@self_made_human's banner p

self_made_human

amaratvaṃ prāpnuhi, athavā yatamāno mṛtyum āpnuhi

16 followers   follows 0 users  
joined 2022 September 05 05:31:00 UTC

I'm a transhumanist doctor. In a better world, I wouldn't need to add that as a qualifier to plain old "doctor". It would be taken as granted for someone in the profession of saving lives.

At any rate, I intend to live forever or die trying. See you at Heat Death!

Friends:

A friend to everyone is a friend to no one.


				

User ID: 454

self_made_human

amaratvaṃ prāpnuhi, athavā yatamāno mṛtyum āpnuhi

16 followers   follows 0 users   joined 2022 September 05 05:31:00 UTC

					

I'm a transhumanist doctor. In a better world, I wouldn't need to add that as a qualifier to plain old "doctor". It would be taken as granted for someone in the profession of saving lives.

At any rate, I intend to live forever or die trying. See you at Heat Death!

Friends:

A friend to everyone is a friend to no one.


					

User ID: 454

I'm not particularly disputing the definition of "good" or "hard" times. I'm saying that they're highly reductive ways of describing something as large and complex as a whole civilization. Reductive doesn't necessarily mean useless, but there are better frameworks. It's not the point of contention, what we started off debating was what happens as a consequence of the goodness of the times (or lack thereof).

I'm not talking genetics. The nature of the times allows men with specific traits to rise to prominence and become characteristic of their civilization, whereas people of a different temperament are marginalized.

Given the population of the globe today, I think it's entirely possible that there are generals of the caliber of Napoleon or Caesar around, without an opportunity to demonstrate their tactical acumen. They've got better things to do, they're probably CEOs, or like an actual descendant of Napoleon, making millions at a hedge fund. They probably play HOI4 in their free time.

I think this is good. I think it's great! I'd rather they make billions as quants instead of launching invasions into Egypt. I think that talent is general and finds a way to manifest in both good and bad times alike, it just tends to be more... violent in the latter case. That does not mean that the times create the men, the men were always there, talents intact, they just went about applying it differently. I do not know if we actually disagree about that.

I think the game, as of Today AD, has about 15-30 hours of content in it before you're really scratching the bottom of the barrel. That is honestly not bad at all, compared to many other titles, but it's clearly unfinished if not quite barebones.

That's fine. I bought it in EA knowing that it's maybe half-done. I had fun, though I'm beginning to notice the burnout. The devs seem competent, I'm not too worried.

I wouldn't worry too much about the fixed characters. I'm part of the largest modding discord, and we've got mad motherfuckers making brand new SLs, units and laying the ground for entire total conversions (WW2 and 40k, because those settings are the obviously correct choices). All of that before we have an official modding SDK.

I intend to play till I feel like I've seen it all, and then I might either download one of the many balance mods or the few ones that currently add new content. Or I can wait a few months and come back to a lot more, including both official and modded content. If you're worried you'll pick the same SLs again and again, then there are already mods that randomize the starting pool, and it takes at least a dozen hours before you have the majority unlocked and can fall back into old habits.

I think this is orthogonal to my point. I find the Somali fraud situation to be somewhere between ludicrous, shameful and funny. Given that I have no skin in that game, I have the luxury of finding it more funny than infuriating.

But that's neither here nor there. The reason Somalis can leech off what seems like staggering amounts of wealth in absolute terms is because America is so wealthy that it's not particularly impactful in relative terms. They're a mosquito treating an elephant as a buffet, at some point they might get swatted for the inconvenience, but they aren't debilitating.

They stole a few billion dollars (I don't know the actual figure, I'm erring on the side of being generous to their talents at welfare fraud)? No big deal. The SF to LA rail line was originally budgeted at $35B and that ballooned to $128B with no projected date of completion. I don't know how much of that money was actually spent, so consider the F-35 program: started off at around 244B, and is now estimated to now have lifetime costs of 2 Trillion dollars. The humble James Webb went from a sticker price to a single billion to a cool ten of the same.

That is a lot of money. That is probably a lot of waste and graft. Yet it didn't collapse the country. A big strong healthy nation has proportionately bigger parasites, and can tolerate those parasites for far longer before it decides it needs that blood for better things. Before it is forced to tighten belts and actually check the accounts for the month.

So you have the equivalent of spare change from a not particularly important government program funding an entire civil war involving millions, without anyone taking action for years. The harm to the average Minnesotan, while real, was mostly invisible because how goddamn rich you lot are. Most countries would kill to have billions stolen from trillions, if that meant they had trillions in the first place.

What is that if not an example of degeneracy and decadence, where this nuclear superpower is getting humiliated by a shithole country?

I mean, what did the Somalians in Somalia do? They aren't on the autism payroll. Your ire should be directed at their compatriots in your country, who I am told mostly hold full citizenship. Bend them over an electric rail, I have no objections. But I think it is a mistake to claim that this is some sign of imminent collapse (assuming that is what you meant, you might have raised it as a single illustrative example).

For all your corruption, graft, poor budgeting and cost overruns, the US military is still the best in the world. The F-35s might have an eye-watering price, but they're good fighter jets sir. The telescope does telescope things which I read about on Quanta. Californian HSR? Now that I can't really defend. Anyway. Your hegemony is only now being challenged, with no clear victor, and only in a very constrained geopolitical theater. While a few carrier groups might be very unhappy in the Pacific should the missiles fly, the average American doesn't need to really worry unless some of those missiles are nuclear tipped. At which point we're all kinda screwed, if that's any consolation. I think it's comforting, in a grim way, that the only way to make life significantly worse for you in absolute terms is to potentially kill a quarter of the planet. Otherwise a decisive Chinese victory at Taiwan means you're looking at something like going from 2026 GDP figures to 2005. That's really not the end of the world. It's probably not even the end of America.

When Byzantium has its 600th civil war and loses provinces to Bulgaria or the Arabs, that's a real defeat for the Byzantines. When Korean court intrigues result in them letting their army rot and constantly imprisoning their best generals, that's genuine military failure. Letting other nations infest and parasitize your politics is just as bad as being humiliated on the battlefield, in so far as the results are the same. Payment of tribute is an ancient custom of defeated nations.

Humiliating? Yes. Annoying? Yes. Bad optics? Yes. Lethal? No. Somali fraud or minor ethnic bickering and identity politics won't bring down the American empire. It won't even make Minnesota flip away from being a Blue state. The Weak Men Hard Times theory focuses on civilizational trajectories as a whole and their failure modes, as does my critique, this isn't quite on the same level of concern.

I'll keep it short (and I'm the person with ADHD):

"Hard times create strong men" is the historical equivalent of saying that forests must need more fires, because the trees that survive fires are so strong.

That is the bailey, where many frolic happily. When pushed, they point out the fact that pyrophytes are a thing, and use that to justify a much stronger statement. Unfortunately, if you put that into practice, you end up deforesting the Amazon, ending up with nothing but eucalyptus groves in California, and make the local conservationists mad/sad for no good reason.

All times make men adapted to those times.

I wasn't aware that Lamarack (somehow) returned. Welcome back, there's a lot to catch you up on. Look, it's obvious to anyone that the typical human is "adapted" to their environment. The problem is with describing their environment with something as reductive as "hard times" and "good times". The nuance is important.

