Hacker News Reader

Comment by ezion | Hacker News Reader

I'm still interested in any viable theory as to how random chemical processes are able to garner enough information in their own structure to replicate themselves.

That is, Von Neumann showed us that to have a self-replicating automaton we need both the factory and the information for the factory (saved within it). We know that's what a cell is. But how do random chemical processes get there?

Note: time works against chemical processes since the organic chemical compounds decay, keep that in mind when creating a hypothesis.

Note 2: remember that in this case, both the factory & the information gets randomly mutated until you have a factory and information that's complex enough to error correct.

View on HN

Imagine a molecule that can create identical copies of itself from some other molecules floating around; not by any "factory" but just by "lego mechanics". This is more or less what prions do.

At some point a slightly different molecule appears which is better at converting the molecules, or perhaps they can convert other types of molecules for replication: the first evolutionary step. Perhaps some other mutation allows molecules to "harvest" the other molecules to replicate themselves: the first predators.

Over time this becomes more complex; from singular molecules and proteins to lots of molecules and proteins.

I don't know what Von Neumann "showed us", but IMHO it's not hard to imagine life starting from very simple and humble origins given the enormous timeframes involved (IMO the biggest challenge in understanding these sort of things is understanding just how long a million or 10 million years is, insofar we're even able to do that in the first place).

>This is more or less what prions do.

I feel this is a bad example to use because prions only do so due to the existence of plenty of things as complex as prions. We would need an example that creates more complex structures out of simpler building blocks, not equally complex building blocks.

That's fair, but it's the best example I know of to demonstrate how this could have worked, even though it's not perfect. Happy to hear of other better examples.

Crystal can grow itself if you put it into or let it touch a solution of its own kind.

I would say solid state is more complex than fluid.

Crystals have very little information content, though.

What do you mean by information content.

The entropy of a perfect crystal approaches zero (at absolute zero). All the positions of the atoms are completely predetermined, so there is no information stored.

You remind me this thing. Those non perfect crystals which are dominant in the real world can have some of the most weirdest behavior in term of quantum chemistry e.g. photo synthesis.

Advanced material science is making use of those crystals to make semi-conductors and stuffs because they can just absorb sunlight and then react with other substances.

You do know that organic chemistry will quickly degrade right? And they have no memory, correct?

So you're telling me that compounds floating around self-assemble into a factory and the information necessary for the factory, including error correcting code, in a short amount of time, randomly?

I don't think you've thought this through.

You're approaching this from a very high-level abstract theoretical point of view; things probably were significantly simpler than you're imagining. There was no "factory" or "information" or anything like that: just basic mechanics on the molecular level. This is still how things like DNA work today, as well as prions as I mentioned before, albeit these are much more complex than earlier life of course.

Perhaps you could detail how organic chemicals without exception degrade rapidly. You may want to share this groundbreaking insight with those that study PFAs, some of which have estimated environmental half-lives of over 1000 years. It's also worth noting that a lot of environmental degradation is caused by the processes of living organisms, which is obviously not an issue if life has not evolved yet.

Additionally there are some theories that rudimentary cell like structures formed in clays, which could protect initial unstable molecules. Then it turned out that adding a biofilm improved the resilience of the barrier.

Cell wall is a really amazing thing.

Even modern RNA and viruses can last for days which is frankly a lot of time. A quick Google indicates you can probably get weeks out of some RNA molecules in reasonable environments.

If you’ve already come to the conclusion you want, why bother asking questions about it?

>You do know that organic chemistry will quickly degrade right?

I'm not sure if I would count days (or possibly weeks) as quick.

> I don't know what Von Neumann "showed us", but IMHO it's not hard to imagine life starting from very simple and humble origins given the enormous timeframes involved

It’s not hard to imagine a fire-breathing red dragon or a unicorn either.

What is hard is coming up with an even remotely testable hypothesis for how non-living molecules became alive.

https://en.m.wikipedia.org/wiki/Abiogenesis

"The prevailing scientific hypothesis is that the transition from non-living to living entities on Earth was not a single event, but a process of increasing complexity involving the formation of a habitable planet, the prebiotic synthesis of organic molecules, molecular self-replication, self-assembly, autocatalysis, and the emergence of cell membranes. Many proposals have been made for different stages of the process."

