C. Elegans: The worm that no computer scientist can crack
139 comments
·March 27, 2025resource0x
This podcast will be a shocking experience for anyone who still holds reductionist views on life.
https://www.youtube.com/watch?v=c8iFtaltX-s
Michael Levin's channel: https://www.youtube.com/results?search_query=michael+levin+a...
His blog: https://thoughtforms.life/
red75prime
> a shocking experience for anyone who still holds reductionist views on life.
As far as I understood he doesn't state that life cannot be explained by physics. But that it might be more productive to choose a more high level view.
Makes sense. If a set of physical laws allows life at all, life will maintain its own set of rules (related to reproduction, goal-directed behavior and so on) regardless of low level "implementation details" that are dependent on particulars of the physical laws.
resource0x
Please read this essay and make your own conclusions. (He rejects physicalism in no uncertain terms). https://thoughtforms.life/platonic-space-where-cognitive-and...
red75prime
I don't think that there's a contradiction. "Ingress of platonic forms" is observationally indistinguishable from a selection effect: "Universes that don't admit realization of certain platonic forms are unobservable because they have no observers."
For example, a static zero-dimensional universe. There's no pi, no Chaitin's constant, no nothing. Or, in other words, there's no processes or objects in there that we can describe as "ingress of platonic forms" (and no observers to notice that).
I do like his ideas (and I wrote similar things about platonic forms elsewhere), but it's not a solid refutation of physicalism. It's an attractive framework, but as almost everything in philosophy it can easily be challenged.
A physicalist can say "Physical processes that follow a certain equation trivially have properties corresponding to the properties of the equation. So what? I can measure physical process and I can think about the equation (thinking is a physical process too), but why should I postulate independent existence of a/the platonic form of the equation?"
The fact that my subjective experiences undeniably exist makes me reject physicalism, but I can't prove their existence to anyone else and I can't use their existence as a solid basis for some philosophical view. After all it's just one bit of information. Or zero bits? I wouldn't have noticed my own absence.
fouc
Reminds me of Neal Stephenson's Anathem with the higher plane of truth. Neat! I mean, I figured Neal Stephenson got his ideas from somewhere.
didericis
"The Matter with Things" by Iain McGilchrist has the same effect. Highly recommend it.
resource0x
By any chance, have you watched the videos from the recent conference? https://ctr4process.org/conference/metaphysics-and-the-matte...
didericis
No, but I will. Thanks.
(link isn’t loading for me on mobile, will try again later)
Willingham
This project seems a bit funny to me. Cell level simulation is the current bottleneck of the project, why would we not start at the cellular level, and then move up to multicellular organisms? Of course this ultimately leads to the question, can we truly simulate anything if we can’t simulate at the quantum level?(at scale) It would be quite profound to find out that we can fully simulate complex life while abstracting the lower levels, if that were the case, maybe we don’t need them either! Good riddance to all those quantum conundrums (;
VirusNewbie
I think the assumption is some abstractions aren't leaky.
acchow
The abstractions are fictional. The system wasn’t “designed” with separation. We just pretend they are there for tractability.
VirusNewbie
Certainly this isn't true at every level of granularity unless you're making some Penrosian claim about QM affecting say, one's decision tree? Which is a valid though quite disputed claim, say more about that if you believe it though.
Animats
It's good that they are still trying. It would be really nice to know how biological neurons really work. That it's so hard, even with the connectome available indicates something important.
I used to think Open Worm was going to be a big deal. Top-down AI seemed to be stuck during the AI Winter. But machine learning got unstuck and Open Worm is still stuck.
melagonster
When last time I hear about the neuron simulation project, they can not get new budget.
the__alchemist
> His goal is nothing less than a digital twin of the real worm, accurate down to the molecule.
I am nitpicking this: The state of computational chemistry is not at a level to support this. I'm optimistic we'll get there eventually; need to find novel approaches, and current ones are too imprecise, or too slow.
