The "right element for the right meaning" crowd is always going to fall when they charge against the sheer walls of "but I like using Caution for my bulleted list" crowd. Or the "of course we put tire patches in the wiring element, it's when we use the tire patches".

Or a million other examples I've wrassled with over the literal decades.

Whatever sophisticated semantic scheme you move into the markup layer, is gonna get messed up. And then your fancy schema is just noise, very expensive noise. Markup layer needs to glue as closely to natural language constructs as possible, and then glue a little more closely than that, and let the bulk of the information derived come from natlang. It's boring, but it works, and it's easier than ever to get quantitative information out of natlang.

Keep the domain stuff out of the markup layer.

All that said, Asciidoc forever baby. Write actual books in it. Not really more complex than Markdown[1]. Beats up DITA, takes its lunch money. Win win win.

[1] Unless you go legit insane with `include` and `ifdef/ifndef/ifeval`

Taking the long term view, our system's failure will be seen as distinctly parallel to the failure of the Soviet system[1] - dual headstones of the European Era. I suspect that the common collapse will be seen not a result of any inferior ideological or economic systems, but a common lifecycle stage for all nuclear powers.

There's a mechanism deep inside the fundamentals of strategic nuclear weapons that has permanent and profound effects on the society and nation holding them[2]. One obvious thing that is easy to perceive even from our own vantage point: nuclear powers suffer from an always-expanding executive branch. Nuclear weapons make the Westphalian state system structurally obsolete . . but we have no replacement, so we're left with zombies. The nuclear states inevitably evolve into these sort of executive-branch security zombies, able to fixate on nothing but destruction.

I'm not going to pretend I have the answers here, but, well, we need to stick around for the future histories, don't we?

[1] The Soviet system having the disadvantage of being beaten the hell up by the great wars of the 20th century. You don't lose 2/3rd of your working age men and just go back to the mines like normal.

[2] Vaclav Smil has written extensively on energy transitions, and the quantitative point is incontrovertible: the jump from chemical to nuclear energy release is on the order of a million-fold, comparable in magnitude to the jump from metabolic to combustion energy. When genus Homo tamed fire, it forced speciation. We would be fools to expect less from the power of the atom.

Taiwan might be indefensible at any rate of production the United States could conceivably spin up within a practical time horizon.

The volume of fire that can be generated inside the 100km line by PLAN/PLAAF/PLAARF forces is nothing less than breathtaking. Even if you parked three Ford class carriers inside optimal mission radius, and their entire complement could take off and land in near-training conditions, and they don't need DCAP or electronic warfare coverage, *AND* if every single bomb and aircraft has a glassy-perfect mission right weapon to right target - EVEN WITH all these impossible conditions satisfied . . you're still not generating enough weapon effect to suppress even half of the PRC fire generation complex vis a vis a Taiwan situation. And they don't need half.

And that is not going to be the operational situation for USN. No, not by a long shot. If we're particularly unlucky, we might not ever know for sure what happened to the USS Whoever - just that it sailed into an electronic fog past Zamami and was gone. Rescued sailors could add little more.

The powers-that-be know this, the elected politicians know this (but don't care because they often have pockets stuffed from Chinese interests), but still we have chest beaters of the unstoppable American juggernaut. Yes, we do have a very big military - it's true! - but it's a military that the largest economy in the world has spent a good deal of its resources working to counter. For twenty years.

My greatest fear is that the chest beaters do assert direct control, court disaster, and have the worst possible reaction. I'm not confident in a sane response to a major surface asset being sunk; these are not people mentally geared to handle humiliation.

We're flirting this line already with Iran - with goddamn Iran of all people - where many People Who Should Know Better have already been flapping their mouths about breaking the taboo on First Use. For Iran.

Where the heck is the mild-mannered retail investor going to go if they bork up all the indices?

Borking up the indices has been inevitable since index investing rose. Think of index investors like a bunch of entirely undefended prey surrounded by predators. If we had a working SEC somebody might do something, but we don't.

Using your metaphor, this rule change represents something more along the lines of an invasive predator, new disease or parasite.

I'll stop there, because Musk is much-loved by the internet, but to pull on this string of systemic risk, let me see if I can walk through my thoughts here:

Nasdaq Inc. (the exchange operator) and Nasdaq-100 (the index) are not independent entities with separate interests. Nasdaq owns and operates the index. A SpaceX listing would be a major win for Nasdaq, reinforcing its dominance in IPOs and billions in licensing fees - Invesco QQQ Trust alone has over $300 billion. Every basis point of licensing fee on that is enormous static income. When SpaceX says "change the rules for me or I list on NYSE," that's a threat to both the exchange listing fees and the downstream index licensing revenue.

