Why Pure Information Gives Off Heat (Jade/upandatom)

Jade produces some interesting brain teaser videos. Here’s one where she talks about the Landaur Limit.

My sticking point throughout the video is the concept of “pure” information - is actually discussing the manipulation of ones and zeros on a silicon chip. Subsequently the ones and zeros create information, like a steam turbine create information in the form of electrical energy - that we use for lighting (information ?) and to transform into physical work (information ?) and heat (information ?) .

Quite often, “Information” is never really formally defined beyond ones and zeros on a chip. But the biological world doesn’t act that way, every photon bouncing around conveys . . . what kind of information? Beyond that, within biological creatures, let alone their interactions, there’s no analogy to this binary foundation of the “information” our computers create.

It’s exactly what I mean about that divide, between the mind, and our thoughts.
The mathematics vs. the actual factual physical world that we struggle vainly to grasp.

Going to the grand scale, it seems to me most great thinkers have never really wrestled with that mindscape/reality. Why? Because it didn’t fit in with the human condition? Survival of the fittest since the fittest could conquer and destroy other peoples and dominate their lands and resources?

Then came Abrahamic thinking with that self-absorption that produced our God vision - our Ego incarnate. Why do so few face that personal reality of the religions we create to wrap around ourselves, like tailored god-impregnated armored cloaks?

With a demand for perfection and certainty, which has permeated westered thinking, up to and including the dualism that is still a thing, rather than part of history.

Yes, yes, this is overstating, there were other ideas, but it’s impossible to consider the whole, by nitpicking nuances, because the whole has been dominated and driven by that “truth”. J.C. look at how we humans continue acting, and the things we’ve been able to once again allow, look at Putin channeling Genghis Khan without the brains. Or, we informed intelligent Americans made Trump our nation’s president because he best represented the general publics disregard for anything beyond one’s own needs. {There was more than voting for or against him - the normalization of an objective baldfaced lie, with the tactical intent of slandering and delegitimizing another American. And we all pretty much stood by silently, as though it just didn’t matter.}

1:50 - you say information “has its own energy cost”, but aren’t you talking about storing & moving bits of information on a chip? In which case, why isn’t it similar to your other examples, except that it happens on a microscopic (quantum) level?

The chip requires energy to store energy. So I’m thinking, it’s not “pure” information, it’s stored/processed “information on a silicon chip”.
Or?
How would you define “pure” information? As opposed to information stored on a chip?

At 11:11 she discusses schemes to supercool computers, which would make the storing/processing of bits to beat the Landauer Limit*
Isn’t that slightly disingenuous, how much energy is require to produce the refrigeration that super cools the computers?

  • This kinda looks like an answer.

"… The challenge in confirming the Landauer principle experimentally for a classical system was that such systems are usually described by a continuum of states, rather than the two discrete states required for a classical bit. The first experimental confirmations made an effective two-state system by using laser light [2] or electric fields [3] to confine a Brownian particle in a controllable double-welled potential.

“Taking on this challenge, Feng and colleagues [4] conducted an ingenuous experiment to verify the Landauer principle in the quantum regime. In their setup, the qubit was comprised of two internal states of a trapped calcium ion. The heat reservoir was supplied by the ion’s own vibrational modes, which were cooled to a few tens of microkelvin. At the start of the experiment, the researchers prepared the qubit such that its two states were equally populated, a condition of maximal entropy known as a classically mixed state. They then “erased” part of this information using a laser, which coupled the qubit to the reservoir and enabled the conversion of entropy into heat (Fig. 1). By taking multiple measurements of the ion, the team recovered the population of the atom’s two levels and the population of the vibrational modes. Comparing populations before and after an erasure, the team determined all terms appearing in the second law (Σ,Q,ΔS) and showed that the Landauer principle holds in the quantum regime. The team also found that achieving close to zero entropy production with a quantum reservoir was very difficult, particularly at low temperatures. …”

On a tangent. In reading articles like this, I’m struck by the notion that personal Knowledge can be compared to a sieve. The more you learn, the more strands are created between the hoop, and the more substance one’s mental sieve can capture.
For me, it’s a postcard where I can make out some fundamental concepts and arguments, that’s a far cry from internalizing (which is what that visceral is about) the full substance of that paper.