I think the connectome aspect is more interesting, as it may be feasible to get there 100% without a computational chem breakthrough.
firtoz
What's the latest and the best so far? Are they using GPGPU? Is quantum computing there yet, or would it help? Heuristics and sampling?
rsfern
GPGPU is definitely mainstream for large scale quantum and molecular simulation. Quantum computing might help speed up electronic structure calculations, but my impression is that it’s still in its infancy
To give a sense of the scale of this problem, the largest frontier simulations I’m aware of are around the trillion atom scale. (On tens of thousands of GPUs [0])
Based on a quick web search, a c elegans cell is between 3 microns and 30 microns in diameter, so if we assume we can count atoms using the density of water then an all-atom simulation of a single neuron would need between 5e11 to 5e14 atoms. c. Elegans has 302 neurons so simulating the full neural network will be 2-5 orders of magnitude larger than current frontier simulations. Honestly more doable than I thought it would be, though all-atom simulation of a full organism still seems quite out of reach
This is all with classical force fields. Doing this simulation at the electronic structure level is much much harder with our current modeling capabilities
0: https://www.mrs.org/meetings-events/annual-meetings/archive/...
kaferoni
There's also interesting custom-made machines for molecular simulations that don't rely on GPGPUs and are significantly faster, e.g.: - https://arxiv.org/abs/2405.07898 - https://www.psc.edu/resources/anton/
throwawaymaths
is there any reasoning that for besides highly reductionistic and repetitive systems like crystals, quantum computing can compute quantum properties of molecules?
it seems to me "the quantum computer you seek" is the molecule + the medium (especially the medium) itself
mkoubaa
And I don't think we will get there with digital computers either.
nilslindemann
So, with what?
mkoubaa
If we get there, and it's not guaranteed that we do, it will be a different kind of machine.
FrustratedMonky
[flagged]
StopTheWorld
> flipping quantum coins...Newsflash: neurons and digital computers both have to play by the universe's stubbornly deterministic rules
I don't really understand what this means - you obviously know at the smallest level things happen by random probability - because you mention the quantum world - but then you say the universe is deterministic.
dingnuts
you're making a straw man argument and being an asshole about it to boot.
neurons aren't binary. who knows if they're special in a way digital computers can't simulate -- they aren't well understood.
nobody said anything about determinism besides you
cb321
Due to its technical content, some may be unaware that the author of TFA, Claire L. Evans, is a bona fide rock star: https://en.wikipedia.org/wiki/Claire_L._Evans with a grammy nomination and everything: https://www.grammy.com/artists/claire-l-evans/280302 . There are not many out there recognized for both music and science writing.. :-)
mateus1
Interesting, thanks for sharing. I’m going to read her book “Broad band” on women that made the internet.
cb321
You're welcome.
If you are interested in more background on biological brains, c.elegans is really just "the very first breakthrough" (in the phraseology of Max Bennett's A Brief History Of Intelligence -- a recent, worthwhile popular science book).
sitkack
Neat book, https://en.wikipedia.org/wiki/Broad_Band
For a copy near you, https://search.worldcat.org/title/1077130914
1hr long lecture by Claire Evans https://www.youtube.com/watch?v=wLQKp8mHHJ0
4gotunameagain
[flagged]
FrustratedMonky
All that was mentioned was that the author was a singer.
So singer and a writer. That was only point made.
frozenseven
[flagged]
amy214
>There are not many out there recognized for both music and science writing.. :-)
Yea, she's gonna give it to you baby! She's gonna go far, kid [1]
brookst
Very cool. We need more Buckaroo Banzais.
frozenseven
[flagged]
smj-edison
It's pretty crazy just how computationally intense biological simulations are. The timesteps are just so tiny and the number of nodes involved is insane...
jboggan
I think it's because we learn biology with a textbook cartoon illustration of the cell with sparse organelles floating in a clear sea of cytoplasm. The packed reality looks more like the halftime bathroom line at the Rose Bowl. Tons of heterogenous proteins and metabolites and mRNA and everything seemingly diffusing to its destination, shouldering past everything else. The sheer combinatorial complexity of the number of neighbors each molecule is imparting force on, combined with the constantly shifting conformational changes of many proteins which change the forces they receive, it's overwhelming. Recall that only recently AlphaFold was able to decently solve protein structures in isolation - add in dozens of shifting neighbors and that's suddenly a dynamic problem.
