Hey all! This week is going to be a gap week; I have quite a bit of teaching related work along with several projects all coming together at once and something had to give. With luck, next week we’ll start a series on the principles of fortifications. In the meantime, of course, I wouldn’t leave you without things to read, so here are (briefly!) some of the things of note recently:
Our helpful narrator has continued making audio versions of ACOUP posts, complete with accessibility features like image descriptions and so on. The newly audio articles are (links direct to the audio):
- How It Wasn’t: Game of Thrones and the Middle Ages (all three parts) (text)
- Elective Monarchy and the Future of Westeros (text)
- The Preposterous Logistics of the Loot Train Battle (text)
- The Preposterous Tactics of the Loot Train Battle (text)
- How Fast Do Armies Move? (text)
- That Dothraki Horde (all four parts) (text)
- That Dothraki Charge (text)
- Meet a Historian: Michael Taylor on Why We Need the Classics (text)
So check them out!
Over at Foreign Policy, Lauren Teixeira compares the history underneath the Mongolian and Chinese music scenes. Notionally the essay is looking at why Mongolian music has been more successfully exported, despite Mongolia’s small population, compared to the Chinese music scene, which of much more vast but less successful outside of China. What I think is more interesting here – I am certainly not qualified to compare either music tradition for quality or reach (though I do quite enjoy The Hu), the discussion of the impact of events in the political realm on the arts is interesting.
Over at War on the Rocks, I thought “Schrodinger’s [sic] Military? Challenges for China’s Military Modernization Ambitions“1 put some specifics to a general point I’ve found myself mulling over a fair bit: the degree to which the actual strength and capabilities of many of the world’s militaries right now are effectively unknowns. Armed conflict always functions as a ‘clarifying spasm of violence’ in that there are invariably surprises in terms of how militaries and their equipment function under pressure (the inherent unpredictability of war, which is to say, friction, is one of the things that leads Clausewitz – DRINK! – to remove its conduct from the realm of calculations of reason into the realm of ‘genius’) but this is perhaps more true now than at any time in the recent past. One hopes that awareness of that unpredictability will inspire greater caution among leaders but I think experience tells us that such zones of uncertainty, leaders often ‘hope for the best’ and gamble, frequently with catastrophic results.
Meanwhile, over at Peopling the Past, they’re spending October talking to classicists who work on the theme of monsters. They also have a video talk with Dr. Chandra Giroux about the nature of the Athenian democracy and Plutarch’s take on it which is really worth listening to both if you want a primer on how the Athenian democracy worked and also if you want a primer on who Plutarch is and why he is such an influential source. If you are interested in what exciting new work is being done in the study of the ancient world, you really need to put Peopling the Past on your regular reading.
And that’s that. Next week (time permitting) we’ll start looking at some of the basics of how fortifications functioned in the past, beginning with a look at the ancient siege ‘toolkit’ from the attacker’s perspective.
Something I’ve been thinking about, which might be worth saving for a Valentine’s Day post…ancient and medieval courtship customs, particularly among the common folk. I keep seeing the statement that marriage for “love” is a modern concept, and people in pre-modern days married mostly for political or financial advantage. While I can understand that being the case for the ruling classes, I have a hard time believing the lower classes approached marriage the same way.
Honestly that sounds like a delightful idea, medieval courtship and the differences of how it functioned between the classes.
Lower classes can still marry for “her dowry is this plot of land next to ours” or “I need someone reasonably competent around the house”.
If I understand the usual late-Medieval arrangement correctly, there was no “plot of land next to ours”; everybody in the village had a patchwork of plots that were sort of adjacent to everybody else in the village. Unless you can arrange to marry your son to a *much* richer family’s daughter, there’s unlikely to be any great land advantage to marrying him off to one girl over any other. Similarly, I doubt there’s going to be a huge variance in perceived homemaking competence at courtship age.
