I do think it's a lot simpler than the problem Itanium was trying to solve. Neural nets are just way more regular in nature, even with block sparsity, compared to generic consumer pointer-hopping code. I wouldn't call it "easy", but we've found that writing performant NN kernels for a VLIW architecture chip is in practice a lot more straightforward than other architectures.
JAX/XLA does offer some really nice tools for doing automated sharding of models across devices, but for really large performance-optimized models we often handle the comms stuff manually, similar in spirit to MPI.
Nothing fancy. I made these with some pretty simple hand written scripts in javascript rendering to canvas: lots of fiddly little boxes moving around are simpler to script than to hand animate. (If I were to do much more of this I might rewrite these in blender since it has much nicer authoring tooling and export control.)
Quantum mechanics is needed to explain any microscopic phenomena in chemistry and biology - that is not at all in dispute.
The odd set of claims is that somehow biology has 1) figured out how to preserve long-range entanglement and coherent states at 300K in a solvated environment when we struggle to do so in cold vacuum for quantum computing and 2) somehow still manages to selectively couple this to the -known- neuronal computational processes that are experimentally proven to be essential to thought and consciousness.
This more or less amounts to assertions that "biology is magic" without any substantive experimental evidence over the last thirty years that any of the above is actually happening. That's why most biophysicists and neuroscientists don't take it at all seriously.
This referenced paper seems like primarily a theoretical modelling paper (almost all of its figures are simulations?) that contains as far as I can read 3 (!) actual experimental measurements in bulk on a fluorospectrophotometer. The claim is that the observed increased fluorescent quantum yield (QY) of microtubules over tubulin can be explained by the ideas in their simulations.
It's hard to buy that their proposed stories are the simplest explanation for these few measurements. Much more boring phenomena can influence QY. e.g. simply occluding fluorophores from the bulk solvent can have a huge influence on QY and spectra. (I used to design biological fluorescent reporter reagents...)
This seems like a theoretical modelling paper (all of its figures are simulations?) that contains as far as I can read 3 (!) actual experimental measurements in bulk on a fluorospectrophotometer. The claim is that the increased fluorescent quantum yield (QY) of microtubules over tubulin can be explained by the ideas in their simulations.
It's hard to buy that their proposed stories are the simplest explanation for these few measurements. Much more boring phenomena can influence QY. e.g. simply occluding fluorophores from the bulk solvent can have a huge influence on QY.
The static shape of a protein doesn't automatically give you a prediction of its functional properties. There's a hell of a lot more biophysics going on that we have no predictive models for that are needed to understand catalysis, allostery, assembly, etc etc etc. We don't even have good comprehensive data for any of that (compared to sequences or structure) to model with.
Fold prediction is an incredibly useful tool for scientists and genetic engineers to help design new proteins, but it doesn't magically solve molecular or cell biology. Designing new functions and mechanisms is still going to involve a huge amount of labor and brute-force experimentation.
Yeah we used to use this in our older models years ago... I don't recall the details exactly, but I don't think it ever did very much.
I certainly don't think it will help at all with stability. Things like Q/K layernorm are better tricks for softmax stability when scaling: https://arxiv.org/pdf/2302.05442.pdf
This is a dumb argument. Sick animals were probably culled immediately by the farms to avoid getting blamed.
As a 2-decade genetic engineer: there are no genetic "markers" pointing to a lab leak, there's really no sign of unnatural manipulation in the sequence.
You don't need access to a proprietary database to refute that sars-cov-2 wasn't copy-pasted at these restriction sites. We have public sequences of the closely related coronavirus strains. The unnatural SNP pattern would be absolutely obvious if someone patched together different lineages around these specific conserved RE sites. Instead we see a set of conserved RE sites related across the publicly known strains by homologous recombination.
What I've tried repeatedly to impress upon people here is that most routine cloning strategies leave pretty clear signatures, and the idea that a lab would go so far as to eliminate these signatures for such mundane virology work is tantamount to a much more elaborate conspiracy theory.
I'm pretty tired of debating people who don't know biology here. Using seamless cloning methods is super common - but they don't work like the paper authors suggest they do. I misspent my youth doing reactions and workflows like these for over two decades.
What they're observing is homologous recombination between strains - all the sites they're claiming are found in nature.
Again - there would be a genetic signal the strength of the noonday sun burning your eyes out if sars-cov-2 was made by cut-and-paste at these sites. You wouldn't need this ridiculous circular argumentation to prove that point.
As someone who was a genetic engineer for a long while, watching HN talk about dodgy papers like this is painful.
This paper posits a completely crazy cloning strategy that makes no sense (ie doing something far more convoluted than typical bsaI/bsmbi seamless cloning workflows that breaks the whole point of "seamless" workflows), and then tries to use that to make a case for a genomic signature that we could look for. They then look at a handpicked set of viral genomes, but leave a bunch out and duplicate others (I think WIV04 and WHu are the same), and largely seem to be observing without realizing it that yes, recombination occurs among these viral lineages.