So the competition here is between index operators (SP500 is actively managed by an actual steering group, so rules changes aren't as big a whoop) for who can net the biggest pool of new shares so that their passive income on index license fees can get maximally farmed. The actual thing the money comes from - retail index investors - don't figure in this equation at all. It's a systemic risk of replacing actively managed funds with consumer 401k products, risk that requires active legislation and law enforcement for there to even be a fair marketplace; markets require regulations, after all[1].

If you're thinking this sounds a little like the competition between the ratings agencies circa 2008 to see who can award the most AAA ratings, you're in the right ballpark. It wouldn't even be possible if SpaceX wasn't onboarding basically every big player in this travesty: Goldman Sachs, JPMorgan Chase, Bank of America, Morgan Stanley, and Citigroup. In fact, I suspect that this cabal of villains are the ones testing the x{N} float idea that has everyone up in arms - they see a golden plan here to smash the window and hit the road before the sirens go off.

And that's the problem with invasives. They don't participate in the system at hand; they monkey with the fundamentals. Compete with other plants? Nah, I poison all the soil where I grow. Not enough substrate? Kill all the vertebrates with symbiont viruses so the world turns into Fungusland. And even if the retail investor wins, that means a massive flight from NASDAQ, which even I have been double checking in my tiny little pool of casino winnings. But triggering a flight from every index everywhere? That would not be bright. But, well, that's the other thing about invasives, right? They're not long term thinkers.

[1] I suppose this has become a political point too, hasn't it? Well, ride with me for a bit.

Pour yourself a drink, as I have a longish story that might be a useful metaphor.

Back in the day, there were more or less two consumer flight sims: MS Flight Simulator and XPlane. MSFS was and has always been the much prettier one, much easier to work with; xplane is kludgy, very old-school *NIX, and chonky in terms of resource usage. I was doing some work integrating flight systems data (FDAU/FDR outputs) into a cheaper flight re-creation tool, since the aircraft OEM's tool cost more than my annual salary. Hmm, actually, ten years of my salary.

So why use xplane at all, then?

The difference was that MSFS flight dynamics was driven from a model using table-based lookup that reproduced performance characteristics for a given airframe, whereas xplane (as you might be able to tell from the company name, Laminar Research) does fluid and gas simulation over the actual skin of the airframe, and then does the physics for the forces and masses and such.

I caught some flack for going with xplane: "Why not MSFS!? It's so much prettier!"

Unless the airframe is in a state that is near-equivalent with tabular lookup model, the actual flight is not going to be actually re-created. A plane in distress is very often in a boundary state- at best. OR you might be flying a plane that doesn't really have a model, like, say, a brand new planform (like the company was trying to develop). Without the aerodynamic fundamentals, the further away you get from the model represented by the tabular lookups, the greater the risk gets.

And how does this relate?

Those fundamentals- aerodynamic or mathematical or electrical- will be able to deal with a much broader range than models trained on existing data, regardless of whether or not they are LLMs or tabular lookups. If we rely on LLMs for aerodynamics, for chemistry, for electrical engineering, we are setting ourselves up for something like the 2008 Econopalypse except now it affects ALL the physical sciences; a Black Swan event that breaks reality.

I am genuinely worried we're working outselves into just such an event, where the fundamentals are all but forgotten, and a new phenomenon simply breaks the nuts and bolts of the applied sciences.

As for my xplane selection, it helped in other ways. Because often the FDR data is just plain wrong, but with xplane you could actually tell, because a control surface sticking out one way, while the flight instruments say another, lights up a "YOU GOT PROBLEMS" light in the cockpit as the aircraft inexplicably lurches to the right.

Ah, but it's no longer "inexplicably!"

Back in the Aughts a large number of home-schooling and educational reform organizations (leaning heavily on the Fundamentalist side of the Christian spectrum) had apparently determined that Set Theory originated in Socialist / Bisexual circles.

"A Beka Book" (now styled "Abeka") was not just the province of homeschoolers, but made its way into the educational and academic curricula in many higher learning institutions.