And if you are going to try and hook your son up with a much richer family’s daughter, given that you don’t have money or corresponding status to throw into the equation, it’s hard to see a more effective strategy than somehow getting her to actually fall in love with him so that she’ll nag her parents incessantly to let her marry her poor but dreamy hearthrob.
Accumulating land was an important goal for all classes in an agrarian society. Contrary to common belief women could hold land by purchase or inheritance. A village family might have no sons, or a land rich family dower a girl with a piece of land. And of course widows inherited the holding if there were no children or might be left a slice by her late husband. Landed widows were very much a thing and extremely sought after by suitors.
Children were put to work as soon as they were capable of it. A girl in particular had plenty of opportunity to demonstrate skills before the age of marriage. (Men’s work was more difficult because much of it required the strength of a full-grown man.)
Depended very much on area. By late medieval times some peasant families were accumulating enough land to move up the social scale – getting local offices, putting relatives into the church, inventing family trees and becoming ‘noble’. We can map a few. Richer lowland areas were seeing concentrated farms emerge alongside the older mixed-strip farming.
My impression is that while landholding and such certainly played a part (especially for the richer end of the peasantry) other importnat considerations were social relations (getting attached to marriage to other kin-networks) and the economic and social capability of the spouse itself: Their being a good worker and capable of doing all the things a wife (or husband) was supposed to do. For a tleast the poorer strata the most importnat economic consideration in marriage was the labour capacity of the spouse themselves. Which was absolutely essential.
I can totally imagine the lower class using marriage for material advantage. In the articles on ancient agriculture on this blog it is described how important social ties are in farming villages, and how important owning land was. Marrying off your son/daughter to some neighbouring family to secure your mutual investment in eachothers continued survival seems like a good move. Or if you only have daughters, you would want to use marriage to make sure you have some son-in-law to take care of you in your old age.
Plus of course she needed to labor. Many signs of beauty are dangerous because if she is not actually rich, she’s lazy.
There is a Finnish proverb Riihisievä, aina sievä: “beautiful when threshing, always beautiful”. Threshing was hard work, and done inside a threshing cabin that was warmed with a stove that had no chimney, so all surfaces would be covered with soot. After working for a while, you would be covered with seat, soot and dust, and your shirt (typically, the only garment worn during threshing) would be a sooty rag. So, if a woman is beautiful even then, she is actually beautiful, and in fact, also in good physical condition, because otherwise, she’ll look ragged from exhaustion. So, the proverb directs to a situation where the capability for hard work and beauty are combined.
In Odds and Ends, from the Grimm Brothers, a woman throws away knots in flax when spinning. Her maid gathers them up and spins until she has a dress.
When her mistress had a match, she made the mistake of jeering at the maid for wearing her odds and ends to her bridegroom. He dumped her and married the maid.
They certainly took this into account – as well as family opinion. But the abundance of love songs, tales of love foolish or otherwise, love tokens and symbols and so on tell us that love (however defined) was a large part of the mix.
There are many folktales about how the sage man picks the thrifty bride or the spinner among potential brides.
True – but folk-tales are often the old advising the young – to no avail.
Actually the advice about a bride ones usually have the bridegroom listen.
The character in the story listens. The fellow listening to the story might not.
It would be accurate to say that financial conditions carried great weight with all classes but that certainly didn’t mean people routinely married partners they found repulsive just for advantage. The goal was somebody you liked who was of the right class and had a bit of money or land. Individuals looking for a mate generally had the same goals as the parents of younger people. Romantics were few and far between. Look at Vess of Hardwick, married four times with each husband better off than the last. In justice to Bess she seems to have been an excellent wife. Her final marriage ended badly but started out happy and affectionate.
My understand is that in medieval Europe, arranged marriages with no input from the bride and groom were purely a custom of the very upper class. That said, my further understanding is that then as now, young men and women considered a host of factors in selecting a spouse, including ability as a parent, provider, and housekeeper and wealth and status, along with physical attractiveness and personal congeniality. Also, then as now, most people spend most of their social life with members of their own class, so that is usually where they find their marital partners. Another factor is that back then, and even now in certain circles, pregnant girls are expected to marry the father of their unborn child. (We had two associates at our law firm in that position within the past year.)