I am not sure where whole-cell simulation is at the moment since I've been away from the field for about 15 years, but I recall a rather difficult multi-month simulation that was trying to model an "empty" volume of cytoplasm away from the organelles, about 1/50th of total cell volume, and with all proteins and metabolites replaced by hard spheres of varying "stickiness". It was considered a huge success to just get a few of the diffusion rates for various compounds in the right order of magnitude. I mean, if you really want to get the fleeting interactions right you need to be modeling individual water molecules. I know there have been large advances in computing in the intervening years, but this was on I think #20 in the TOP500 at the time. Unlike AlphaGo I don't see any immediate avenue for AI to help with this because unlike protein crystal structures there is no wealth of quality training data for cellular dynamics at the molecular level.
pakue
It’s not simply diffusion complexity. A lot of cellular transport is dependent on directed transport via the cytoskeleton and myosin. For example a 1m long neutron would need years to move proteins from one end to the other if it just relied on diffusion.
smj-edison
Great point on the chaos! I'm planning to pursue atomic simulation in college, so I've tried to read as much as I can about the field short of the advanced math (hence college, lol). It seems to me that cross cluster synchronization is a massive scaling issue, since you essentially have global state both reduced and broadcasted every couple timesteps.
I've been thinking it would be cool to design chips to be realtime safe—that way there's no need to synctronize—and have further away information delayed (just like relativity) to deal with speed of light communication issues.
Never heard of fleeting interactions, would you mind to elaborate?
jyounker
"Fleeting" in this context is just a synonym for "short lived" or "momentary".
throwawaymaths
the heterogeneity and packedness scratches the surface of the oversimplified complexity. particles are moving (usually with 6 degrees of freedom -- though biology can cheat this) randomly. lets say a protein docks with another protein... the number of unproductive collisions per docking is on the order of millions.
null
FerretFred
Yeah, I found this when I was Boinc protein folding for Covid. The work units ran for hours, my towercPC nearly melted, and then I found that this produced mere milliseconds of simulation!
jxjnskkzxxhx
Now consider that the Planck time is actually 1e-44 secs.
dekhn
Bonds jiggle and wiggle with characteristic times in femtoseconds, and most people believe it's not necessary to simulate at higher frequencies than that.
atemerev
Hardly any actual physical simulations simulate every possible moment in time. We just calculate consequences of current events and put them in the queue to happen sometime in the future. And there's evidence that the real world works in a similar way.
xattt
> nearly melted
I’ll be that guy, but a computationally complex problem won’t push your computer to a temperature beyond design limits.
FerretFred
You're welcome to come to my attic office in summer, as they ambient temperature reaches 200F m a sunny day and my obsolete-but-powerful tower PC consumes 350W power and adds to the tropical heat :)
userbinator
It may if components are marginal... https://news.ycombinator.com/item?id=43008879
JohnHaugeland
It’s just a turn of the phrase.
When someone says that the sky is falling, that also doesn’t need to be explained
LoganDark
> a computationally complex problem won’t push your computer to a temperature beyond design limits.
Allegedly. It allegedly won't do that.
Practically, running your CPU at design limits for a very long period of time tends to cause the temperature of the rest of the chassis to want to equalize to that temperature, which can be above comfort limits.
3np
That's life.
mkoubaa
Except this isn't computational biology, it's fluid dynamics with an agent inside running a neural network.
You'd think they'd try something like smooth particle hydrodynamics so that it can be done more efficiently on a GPU cluster but I think they are just doing classic CFD
monero-xmr
My (religious, spiritual, please-don’t-downvote-because-of-hate) opinion is the aspect of free will which extends even to the lowest creatures means computers will never simulate them accurately
danwills
No hate from me! You can believe what you like but I do disagree I'm afraid.
I think any rule-following (ie ~computational) system is either below, or at the same level of dynamic complexity as all others.
I also think that it's truly impossible to draw a sharp line between alive and not-alive, so the idea of 'free will' would have to extend all the way down to the smallest things like atoms/quarks/space etc.
Is it really meaningful to suggest that it could be anything like 'free will' at that size? Sure some quantum stuff seems indeterminate but could be fed by something that looks random to us - it might not actually be (dunno!?), but the idea that things at a larger scale could be meaningfully controlled by feeding-in different values (that still somehow have to look perfectly random!) seems extremely far-fetched to me.
