularly insightful. But this is what humans do: make confused and confusing statements, often wrong. But knowledge does accrete. The avalanches, these we term the “scientific revolutions”. The old edifice collapses, as it no longer can bear the weight of the new knowledge. The small avalanches, aka “wrong papers”, “wrong insights”, “incorrect proofs” are commonplace. 1/f noise.

Science works (or worked? Should I use the past tense?) because academic journals were an early form of social media, allowing the rapid spread of new ideas, while also moderating the truly bad ones via peer review. There’s no peer review in twitter, facebook, bluesky. There’s only hearts and likes. I fear that the global brain will continue to edge towards delerium until we find some effective way of modulating the delusions. Sorry, anti–vaxxers.

Another place where we see accretion is in what I call “the here and now”. In case you haven’t noticed, we are permanently jailed in the present, and are propelled at light speed from the past to the future. Please observe that the past doesn’t exist: we cannot send any device into the past to take some photograph of how things used to be. We can only alter the present so as to fashion a record of the past, e.g. by inscribing letters into clay tablets. Or by inscribing memories into our neurons. The future also doesn’t exist … yet. The so–called “many–worlds” of quantum exists only in the here–and–now, in vacuum chambers and fiber optics, with wave-functions collapsing into the past at a furious pace.

Is the past a form of “platonic reality”? I suppose. It’s not like the platonic reality of mathematics, where anyone can infer 2+2=4. Sometimes, you can infer the past: this is what police detectives do for a living. Usually, you cannot: CIA covert ops will remain secret forever, as will the events on the Mongolian steppes from some millenia ago. Although these are in our past causal light–cone, direct inference is blocked.

Is the past some large cardinal axiom (literally) ? Well, either it is, or it is not. If it is not, then you have to posit something that exists outside of the von Neumann universe: call it “God” or “machine elves” or “the human soul”.  But then you have a recursive reductionist problem: if physical humans are “radio receivers” for souls from the great beyond, well, how does that work? And if you can start to explain that, you inevitably slip back into the domain of mathematics. So it seems that the past is some inaccessible, ineffable large cardinal.

Exactly how we use free will to shape the future and convert it into the past is perhaps the great unsolved problem in statistical physics. I ponder it intermittently. I should point out that quantum mechanics (quantum field theory) exists only in the here–and–now; the wave of freezing, moving from future to the past like ice-9, seems to have a “thickness”: the range of time over which quantum is valid. This is femtoseconds for chemistry, minutes for vacuum chambers in physics labs, and billions of years for light reaching our eyes from distant galaxies (which arrive here in “zero time”: the photons lie on the null light–cone.)  For example, in the famous “twin paradox”, the twin arrives younger, but he arrives in the here–and–now, and not in the past. When the twins reunite, they reunite, co–exist, right now.

Another thing that exists “right now” is my consciousness. I can remember the past, but I am not conscious in the past, I am conscious only right now. My self–awareness also appears to exist in some noosphere: albeit in the kitchen, in front of my laptop, where I type this, but also in the space of ideas. What is this space of ideas? One answer is mumble mumble mumbo jumbo quantum. Another is to note that LLM’s seem to “exist” on the surface of a billion–dimensional hypersphere (aka “the weight matrix”). This weight matrix consists of floating point numbers (aka “real numbers”). I would very much like to point out that quantum mechanics exists on the surface of a very high–dimensional complex–valued hypersphere (aka complex projective space CP(n) for n=10^80 particles in the visible universe)  So I’m quite happy to say “ah hah, that is the seat of consciousness: it is this ultra–high–dimensional hypersphere” (of course I would claim something like this, I’m that kind of guy.) But, and this is a big but: does it require the space to be complex–valued? Or can it be real? If it must be complex, and cannot be real, then LLM’s can never have the sensation of consciously existing in the present. Kind of a hot topic, you know.

I’m entirely happy to ascribe the (very human, personal) sensation of “self” aware and alive in the here–and–now, to animals: certainly all of those animals in all those youtube shorts, but also, well, that octopus in that Netflix documentary “My Octopus Teacher” friggin awesome movie. Go watch it. I suspect that amoeba have consciousness as well, but assholes like to argue with me. I don’t like arguing with them, because my communications style insults people. Oh well. Qualia, qualia.

Should I ascribe consciousness to a rock? Well, no. Perhaps consciousness only arises on one side of a phase transition, when a system is complex enough. Recall my description of “percolation” (the percolation of natural gas through fractured rock) in my earlier email. The percolation happens only when there is sufficient connectivity from one side to the other. So perhaps consciousness is like this too: some minimal amount of network connectivity is required. This is the statistical mechanics rebuttal to the philosophical idea of panpsychism.  The descartian notion of panpsychism says that its “qualia all the way down”. (DesCartes called them “monads”) The stat–mech reply is “its percolation”.  Percolation of what? Of the future into the past?

The problem with “mechanics” is that, well, it’s so “mechanical”. I’m trying to figure out how to slot “free will” into this wave of freezing from future to past, that we call the inescapable prison of the present. I’m working on it. I’ve attempted to staple statistical physics to quantum physics in a dozen different ways, none are satisfactory. My ideas are probably wrong.  But I’m working on it.

So there you go, Anton — my humble and belated submission for your conference. Best I can do on short notice.

— Linas

Postscript / Footnote

I’ve been pondering the “Here and Now” for a few years. If you search diligently, you will find my disorganized diary online in a very public place. That diary includes various attempts to make the idea mathematically concrete, including a variety of mathematical formulas and derivations. These should be accessible to anyone with a PhD-level formal education in theoretical physics. So you’ll need an advanced degree. But, in any case, none of the various ideas, attempts, ruminations and philosophizings work out. I mean, I learned something, but that something is not particularly Earth-shaking. Caveat emptor.