That’s pretty much correct. Generally speaking the more property involved the less the youngsters had to say about it. Girls generally married younger and had limited social contacts and so usually less to contribute but young men often took an active part in seeking out a suitable wife. Older spinsters and widows also did their share of husband hunting.
Romantic love was believed to be a very dangerous way of choosing a mate, ideally love came after a sensible match made with due attention to financial and social matters.
Was reading a modern Chinese woman’s account of getting married. Both she and her mother were looking and what’s striking is how well they agreed on what to look for, they didn’t even have to talk about it.
Same with the “everyone in Medieval Europe married in their mid-teens” stereotype. Marriage patterns among the nobility, especially as you went higher up the hierarchy, were weird even by their own societies’ standards.
Very much so.
https://en.wikipedia.org/wiki/Hajnal_line
Note that the criticisms of Hajnal are not that he put Northwestern Europe’s average marriage age too high, but that he put Eastern and Southern Europe’s too low.
“Class differences played a great role in when a couple could marry; the wealthier that a couple was, the likelier that they were to marry earlier. Noblewomen and gentlewomen married early, but they were a small minority;[40] a thousand marriage certificates issued by the Diocese of Canterbury between 1619 and 1660 show that only one bride was aged thirteen years, four were fifteen, twelve were sixteen, seventeen were seventeen, and the other 966 of the brides were aged nineteen years or older when they married for the first time. The church stipulated that both the bride and groom must be at least 21 years of age to marry without the consent of their families; the most common ages of marriage were 22 for women, 24 for men; the median ages were 22.8 for women and 25.5 for men; the average ages were 24 years for women and nearly 28 years for men. The youngest brides were nobility and gentry.[41]
The moderate rates of fertility, mortality, and marriage within the region were tied to the economy; when times were better, more people could afford to marry early and thus have more children and conversely more people delayed marriages (or remained unmarried) and bore fewer children when times were bad. This contrasts with societies outside of this region, where early marriage for both sexes was virtually universal and high fertility was counteracted by high mortality;[42] in the 15th century, a Tuscan woman 21 years of age would be seen as past marriageable age, the deadline for which was 19 years, and easily 97 percent of Florentine women were married by the age of 25 years while 21 years was the typical age of an English bride.[43]”
Of course, the important thing to remember about the nobility and royalty is that part of their job is to secure the succession so that there isn’t a big to-do about who gets what after you kick it. Therefore, getting an heir and a spare or two ASAP is kind of important, which means that you need to get hitched ASAP.
“heir and spare” was a luxury for the 19th century. Before then you wanted as many heirs as you could get, and there was no guarantee you had assured the succession even if you had half a dozen sons.
Of course, part of the reason you wouldn’t have an assured succession is if you had too many sons, who would then proceed to fight and intrigue over the succession. The ideal was to have exactly one son who is alive and in good health by the time the landholder kicks the bucket. Figuring out how to get there in the face of uncertain mortality was fairly tricky.
Reading that War on the Rocks piece, I can’t help but remember your comment about how dictatorships struggle with the flexibility and devolution of command required for the Modern Style of Warfare. Giving that kind of autonomy to military commanders requires a lot of trust that they won’t use it to usurp civilian power.
If the services are constantly at each other’s throats, maybe they should have them develop effective combined arms within themselves. Have the PLAN further develop its own force of marines, have the Ground Force have its own air operations to support it, and so forth.
That’s more or less what happened both in Imperial Japan and the Third Reich during the lead-up to WWII, in different ways and to different extents
I feel a duty to link to this Drachinifel video, Ships of the Imperial Japanese Army: https://www.youtube.com/watch?v=mZ9AFbnY4To
Surely then there would just be fighting about whether they should create the People’s Liberation Army Navy Army or the People’s Liberation Army Air Force Army or even the People’s Liberation Army Navy Air Force Army.