You can always start your search on matters like this on Wikipedia.

The replies to this comment are zoning in on the epistemological dilemma it poses. Soliciting theories for one of the most difficult scientific puzzles, on HN of all places: do you expect someone to serve you up a neatly-wrapped viable and plausible mechanistic theory?

And yet, I think it’s a serious comment, not a troll, because this is the right epistemic stance. We don’t know how life emerged yet and this is rightly uncomfortable.

As for the object-level topic: I’d direct my attention to any research that investigates the relationship between the cell wall and the inner organelles. I suspect that the viability of the cell factories was dependent on its co-evolution with the cell wall, which, by creating a semi-closed (permeable) system, would change the entropic conditions inside the cell.

We don’t even have a good grasp on what life is! Try to get an explanation of why a living thing is alive, but when it dies despite having the same observable material condition its corpse is not alive. I’ve yet to hear anything beyond the most obvious begging of the question; it’s always some variation on living things are alive because they have life processes and nonliving things aren’t because they don’t. And what’s a life process? Well it’s a process that occurs in living things of course.

I'll take a stab at it: a multicellular organism dies when the processes that are able to sustain the operation of the cell community cease, thus causing the individual cells to die individually or to live as single-celled organisms in the wild for a short while. The main processes that are essential to the life of an organism as a whole are the supply of nutrients and oxygen, the elimination of waste, and the coordination of various support activities between different groups of cells in the body. They're essential because most of the cells have delegated the capacity to perform these functions to the organism, allowing an increase in efficiency at the cost of tying their survival to the survival of the organism.

A cell dies when it's damaged mechanically or chemically, or when it's unable to sustain its metabolism thus causing chemical damage, or when its programming instructs it to destroy itself for whatever reason.

I agree I think the cell is key as its the foundation of the energy dynamics of the cell.

And re: the epistemological questions, etc: I think there's a temptation for people to look for a transcendent, teleological prescription for the 'why' of life; which can range from either 'God' to some abstract 'progress' concept where nature proceeds to 'higher stages' etc (panspermia, gaia, whatever). It's deeply part of at least western culture since at least Aristotle and Plato to use these kinds of tools.

But I believe the 'answer' here is just a big messy 'immanent' one rather than any transcendent order. Life happened as it happened because it happened... and it will unhappen someday, too.

And of course we would not be here to ask questions about it if it hadn't, so is it really a puzzle that needs a formula to answer it?

Humans look for "reasons" for things because that's a useful conceptual tool for understanding why other people in our kin groups do things, or why a herd of gazelle we're hunting is in a particular place, or why a plant we harvest from didn't grow well this year.

But there's no reason to assume that this conceptual tool makes any sense for understanding "life" or "existence." It's a crude instrument. At that level, there is no "why", there is only "is".

(And I'd posit further we should not make assumptions like "oh it happened here [us] so it must have happened elsewhere because <xxxx principle / transcendent order / natural progression> ... so ... Star Trek!"; the universe is massive, yes, but that doesn't mean the precise events duplicate multiple times and it does seem like complex multicellular life was a kind of ... fluke ...)

Until a synthetic agent will be here we can only guess. The current best guesses seem to have to do with the free energy principle, active inference, Markov blankets [1]. It might be just that simple: it's what happens when some proton gradient reduces ferredoxin [2]. Also related [3].

