timr

Well, you sort of expect the pop-science press to be crap. It's when you read papers in Science and Nature and come to the same conclusion that you start to go galaxy brain.

> I'm implying they may be biased in what they chose to report on and how they frame it, yes. Deciding which figures to turn into a chart is an editorial decision.

I think the Apollo thing is a fairly dumb hot take based on a near-term wiggle in a graph, but it's hilarious that you are calling them biased.

Pot, kettle, black, etc.

Substantively, I don't think it's surprising at all to see those lines go up -- it would be shocking if they didn't go up, given the intensity of the hype. The slope is what matters. In particular, I wouldn't argue that your graph shows any kind of acceleration in adoption over time.

The times when internal production is insufficient are basically neonatal and diseases. Humans, in general, do not need to supplement arginine.

Not my area of expertise, but my understanding is that this is an extremely limited claim (i.e. unborn babies / neonatal supplementation, or children with genetic defects), and not something you can even remotely generalize to "children" as a whole. The way we teach this in biochem 101 is "arginine is conditionally essential", with the "conditional" being "periods of stress or sickness".

In any case, I seriously doubt the author is talking about this, given the imprecision of the language. Also, eating this wheat specifically for arginine supplementation is goofy gym-science. Arginine is trivial to supplement.

https://www.sciencedirect.com/science/article/abs/pii/S03783...

https://www.sciencedirect.com/topics/medicine-and-dentistry/...

> It also had more arginine, an amino acid that children need to grow.

My kingdom for even a tiny bit of scientific literacy amongst people who are reporting on science.

Not only is this absurd ("children" need all amino acids to grow...as does anyone else), arginine isn’t even one of the essential amino acids. Your body makes it unless something is seriously wrong.

Not even close to a relevant metaphor. Compared to drug development, coding is trivial. It was always obvious that coding would be the first to fall: there’s a corpus of all the code ever written, the system is fixed in scope and relatively small, the data is text, and it’s all easily validated by machine for correctness.

There is no objective function you can optimize your way to a drug discovery machine, let alone a fundamental understanding of how the underlying system works.

Well first off: molecular screening (which is basically what you're describing) is inadequate for drug discovery. There are too many other factors that matter: water positions, molecular flexibility, short and long-range electrostatics, entropy loss, coordinating ions, and more. You need not only the "pocket", but an accurate model of the system's conformational change over time, to have any hope of understanding binding energetics. It's a bigger problem for small molecule drug discovery than biologics, but it still matters.

That said, we have extremely well-refined force fields that attempt to model the underlying physics, and still fail to do this well. There's absolutely no reason to believe that a generic LLM will bring anything to the table that improves the situation. Even with the hand-wavy assertion that we (somehow) get "lots and lots of data", and somehow train a purpose-built model to do the same, it's kind of like saying that we could turn a minivan into a formula one car with enough effort. Different tools for different jobs.

I actually do believe that informatics approaches will be helpful given lots of system-specific data, but when I say that, I mean "a bespoke neural energetic model that is trained specifically for some molecular system down to customized force-field parameters", not "Claude".

So sure, maybe Anthropic will invest lots of time and money and energy into making these kinds of molecular simulation models, which they can then use for high-throughput screening experiments. But even if they do all of that, there's still no guarantee that they'll bring anything to the party that hasn't been done before...because this is the lowest-value step of drug discovery, and a model like that won't tell you if your drug is going to be toxic, or have off-target effects, or have metabolic problems, or...a million other things, which are what kill drug programs in phase 1 and beyond.

That's all true, but they don't mention the relative costs of those different tasks, or how AI helps them.

Optimizing molecular properties of a lead is important, but ultimately not a huge source of costs. In any case, LLMs probably have as much chance of predicting outcomes as anyone else (i.e. very small).

Manufacturing is somewhat expensive, but most of the cost sits in actually creating a production process, which is not a thing LLMs can do.