Not because they were born into "bad times" in prosperous Roma, but because their military service put them into "hard times" in the Roman military. What was it, twenty years of eating porridge, lugging around your whole kit and then some, and most likely getting clubbed to death in an ambush by some barbarian Gauls? That's not the decadent life - and if their organizational and technological advantages let them remain a little softer than their opponents, then well, they had those exact advantages to more than bridge the gap.

As you correctly note, hard and soft are relative. Yet the whole Fremen Mirage wasn't focused on the austerity and discipline of the kind of military drills standing armies performed, it focuses mainly on the civilization as a whole because that's a major focus for the Strong Men types. The typical Roman citizen had it good compared to the barbarians on the frontier. In the meantime, their soldiers, who lived rough out of both necessity and for training, beat the snot out of the tough barbarian folk for centuries; and only then were beaten by Romanized barbarians who adopted their tactics and equipment to a large degree.

Similarly, in a modern context, it doesn't matter if the average American has it easy compared to the North Koreans, the Chinese or the average Russian vatnik. Boot camp and ozempic are both helluva drugs. And the physical fitness of an individual soldier isn't particularly important or decisive compared to logistics, state capacity and industrial output. The US is at some risk from China because of the latter factors, not because its populace has it too easy to win a serious/existential fight.

The Strong Men advocates tend to want the Hard Times for everyone, no matter how pointless that might be.

All I'm saying is this - cushyness softens, and hardship hardens, and that much should be beyond debate.

Neither I nor Devereaux dispute that, as far as I'm aware. It simply doesn't matter very much on the scale of civilizations, and it's piss-poor at predicting anything except in hindsight.

Paintball is even more divorced from my interests I'm afraid, but thank you for the suggestions!

Obviously bad times create hard men. Obviously good times create soft men.

This not at all obvious to me. Do good times not make hard men? Did Caesar not rise from a Roman society that was doing pretty damn well for itself, all things considered? Do bad times not breed no end of weak men who make things worse or limp along maintaining the status quo? Do Strong Men not defeat other Strong Men?

It's mostly selection and availability bias, as well as (unconsciously) motivated stopping. You only count the phase shifts and ignore the status quo, even though that's just as important. You should read the original ACOUP series, if you haven't already. It's great stuff.

The first question you have to ask about any cyclical theory of history is: what breaks the cycle?

If hard times reliably create strong men and good times reliably create weak men, you would expect history to look like a perfect sine wave. Rome gets hard, Rome gets strong, Rome conquers everyone, Rome gets soft, Rome falls, barbarians have hard times, barbarians get strong, barbarians become the new Rome, repeat until the sun explodes. Clean. Elegant. A little depressing, but hey, at least it has narrative structure.

But this is not what history looks like. History looks like someone handed a child a pen and said "draw a graph of everything." Some civilizations collapse and stay collapsed. Some have good times that somehow produce further good times. The Mongols had extraordinarily hard times for generations and produced a brutal conquering empire that then also had hard times and collapsed without producing anything the theory would recognize as "good times" at all. The Dutch Republic had comfortable mercantile prosperity for most of its peak period and somehow found the strength to fight Spain, England, and France more or less simultaneously. The theory predicts that comfortable merchants should have produced feeble grandchildren who surrendered immediately. Instead, the comfortable merchants were the people fighting Spain.

The adherent of the theory can always reply that I'm missing the true scale at which the cycle operates, or the true definition of "strong," or the true timing of when good times become corrupting. This is the theory's second line of defense: it is unfalsifiable in a way that wears the costume of historical wisdom. If someone was strong, they must have had hard times, and if they had hard times, they must have become strong. The apparent exceptions are just cycles we haven't identified yet. The theory is so elastic it can absorb any data point, which should cause us to wonder whether it is a theory about history or simply a template we are laying over history after the fact, the way you can find a face in any cloud if you want to. Hardship made the Mongols, someone says. Well, the Inuit have it even worse but they haven't overrun civilization yet. What's that? A lack of pasture, a nomadic lifestyle in a marginal climate that buffers them from more centralized states? Horses? Wait, where does the moral character come into the picture again?

But let's say the theory is directionally true. Let's say, all else equal, some adversity produces some useful capacities in some people. This seems plausible. People who struggle with things sometimes learn from the struggle. There's probably a study.

The theory's problem is that it treats "strong" and "weak" as unitary properties that civilizations have or lack, when they are in fact thousands of orthogonal variables pointing in different directions! A generation raised in agricultural famine might become extraordinarily good at food preservation and resource management. They will almost certainly be shorter and have worse teeth. They will likely have high rates of anxiety, intergenerational trauma, and various diseases caused by malnutrition. They will maybe, if the theory is right, have a certain toughness of spirit. That's a strong maybe and a big if.

Are they "strong men"? By what metric? Note that half the reason I'm annoyed by FC's thread is that he handwaves these definitional problems away. Compared to the generation that grows up with antibiotics and universities and compulsory education, they will be worse at almost every measurable thing except some vague spiritual fortitude that we mostly admire in retrospect because we've romanticized it. The people on X with Greek statue profile pictures romanticize it, but they're usually idiots. FC probably romanticizes it too, for what I can only hope are better reasons. I just do not see them.

Consider: the populations that by any historical measure have had the hardest times in the last two centuries have not subsequently had the best outcomes. The populations that have had relatively good times, with strong institutions and widespread education and reliable food and medicine, have been the ones generating technological progress and stable governance. "Hard times" in the form of poverty, war, famine, and institutional collapse tend to produce more poverty, war, famine, and institutional collapse. The sociologists call this "path dependence." The supplement-seller calls this "weakness," but only when it's happening somewhere he has already decided was weak. You bet your ass those same Greek statues are going to shill supplements when they can get away with it.

If good times create weakness, and you want strength, then the policy implication is either that you should create hard times, or that hard times are secretly fine, or both. You occasionally hear this stated openly. A certain kind of person says that softies and their participation trophies are producing a generation that won't be able to handle difficulty, and the proof is that they can't handle difficulty, which we know because they are soft, which we know because they have participation trophies.

But if you actually believe this, consistently, you should be in favor of allowing preventable suffering to occur, because it will produce strong men who will produce good times. You should oppose medicine that reduces childhood hardship, because hardship builds character. You should be suspicious of social safety nets, not just for economic reasons but for civilizational-strength reasons, because cushioning people from consequences will produce people who cannot handle the world.

What it actually produces, reliably, in the present tense, is a feeling: that your discomfort is cosmically significant, that comfortable people are building toward their own doom, and that you are the hero of the next cycle. This feeling is available to everyone simultaneously, which should tell you something.

I suspect that this is the actual psychological drive behind people advocating for such a prediction-free Theory of Everything. Unfortunately, it's closer to astrology for Manly Men than it is anything remotely actionable or interesting. Yeah, empires collapsed because of:

  • Governance failure or
  • Someone stronger whacking them over the head and taking their toys

No shit. There's a bunch of other stuff, like plagues, ecological collapse, famine or a supervolcano going off in Indonesia. What does that theory offer in terms of actionable insights? Nothing. You can look at a switch flicking between two binary states based on a gazillion different factors, and make up all the stories you want about some grand cycle. But you can't predict when it'll flip or how long each state stays stable, and when you actually try to analyze the past with historical rigor, you'll find that a whole bunch of other things matter more than the general "hardness of the times". There's no Napoleon coming out of Somalia because the country sucks. The US is not doomed to collapse because the country has become a collective of godless hedonist commies who don't salute the flag enough and don't beat their children.