Beyond that, within biological creatures, let alone their interactions, there’s no analogy to this binary foundation of the “information” our computers create .

Down in the subatomic, sure guess that would be quanta, but here’s another case where scaling matters.

The Quantum world creates the substructure, skeleton, frame, but when we start scaling up, that “quantum” (a discrete quantity of energy proportional in magnitude to the frequency of the radiation it represents.) recedes into stuff that recedes into coordinated stuff, that becomes biological units, that build up to more complicated capable biology, that are capable of way more, than those original quantized building blocks.

I believe that organisms do have a related method of information transmission. It is not by zeros and ones, but via “differential equations” across cell membranes that trigger voltage differentials in the neurons of organisms which are then transmitted to the brain that responds by issuing “action potentials”.

Physiology, Action Potential

Michael H. Grider; Rishita Jessu; Rian Kabir.

Introduction

An action potential is a rapid sequence of changes in the voltage across a membrane. The membrane voltage, or potential, is determined at any time by the relative ratio of ions, extracellular to intracellular, and the permeability of each ion. In neurons, the rapid rise in potential, depolarization, is an all-or-nothing event that is initiated by the opening of sodium ion channels within the plasma membrane. The subsequent return to resting potential, repolarization, is mediated by the opening of potassium ion channels. To re-establish the appropriate balance of ions, an ATP-driven pump (Na/K-ATPase) induces movement of sodium ions out of the cell and potassium ions into the cell.

Cellular

(Physiology, Action Potential - StatPearls - NCBI Bookshelf)

And of course, we know how those voltage differentials are transported through the neural network; microtubules!

Yeah, okay that’s what the evidence is telling us.

So, not like a computer, with its information confined to 0s & 1s.
______________________-
Maddy is mad at so gotta run, then there’s wood to deal with before the promised storm tonight.

I don’t want to dis microtubules, I learned about them all ago and you’ve increased that understand m quite a bit. But when I hunt around, I see them mentioned but never elevated the way Hameroff (and you do). I won’t have time till late, but take a glance at these two article, why are they missing it?

https://organismalbio.biosci.gatech.edu/chemical-and-electrical-signals/neurons/

1 Like

True, although IMO, that distinction is only expressed in degrees of sensitivity.

I am not sure of this but the new series of GPT AI are now using clusters of data called “tokens” which are more gradiated than a plain on/off state.

What is a GPT token?

The GPT family of models process text using tokens, which are common sequences of characters found in text. The models understand the statistical relationships between these tokens, and excel at producing the next token in a sequence of tokens.

You can test the predicitve ability in the link above.

and still no acknowledging about biology forming in environment, community, etc.

It’s like the artificial (created within our minds) is more fascinating - sure but it’s an inhuman reality, it’s mathematical, and I am surely a human, not a mathematician, though I do use math all the time.

Because we never talk about the copper wire in electrical circuits either. It is accepted knowledge that when we talk about electricity that copper wires are the common transport highway.

And so it is with microtubules that are the information transport system in all of our cells, including neuronal cells (neurons are cells).
The term cyto means “cellular” and that means microtubules as the cellular data transporters.

In your link th term “astrocytes” is mentioned in the first paragraph. That means microtubules.

Moreover in the excellent link you provided, type in microtubules in the search menu and you’ll get 20 books written on the function of microtubules in physiology.

Science has become so fractured that in many areas of research in tangentially related subjects there are some 5 different terms used for “microtubules”. This is a result of the competitive rather than cooperative nature of modern science… damn!