Clinical trials are where most of the costs lie, but unless you can create a magical classifier that can tell me if my drug will make it past phase 3, you're not significantly bending the cost curve.

It doesn’t. That’s what I mean when I tell you the bottleneck is somewhere else.

Re-ranking pre-clinical leads is the scientific equivalent of that old saying about opinions.

Sort of. There’s also a big helping of intellectual arrogance, wherein the tech people think they can apply their same skills to biology, and get the same results.

(I say this as a person who has spent a non-trivial portion of my life as one of those people.)

> Seems like they're improving a $20 part of a billion dollar process, and a part that was already heavily using state of the art AI/simulation at that. The computers already spit out more potential candidates than what can be reasonably tested in vitro.

This is the fundamental problem with nearly all tech-meets-drugs startups. They almost always focus on the part of the problem where there isn't a bottleneck, because the parts where there is a bottleneck are the parts we don't understand at all.

Yeah, I mentioned this in another comment. There's clearly a trend toward equalization over longer distances, but even here, you see differences in performance out to 100 miles. But there are also no error bars, so we have no idea how significant these differences are.

Truly, you have a talent for ignoring the mountain of evidence in front of you, to fixate on the one thing that validates your biases.

The sentence you re quoting is from a site showing an across-the-board advantage for men, which gets smaller, but does not disappear, as distances increase. Literally every graph on the page shows that men are faster, in both swimming and running events, at all distances. But somehow you managed to read past all of that to find the one sentence that kinda-sorta says what you want to hear.

Impressive.

https://swimswam.com/comparing-mens-womens-records-in-swimmi...

No, it doesn't. Read the links. I'm giving you papers, and you're citing anecdotes from CNN and Women's Health.

There are literally thousands of female participants in marathons around the world every year, and you have to be in denial to ignore the obvious sex-based difference in performance.

For ultramarathons, in particular, the performance difference is smaller, but it still exists. The most you can claim is that at ~200 miles there's a slight average performance advantage in pace for women in specific studies.

https://www.nature.com/articles/s41598-026-46032-9?fromPaywa...

https://trainright.com/women-faster-than-men-ultramarathon/

https://pmc.ncbi.nlm.nih.gov/articles/PMC3870311/

> You've been posting one bad take after another and you clearly know nothing whatsoever about distance running.

I was a distance runner.

It helps a lot when you don't cherry-pick data that supports your pre-existing belief.

https://runninglevel.com/running-times/marathon-times

https://swimswam.com/comparing-mens-womens-records-in-swimmi...

https://link.springer.com/article/10.1186/2052-1847-6-7

https://results.nyrr.org/event/M2025/overview (pick any year you like by changing the URL)

https://www.marathon.tokyo/en/about/past/2026/ (Same)

https://www.nature.com/articles/s41598-026-46032-9?fromPaywa...

The overall story is one where the difference in performance decreases with distance, but is very real.

> Where did I say “for the whole marathon”?

You didn't, which is exactly the point. It's a critical piece of information for evaluating the truth and relevance of your argument.

> And yet, it seems you do not know that the gap between men and women shrinks the longer the distances get?

So what? The gap exists. Saying that it diminishes with distance is interesting, certainly, but it's a far cry from evidence that the gender gap is due to something other than biology.

> The article makes the claim that on average, women would perform better in ultra-endurance events, supporting their thesis that women actually did the hunting back then.

This is sophistry. In reality, even on average, men are faster than women.

> Generally they can’t - the pace setters run at a pace above typical max for the distance (but within max for the shorter distance they are running), sheltering the people they’re pacemaking for.

Come on. Think things through to the end. If it were actually true that the pace-setters could run faster than the competitors for the full marathon, then the pace-setters would win the event.

What you want to say is that the pace-setters are capable of running a faster pace than the competitors for a shorter distance. Maybe that's true (I doubt it), but it's not responsive. Pace setters are there to set a pace, like a metronome. If the competitors feel this pace is inadequate, there's absolutely nothing stopping them from running faster, or demanding a faster pace-setter.