And often enough, the poor but hardened barbarians did overthrow the decadent and softened empire. Persia, Rome, Byzantium, Granada, China. And equally obviously, these are relative strengths and weaknesses, two factors among many in any contest, and on their own they do not suffice to predict the course of history. This hardened/decadent distinction matters in near-peer conflicts; obviously the discrepancy can be overcome by greater economic, technological or military advantages. It all just seems...obvious, and simple, and I don't get how such a lengthy discussion about the fundamental truth or falsehood of it can take place.

Once again, I strongly suggest reading the original blog. It covers at least the Roman case in exhaustive detail, and notes most of the others. You say "often enough", the Strong Men say all the time, and Devereaux shows that the big cushy empires win an overwhelming amount of times, until they don't.

(It's not like I want to be having this argument, I've just been nerd-sniped by people Being Wrong On The Internet. Happens to the best of us.)

MENACE is not XCOM, but I'd say that a good fraction of XCOM players will enjoy MENACE.

Most skills transfer, but the scale expands. You're going from micromanaging individual soldiers in a squad to controlling a whole mechanized platoon, but each squad is effectively just a character. I honestly prefer MENACE over XCOM (or the final finished version in my head, it's still a good game rn), since you get a lot more toys and playstyle variety that isn't just knowing which special abilities to fire off when. There's this YouTuber called Perun who usually discusses military strategy and geopolitics IRL, and he's doing a play through where he applies said tactics in the game and finds that they transfer over pretty well. What more can I ask for?

A Strong Argument Against Weak Men

I wrote all of this as a reply to @FCfromSSC 's defense against my critique of his critique of Devereaux’s critique of the "Weak Men, Hard Times"/"Fremen Mirage" thesis (try and keep that straight). It got out of hand. I put a lot of effort into it. So top-level post it is, even though it's primarily directed at him. You likely want to follow the link first.


Fair enough, I asked for concrete arguments and you've made an attempt at providing them. Forgive me for saying this, but I find it rather light on detail.

I'll start with "inevitable," because that word is quietly doing the work of three different arguments while dressed as one. You have two options. Either you mean it literally, in which case the counterexamples are immediate and fatal (and I'm going to share them), or you mean something like "tends to produce," in which case welcome to the world of probabilistic claims, where you owe us a measurable prediction: increases the probability of what outcome, by how much, measured how. Right now "inevitable" floats above the empirical fray like a philosophical dirigible, too vague to be shot down, too vague to be useful.

It also helps to be explicit about the three claims that keep getting quietly bundled together. As I see them:

  1. Morale and cohesion matter militarily. Correct, well-attested, Clausewitz built it into the load-bearing structure of his framework under “moral forces,” nobody serious disputes this. It's acknowledged in official US military doctrine.

  2. Culture can shape military performance in meaningful ways. Also largely agreed, also case-specific, also entirely compatible with Devereaux’s thesis.

  3. Hardship reliably produces superior fighters and drives predictable civilizational cycles.

That third claim is the contested claim. That is the meme. Conceding the first two does not move the needle on the third by even one tick, and yet your argument keeps treating “morale matters” as if it's a portal that gets you to “hard times generate conquerors.”

(How exactly do you rule out the possibility that Good Times also generate Strong Men at similar base rates, but those men don't get the opportunity to do much raping, pillaging and conquering because of the restrictions put upon them by a functional society? Or because they're quite happy with how things are going?)

I will admit to a certain sympathy for the "Hard Times" thesis, one that was beaten out of me by reading actual history. It appeals to a very specific, conservation of energy intuition we all seem to share share. It feels like physics applied to sociology. If you put a human being under immense pressure, they should turn into a diamond. If you put a human being in a warm bath with a steady supply of peeled grapes, they should turn into soup. The universe seems requires a trade off between comfort and capability, so it makes sense that a civilization maximizing comfort would bleed capability.

But if you're going to accuse Devereaux of baking absurdity into his arguments, it's desirable to at least check whether the hard-times model actually predicts anything.

If hard times make strong men inevitable, then the places experiencing the hardest times should be the places producing the strongest men, and by the meme’s own implication, the most formidable military actors.

Hard times are not hard to find. Somalia has had a shit time since 1991. North Korea is so chronically deprived that the population is physically shorter than their southern neighbors. Haiti is a recurring disaster. Eastern Europe is literally eating trench warfare again.

Does the hard-times generator produce strong men in the way the meme implies?

It certainly produces tough men. I do not doubt that a survivor of the Somali civil war has a tolerance for pain, a resilience to trauma, and cockroach-adaptability that would shatter the average San Francisco software engineer in 48 hours. If “strong” means “hard to kill individually,” then yes, hard times can produce hard shells.

But the Fremen Mirage is not about individual survival. It's about civilizational competence and military victory. The meme posits that these hard men will inevitably conquer the weak men of the good times.

Yet Somali pirates do not win fights against destroyers, probably not even against a PT boat. Beef-feed American boys grow up playing Halo with an Xbox controller and have no issues using another Xbox controller to turn them into a halo of fine mist. North Korea is unlikely to win any military victories in the foreseeable future (maybe if all the South Koreans end up in retirement homes, but isn't the grind and rigidity of the latter culture itself a form of "strength"? They've got mandatory conscription, how hard is that?). Ukraine is a meatgrinder that turns sunk cost into well-fertilized and mined sunflower fields.

And here is where the model fails to predict the past two thousand years of warfare better than chance (and in fact, worse than chance), and fails even harder in the modern era.

When the soft, decadent, weak United States military engages with forces bred in the hardest of times, the exchange ratio is often lopsided to a degree that defies belief (and draws accusation of war crimes). The hard men die in droves, often without ever seeing the "weak men" who killed them from an air-conditioned control room. The North Korean soldier, bred in the hardest of times, is not a super-soldier. He's a poor bastard with intestinal parasites, brittle bones, and a rifle from the 1960s. In power-projection terms he is not strong.

Look at Venezuela. I strongly suspect that you are rather focused on American affairs, so let's look closer:

Venezuela got clapped. No contest, no lube. It's a historical humiliation of near unprecedented chutzpah, and a military masterstroke. Putin probably believes he's a hard man ruling hard men, and he wishes his 3 day special operation was a tenth as effective. Grit ain't nothing compared to total air superiority.

Going back, there's an even more glaring example in WW2:

The Japanese were the archetypal hard men.

If any modern society ran the Fremen playbook sincerely, it was Japan. Institutionalized martial culture, explicit civilizational contempt for comfort and luxury, bushido as doctrine, genuinely formidable individual bravery. No proxies. The hardness was real, rigorous, and deeply embedded.

Their men expected no quarter, and gave none. Their pilots showed a willingness to convert their planes and bodies into guided missiles without a parachute. They dug in like Alabama ticks and fought like hell till their blood pooled in the mud.