The key here is the devolution of command to the lower officer ranks and upper non-com ranks. Allowing folks at those levels to make critical combat decisions without having to ask their commanding officer is what is critically important. If you look at Arab-Israeli battles in detail, particularly the Jordanian 40th armored division (IIRC), then you will see good examples of this factor.
Joint operations are critical, but they are always difficult. The devolution is much more important.
Worth noting that the VC and PAVN fought a very flexible, devolved war, as did the CCP against the Nationalists.
As I recall, Michel Goya suggested that the democratic French Third Republic deliberately fragmented the command structure of the French army to reduce the danger of a coup by monarchist generals.
The point is that the issue is not so much the difference between democratic and dictatorial regimes, but the difference between regimes in which the army is, or is not, loyal to the political leadership. If the generals are not that loyal, anything that makes it possible for them to independently cooperate with each other, makes it possible for them to cooperate against the country’s leaders. (And the reverse: If the generals are that loyal, anything that makes it possible for them to independently cooperate with each other, makes it possible for them to cooperate against a potential traitor.)
So Germany’s authoritarian Second Reich could probably trust its generals with more authority than France’s democratic Third Republic.
These days democracies tend to be seen as more legitimate than dictatorships, so they probably can trust their generals more, but it does not have to be so. If it comes to be the case that Chinese generals are more trusted by President Xi than America’s generals are trusted by the Oval Office, then China might easily end up with a more flexible, devolved command structure than the US.
And I would not be sure this is impossible, in the coming decades. If January 6 had gone a little differently, we might well have seen American generals having to decide who the legitimate President was. And we might yet see such a thing, at the next election, or the one after.
Yes, in an alternate timeline.
You are correct about the spelling of Schrödinger/Schroedinger. (Source: non-native but fluent German speaker)
Even 10 or so years ago, handling diacritics in anglocentrically primitive software was quite a major hassle so the ae, oe, ue forms were common. I don’t see any obvious reason to use them nowadays.
Odd exceptions: “Goethe” is never spelt “Göthe” these days, although Johann Wolfgang himself apparently sometimes did. Writing Händel as Handel is imo ok, as a rare exception, because (a) he lived and worked for much of his life in England and (b) “Händel” pronounced correctly sounds just like “Hendl”, which is “roast chicken” in Bavarian.
German native speaker with umlaut in his name here, I can confirm.
Good thinking with Goethe (I am actually impressed!), I wouldn’t even have thought of him, although I have a book here by Schopenhauer, where he is consistently referred to as “Göthe”. Today, this would be considered incorrect.
In email addresses or user names, where only ASCII characters are allowed, you virtually always see the ae/oe/ue (and ss for ß) transliteration.
Thanks for the confirmation. And just chipping back in on the primitive software aspect – a possibly unintended but interesting effect of the GDPR is that it gives people in the EU a legal right to have their names spelt correctly in computer systerms.
Does this extend to names like 山田花子?
Physicist here. Schrodinger is much more common than Schroedinger, although the Schrödinger is preferred if you can do it. The English pronunciation is also Schrodinger, with no attempt at the umlaut.
Schrödinger’s cat entered popular culture in an extremely distorted fashion. Originally, it was a reductio ad absurdum, not an explanation of how the world works. It is specifically referring to superposition, not just any type of uncertainty. If your uncertainty does not require the use of complex numbers (involving ‘i’), then Cardano (or Fermat & Pascal) is a better reference than Schrödinger.
Handel himself adopted England as his homeland, and Anglicized his name accordingly (from Georg Friederich Händel to George Frideric [or Frederick] Handel), so without the umlaut is perfectly fine.
Gödel is today usually spelled Godel rather than Goedel, I believe. (e.g. here: https://www.youtube.com/watch?v=i2KP1vWkQ6Y ) Nobody knows what the e means.