This isn't even getting into the fact that a restriction-ligation based cloning strategy would leave glaringly obvious fingerprints behind in the form of the hundreds of nucleotide differences that are present outside the cutsites across the lineages... it would be blindingly obvious if someone just cut-and-pasted sars-cov-2 from other studied genomes.
There's a big difference between not caring about stability, and being willing to trade precision for better memory bandwidth for an application that doesn't benefit from increased precision. When doing large training jobs on TPUs, stability is paramount! It's true that you have to know more about what you're doing when you reduce bit-depth - the horrors of floating point are harder to ignore, and it's wildly inappropriate for many scientific computations. However the reduction of bit-depth is likely to continue as we seek to make modern models more efficient and economical to train and use.
To add to this comment (from someone who used to engineer proteins, and long ago DFT as well): DFT is only really decent at ground state predictions, computational chemists often have to resort to even more expensive methods to capture "chemistry", i.e. correlated electron-pair physics and higher-state details. Simulating catalysis is extremely challenging!
In the absence of any concrete scientific evidence for a lab leak (and an enormous amount of real, arguably dispositive, evidence against it), I see this doubling-down of articles attacking the character of anyone peripherally involved in work on coronaviruses. There's nothing at all new here.
A lot of people weren't even on this site back then... but this is exactly like watching the "ClimateGate" scandal play out here on HN over a decade ago, and in fact even involves some of the key players like Matt Ridley! Now, over a decade later, it turns out some messy dendrochronological paleoclimatology did not in fact represent a conspiracy invalidating the entirety of climate research and global warming is an even clearer threat.
In another decade we're going to have a much better understanding of horseshoe bats, asian animal markets, and coronavirus evolution... and not a single one of the breathless accusers involved in these screeds will ever apologize for flogging conspiracy.
Yeah, science is messy, scientists are flawed humans, desperate for money and generally terrified of bad PR and the mob. But the entirety of viral reverse genetics can't be thrown into an ill-defined "GOF" bucket. What constitutes acceptable risk and GOF in these areas has been an active debate in virology for as long as I can remember, and the quality of this dialogue is not going to be aided by this circus.
Sorry? I was a genetic engineer for two decades who has done plenty of mammalian cell engineering and viral design work. There's simply nothing about the original SARS-COV-2 genetic sequence that looks artificial.
Alina chan's original "evidence" put ridiculous weight on a few SNPs before we really had a lot of data or understanding of sars-cov-2 evolution (esp. it's enormously broad mammalian host range). It looks ridiculous in retrospect... and the rest of that book is conspiratorial storytelling without much in the way of evidence to back any of those fever dreams up.
Occam's razor does not mean every coincidence is meaningful - that's just pareidolia. The simplest scenario is that this viral pandemic is like every other one in known history - including the last coronavirus pandemic - a zoonotic crossover. Something that most professional biologists like myself has have been trying to stress.
Yeah, I used to engineer viruses and immune cells, etc. I've read that grant a few times. Again, a rejected _American_ grant for work to be done in Baric's lab in North Carolina. Most of the work outlined in it involved chimeric genomes that would be obvious. There's a few sentences about verifying properties of full-length quasispecies isolates, but the QS experiments aren't well outlined - a lot of chimeric experiments and re-syntheses, all which would likely leave pretty tell-tale signs.
Of course it's possible to describe a physically possible route to sampling, lab accident, and infection. Anyone can come up with these just-so stories. Professors too! Making a chimera of random isolates is crazy story. It's incredibly annoying to mess with 30kb RNA viruses. You have no idea. That's just not how a skilled practitioner would go about asking these questions. There's a reason most work on these things are in characterized strains.
These stories remain ludicrous when compared with the unambiguous epidemiological evidence for zoonosis that we have in hand. You're simultaneously alleging this tight collaboration between western scientists and WIV, but positing that somehow all these western scientists were in the dark about a conspiracy to do this massive amount of work without our knowledge, or an active conspiracy on the part of a significant percentage of western virology to hide it. All in a field that before the pandemic was a completely undramatic backwater of science!
The power of "lab-leak" isn't its strength as an evidence-backed scientific hypothesis, it's its power as a compelling work of fiction, and I doubt any amount of hard evidence will kill it.
The damage it does is distracting us from the unregulated wild-meat and fur industries that were the overwhelmingly likely cauldron for evolving these strains (as in SARS-1!). If we yet again fail to shut down those sources, we risk another pandemic just like this happening again in a few decades. That's where international attention and pressure should be applied, not this cockamamie distraction.
Sigh. As I've said in other posts, if any experiment of this sort were the source of sars-cov-2, it would be -blindingly- obvious. The synthetic metagenomic experiments looking into spike, etc. leave completely obvious traces, number one being the use of known, characterized viral vectors. Using some random isolates as experimental vectors is not only methodologically pointless and counterproductive, it's vastly more difficult to do.
JAX/XLA does offer some really nice tools for doing automated sharding of models across devices, but for really large performance-optimized models we often handle the comms stuff manually, similar in spirit to MPI.