Among the most promising military applications of AI is staff work. Tons of routine products—intel summaries, orders, etc.—can be generated much faster by machine. Does this mean staffs will reverse the historic trend and begin to shrink?
No: they’re about to explode in size.🧵
In the Napoleonic era, a divisional or corps staff was never more than a dozen soldiers, whereas today it’s pushing toward a thousand for formations of about the same size. Part of a general trend in tooth-to-tail ratios.
The reasons are fairly obvious: modern armies are more complicated, requiring more logistical coordination, fire control, etc.
BUT. There’s a subtler effect at play too: Jevon’s paradox. Simply stated, the more efficiently a resource can be used, the greater the demand.
It’s the story of Eli Whitney and the cotton gin. He thought he could reduce the demand for slavery by creating a labor-saving device for processing cotton. But by increasing the cotton each slave produced, he made them much, much more valuable.
Same story with staff work. The more valuable data/products/whatever that each staff member can generate, the greater the demand.
The typewriter, for instance, did not reduce the number of clerks (secretaries); it greatly increased the volume of correspondence.
This came at a convenient time, when more information needed to be sent over greater distances. But typewriters also *enabled* more complex operations, requiring more detailed orders, greater coordination, etc., and thereby fueling demand for larger staffs.
As an example, consider the situational awareness that persistent surveillance gives HQ—often better than the ground troops. Pair it with AI for threat ID, predictive firing solutions, etc., and you have several staff members micromanaging a single squad.
This is just one example, and not an especially good one—the entire point is that it’s hard to predict new uses for technology until its available in abundance. The one certainty is that that abundance will only grow demand, not shrink it.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Lots of focus on the difficulty of creating a breakthrough in a long-range precision strike regime, but that’s only half the problem. Assuming it can be done, pushing exploitation forces up from the rear would be another challenge in itself.
🧵
A basic element of combined arms is their simultaneous offensive and defensive function. Hitting an objective in several ways increases the odds of success, while making it harder for the enemy to hit vulnerable troops at the point of attack—an overwhelming pulse of combat power.
LRPFs frustrate this by being harder to suppress and quicker to counterattack. Successful offensives require either extreme incrementalism (what we see in Ukraine), or an absurd concentration of deep strikes, AD, EW, artillery, etc. that exceed any current military’s capacity.
Saltpeter, or crystallized potassium nitrate, was the most important component of gunpowder. It naturally occurs in deposits around the world, but for countries like Sweden that lacked them, it could be created through an involved and somewhat disgusting process.🧵
Saltpeter provides the oxygen for rapid combustion, giving gunpowder its bang. Early formulas were more fast-burning incendiaries than explosives, with pitch or oil mixed in, until the optimal ratio was discovered: 75% saltpeter, 15% charcoal, 10% sulfur
Europe as a whole had little naturally-occurring saltpeter: much of it came from the Americas, where it formed in caves from bat guano, and India, where it was refined from certain soils—this drove Dutch, English, French, and Danish colonial trade with India in the 17th century.
OTD 1814, the Battle of Bladensburg: the British defeat an American force defending Washington, DC.
Although not a terribly interesting battle in itself, American tactics resemble Hannibal’s at Zama—and both lead to the sacking of their respective capitals (sort of).🧵
The first two years of the War of 1812 overturned expectations: on land, the American invasion of Canada made no headway, but at sea her privateers & frigates had great success—the Napoleonic Wars were raging, and the Royal Navy could not spare the effort.
But following Leipzig, Napoleon was all but defeated, allowing the British to redeploy forces in 1814. They extended their blockade to the entire eastern seaboard, occupied Maine, sent an expedition to the Gulf of Mexico, and launched an overland invasion via Lake Champlain.
Despite being small and poor, Sweden became a major player in the 17th century through the feats of its superb army. But military power depended in turn on the extremely efficient organization of all society and expansive investment in industry.🧵
Sweden's disadvantages were partly offset by some critical natural resources. Much of the crown’s revenues came from raw material exports: iron, timber, naval stores, and above all copper.
The enormous Falun mine supplied up to two-thirds of Europe’s copper in the 17th century.
Swedish kings tried to move up the value chain by investing in smelters, forges, and eventually cannon foundries. For this they relied on commercial connections with the Dutch, whose merchants were very active in the Baltic.
Chinese bureaucracy was a sort of intangible infrastructure that made their military more effective—preserving and disseminating techniques, driving weapons development, etc. But Europe was developing its own intangibles which by the late 1700s completely surpassed this.🧵
Western writers began seriously recording military knowledge only a little before the Chinese developed countermarch musketry (although at first much of this was more humanists’ naively searching for classical precedence than practical techniques: )dispatch.bazaarofwar.com/p/paperback-ed…
Over the following two centuries, however, Europe saw an explosion in books on military organization, weaponry, and techniques, which were ruthlessly copied and modified by rival military establishments.
Andrade takes seriously the hypothesis that endemic warfare among semi-stable states is what drove European military advances, and applies that model to China.
In particular, he identifies three "warring states" periods in which China saw great leaps in military technology:
1. The "Song Warring States Period" (960-1234): northern China was controlled by various steppe dynasties (Xi Xia, Liao, Jin, later Mongols), south by the Song.