Unlike "modern math" theorists who believe mathematics is a creation of man and thus arbitrary and relative, A Beka Book teaches that the laws of mathematics are a creation of God and thus absolute, and that A Beka Book provides texts that are not burdened with modern ideas such as Set Theory.

It would have made a great deal less fuss if it didn't turn out that Abeka books were being bought in their thousands with tax dollars. I suppose this sort of thing would barely raise an eyebrow these days. I've been seeing far more avante garde ideas flowing forth in the public-funded wells of the former Confederacy of late.

The difference in response time - especially versus a regex running locally - is really difficult to express to someone who hasn't made much use of LLM calls in their natural language projects.

Someone said 10,000x slower, but that's off - in my experience - by about four orders of magnitude. And that's average, it gets much worse.

Now personally I would have maybe made a call through a "traditional" ML widget (scikit, numpy, spaCy, fastText, sentence-transformer, etc) but - for me anyway - that whole entire stack is Python. Transpiling all that to TS might be a maintenance burden I don't particularly feel like taking on. And on client facing code I'm not really sure it's even possible.

So, think of it as a business man: You don't really care if your customers swear or whatever, but you know that it'll generate bad headlines. So you gotta do something. Just like a door lock isn't designed for a master criminal, you don't need to design your filter for some master swearer; no, you design it good enough that it gives the impression that further tries are futile.

So yeah, you do what's less intesive to the cpu, but also, you do what's enough to prevent the majority of the concerns where a screenshot or log ends up showing blatant "unmoral" behavior.

This door lock doesn’t even work against people speaking French, so I think they could have tried a mite harder.

The up-side of the US market is (almost) everyone there speaks English. The down side is, that includes all the well-networked pearl-clutchers. Europe (including France) will have the same people, but it's harder to coordinate a network of pearl-clutching between some saying "Il faut protéger nos enfants de cette vulgarité!" and others saying "Η τηλεόραση και τα μέσα ενημέρωσης διαστρεβλώνουν τις αξίες μας!" even when they care about the exact same media.

For headlines, that's enough.

For what's behind the pearl-clutching, for what leads to the headlines pandering to them being worth writing, I agree with everyone else on this thread saying a simple word list is weird and probably pointless. Not just for false-negatives, but also false-positives: the Latin influence on many European languages leads to one very big politically-incorrect-in-the-USA problem for all the EU products talking about anything "black" (which includes what's printed on some brands of dark chocolate, one of which I saw in Hungary even though Hungarian isn't a Latin language but an Ugric language and only takes influences from Latin).

I just went through quite an adventure trying to translate back and forth from/to Hungarian to/from different languages to figure out which Hungarian word you meant, and arrived at the conclusion that this language is encrypted against human comprehension.

dark chocolate is "étcsokoládé" literally edible-chocolate in Hungarian.

i heared the throat-cleaning "Negró" candy (marketed by a chimney sweeper man with soot-covered face) was usually which hurt English-speaking people's self-deprecating sensitivities.

En toute honnêteté, je pense avoir dit "damn it" plus d'une fois à chat gépété avant de fermer la fenêtre dans un accès de rage

Nom de dieu de putain de bordel de merde de saloperie de connard d'enculé de ta mère.

It's like wiping your arse with silk.

Yea.. but.. in English only.

Fortunately I can swear pretty well in Spanish.

Only a native speaker can tell if you swear well in a foreign language.

And Claude can't tell at all.

That's like saying you can use a chisel for woodworking.

If it’s good enough it’s good enough, but just like there are many more options than going full blown LLM or just use a regex there are more options than transpile a massive Python stack to TS or give up.

> Someone said 10,000x slower, but that's off - in my experience - by about four orders of magnitude.

You do know that 10,000x _is_ four orders of magnitude, right? :-D

OP is saying that in their experience it is more like eight orders of magnitude

I guess I need reading glasses ... :-D

They're sending it to an llm anyway tho? Not sure why they wouldn't just add a sentiment field to the requested response shape.

because a regex on the client is free vs gpu compute is absolutely not.

BUT THEY'RE ALREADY RUNNING IT THROUGH THE LLM.

"If you're only going to be good at one thing, be good at lying, cuz if you're good at lying, you're good at everything"

- Do not remember where I heard this and desperately hope it did not come from my own corroded soul.

That hits. I've always been terrible at lying, so I just try my best not to. I'm so bad at it that I wear it on my face. I've been passed up for many opportunities in life that went to someone who was better at it.