Entropy and Life
en.wikipedia.org
en.wikipedia.org
I read another article recently about the unexpectedly large role that randomness plays in embryonic development, and an idea clicked into place:
Life is about sustaining order amongst chaos, negentropy in a sea of entropy. But how does evolution lead to larger and larger pockets of negentropy that are capable of sustaining in increasingly hostile environments? How exactly does evolution lead to more and more “advanced” life forms?
Enter the magic of randomized algorithms. Randomized algorithms can often solve hard computational problems very efficiently, with the tradeoff that they have a small chance of failure. We can envision evolutionary leaps as computational problems, such as finding just the right folded protein to catalyze a particular cellular reaction. The magic of evolution is not just in building stable order, but also in harnessing randomness/entropy to solve environmental problems and then bootstrapping those solutions to solve higher level problems. Think about how just enough randomness is allowed into the process of meiosis to create perfectly functioning new humans that are wonderfully unique.
DNA and RNA are the non-volatile memory of the biological computer. Central nervous systems eventually reached a level of complexity that allowed them to persist memories, which opened up an even higher order problem solving mechanism. We humans have taken it even further with a cerebral cortex capable of abstraction, leading to complex language and the technology to record that language permanently.
To be honest, I've tried to wrap my head around entropy a few times (both in information theory and physics), but I've never really understood it well. It's related to but not (completely) the same as a measure of chaos, and it is related to but not the same as the number of potential states- and so on.
Could people direct me to a good introduction/explanation to entropy in an information theory sense? I feel like I'd really enjoy biting into this topic, but I haven't found a good entry point yet
Wiki: https://en.wikipedia.org/wiki/Free_energy_principle
HN Discussions: https://news.ycombinator.com/item?id=17529408
Capitalism is to Natural Selection as the Brain is to the Genome. Capitalism is the same Darwinistic entropy engine as Natural Selection, just abstracted to the plane of higher thought rather than raw biochemistry.
I think we can say that natural selection operates across all levels of abstraction, because in the end the only objective reality is the biochemical one. Everything else is simulation. There is certainly a strong parallel between capitalism and Darwinian natural selection but I can’t immediately see a way to state that relationship in a clear analogy.
A company's success is the result of a meme (or set of memes) the founder created (or borrowed) and convinced others to buy into. Others only buy into the meme if they believe it would be beneficial to themselves as well. Thus the system as a whole selects for memes that provide high benefit to society.
At least that's what I thought until I did a lot more reading over the next couple days after my trip and discarded the idea. Entropy and how life interacts with it turned out to be significantly more complicated than I thought.
Paraphrased from what a physicist doing "systems biology" (more or less: biology from first principles) said to me. I have also had a few lectures about thermodynamics.
Then the complexity goes down again, as it becomes uniformly mixed.
Yet, the coffee and cream are not alive.
the low entropy starting initial condition is the same in both cases, the glass is meant to be an analogy of the big bang.
This part is incorrect. It requires the intervention of energy by motivated life forms to to un-mix a mixture of coffee and cream, therefore it is very high entropy.
No entropy (absolute zero temperature) leads to the absence of Brownian motion which consequently leads to inability of atoms to recombine and form new matter. All chemical reactions would totally stop at 0 degrees.
Life would not be possible without chemical reactions. Entropy is invisible fuel that powers everything and makes things tick.
All life is preoccupied with removing free energy in order to get greater predictability over the future, and that encompasses literally everything we do in one way or another. Therefore, the theory that life reduces local entropy is not in any way useful, because it doesn’t guide us to do anything differently than we are already doing. It’s a bit similar to the anthropic principle.
Example: let’s say you need to clean up your room. You’ll find that even doing a very quick straightening of debris on a table will make a space feel more orderly and produce some level of satisfaction. But if you’re not the type to do that regularly, you might not bother the next time because clutter is the norm and it’s more dependable.
Another example: tell a businessman that life is a process which reduces entropy locally. Ok, so he has to get greater predictability over his circumstances, how to do that? More money and power. That was already taking place.
Thus, this theory hasn’t really taken off for lack of utility.
When we have a magical entropy measuring device, then yes, we can use it for all sorts of things...
In fact, I believe that life is not exclusively about order, but rather about harnessing randomness (entropy by another name) and capturing its output to bootstrap solutions to environmental problems. Our entire manner of reproduction is about allowing just enough randomness to produce interesting new results while maintaining the integrity of the life-form.
I’m a cryptographer, by the way, so what got me thinking about this was how we put a nice box around randomness to produce cool stuff like encryption and signature schemes.