Since as far as I can tell the German letter “ä” is exactly equivalent to “e”, right down to the pronunciation of “äu”, I’ve been wondering why it would be Anglicized as “ae” and not just “e”.
German ä sounds like German e. It doesn’t sound like English e. It sounds like the ae in “fate”.
More to the point, I think, it’s our custom in English to use foreign spellings of foreign words when our alphabet allows it. And changing ä to e would be a misspelling in German, even though they sound the same.
Note that there are official, military, alternate spellings of umlauts and other special characters for telex/Morse/all-caps writing. These are not common, nowadays.
ä = ae, ö = oe, ü = ue, ß = sz (not ss, that’s slightly different)
The reason the Ö and Ü umlauts tend to disappear without trace is IMHO that English does not have those sounds; ö is sort of like the u in “murder”, and ü simply does not exist.
Spelling of names is weird, and follows no particular rule – it’s Goebels, but Göring, for two famous examples,.
There is actually a swedish joke about that I think would still work:
“Goethe was a great poet” “Goethe is pronounced “Göthe”, the OE is an Ö” “Göthe was a great Pöt”…
That was an interesting article about China’s military. However, if AI does come to dominate the battlefield, it will hopefully look nothing like Star Wars. The one time the Rebels did try to use a targeting computer, it failed to hit the target. And the Seperatists’ droid army didn’t impress much during the prequels. Still, I’d prefer that to it turning out like it did in Terminator.
Question for the comment section. Why did modern cavalry in the west (england, france etc) use sabres instead of lances?
Spears from horseback are generally superior given their range, in fact polish lancers where quite well regarded and helped popularise their re-introduction. But then in that case why would they ever fall out of favour, the only reason I can see is that it was easier to train mass cavalry with swords than spears.
I believe modern cavalry in the west used firearms instead of lances, and continued to use swords as a sidearm.
I’m more into ancients and medieval, but my understanding is that the late medieval lancers were largely replaced in Renaissance era Europe by pistol armed cavalry. Range and accuracy for a pistol on a charging horse isn’t great, but you can still shoot a lancer before they’re in reach of you.
In the follow up melee more pistols (stuffed into boots, saddle holsters, where ever you can) will be handier than a lance, and after that you’re back to body armour and swords. Shields get dropped because they can’t protect you from all the bullets flying around on a Renaissance battlefield.
That’s right. In the various 16th century Wars of Religion in France and England, there were a few encounters between lance and pistol/sabre. Pistol/sabre won. That lance cavalry tended to fight in thin lines (‘en haye’), while pistol/sabre rode knee to knee in deeper formations may have been a factor.
Eastern armies did not employ pike formations for their infantry. Pikes, spears and pole weapons were used extensively but în open rănesc. Cavalry could easily penetrate such formations once they have been affected by a short march, fatigue or light casualties. The lance was ideal for cavalry vs infantry fighting.
The Western cavalry could engage only other cavalry or infantry on the run. The saber was more effective in both cases.
Western cavalry did sometimes use lances. The 17th Lancers, for example, charged with the Light Brigade at Balaklava. And this fellow, who looks like he should come from Warhammer 40,000, was actually in the German army of 1918: https://rarehistoricalphotos.com/german-cavalry-lances-1918/
I get the impression from Richard Holmes that lances were regarded as a less flexible weapon: What happens after your lance-point has gone past your adversary? Probably you drop the lance and try to draw your sword…
As late as 1914, the German army issued lances to the cavalry. Not much used, in the event, but very much a thing that happened.
Uhlans, at least, routinely trained with the lance, although that tended to get expensive – lances have a habit of breaking wen they hit things, whether that’s an enemy soldier or a training dummy. At which point, the Uhlan was expected to discard the roken lance and continue with the sword.
The traditional narrative is that lance kinda came and went, almost disappeared in the 1600s in favour of pistol-and-sword only to come back again in the 18th century (partially inspired be “eastern” fashions)
The usual narrative is that it had to do with pikes: Lances quickly become difficult to use against pike-and-shot formations, (where pistols at least would let you do some damage) but as pikes were replaced by bayonets lances once more got the range advantage and could be useful again.