But just remember. When the word cyto is used we are talking about microtubules as one of the major players in cell biology.

https://www.ncbi.nlm.nih.gov/books/?term=microtubules

Do you believe in abiogenesis? If so that answers your question.

But you are contradicting yourself. Your mind uses maths all the time but it does not function mathematically?
Mathematical biological patterns are all around us. Animals, plants, chemistry, weather. It all functions in the language of logic which is mathematics.

Everything in your body has a mathematical aspect to it. Everything in the universe and nature has a mathematical aspect to it. If it didn’t we would not be able to use mathematics to understand how things work!

This is axiomatic. It requires no other proof than its functional application in RW.

Nah, I’m not contradicting myself, you are looking through the wrong end of the microscope.
Thing is, it’s not my poetic flourishes that are at question. It’s little matter of philosophical bias and overstating the known facts.

Why play the silly?
Abiogenesis was a one time happening and it had more to do with chemistry than biology!
Biology only stepped up, after that initial spark took hold in Kreb cycle, and since then it’s been all about evolving cells in various broths, then by and by living creatures within evolving environments.

Does it really?
Isn’t it more a case of microtubules being components of astrocytes?

Hameroff makes microtubules the SuperStar, you amplify that SuperStar status. I never did any premed, oh I did pass an EMT course way back when, but that’s all common sense and learning to following protocols and communicate, not medical science, so I don’t pretend any sort of authority but I do know how to do some research, and when I do look up papers by genuine authorities, though I don’t follow the full scope of what’s being said, I can recognize that systems are always more complex. Microtubules are turning out more important component than previously thought, but they aren’t doing it alone, they are components of systems.

Direct visualization of microtubules using a genetic tool to analyse radial progenitor-astrocyte continuum in brain

… Astrocytes contribute to the emergence and maintenance of mature brain circuitry through their function as modulators of neuronal activity, neurogenesis, and as potential neuronal precursors 3,4,5,6,7,8,9. Radial progenitor development and astroglial polarity depend on dynamic modulation of microtubule cytoskeleton of these cells 10,11,12. …

I had another couple papers I was gonna quotes, but then YouTube occurred to me.

Astrocytes - for beginners

Neuroscience Alliance at UCF

Yes, yes, yes.
I’ve never claimed that it didn’t.
How could it be any other way, you think what happen on this planet, just fell together willy nilly? If the folds within folds didn’t harmonize ever so exquisitely perfectly, all this would have dissolved before it ever started.

For me, it’s like, okay, of f’n course math courses through it! So? What’s to poop our pants over, same with the wave particle duality, so what, why shouldn’t it be that way at the very tiniest of tiny?

I keep trying to explain, my heads up here in the magic realm of being a thinking sensual aware creature of roughly 40 trillions of cells, each containing an average 100 trillion atoms. An evolved sensing organism within a world of evolved sensing organisms.
That’s what? 4,000 trillion times removed from that weird quantum world that so many love getting lost within? While our world turns.

look at what you wrote below!

You have just described the advantage of continual growth in naturally communicating neural networks such as found in brains and forests vs limited growth in AI, that are restricted by size and available resources.

Although the new GPT 4 can boast some awesome features.

GPT-4 Is Coming Soon. Here’s What We Know About It

Official info, current trends, and predictions.

Multimodality: GPT-4 will be a text-only model

The future of deep learning is multimodal models. Human brains are multisensory because we live in a multimodal world. Perceiving the world one mode at a time greatly limits AI’s ability to navigate or understand it.

However, good multimodal models are significantly harder to build than good language-only or vision-only models. Combining visual and textual info into a single representation is a daunting task.

We have very limited notions of how our brain does it (not that the deep learning community is taking into account insights from the cognitive sciences on brain structure and functionality), so we don’t know how to implement it in neural networks.

more… GPT-4 Is Coming Soon. Here’s What We Know About It | by Alberto Romero | Towards Data Science