And yet, once again, they got clapped. By the richest and most "decadent" nation around. While the Japs ate thin gruel and Germans were making ersatz-coffee, the US had entire ships dedicated to ice cream for its troops. By virtually any Fremen metric, the United States was the most materially comfortable, consumerist, pluralist fighting force in recorded history. The kind of civilization the meme would predict to crumble at first contact with anything properly hard. Dare I say, the most decadent?

Huh. Funny how that works.

Before you bring up Vietnam or Afghanistan, I will keep it simple: in the past century, the US has not lost a single war that mattered. The safety and comfort of the homeland, the F-35 flybies over college football stadiums? None of that was ever in jeopardy. Your wars were toy-wars, usually fought with one side following gentleman's rules (you). You pissed away the GDP of entire continents on dirt that barely grows poppies and didn't even get dehydrated in the process. Even victory would have, in hindsight, been largely inconsequential.

It is far from obvious that any other force can overturn this, if your closest competitor, China, has any hope of a victory in the medium-term, it'll be on the backs of the industrial capacity to produce a bazillion missiles and drones, not the grit of their soldiers or the genius of their generals. And yet they don't even dream of landing an expeditionary force on the US mainland. A potential victory (emphasis on potential) hinges on industrial capacity, not moral character.

Looking at your claims about “free energy,” which I think has the causality backwards: Good times are not a reduction in the energy available to strong men. Good times are the surplus that makes strength scalable.

Good times, meaning surplus calories, high GDP, technological capacity, lack of immediate existential threat, allow a society to take a percentage of its population and tell them: you don't need to farm. You don't need to hunt. You don't need to worry about where your next meal is coming from. You will spend twelve hours a day lifting, drilling, training, learning complex machinery, building unit cohesion, practicing logistics, rehearsing doctrine.

That is how you build a Roman legionary. That is how you build a US Marine. The hard man from the hills has to spend his energy surviving. He is a part-time warrior because he is a full-time survivor. The soft man from the empire is a full-time professional killer subsidized by the very economic complexity you are sneering at.

But in actual fact it is obvious that Morale and Morals/virtue/character, are pretty clearly linked, and that even central examples of Moral Purity in the sense he frames it have in fact been used historically to build winning armies.

Your Cromwell example does not prove as much as you want it to. Yes, moral conviction and ideological zeal can be force multipliers. Asabiyyah is real. Morale matters. Nobody disputes this. I'm not disputing this, Devereaux isn't disputing this.

But Cromwell’s army was effective not because they were ragtag survivors of collapse. They were effective because they were paid, equipped, drilled, and organized according to cutting-edge military science. The New Model Army was a product of high coordination inside a relatively wealthy society. Zeal multiplies competence. It doesn't replace it.

Now, decadence.

You accuse Devereaux of refusing to engage with the semantic content of decadence. That is not quite right. Devereaux engages a specific operationalization: luxury softens bodies, literacy softens minds, complexity demilitarizes populations. Your complaint seems to be that this is too crude and that a richer version is hiding behind it.

Fine. Name it.

I'm serious. What is your definition of decadence, and is it of any use for predicting the course of history?

As I say (because it's true), "all models are false, some models are useful". The better models are differentially useful. They cut reality at the joints and serve as useful compression of complex systems, and more importantly, predict future events. At the very fucking least, they should describe history.

Back to Rome (it's my Roman Empire):

Decadence, as it is usually deployed, is almost always defined tautologically:

  1. Rome fell.

  2. Therefore Rome was decadent.

  3. How do we know Rome was decadent?

  4. Because it fell.

If Rome had repelled the Goths (and then were wiped out by a convenient asteroid so I don't have to write a full alternate history), as they had repelled Germanic tribes for centuries prior, often while being just as wealthy and just as bath-loving and just as bookish, we would currently be writing essays about how civilized discipline triumphed over disordered barbarism.

As I said in my original reply, an empire needs an unbounded number of victories to survive. In the worst case, it only needs a single defeat to crumble.

This is also why the unfalsifiability problem runs like a hairline fracture through your whole framework (or at least the HTHMWTWM theory as popularly understood, which is what Devereaux is dismantling).

Rich state wins? Material advantages masked decadence temporarily. Poor state wins? Fremen thesis confirmed. Rich state loses? Decadence, obviously. Poor state loses? Not hard enough, or the material gap was too vast. Every outcome is accommodable. Every counterexample has an epicycle preloaded. A theory structured this way is not doing historical work. It's a just-so story that could be replaced by a well-decorated rock. At least I can kick a rock.

It's the equivalent of saying "what doesn't kill you makes you stronger." Sounds great, doesn't help when you transect your spinal cord in a car crash, or ruin your knees after too much PT. A claim being "inspiring" is very different from it being true.

The more I squint, the more this becomes the Just World Fallacy wearing combat boots. It suggests that suffering has a purpose (to make us strong) and that comfort is a sin (that makes us weak).

The universe is often far more cruel than that. Sometimes hard times just break you. Sometimes suffering is just suffering. Sometimes the "Hard Man" fights the "Soft Man," and the Soft Man presses a button and deletes the Hard Man from existence, because the Soft Man spent his "Good Times" studying physics and engineering instead of learning how to endure hunger.

What's probably true: character matters, organizational culture matters, genuine commitment to a cause produces measurably better outcomes under certain conditions, and a civilization that cannot articulate what it's fighting for faces real disadvantages at the margin. I'd go further and say that the Athenians, the decadent, democratic, philosophizing, play-writing naval-gazers, are actually the better case study than the Spartans here. They bounced back from defeat after defeat because their "Good Times" culture was dynamic enough to reinvent itself. The Spartans, the hardest men in Greece, ossified and collapsed precisely because they were so committed to their own hardness that they couldn't adapt to a changing world. They were too busy attacking their own allies, diddling boys and randomly killing the slaves that tilled their fields. The very quality the meme valorizes became their brittleness.


Miscellaneous thoughts, because I'm tired after studying all night:

It is helpful, and I would say good practice, to operationalize and define terms, especially those in contention. Devereaux does this well, he lays out a specific argument commonly found in the wild, and musters an offensive on every front. You do not do this. You haven't defined terms, at best you wave at your (implicit, unknown to us, or at least me) definition of "decadence", of "weakness" and "strength". A proper debate requires that both sides leave enough rope to hang each other with, and that the loser is sporting enough not to offer shoe-laces instead.

Is self-sufficiency and flexibility a bad thing? Is there such a thing as overspecialization or excessive complexity as legible cultural problems? Are the average men in societies, populations, or tribes more or less capable of becoming soldiers en masse, due to the culture they've been shaped by? Does this problem show up even from the perspective of men who appear to, in fact, be quite strong? Fuck that noise, questions are for dweebs! Let's round it to "all men in the society are warriors", that sounds way less complicated.

You are making up people to be mad at. At the very least, you are putting words in Devereaux's mouth, and as far as I can see, he never said what you claim he says, nor does he imply what you think he implies. Please, a quote where he even implies that asking such questions is "for dweebs" would go a long way. The man is a military historian focusing on classical history, I'm sure he's on the side of the dweebs and nerds.

To reiterate: he lays out an argument. He shows that the argument is shite. He is attacking a strawman, but unfortunately, the majority of people making that argument have straw for brains. The actual Motte doesn't need defending because it's nigh impregnable. You can't attack a critique of Lizardman conspiracy theorists on the grounds that he hasn't addressed the steelman version: humans and dinosaurs have a common ancestors, at some point before the split between synapsids and diapsids. Nobody disputes that. Nobody cares. A non-negligible number of people adhere to the batshit crazy version.