0: https://0134340.blogspot.com/2019/02/playing-tenth-man.html
There were some simulations, but even then it was messy. For example, there were setups where energy was supplied by a driving force, and you could evolve "resonators". A naive application of Jeremy's ideas would predict a lot of structures in resonance with the drive, but in practice you often got less, because too good resonance actually made the structures break apart. So you couldn't get a sharp prediction about the final state even for a really simple system.
The issue is that thermodynamic notions are inherently "coarse", they only give you a high-level view of what's going on. Sometimes it's not powerful enough to say much useful unless you also know the detailed dynamics. This is also why, even if sharp and correct predictions were made for simulated systems, it's quite likely that biologists wouldn't view this as a true explanation for life. They would, perfectly legitimately, want to know about the detailed dynamics, the specific chains of chemical reactions that actually happened.
Now I'm in particle physics, so I'm not up to date, but the same dynamic happens here too. There are lots of beautiful ideas that come up, generate a flurry of excitement, and then get stuck for lack of sharp predictions, or feasible tests of those predictions. We file them away in the hopes that their time will come, possibly generations later. Science is hard!
A New Physics Theory of Life - https://www.quantamagazine.org/a-new-thermodynamics-theory-o...
Here's his website with a list of recent publications: https://www.englandlab.com/publications.html
---
This one stands out for me:
Design of conditions for emergence of self-replicators - https://arxiv.org/pdf/1709.09191.pdf (2018, pdf)
One comes from the work of Jeremy England[1], who has published some excellent papers showing what he calls 'dissipation-driven adaptation', natural selection working on phsyical and pre-biological systems. Englands work is really eye-opening and I have feeling if he keeps it up he may be a likely candidate for a Nobel in physics or biology some day.
The other line comes from Adrian Bejan's formulation of what he calls the "constructal law"[2]. This describes some specific patterns according to which physical systems evolve to dissipate energy more efficiently. It seems to me that Bejan is a bit under-appreciated in academia, although apparently people have found quite a few practical applications of his ideas.
I think this mechanism actually explains why there is anything at all. Are there any theories about the evolution of physics itself? Or how could a universe be evolved out of randomness? Or if there's a 'base reality' which can support 'simulated reality' be evolved in it? I think something really simple could evolve into supporting simulated child-universes with more complex rules.
The speed of light is also the speed of causality. Information can be sent out at that speed. So information sources have a sphere of expanding influence. Information that one node sends can be taken in by another node, modified, then retransmitted. This is general enough that it could apply to many contexts, but I am interested in the most general context. Information that replicates itself: memes. Is our existence creating a meme sphere? Will our meme sphere(s) collide with other meme spheres? Is this all just an unnecessarily strange way to look at things? I wish I had a firmer grasp on interpretations of quantum mechanics, since that may inform some of my questions.
> Will our meme sphere(s) collide with other meme spheres
Of course, "meme spheres" already collide all the time, but I think they are subject to too many intervening forces to be idealized as actual physical spheres like the propagation of light in a vacuum might be.
Some ideas get retransmitted a lot more than others. I wonder if a universally retransmitted (maximum strength) meme is possible.
There’s one expanding into space right now at the speed of light representing the comment you just posted.
I find the idea of life as an organizing force that locally reduces entropy and 'offloads' disorder on the environment deeply compelling, and what I wondered was, does this not put a pretty big nail into reductionist scientific worldviews that argue purely in terms of bottom-up physical explanations?
Is there some account of life at the level of particles that could ever give a reasonable description of the behavior we observe, or is the existence of living organisms evidence of some sort of genuine top-down causality?
The reason why I was looking into this was the TV show Devs that toys with the idea of determinism and the idea that even though someone could look at a prediction of their future, they could not change it, and to me this made sense if the world could be purely explained in terms of bottom-up physics, but I started to scratch my head if decision making could actually go top-down.
Anyway slightly rambly post but if someone has a book, or essay or some other reading recommendation on this I'd appreciate it.
https://www.quantamagazine.org/a-new-thermodynamics-theory-o...
https://www.quantamagazine.org/first-support-for-a-physics-t...
It's just generalized enough to be vaguely descriptive, not prescriptive, at least not yet. And I don't think it ever will because just as we have biologists, we still need medical professionals that can diagnose and understand more specifically the bottom layers.
Second question, someone mentioned Into The Cool, a book I'd like to also read that seems to take on this question.
Third question... to evoke the previous answer, why not both?
I mean that with that huge size and so number of opportunities, "invention" of self-reproducible forms is inevitable. And so we can treat the life as yet another form of matter.
There are quite a lot of possible consequences of this axiom.