> Inequity between male and female athletes is a result not of inherent biological differences between the sexes but of biases in how they are treated in sports.

But that's obvious nonsense. It doesn't even pass a sniff test, if you know even the first thing about running. We're talking about many, many data points, all pointing in the same direction: in practice, men outperform women in distance sports [1].

To even begin to argue against this mountain of evidence, you have to resort to slippery rhetorical tactics, like this:

> As an example, some endurance-running events allow the use of professional runners called pacesetters to help competitors perform their best. Men are not permitted to act as pacesetters in many women's events because of the belief that they will make the women "artificially faster," as though women were not actually doing the running themselves.”

So, the argument being implied here (but not actually made!) is that the performance difference we see throughout the history of the marathon is due to...pace setters?

Y'all realize that the competitors can run faster than the pace setters, right?

This reminds me of the old expression: "Keep an open mind, but not so open that your brains fall out."

[1] I'm talking about distributions here. Obviously, there are going to be many women who outperform many men on an individual basis.

The Atlantic did not shroud itself in glory during the pandemic. Better than SciAm, certainly, but they publish a lot of dubious stuff right next to decent articles.

For example, they publish this, now:

"The Revelatory Cringe of Anthony Fauci"

https://www.theatlantic.com/books/2026/07/anthony-faucis-emb...

Which, regardless of what you feel about Fauci, is at least against type for the Atlantic. But in 2020, they had no problem publishing garbage like this:

"Georgia's Experiment in Human Sacrifice"

https://www.theatlantic.com/health/archive/2020/04/why-georg...

I've been thinking a lot about this example of decline, and the conclusion I've come to is that as print media has lost relevance and power, the only people willing to fill the remaining roles are those people who have ideological axes to grind.

Basically, formerly great media empires have declined to the point where only people willing to work for low wages in uncertain careers will accept positions writing for those institutions. So we mourn the loss of the brand, but if anyone actually cared about the brand in practice, it would not employ these people.

> I liked this article. I appreciate that it didn't try to say it is or isn't AI.

...because nothing in it supported the thesis?

Print and audio/video have been showing clear declines since the 90s, thanks to the death of print, CDs and cable. Maybe AI is part of this, maybe not, but if it is, we're talking about a slight change in slope. These industries were walking dead before CNNs were a thing.

The one industry where you'd expect to see the clearest impact (coding) is basically flat.

Japan is unique in the world for the stringency of its service standards, in my experience. However, if you use western review services instead of their domestic versions, all bets are off. Google Maps, Yelp, etc., are packed with stuff tourists love, which no local would touch. Things that locals find unbelievably rude get 5 stars on Yelp.

Tourists think that this stuff is still higher quality than what is found outside of Japan (which it usually is), but things have definitely begun to degrade with the tourist boom.

This was less bad than that (a rigorously researched novel), but yes, the same basic idea. The fundamental problem with that kind of narrative non-fiction is that you don’t know what you’re citing - is it the fiction part, or the non-fiction part? Both?

The fact that a “legitimate” academic text cited this novel as a sole reference is an indictment of the entire book.

I’m not in an edit war. I’m reporting what other people have done.

When I saw the edit “war” (in reality it was the same editor and multiple different people making the same edit over years), I did research into the sources being cited because I was interested.

Based on what I found in the primary literature, I would have made the edit, but I gave up when I saw the situation.

Plainly less. It’s not a question.

Every source has bias, but wikipedia has the additional problem that anyone with a torch to burn, regardless of experience or knowledge, can become a dictator for the content on a topic.

Say what you will about encyclopedias, but they generally farm their writing work out to subject-matter experts, and at least try to be neutral in their editing.

> I think you have one thing backwards: the more obscure and niche a topic is, the more likely you'll find it biased, since it has fewer editors looking at it.

Maybe, but the point remains: this bias has fully permeated wikipedia.

I don’t want to get accused of brigading, and also, I think it has the potential to be a distraction from my point.