Devereaux is careful to avoid cherry picking evidence. His analysis spans roughly the entire history of the united Roman Empire, and then its Western successor. That is hardly cherry picked, both because Rome is usually held up as the example of decadence killing a civilization and because that's literally his field of scholarship. That is the breadth and depth of scholarship to aspire to, when discussing something as complicated as grand patterns in world history and the rise and fall of empires. I don't expect you to do that much, but come on.

More recently, we have the truism that "no one is going to fight a war on behalf of an economic zone." While we haven't tested this principle hard yet, I know which way I'd bet.

I recall a lot of blood spilled in the Age of Colonization, over what can loosely be termed as economic zones. It is unclear whether the US would, say, bat for Taiwan in a hot war with China. But it is unclear, and even then, a single sparrow does not a summer make. Devereaux throws a net over the whole swarm, you don't. I'm not aware of a rebuttal that does. Besides, I think China would have words for someone invading Shenzhen or Hainan, and what are those but economic zones? Rather special ones even.

By contrast, it seems to me that Devereaux aims to convince his readers that military affairs are largely deterministic, with a layer of luck on top. Therefore, empires are born because they got a streak of good RNG hits, and Empires die because they got a streak of bad RNG hits, and human decisions are not really terribly decisive either way.

And this is incorrect because? I do not see Devereaux putting numbers on the relative importance of "RNG" versus determinism. Rome might have been very different if Justinian didn't have to face one of the worst plagues in human history, and if Belisarius's wife hadn't been such a hoe. What might the world look like if Barbarossa took swimming lessons? He notes that macroscopic factors like population size, wealth and military metrics matter, he does not claim that nothing else does. I do not see why you consider it an excuse to insert your own interpretation and then get mad at him. My understanding is that he sees those metrics as important, often decisive, which is not the same as what you seem to believe. And randomness only adds variance. Devereaux believes systems (logistics, tax bases, agriculture) determine the probability of victory. It isn't just "luck"; it is that a "decadent" society with a 90% win rate due to logistics will eventually crush a "hard" society that relies on a hero rolling a natural 20 every time. (Numbers my own, and made up).

Finally, you accuse him of being a propagandist. On what basis? What basis in fact? Do your facts weigh up to his? He's got plenty. I've got plenty. Propaganda can be both propaganda and true (I do not agree that his approach to the Fremen Mirage constitutes propaganda). You need to demonstrate that it is both misleading and factually incorrect.

(Posters advising people to wash their hands are propaganda posters, but you're better off doing as they say)

If you insist on treating your arguments as soldiers facing off against Devereaux while declaring him an enemy propagandist, you had better hope your arguments are good soldiers. I remain unconvinced that they are.

I think I have put enough effort into concrete disagreement to risk slightly uncharitable psychoanalysis: you disagree with his thesis because it goes against your values, and this is more the cause of your discontent and disdain than its actual bearing on the truth. I suspect you are deeply unhappy with the status quo, and see bloodshed, strife and suffering as necessary for a phase transition to your ideal sociopolitical system, and that it helps to imagine that that suffering is inherently or terminally good in of itself. I hope to be proven wrong.

TLDR:

If decadence is defined as the cultural conditions that cause military decline, then the Fremen thesis becomes true by definition and empty of predictive content. For it to be a real historical claim, decadence needs to be identifiable independently of military outcomes, and then shown to strongly correlate with them. I do not believe this has been demonstrated, and I strongly doubt it can be demonstrated (because it's not true).

I've had my eye on it, but my milsim buddies who do play it live rather far from me :(

My impression is that it's less developed as a hobby in the UK, as compared to the States. Sure, it has a following, but it's quite limited.

I'm open to it, but it's not quite what I'm looking for:

  • I do enjoy a bit of the old LARP, but I actually am quite happy doing it in video games. There is a non-obvious tradeoff here: video games explore some aspects of the modern military experience better than most things short of an actual military training exercise can manage. It would be pretty hard to bring a tank or helicopter into play in airsoft, or the experience of being under sustained artillery bombardment while your buddy bleeds out next to you. Video games scratch that itch pretty well for me. Airsoft is great for other things, but I'm not that set on hiding in bushes or kicking down doors in real life.
  • I like actual guns. I'd like to own a Gucci d out AR-15 and a AR-10/DMR. I'd like to take it down to the range or go shoot tin cans with friends. It matters to me that the gun is an actual gun that shoots real bullets, even if I don't particularly want to use it in anger. I could, like many people, spend a ton on accurate parts for an airsoft kit, but it wouldn't make me happy. For that much money, I want the real deal.

Otherwise, I don't think it's totally impossible to shoot a shotgun in the UK, is it? That's fun. One of my four guns is a pump shotgun. You'll have to visit a Mottizen in a freer state when you eventually visit America. You like pistols better, or rifles? You like milsurp or modern AR bullshit?

I must (with slight shame) admit that I like "cool" guns. Tacticool, with all the drip. I could, in theory, get a gun license here (it's an enormous PITA) but I don't want to shoot a hunting shotgun or some ancient bolt-action rifle. I want a laser with a pad, I want an LPVO, I want to fuss over the perfect foregrip even if I'll end up using a c-clamp. Ideally, I want to go to Vegas and shoot an M60 while doing my best Rambo impression (before my limp arms or the rangemaster get the better of me).

I was actually planning on getting my hands on an actual firearm in the US. A dear friend of mine was getting married in Texas, which would have been the perfect opportunity. Sadly the visa officers at the London embassy disagreed with my ambitions. I still do plan to visit the States when I have a decent excuse, and you bet that going to the range is very high on the bucket list! I'm going to shoot hogs with a Barrett from a helicopter at some point in my life, or I've never really lived.

I'll keep an eye open for airsoft events in my end of Scotland, and if I can rent a kit for not too much money, why not? Thanks for hearing me dream aloud, haha.

I only played it for ten minutes over ten years ago, so my personal experience is limited. I have tried to watch YouTubers (that I otherwise enjoy) play it, and I'm somewhere between confused, bored and unsure of the appeal. I'm not really a roguelike person, but I do play them on occasion. Of course, not being familiar with the game mechanics does limit the enjoyment of observation, but I can name plenty of games I wouldn't play myself where I still greatly enjoy play throughs. Isaac just isn't one of them I'm afraid.

I promise you that it's more than "a couple of guys". My Twitter is schizophrenic enough that I find myself looking at their posts more often than I consider ideal. Oh well, it's good ethnography if nothing else. They're thriving out there, posting inspirational quotes and bad history takes when they aren't recovering from parasitosis after the consumption of raw meat.

You'll find plenty of examples on this very forum, if you use the search functionality.

burning down the bailey, and claiming he's destroyed the motte.

The thing is, there is no Motte! Or rather, there is no interesting Motte. Empires rising and falling because {many reasons} is the boring yet correct explanation.

I do not blame Devereaux for targeting the version found in the wild, the meme tuned for maximum virulence. If there is a counter-thesis of comparable scholarship arguing in favor, well, I haven't found it yet. Sometimes, one side of a debate really does have a disproportionate number of idiots alongside little factual merit, see the Flat Earth community for an existence proof.