Larger things like a molecule, a transistor or a human! generally stay as coherent and predictable/controllable (or self-controlling) things, even in the face of indeterminate-ness at the smallest scales underneath. I just can't see how there's any utility in connecting the idea of free-will/consciousness to this indeterminacy.. it's like saying free will is just fuzziness, and how does that really help in understanding life, the universe and everything?
mromanuk
isn't free will, more about inner control, in opposition to external control?
jxjnskkzxxhx
Interesting. This kind of dont-understand-therefore-fill-in-the-gaps-with-magic is the same mental shortcut that would lead you to believe that cryptocurrency is a good thing.
vixen99
Well of course Shakespeare got it wrong with ‘There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy.’. At least, that's seems to be the acceptably prevalent view.
null
monero-xmr
My primary reason I like cryptocurrency is it takes money out of the hands of government, which is run by people and is therefore fallible
GuB-42
There are many things that we don't consider to have free will that is very hard to simulate accurately. Like the weather, and generally any chaotic system. We can't even simulate interactions between small molecules accurately.
I don't consider free will to be a scientific concept. It is philosophical, or religious. But if we give free will to things we can't simulate accurately, maybe planets like the Earth have free will. It is a common religious belief, and if you believe that simple life forms have free will, not too far fetched. The Earth, with its atmosphere, magnetism and geology is a really complex system we have a hard time understanding.
sitkack
Life is already an analog computer running on hardware (the universe). You could be correct that if life is using the hardware to the fullest extent, simulating it again would you a horrendous double virtualization penalty.
We could already be living under that same double virt penalty.
I think we have an outsized focus on free will, partly because free will is a trick played on us by evolution. It takes an exorbitant amount of effort to exercise anything close to what most people would consider free will and the phase lag is measured in hours if not years.
GoblinSlayer
Nature being continuous is exactly the reason why our universe is not a simulation.
kalb_almas
Can't continuousness be simulated by lazy evaluation? Also you're assuming the simulator is bound by the same physical limitations that exist inside the simulation which seems unreasonable to me. Simulations are usually vastly simpler than the substrate they run on.
viraptor
Has this been actually proven? I remember multiple experiments trying to prove the opposite - both effectively a discrete-like and a grid-like universe. Wolfram for example has some ideas about discrete graphs https://writings.stephenwolfram.com/2020/04/finally-we-may-h...
aswegs8
Quantum mechanics teaches us nature isn't continuous though.
rocqua
Does that mean you believe there is something beyond the physical realm that gives free will? Otherwise, it seems to me like another physical process should be able to simulate a creature.
There is still the general fact that plenty of processes are 'chaotic' making them impossible to accurately simulate (because errors in the initial state compound exponentially). But it seems to me like a good enough simulation of a chaotic system is phenemologically indistinguishable from the real thing.
brookst
I’m curious if you think plants, single-cell algae, or single-cell animals like amoebas will be simulatable?
monero-xmr
But simulate-able how? Not like they would exist in reality, interacting with the rest of the universe
JR1427
One interesting fact about C. elegans is that the way the grow from 1 cell to adult (~1000 cells) is almost completely reproducible between different worms.
I.e. the pattern of cell division and the positions and identities of the cells is very fixed, with known timing etc.
Compare this to mammalian development, where even at the early stages, development becomes less rigid and operates by "course correction", rather than by following strict sequences.
(I studied the early development of C. elegans for my PhD)
fastaguy88
Let's see: C elegans -- the worm no computer scientist can crack S. cerevisiae -- the yeast no computer scientist can crack E. coli -- the bacterium no computer scientist can crack HIV -- the virus no computer scientist can crack
Has a computer scientist cracked any complex system that was not engineered?
qrios
Interesting article, but no reference to "DEVS"? Prediction of the worms behaviour is a key scene[1] in the opening episode.
ddtaylor
I was reading the article and it disappeared entirely. Is there website functional? Is this intentional?
0manrho
What do you mean by "disappeared"? Did the page just go white/black or something?
I didn't experience that while reading the article (waterfox browser), and can't recall that ever happening on Wired, and I somewhat frequently read wired articles (few times a week, not a registered/paying member).
I tried it in chromium (no extensions) and if you scroll there is a fullpage ad where the article seems to "disappear" but it's essentially an ad fold, and if you keep scrolling the rest of the article is there. Otherwise not sure what you mean? What browser you using?
moffkalast
> But there’s a difference between schematics and an operating manual. “We know the wiring; we don’t know the dynamics,” Cohen said.
They have the architecture, but can't read the weights and biases as it were? I was under the impression that they somehow mapped the activations, but if not then they're really just flying blind.
https://archive.ph/2025.03.27-055757/https://www.wired.com/s...