[1] On-going series of lectures, "Physics as Information Processing" - Chris Fields

Lecture 1 - https://www.youtube.com/watch?v=RpOrRw4EhTo

Lecture 2 - https://www.youtube.com/watch?v=WkWIqpxWRM4

Lecture 3 - https://www.youtube.com/watch?v=TOZp_XNYijQ

[2] "How does chemistry come alive?" - Nick Lane, Alkaline hydrothermal vents at 18:44, https://www.youtube.com/watch?v=QmvS7tgvy6U TL;DW at 52:00 "How does chemistry come alive? It happens when a focused, sustained environmental disequilibrium of H2, CO2, and pH across a porous structure that lowers kinetic barriers to reaction continuously forms organics that bind and self-organize into protocells with protometabolism generating catalytic nucleotides which promote protocell growth through positive feedbacks, favouring physical interactions with amino acids, a nascent genetic code where RNA sequences are selected if they promote protocell growth so genetic information has meaning from the beginning"

[3] "Origins of the RNA-Protein World – Lost in Translation?" - John Sutherland, https://www.youtube.com/watch?v=bSjIDStlZg8

> Note: time works against chemical processes since the organic chemical compounds decay, keep that in mind when creating a hypothesis.

This caveat does not seems to make sense to me, the advantage of time is that you run the experiment over and over, if one particular run of the experiment fails and then the organic chemicals involved decay, that doesn’t mean time isn’t still on your side.

Also keep in mind that the experiment is run on an absurd number of planets and of we can only be around to ask the question on the planet where it worked out.

Life didn't start as cells. It started as "naked" self-replicating molecules with no supporting infrastructure. Think a virus without its protective protein coat. Cells only came along much later, because they are much harder to evolve.

We already know that an energy input, such as solar energy, to a large collection of simple molecules like methane and ammonia can produce the building blocks of life: amino acids and nucleotides. Once a short chain of nucleotides is formed, it will (slowly) replicate itself by attracting other nucleotides in the environment. That's all the original "life" on Earth consisted of. The simple chains of nucleotides were both "factory" and "information".

We don't know all the specific steps that life took between those first simple replicators and cells. We may never know for sure because data left over from that time is so sparse. But that's not the same as not having a viable theory. We have that.

That's a good story, but is there any evidence for it?

We know simple building blocks can be created, in low concentrations, with large amounts of other components that don't end up in life. The concentration of any given chemical in this mix is exponentially dependent on its complexity. The "once a short chain is formed, it will (slowly) replicate" is entirely unsupported by experiment. Indeed, it's difficult to see how it could work, when it's replicating in a soup loaded with junk monomers.

If it is random, and the universe is infinite, as we assume it to be, it will happen somewhere, an infinity of times, no matter how unlikely the event is. There should be exact copies of you popping into existence somewhere from quantum fluctuations alone.

Current studies are about finding the most likely pathway, rather than knowing whether of not it is possible. We know it is possible because we are here.

As for the technical details, look up "abiogenesis" ( https://en.wikipedia.org/wiki/Abiogenesis )

It's hard to go beyond: "Even highly unlikely events will happen occasionally if you let the system run long enough".

This is clearly a troll. Can we not feed the trolls, folks?

It has come to this. Asking real scientific questions is trolling.

They're not actually interested in the answer. It's like asking "If humans evolved from monkeys, why are there still monkeys?!" You explain what a common ancestor is but it doesn't matter because they've already moved on to another bad faith question and aren't listening.

There's real interesting questions to ask here, you're right. They're not getting asked in this thread, though.

This is a convenient excuse for not answering the question. No, science doesn't excuse you if you decide the question wasn't answered in good faith. That's because if your theory cannot answer a question like that, it's a bad theory, regardless of the motivation of the person asking the question.

What's I see here is people defending conventional OoL as a kind of quasi-religion, with dogma that does not need to be experimentally demonstrated and must not be questioned. I treat all religions with disdain, including yours.

This isn't a place to make formal scientific inquiry, and that's not the way to ask scientific questions formally. The way to do that is to write a paper to a journal citing findings that appear to contradict whatever hypothesis is in dispute.

If you ask a question about science to a random person on the Internet they're entirely within their right to say "I don't think you actually want to learn something you don't know, I think you're asking your question to put forward an argument that I've seen many people before you put forward and that I'm tired of hearing, and therefore I'm not going to humor you by answering your question", until you convince them that talking to you isn't a complete waste of time.