Deleuze

They're preparing a padded cell for me already, I've booked one with good wifi reception.

Appreciate the recs! After a bit of searching, I've discovered that the genre for that song is best described as "martial industrial". I don't know how accurate that is, but the playlists I found are boppers. If you like that grungy industrial orchestra with ominous chanting vibe, I'd recommend the entire OST for Mechanicus (a Warhammer tactics game).

I do mean to check out Mewgenics, the reviews are overwhelmingly positive. I didn't particularly like the Binding of Isaac, but it seems interesting enough to give it a try once I've exhausted all the content in Menace and need a fix.

"Hard Times Make Strong Men" exists as a thesis because we can directly observe that rich, powerful states often actually do decline, that states are defeated by non-states, and rich states are beat by poor states. Not all the time, not as the expected result, but often enough that very clearly wealth, population, or whatever other technocratically legible KPI one prefers are not deterministic.

Notice the caveats, that you wrote yourself. Notice that ACOUP argues against the typical way in which the "hard men" theory is presented. The dudes with Greek statue profile pictures aren't doing nuanced historiography, they actually want to camp out in the bailey. They want to claim that moral rigidity/orthodoxy, avoidance of "luxuries" and a focus on martial prowess uber alles is an easy short-cut to civilizational dominance. Setting the bailey on fire, as Devereaux does, means there's little Motte left to defend.

His points are, broadly:

  • "Hard men" are poorly defined as a class. In your first quote, he notes (correctly) that people cherrypick whatever aspects of a civilization make for the most rhetorically convincing argument.
  • The factors mentioned above are far from decisive or notable in understanding the decline of empires.

Hell, I'll give up on summarizing it, and focus on his own definition:

what do I mean by the Fremen Mirage? I think the core tenants run thusly:

  • First: That people from less settled or ‘civilized’ societies – what we would have once called ‘barbarians,’ but will, for the sake of simplicity and clarity generally call here the Fremen after the example of the trope found in Dune – are made inherently ‘tougher’ (or more morally ‘pure’ – we’ll come back to this in the third post) by those hard conditions.

  • Second: Consequently, people from these less settled societies are better fighters and more militarily capable than their settled or wealthier neighboring societies.

  • Third: That, consequently the poorer, harder people will inevitably overrun and subjugate the richer, more prosperous communities around them.

  • Fourth: That the consequence of the previous three things is that history supposedly could be understood as an inevitable cycle, where peoples in harder, poorer places conquer their richer neighbors, become rich and ‘decadent’ themselves, lose their fighting capacity and are conquered in their turn. Or, as the common meme puts it:

“Hard times create strong men. Strong men create good times. Good times create weak men. And weak men create hard times” (The quote is originally from G. Michael Hopf, a novelist and, perhaps conspicuously, not a historian; one also wonders what the women are doing during all of this, but I have to admit, were I they, I would be glad to be left out too).

That is what he's arguing against. That is actually how people use the phrase.

More importantly, I did not get the impression that:

This is a fascinating question, but Devereaux appears mainly interested in cauterizing such interest in anyone he can, and is enthusiastically willing to employ the argumentative dark arts in doing so.

And I've read the whole series. Devereaux does excellent scholarship, studies a variety of different cases, and provides citations. Rome is typically used as an example in "favor" of the HTWM theory, and luckily for us, he's a classical historian. He covers several hundred years of Roman history:

He uses shipwreck archaeology, ice core analysis of atmospheric lead, and epigraphic evidence to track Roman wealth over time. He outlines a clear pattern (supported, as far as I know, by other period experts): a period of rising affluence in Italy in the Middle and Late Republic, followed by a long period of prosperity in the early empire, disrupted by the Crisis of the Third Century, with another period of economic stability (but at a lower level of prosperity) in the fourth century.

Guess what? : no part of Roman military 'decline' follows this patternz. Rome's military power was greatest when it was getting wealthier and more urban was growing, and began to decline in a period where the empire seems to have become somewhat more rural and poorer.

Even better: Romans were complaining about decadence the entire time! Polybius, Cato the Elder, Sallust, Tacitus, all moaned and bitched about declining Roman virtue.

(Sallust wrote about decadence two centuries before the peak of Roman power under the Nerva-Antonine emperors.)

Over eight centuries, Rome fights dozens of "Fremen" peoples. The Samnites fought three wars with Rome all of which were tough and in many cases the Romans lost battles and struggled, but Rome ended up winning each war.

The Gauls in Cisalpine Gaul? Crushed at Telamon, then systematically smashed one by one after Hannibal's defeat. Caesar had a great fucking time up there. The Celtiberians in Spain? Three wars, all Roman victories. The Germanic Cimbri (Marius stomped) and Teutones? Effectively annihilated. The Helvetii? Near-total genocide. All we've got left of them is a font, the poor bastards.

The successful "Fremen" invaders at the end of the Western Empire make a relatively short list: Senones, Visigoths, Ostrogoths, Vandals, Franks, Angles/Saxons/Jutes, Alamanni. That's seven successes against dozens of failures. Most of them were Romanized too!

If it's not obvious, to survive, an empire must continue winning indefinitely. To lose, it can take as little as one war. It's the KDR that counts dawg, if you live long enough, common cold, cancer or a car will end any winning streak, and we've got several thousands of years of history to measure the life expectancy of empire.

The quotes you picked do not demonstrate Devereaux “baking in absurdity.” They are him accurately describing the common version of the claim, including the decadence framing, then openly mocking one specific implication (virtue leads to battlefield performance).

He then does the opposite of propaganda: he tells you to watch out for selection effects and to ask about win-rates, not vibes. And he summarizes his conclusions in a way that is falsifiable: if the “Fremen” were systematically superior, you should see them winning more often, not losing more often than they won.

That destroys the bailey and salts the fields. If there's a more sophisticated version hiding in the Motte (one that merely says "states can decline for complex reasons including but not limited to overextension, internal political dysfunction, and occasional bad luck") then congratulations, you've described basic history. And that version doesn't need the "hard men" framing at all.

If you disagree with his central thesis, then I welcome actual arguments.

The link to the fanfic doesn't seem to load for me.

Huh. Apparently it's blocked by my ISP in India, but works with mobile data or a VPN? The fuck?

Actually reading the stuff, it does seem to be a marked improvement from the last set you posted, barring the fact that there's no actual NSFW in it. OAI is planning to launch an adult only model with relaxed guard rails, and I suspect Anthropic will follow suit at some point. The demand is there.

MENACE came out in Early Access on Steam, and I've been enjoying the hell out of it. Sure, it's got rough edges and needs some balance tweaks, but the devs are absolutely cooking.

That includes a banger OST. Just listen to this:

https://youtube.com/watch?v=Xt55wnTWt60

Hoorah Major. Every time I hear it play, I absolutely lock the fuck in. It's going in my gym playlist, I don't even know what genre this is, but I want more. Inject it into my veins with a fire-heated needle.

What are your thoughts on resting a pistol on the crook of your forearm (elbow bent into an L horizontally, gun resting on the elbow)? I've done it in VR because it looks cool, and kinda works to help with stability and arm fatigue, but I don't know how that translates to an actual gun. The Weaver stance doesn't really translate, because most VR controllers would physically intersect. Sigh, I wish I had a real gun, preferably three.