Here’s a fun question: According to some apocalypticists, China will move its 200-million army – equipment and all – through the Khyber Pass (or at least a significant portion of it). What do you think of the logistics of such a maneuver.
Uh, China’s army is not 200 million people. Also, sorry, why would anyone do this?
But if China tried to move a seventh of its population into Afghanistan, increasing the population of Afghanistan six times over, I don’t think there would be sufficient food/water in the area to feed everyone, or infrastructure to support the needed imports. I am not at all an expert, but most places aren’t set up to support that sort of population increase overnight, even if they have decent infrastructure.
The people who suggest this are people who believe the Book of Revelation predicted helicopters. We are not talking the cream of the crop.
Wait, is this another take on Gog and Magog? Have I missed some new developments in really stupid versions of apocalyptic Christianity? Do you have a link for us to laugh at?
It used to be a thing for biblical prophecy nuts – notably, the Left Behind books – to say that the prophecy of Gog and Magog referred to Russia invading Israel with its entire army, a move that’s equally dubious in its logistics. But I haven’t heard anyone talking about *Afghanistan* as a Biblical battlefield, and I’d love to hear what sort of logic leads to that conclusion.
Don’t you see! China has to go through Afghanistan to get to Israel!
Good luck with the teaching stuff.
“Clausewitz…”
DRINK!
rofl . . . good catch! Our esteemed host missing his chance there!
Fixed!
I haven’t checked the previous comments yet, but–in the realm of proofreading–here are 3 “gaps” I think you ought to consider:
which of much more vast > which is much
more interesting here . . . on the arts is interesting (grammar and punctuation of sentence)
us that such zones of uncertainty > (is there a missing in after that?
Anybody know good resources for the Greek War of Independence? Asking for a friend.
Is the friend a ghost or reincarnation of Lord Byron?
No comment
Unfortunately, I’ve never even heard of Greek War of Independence-themed Rule 34. But do not despair, for I remain sure it exists somewhere out there.
It exists and has for a long time. https://www.nga.gov/collection/art-object-page.166484.html
The situation with China’s music scene very much brings to mind a situation last year with that erupted in the vtuber scene and led to Hololive, the biggest agency, being forced to abandon the Chinese market and fold operations there. The political situation in authoritarian regimes means that making culture is inherently dangerous, and predictably that means culture becomes stifled. You can’t make art without risk (even in a free country the best art is challenging), and if the risk penalty is high enough, the art will not get made.
Mongolia is a rare bright spot in many ways. It’s a big contrast to the political situation in the rest of central Asia. I wish I knew more about how the managed that and how they maintain it.
Well it looks like they’re going to lose it since the CCP has been looking at them in a similar light to the Uygers now, won’t be surprised if they try and pull a culteral genocide.
That would be somewhat difficult without an invasion.
There are ethnic Mongolians within China’s borders. Mongolia proper is safe, or at least in less danger.
Relatedly, I just recently finally read The Three Body Problem and was absolutely fascinated by the “Author’s note for the English edition” at the end, wherein the author declared that he was absolutely not making any social commentary. (Uh-huh, right; in a book that starts with the brief 1967 civil war during the Cultural Revolution and ends with pro-alien insurrectionists forming factions that are at each others’ throats.) I strongly suspect it’s that same phenomenon of making culture in an authoritarian regime is inherently dangerous. The author is the single most popular living Chinese SFF writer, and he still felt a need to make that declaration when exporting to a society that’s not like that.
A serious and thorough assessment of the correlation of society and the art it produces would certainly be an interesting read. But the results would probably be a little bit more complex. After all, the likes of Tolstoy or Mozart also came up to be in fairly authoritarian societies…
Not to mention Eisenstein or Riefenstahl.