Let me make it clear: OoL science has not answered the question of OoL. There are enormous gaping holes in the story. Asking about these holes, or pointing them out, is allowed. Pretending OoL has filled in these holes when it has not, and bullying those who ask the questions, questioning their motives, is not.

First of all, we will probably never know how life originally started, because it was a process at the molecular level that happened billions of years ago. Even if we found a chemical chain that leads to life, we would have no way to know if that particular one is the one that led to current life.

Second, anyone can ask any questions they like and anyone can answer or not any questions they like. No one is obligated to engage with you just because you asked a question. Actually, if someone tells you that they think you're asking your question in bad faith that's a courtesy they're having towards you, because they're giving you the chance to convince them that you're actually being sincere. They could have simply ignored you and gone on with their day.

Asking questuon without cost and put nothing into effort to work toward an answer is not doing science.

You're dismissing the scientific process that you claim to support by calling it a "quasi-religion". It's not, you're tilting at windmills, and creationists don't have anything interesting to offer. Here's a great resource when you're ready to learn:

https://www.talkorigins.org/indexcc/

Sorry to be short, but you're doing a very aggressive enlightened centrist bit. Chill out for a bit, and read through the above when you've got some time.

Also, the question was answered in this thread. If you weren't trying to do so much name calling, you'd probably have read the answer and been informed.

Garbage and nonsense, sir. The quasi-religion is claiming certainty when the science has not delivered anything close to it and rejecting attempts to question this false certainty. I see this all too often in OoL and SETI discussions.

This is such a great resource, thank you for sharing! I'm going to bookmark it.

> random chemical processes

You're begging the question here. How do you know it was random and not inevitable?

Chemical processes are random because they consist of events that happen when atoms and other things randomly interact with one another as they move around and transition between states.

What would it mean for a chemical process to be "inevitable"? Merely that it has a very high probability of happening?

Me feels like the usage of "inevitable" is to try to push the discussion toward religious and god or something like that.

Maybe not.

When we look at the evidence, it seems like life began pretty much as soon as it could have as the earth cooled. It could be that life starting on planets like ours basically just happens as a byproduct of the planet building.

Or, it means that the conditions under which life could arise are not persistent. Life either originated quickly or it would not originate at all. For example, if life originated in a small warm planetesimal, it only had maybe 10 million years before the short lived radioactivity in the early solar system (which we know was there from the decay products) died away and the planetesimals froze.

No, not at all. It was to try to push the discussion toward questioning our ability to predict the probability of life arising. The "random" argument is typically an argument from incredulity, "how could random processes make such a specific thing? The probability is astronomically small." But with only one universe observed, and only one instance of life observed within that one universe, we don't have any way to say that random processes just accidentally formed life vs the probability that it is likely, or even inevitable, that life would form given the properties of the universe.

Adding a god to the discussion doesn't really advance anything at all, honestly. It just explains the mystery with a bigger mystery, so it doesn't explain anything.

Given enough time, even with a random process you will have iterated through every possible result some day. So inevitable here means, it would have to happen at some point, simply because it is possible. So, math.

That highly probable events actually happen can be a reasonable practical assumption, but definitely not a logical certainty. It is the hard, probabilistic version of the sorites paradox: with few atoms and little time, any chemical process is unlikely; with a whole planet and a few billion years, chemical processes are very likely; at some intermediate sample size you can feel justified to assume that what you are interested in happens, but it is only a feeling, not a real qualitative threshold.

I think Von Neumann suggested what might be needed to engineer a complex self-replicating system. He didn't suggest a process that might incrementally lead to that though (abiogenesis in the case of biological life), which might be able to rely on simple chemical behaviors to start accumulating complexity.

Are you implying that the only viable theory is that of a creator god to you or do I not understand your post?

I never implied that in my comment.

OK, so what is your favorite explanation then? You seem to imply /something/, maybe just be explicit about it?

In his defence, just that abiogenesis sounds impossible to him doesn't necessarily mean that he has an alternative explanation.

Agreed. But that sounded like creationist dog whistling to me, and I'm curious why OP doesn't want to be explicit about what he thinks. Being explicit usually leads to improving the quality of a discussion IMHO.