But if you could hold other factors approximately equal, I'd still bet my life on the guy whose' seen 10,000 cases and performed a procedure 8000 times over someone who is merely younger but with 1/3 the experience.

https://www.nature.com/articles/s41598-022-15275-7

The mortality in patients undergoing surgery by old-aged surgeons was 1.14 (1.02–1.28, p = 0.02) (I2 = 80%) compared to those by middle-aged surgeon. No significant differences were observed according to the surgeon’s age in the major morbidity and subgroup analyses. This meta-analysis indicated that surgeries performed by old-aged surgeons had a higher risk of postoperative mortality than those by middle-aged surgeons. Thus, it necessitates the introduction of a multidisciplinary approach to evaluate the performance of senior surgeons.

I don't think 14% is a big deal, there's already a great deal of heterogeneity in terms of surgical outcomes for all surgeons overall, but it does exist.

Yeah, I suspect that even if LLMs are a full standard deviation IQ higher than your average doctor, the massive disadvantage of only being able to reason from the data stream that the humans have intentionally supplied, and not go in and physically interact with the patient's body will hobble them in many cases. I also wonder if they are willing/able to notice when a patient is probably straight up lying to them.

While it's frustratingly hard to find actual sources on the average IQ of doctors, most claim an average of 120-130. IQ testing LLMs on human IQ tests like the Stanford-Binet or Weisler-IV is fraught, but I've seen figures around 130 from o3 onwards. If I had to wild-ass-guess, 130+ is a fair estimate for GPT 5.2T, and versions with enhanced reasoning budgets are making novel discoveries in mathematics, so...

(Did I say that IQ research in doctors is bad? Oh boy, just see what it's like for LLMs. There are papers still awaiting peer-review that use Claude 3.5 Sonnet. The field moves *fast".)

https://www.trackingai.org/home seems better than nothing and claims 140 IQ for GPT 5.2T on the public Mensa Norway test and 129 for the so-called offline version.

Note: The "Offline" IQ quiz is a test made by a Mensa member that has never been on the public internet, and is in no AI training data. Mensa Norway is a public online IQ test.

Make of that what you will.

I haven't specifically tested models on their ability to catch lies or inconsistencies, but I think they'd do okay, but probably worse than a decent human doctor. This is a moderate confidence claim, and could be ameliorated by giving them previous clinical records, but a video feed would be even better (doable today). I'm already zooted on stims and typing up massive comments instead of studying, or else I'd try it myself.

And yet, they're finding ways to hook the machine up to real world sensor data which should narrow that advantage in practice.

The Transformer architecture is a universal function approximator, LLMs are already multimodal despite the name, and in the worst case, they can ingest text instead of raw sensor data like humans often do. I don't look inside the SpO2 probe, I read the number.

And as you gestured at in your comment... you can very rapidly get second opinions by consulting other models. So now that brings us to the question of whether the combining the opinions of Claude AND Gemini AND ChatGpt would bring us even better results overall.

I've seen some evidence that diversity of models is good, for models of similar general competence. Even so, just putting the same info into another instance of the same model is highly effective, and I wouldn't yell at someone who did that.

I didn't do anything special. I started off with oral semaglutide at 14mg, instead of following the standard recommendation to titrate up from 7mg after 2 weeks (I'm lazy). No side effects worth mentioning for me, though I could barely finish a full meal for the first few days (I can hardly complain). Hell, no side effects at all.

My mom did experience the transient nausea and constipation for the first few weeks, so that recommendation exists for a reason. Caveat emptor.

Otherwise? It's pretty safe. If you're super scrupulous, you could get enzymes tested because of the (small) pancreatitis risk. Overall, it's a remarkably unfussy drug.

I'm a guncel, in the sense that I'd very much like to own a gun but have the unfortunate reality of living in jurisdictions where that's impossible or a PITA.

Shooting with a red dot on a rifle is a much smoother experience than shooting with iron sights on a pistol, I guess mainly because you have many more points of contact with the gun so it's easier to keep it steady. Even though it's not sighted in perfectly, I have never managed to get that many hits in at 50 yards, though it's been like, ten years since I ever tried it.

I noticed this quickly in VR. Pistols are counterintuitive in a manner that rifles aren't, when I try and use red dots with them, I realize that I'm always holding them too low by default, and that the sight picture is an absolute pain to maintain (H3VR, the VR gun sim).

Hell, rifles are so much better in general. Point shooting is so much easier when you have the gun shouldered (even when that's simulated) and the longer barrel gives your eyes and brain a much better picture of where the muzzle is pointing.

I doubt anyone has tested reaction speeds specifically for LOL, but it's well-established that there are significant differences in reaction speeds between the sexes, and I don't see why that wouldn't hold here.

https://psycnet.apa.org/record/2006-03906-007

There are significant sex differences, most notably for choice RT variability.

I don't have full access, but theh found males had faster reaction times at all ages, with the sex difference relatively stable across adulthood. Other research suggests a 15-30ms gap, which all but precludes women in the highest echelons of twitch-reaction games.

@omw_68

I tried getting GPT 5.2T to look for examples:

I went looking and I cannot find an older “canonical” page for that exact Empire State Building + 300-mile legs wording. The only clearly indexed hit I’m seeing is a very recent mention in a TheMotte thread (posted Feb 16, 2026).

Lol. Lmao. I suppose Google or Bing has very fast crawlers?

I would guess a doctor with 50 years of experience (and good enough recall to incorporate all that experience) can still make important insights in tough cases, that would elude an AI (for now).

As an aside, older is not better for doctors. It's a common enough belief, including inside the profession, but plenty of objective studies demonstrate that 30-40 year old clinicians are the best overall. At a certain point, cognitive inflexibility from old age, habit, and not keeping up with the latest updates can't be compensated for from experience alone.

(This doesn't mean older doctors are bad doctors, it just means they aren't the best anymore, all else being equal)

Taking this at absolute face value, I wonder if this is at least partially because the specialists will have observed/experienced various 'special cases' that aren't captured by the medical literature and thus aren't necessarily available in the training data.

I think there's a component for this, but if pushed I wouldn't say it's the biggest factor. An ophthalmologist hasn't studied cardiology since med school, they might remember the general details and interactions when it comes to the drugs they prescribe, but they're still not a cardiologist.

Gun to my head, I'd say that the human doctors who outperform LLMs are still smarter, if not as well read (they can't boast a near encyclopedic knowledge of all of medicine like any half-decent LLM can). IQ matters, and some doctors are just that smart, while having the unfair advantage of richer interaction with a human patient. Plus LLMs don't have the same "scaffolding" or affordances, they can't just look or lay hands on a patient (though they can ingest pictures, that's still an extra step). I suspect the difference diminishes to a large extent when the doctors are given the exact same information as an LLM, say some kind of case overview and lab tests + imaging. GPT-4 was scoring at the 95th percentile level in the USMLE, and these days medical benchmarks are simply not good enough to compare between them (official, graded benchmarks, I'm sure you can make a few more-ad-hoc ones if you really try, though by "you" I mean a competent physician).