Having just the other day attended a seminar about how to get robots to not trip when there’s a box in the way… I really wouldn’t worry about autonomous killbots any time soon. Compiling and computing sensory data fast enough to support an effective autonomous gait is *hard*. We’re probably at least ten years out from bipedal bots that can’t be trivially defeated by a trip wire, let alone all the other crafty things people can come up with to make an environment less than compliant. And in any case, a country of 1.4 billion with a glut of partnerless young men does not want for foot soldiers. Better to save the industrial capacity for making ammunition.
In general, it’s not really clear to me that there’s a lot of problems solved by military AI that we don’t already have better answers for. People talk about speech recognition and object recognition, but China’s most likely near-term enemy is already mostly speaking the same language. As for object recognition, given the choice of something super advanced that uses visual-spectrum data to annotate your eyesight, vs some OTC infrared goggles, I’d expect the latter to be far more useful. Automatically detecting IEDs would be super cool–but unless someone has a million-picture dataset of IEDs (of the kind that would be *actually encountered* and not just the thing your handful of engineering officers come up with when tasked to build IEDs), I would not expect it to work reliably. Object recognition depends on massive amounts of data & generalizes poorly even at google scale; and artificial intelligence *is not intelligent*, it will not put the pieces together to infer anything reliable from e.g. the way the civilians in the area are giving that car a suspiciously wide berth. China has 1.4 billion highly trained object-recognition modules honed through literally millions of years of practical in-field refinements: what do the computers bring to the table?
For targeting, if the US experience is any guide, where AI would be useful is in making targeting decisions *less* trigger-happy. That’s a good thing surely, but a) it’s probably not a key operational advantage for the PLA and b) the users are still going to bomb the bejeezus out of everything that moves funny, on the off chance the threat assessment module got a false negative.
There might be some scope for higher-level strategic decision support, but again, it’s tough to say that there’d be enough training data to really make a good model. And it’s not operating in isolation–if a computer is calling your shots, the other side is definitely going to be working very actively to convince it to call the wrong ones.
Honestly, the realistic applications of AI are less about field combat than about running an effective police state–as China is already doing. It’s about processing large-scale surveillance data, effective censorship/media control, and profiling/tracking the complete daily activity of 1.4 billion people. But this part is not really decades-of-research away; they’ve got a lot of these problems licked well enough (at least for where we are right now in the adversarial co-evolution process), and I expect further improvements to be incremental rather than revolutionary.
Fortunately for the PLA, this is probably exactly the problem they’re likely to face–“How do we do an effective occupation in a hostile country where we might have capable organized resistance.” AI can probably provide some useful labor-saving tools. But I suspect the answer they will actually go with is not “Smart _____” but “leverage our core competency in police-state-ing, and round up the Taiwanese into camps like we do the Uyghurs.” Whatever they pick, they need to be able to deliver abundant adaptive-semi-autonomous decision-making muscle locally to actually act on their surveillance & carry out punishments for non-compliance. That business works great when you’re talking about under-resourced landlocked people in the central Asian steppes where you can ship in plenty of Han cops by train while the rest of the world turns a blind eye; and it’s working fine in Hong Kong where you can bring in ever-more cops by bus over the bridges while the rest of the world appeases you. I’m not sure it still works when there’s lots of coastline to bring in weapons for guerillas, your own (necessarily ship-based) supply lines are vulnerable to missile strikes or asymmetric naval warfare, and there’s a ton of mountains for guerillas to hide in. Now, it might work! (Especially when you have automated computer-vision systems watching the infinite camera network!) But it seems not-guaranteed.
In any case, I hope we don’t find out any time soon where the balance lies–I really don’t think it’s going to be clean and digital and abstract.
There is one capacity where AI could show promise: Logistics and planning. As you say, China has a lot of people to move; digital tools to assist and guide their movement and associated equipment would probably have a good niche, though from what I hear of research in that area, the primary tools are mapping and simulation still, not necessarily AI (optimization of logistics seems awfully difficult to define good targets for)