You are exhibiting a deplorable behavior here, equating questioning of a scientific hypothesis with creationism dogwhistling.

When did asking "how do we know" become forbidden in science? Convention OoL is looking entirely too dogmatic and religious for my taste.

Just to respond to your first point, I don't think anyone here is "equating questioning the scientific hypothesis with creationism dogwhistling".

People are calling out Ezion for dogwhistling because of his writing style. Comments written in a glib manner, with superficial counterarguments presented as gotchas, that is suggestive of dogwhistling.

The alternative is that some exponentially unlikely step is needed, which would imply we're not going to see another independent abiogenesis event anywhere in the observable universe.

(replying here because too deeply nested to answer on the comment)

> When did asking "how do we know" become forbidden in science?

What has been forbidden exactly and are you sure we are "in science" here?

> You are exhibiting a deplorable behavior here, equating questioning of a scientific hypothesis with creationism dogwhistling.

Let me sum this up. Asking "how do we know?" with a hidden agenda is OK, but asking "what do you have in mind exactly?" is "deplorable behavior" and should be... "forbidden"?

OK, troll, bye bye.

It's sad, but not really unexpected, that you responded that way. Cult members don't realize they're in cults.

I found this video to be quite informative and addresses a relationship between the information incorporated into a chemical process and its function.

Nobel laureate Jack Szostak from University of Chicago delivered the Eyring General Lecture on March 17, 2023 at Arizona State University.

"The Origin of Life: Not as Hard as it Looks?"

https://www.youtube.com/watch?v=ZLzyco3Q_Rg

Looks like a good application of Betteridge's Law.

I think the only necessary ingredient for life is sloppy self-replication. I'd have to look at von Neumann's proof but the theoretical assumptions may not all be valid.

On that note, I'm curious about quasi crystals or crystals prone to substitutions in the lattice. I envision a crystallization process that occurs right on the edge between crystallizing/dissolving. Then any small unit block with some particular contaminant that is slightly more stable becomes favored by evolutionary processes and entropic ones.

If such a thing is possible, and has a sufficiently high upperbound on possible variations, I think you can get a lot of interesting behavior. Over a long period of time, perhaps a crystal unit block could develop that encourages more of its own creation.

It's a fine line though, because usually crystallization doesn't have sufficient complexity to keep evolving, and it's usually driven more by external conditions than by local conditions in the lattice, but nevertheless, it's the most plausible bridge to life I've ever come up with.

Hence why the guess is that it was clays not crystals. Clays are more complex mechanically, somewhat porous, kind of work as a prototype cell wall. Consensus is more on the side of biofilms than crystals at that point. There are relatively simple multiphase processes that can create such lipid and biological bubbles. Just check how micelles form from random hydrocarbons in water with additives. The more stable micelles would allow biofilms to form inside, and so it goes.

Generally cell wall first hypothesis over RNA/peptides first.

Through such a biofilm micelle only certain chemicals can get in or out. This favors particular molecules "crystallizing" (or rather polymerizing) inside such micelles.

>Generally cell wall first hypothesis over RNA/peptides first.

I've always felt that RNA was probably a very late addition to the game of life.

> I think the only necessary ingredient for life is sloppy self-replication.

But self-replication has to be very good to be useful. If it makes a mistake 10% of the time, the longest piece of information that be retained is about 10 bits.

It only has to be very good if the things being replicated are very fragile. If you only need 50% accuracy, replication can be very sloppy. And remember, thanks to selection, we can start off sloppy and improve.

How could it only need 50% accuracy? That gives you a genome of 1 bit. That's completely useless and doesn't explain how this process is going to lead to anything with higher accuracy.

There are thousands of examples of proteins that are less than 50% identical, yet fold into exactly the same shapes and catalyze exactly the same chemical reaction. Biological systems are extremely robust to change. And remember, self replicating systems do not only produce one output, they may produce millions. If only one of those million preserves the critical features, things can continue to evolve.

There's vastly more than 1 bit of information in those proteins. If there was just 1 bit, then that would mean that any randomly generated protein would have a 50% chance of having those properties (which is false.)