It recognized it as a "famous puzzle" in its thinking trace. However, I suspect that the common version of the puzzle doesn't account for curvature. I tried looking for it, but didn't find anything, but similar variants (often seen in aptitude or IQ tests) implicitly assume a flat surface.

In fact, on double checking, the model knows that the classic form assumes a flat map. It specifically decides to answer it in more depth.

Why

The most common failure mode in LLM skeptics (and I don't mean to use that phrase to describe people who don't believe that LLMs are AGI, or that they have clear flaws) is to assume that all improvements come from intentional efforts by AI companies to hastily patch such flaws. It's not that this doesn't happen, but it's usually in the context of benchmark maxxing by the less scrupulous companies (and occasionally, when the PR hit is strong enough, they'll add specific instruments, such as the "Rs in strawberry" one, which was specifically addressed in Claude's system point a while back).

The issue with this approach is that it leads to maximal paranoia and complacency, and as excuse to dismiss clear and obvious improvements in all domains. And even in the worst case, patching specific failure modes is still an improvement. LLMs are supposed to suffer on truly "out of distribution" problems (I have my reservations, I wonder how the average human fares) but in principle, if you can actually capture most of that distribution, you've got something that is effective in deployment (though it might be brittle, but once again, we're talking about a hypothetical model that is actually trained on nearly everything).

Finally, I really doubt that OpenAI or Anthropic went to the trouble of patching this specific puzzle on purpose. They didn't even hard code the strawberry example, they just hinted to the model that it suffers from tokenization problems, and that it should try and use code to check instead of parsing it itself (a defensible position). They didn't, as far as I can tell, patch the far more famous "but I can't operate, the boy is my son!" trick question, and it was tripping up the best LLMs for years. I suspect it might do so today.

In other words, if you're famous for maintaining some kind of formal benchmark, it might be worth their while to artificially target your questions. They have better things to do in general, for smaller problems like this.

It lands in New Jersey.

Reason: after flying 300 miles north from New York City, the “300 miles west” leg happens at a higher latitude, where lines of longitude are closer together. That westward leg changes your longitude by more degrees than the final “300 miles east” leg (which happens farther south). So you end up a bit west of the start point, in central New Jersey (roughly near the New Brunswick area).

That's GPT 5.2 Thinking first go. Examining its reasoning traces reveals that it immediately noticed the issues arising from the Earth's curvature, and it even wrote a whole-ass program to compute exact latitude and longitude before outputting its final answer.

I would think that by now, this question is now in the training data, but still, these sorts of quick fixes don't solve the general problem.

That way lies madness.

Somebody like @self_made_human might be able to cut through the chaff and confidently take advice about ocular migraines, but could you? Could I? Hell if I know.

I still saw a real doctor after consulting the models. In fact, I saw a doctor because I consulted the models: they raised the possibility of differential diagnoses like TIA (mini-stroke) that, while unlikely according to both my judgment and theirs, seemed worth ruling out. As I mentioned in the linked comment, Dr. GPT still lacks opposable thumbs. Most medical advice requires actual physical examinations and actual tests to implement.

This doesn't excuse the first two human doctors who misdiagnosed me. The symptoms were clearly inconsistent with their diagnosis, though I'm not confident 2024-era models would have caught this as quickly as today's versions do.


Beyond this specific case, I have thoughts.

LLMs are both force multipliers and substitute goods. "Substitute" sounds pejorative, but it shouldn't. An MRE is a poor substitute for a home-cooked meal if you're at home. But on a hiking trail, you'd gladly take that chicken tikka over nothing, even if your digestive system later files a complaint. A terrible car beats no car most of the time. And so on.

My medical training lets me extract more value from any model. But even without that training, LLM medical advice beats having no doctor at all. It beats frantically Googling symptoms at 2 AM like we used to do. One of my most upvoted posts on The Motte discussed GPT-4, which now lags so far behind the current state of the art that it's almost embarrassing. It was still incredibly useful at the time. Back then, I said:

I'd put their competency around the marks of a decent final year student versus a competent postgraduate resident

Now? Easily at or better than the median specialist.

(This is part of why people not paying close attention miss the improvements in models until there's a flashy new headline feature like image generation, web search, Deep Research, or in-interface code execution.)

At this point, I would trust GPT 5.2 Thinking over a non-specialist human doctor operating outside their lane. It gives better cardiology advice than an ophthalmologist would, better psychiatric advice than an ER physician. Even specialists aren't safe: I know cases where models outperformed my own superiors. I'd already noticed them making suboptimal choices; confirming this with citations from primary literature didn't take long.

For laypeople, this is invaluable, albeit bottlenecked by the need for humans who can authorize tests. LLMs can recommend the right drugs and doses, check for interactions, create personalized regimens, but you still need a human physician somewhere in the chain.

(Much of this reflects regulatory hurdles. See recent discussions about why LLMs giving legal advice lack the same privileges as lawyers saying identical things.)

LLMs serve as both complement and partial substitute for human physicians. Many doctors get defensive when patients quote ChatGPT at them. I try not to. Even the free tier usually gives non-terrible advice. It's eminently reasonable to consult LLMs for help, especially for non-critical symptoms. They're surprisingly good at flagging when seemingly innocuous problems might indicate something serious. For anything important, treat them as an informed second opinion before seeing a human doctor, or use them to review advice you've already received. I'd take any LLM-raised concerns from a patient seriously and double-check at minimum. If your current doctor isn't as generous, I apologize; your mileage may vary.

The Layman's Guide to Using LLMs for Medical Advice Without Shooting Your Dick Off

1. Pay for a state-of-the-art model. Your health is worth $20 a month, you fucking cheapskate. Google gives away their (almost) best model for free on AI Studio.

2. Be exhaustive. List every detail about your symptoms. When I asked GPT 5.2 Thinking or Gemini 3 Pro about my eye problems, I had an annotated Amsler grid and timeline ready. Over-explaining beats omitting details. Unlike human doctors, LLMs don't bill by the hour (yet). Remember that they don't have the ability to pull open your medical records or call your other doctor for you. What you put into them informs what you get out of them.

3. For anything remotely important, consult two or three models. Note commonalities and differences. If they disagree, have them debate until they reach consensus, or get another model to arbitrate. This effectively mitigates hallucinations, even though base rates are low these days.

4. Ask for explanations. Medical terminology is arcane. LLMs are nearly superhuman at explaining things at your exact level of understanding. I wish my colleagues were as good at communicating information, even when the information itself is correct. If you're confused about anything, just ask.

5. Optional: Ask for probabilistic reasoning. Get them to put numbers on things like good Bayesians. Have them use their search tools if they haven't already (most models err toward using them even when not strictly necessary).

6. Remember you'll need a human eventually. But you can enter that consultation well-prepared.

That's it, really. A year or two ago, I'd have shared sample prompts with extensive guardrails (red flags, conflicting treatment protocols, high yield tests etc). You don't need that anymore. These models are smart. They understand context. Just talk to them. They are smart enough to notice what matters, and to tell you when the right move is “stop talking to me and go get checked.” I did just that myself.


Edit:

Humans are hardly immune to hallucinations, confabulation or suggestibility. You might have fallen prey to:

Say silk five times. What do cows drink? Milk. Oh fuck, wait a second–

And that is not very good evidence of humans not being general-purpose reasoners. I invite people to look, actually fucking look at what AI can do today, and the rate of improvement.