The question was, how accurate does replication need to be, in a replicating system with selection. My argument is that if you do not need perfect copies, and if you can make large numbers of copies, replication does not need to be very accurate to produce offspring that can continue exhibiting some property that can be selected for.

It seems like your perspective is more information theoretic, and does not include selection.

Chemical processes are not random, at least not in the way you're using the word. When you mix a base and an acid it's not subject to a roll of the dice whether they will neutralize or not. If you pour oil into water there's not a small chance that this one time they will dissolve into each other.

>We know that's what a cell is. But how do random chemical processes get there?

Work backwards. Think of the simplest cell you can. Now remove the membrane and the little molecular machines are free to wander in the medium arbitrarily far away from each other. The chemical processes are still happening, but they just take a long time to finish. All you need is for the abiotic medium to contain at least one of every of those molecules to have a sort of diffuse protolife.

Even if the probability for this to succeed is really really small, there's a good chance it will happen when you have billions of years time and idk how many orders of magnitude (10^40? 10^50?) of atoms/molecules constantly interacting.

No, that's not clear at all. The smallest known system capable of independent existence and Darwinian evolution has billions of atoms. A mere 40 or 50 orders of magnitude is trivial compared to the unlikelihood of this arising by chance.

Yes! I vividly recall the moment when I realized that the emergence of life of earth wasn’t just the lucky occurrence of a cell coming into being, but a cell that could also replicate itself in a non-destructive, information-passing manner. Truly mind boggling stuff. I feel like “time” is often used to obfuscate just how incredible such a development would be, “well if given enough time, then of course we saw a cell that reproduced itself in a sustaining way that actually passed on non-corrupt information while not destroying itself.”

I've always found "you don't understand, it took a really long time" to be an incredible copout

It might just feel that way because it is a boring answer, but boring answers are usually a good sign. We all want the emergence of life to be special, but there’s every reason to believe it was just random. You can fit some pretty unlikely events in (billions of years)*(however many planets). It is really more space, than time, I think, though. I’m actually not sure how many planets there are in the whole universe (not just the observable universe; if we happened to have sprouted up on another planet we’d just be asking this question from inside a different light cone, nothing special about ours).

No, it's a copout. The handwaving about "a long time/many planets" is very weak, when confronted with the complexity barrier between non-living matter and the simplest known system capable of Darwinian evolution. All that time, and all those planets (that we can see in the observable universe), give a number that is very small compared to the unlikelihood of just randomly surmounting this barrier.

Is there an actual mathematical model of the complexity barrier that I can look up?

I don’t see any reason to limit things to the observable universe. What prevents life from emerging outside of our light cone?

Nothing prevents life from originating outside the region we will ever be causally connected to, but life that can never be observed doesn't present much of a handle for conducting science.

As for complexity barrier: consider the number of possible stable configurations with N atoms. It grows exponentially in N. You need to argue either that the chance of a random molecule being alive is high enough (which doesn't seem consistent with the structure we see in life) or that the sampling of these molecular configurations is extremely biased toward some that can start evolution (which is just begging the question.)

> Nothing prevents life from originating outside the region we will ever be causally connected to, but life that can never be observed doesn't present much of a handle for conducting science.

It is a mathematical argument, and not a complicated one, so I’m confused as to how we could be talking past one another here. If we model the chance that abiogenesis occurs on a planet as a random event for that planet, then the probability that it doesn’t occur on any planets is

1 - product(from i=1 to n_planets, 1-P_i), where P_i is the probability that it occurred on the i’th planet. (Sorry for the ugly notation, it is a limitation of basic text, hopefully it makes sense).

These are independent random events, there’s no need to have them linked causally.

> As for complexity barrier: consider the number of possible stable configurations with N atoms. It grows exponentially in N. You need to argue either that the chance of a random molecule being alive is high enough (which doesn't seem consistent with the structure we see in life) or that the sampling of these molecular configurations is extremely biased toward some that can start evolution (which is just begging the question.)

So this seems like an argument from first-principles in chemistry, I don’t know enough about chemistry to say much about it, but I don’t think enumerating every possible configuration of atoms is how they typically do it. I could be wrong though, it is outside my wheelhouse. I would be more comfortable with this argument if it was some well-sourced/well-known thing.

Anyway, this sort of argument seems like it will just shift the individual P_i’s in the equation up or down. We don’t really know enough about abiogenesis to make good estimates there anyway as far as I know, so I don’t see any reason to write off this explanation as a cop-out.

When I mean "causally connected to" I meant "we could send signals to each other". Most of the universe we can currently see is causally disconnected, in that we can never signal them (the nonzero cosmological constant means they will recede faster than a signal could reach them.)

There could be unlimited volumes of spacetime beyond this horizon. Whether life originates there or not we can never know. There is a finite amount of matter within this horizon, and finitely many chances for life to originate. If OoL is sufficiently unlikely it is unlikely to have happened twice in the volume we can reach.

I definitely would agree that we don’t know enough about those probabilities to say much about the density of life.

> able to garner enough information

"Garner" is perfect here, thank you.

> But how do random chemical processes [make the leap from molecules to cells]?

Ya, that's a kicker.

Another mind bender for me is the concept of "self". When some collection starts to prefer their own company over others. That boundary (or barrier or affinity or homophily or whatever) between "me" and "not me".

>Note: time works against chemical processes since the organic chemical compounds decay, keep that in mind when creating a hypothesis.

The question is how quickly do they decay in a sterile environment. A virus which has no active mechanism for preventing decay can last for days. RNA can have half lives measured in days. That's a lot of time and I'm sure there's more stable variants but in modern times stability is not a prime concern for organic molecules.

The same way Christianity and Islam became prevalent. Pre-scientific people made up many stories about the universe. The ones that spread the farthest are the ones that replicate, by killing and converting opposition like these religions have. Similarly, many compounds have randomly formed that exhibit parts of what we call life. The ones that exist in noticeable quantities are the ones that successfully replicate.

Survival of the fittest, in relation to atomic structures. https://medium.com/@clay.c.edgar/the-basic-building-blocks-o...

Recommend this paper[0] that discusses the 'infinite monkeys' version methodically.

[0]: https://arxiv.org/ftp/arxiv/papers/1304/1304.3381.pdf

My amateur-pop-philosophy-science way of thinking of it is that the transition from low to high entropy creates all kinds of interesting and complicated physical patterns along the way. Enough time, and the right patterns, and self-replication could emerge. But it's the externally driven 'stirring' of the low->high entropy transition that makes this possible.

That is, life exists at the barrier between active and cool energy states. Sun to atmosphere/surface, hot planet core to surface/ocean. It's a (extremely complicated) product of these transitions.

So it's not so much about "how can this just happen by chance" as if you think about what life actually is -- a boundary condition between two entropy states -- it makes more sense: There is a stir stick spinning (solar energy & earth's core) in a very complicated long lasting layered beverage (our earth), and life is the clouds of pretty patterns as the layers mix. The self-replication doesn't happen without that stick being stirred. It's not magic, it's just really complicated.

Biochemist Nick Lane writes quite a bit about the alkaline hydrothermal vent theory and it is very compelling. I don't know (bio)chemistry well enough to fully understand it but it's something like (please people who understand this better correct me...)

a) Very early earth had the 'lucky' combination of active hot core and cool (but not frozen) water covered surface.

b) In addition to 'hot smokers', 'warm' (e.g. closer to cell / body temperature) vents exist and existed on the ocean floor.

c) The ocean water was relatively acidic compared to now. But the warm vents had alkaline chemistry.

d) The specific chemical reactions involved (hand waving here) as the warm vents mixed with the ocean created complex molecules, some organic. And some of the products produced were capsules/bubbles with 'walls' similar in structure cells.

e) There's (more hand waving) reactions that can happen across the walls and molecules produced inside the 'cell-like' things.

f) At some point they 'escape' the vent and its energy source. And in escaping 99.999% of them 'die', but eventually one that escapes does so in a way that it can self-sustain from other energy sources.

g) At the same time at some point those proto-cells develop RNA or RNA-like processes internally such that they can reproduce

It's f->g that I guess is the confusing part for me.

But work is being done in the lab to try reproduce this whole process.

In any case eventually the sun and earth will die, all cell machinery will stop, and entropy will have finally had its way with us. It's not a self-sustaining automaton. It requires an externally driven energy transition.

Very uninformed guess: couldn't you just have a factory with information of a different factory, with information of a different factory, with...?

While the original factory is not self-replicating, if you reached a fix point you would get a self-replicating factory

This is starting to sound like Java...

When the thing is not just simply a molecular but work as a group, that can create some "energy potential" by itself, like attractive to water or nutrition or something. The whole idea of self evolution is not impossible.

I've always assumed it's something like the infinite monkey theorem?

It’s that, but with the efficiency of violent competition to ensure things develop more coherently.

hmm right, so while the monkey needs infinite time... in reality there's a feedback loop that rewards the "correct" output

“Punishes” the “incorrect” output, more commonly.

I tried writing code to simulate the "Monkey's typing on a keyboard === Shakespeare" concept.

It was a simple "hello world", and then with only the characters needed for the program. I didn't see it spit out any usable code when I played with it.

But even if it had, it still wasn't able to run without my code existing already... (and the environment I coded it in)

A closer analogy would have been generating machine code directly.

Now imagine that at every time step you also flip bits randomly on your own code.

That's what needs to be done to achieve what was achieved.

Are you coming at this from a creationist point of view? That's common rhetoric I heard when I was growing up. You'll need to jettison that entire way of thinking, it's built on faulty analogies like "the probability of shuffling a deck of cards into a certain arrangement is astronomically small, therefore evolution couldn't happen!" If you're actually interested in learning and not just sealioning, IMO this is a great resource:

https://www.talkorigins.org/indexcc/

There's a section specifically on abiogenesis: https://www.talkorigins.org/indexcc/list.html#CB0

To answer your initial question, it's really difficult to mentally grasp the scales involved with this. Billions of years, quadrillions of watts of power from the sun, some astronomically large number of chemical interactions happening simultaneously all over the early Earth. These numbers throw off one's naive sense of scale. The basic recipe of "hydrogen + time" will produce some very interesting results, however. Eventually we'll reach the heat death of the universe, or some other sort of end[1], but until then all the matter and energy bouncing around will get up to some cool stuff.

[1]: Good reading here: https://en.wikipedia.org/wiki/Ultimate_fate_of_the_universe

Are you equating any questioning of OoL dogma with creationism? That's not how science works, buddy.

Sigh, no. You're asserting that a lot in this thread, but that's not what's happening. In my comment above, I linked to a great learning resource that directly addresses the questions ezrion raised, and is good reading if you're actually curious. Separately, I addressed the likely context in which the question is getting asked, and pointed out that the worldview that starts with creationism is thought-terminating, and should be reassessed as a foundation.

It's great to question any existing consensus. But do so with the intent of learning, not trying to sneak in a religious worldview. If you're going to push a religious agenda, be courteous and at least be up front about it.

No, that's exactly what's happening. Ezrion is asking good questions. Dismissing those questions is not valid science. It's always valid in science to ask "but how do you know?" It is then necessary to accept a good explanation in good faith, though. However, OoL has gaping holes that are currently papered over with handwaving. It's not required to accept handwaving.

What are you talking about? Adding random noise to your random thing do not make it "more random". It remains the same.

Not trying to argue one way or another, but this kinda sounds like trying to reason about rock erosion caused by water by watching a mountain stream flow for 5 minutes.

We have a pretty detailed snapshot of life now, some fossils, and a general understand of biology.

I think you could actually get pretty far at understanding erosion if you had a pretty detailed snapshot of interesting river basins, some landmarks (this rock was over there, etc) and a general understanding of physics.

And yet, self taught ML algorithms have mastered Chess.

How about fire?

Reply on news.ycombinator.com