Transcription
Everybody, Peter Zeihan here, coming to you from Colorado. Today, we're going to talk about South Korean defense policy. Specifically, in June, the Koreans decided that they're going to make every soldier a competent drone operator. Five big things come from this.
Number one, why they needed to do it primarily is a demographic situation. Korea has been one of the great success stories for turning a backward agrarian economy into a forward-looking technocracy, doing it really in just a generation, generation and a half. Fastest infrastructure build-out in human history, and a fast urbanization rate that was surpassed only by the Chinese a couple of decades later. Anyway, you combine that with the heavy drinking male culture and women deciding that, you know what, maybe there's an option for me other than the home, and you've got a high divorce rate, a low marriage rate, and arguably the lowest birth rates in the country. And so this has been going on for over half a century. And so the Koreans now have more people in their 60s than their 50s, than their 40s, than their 30s, than their 20s, than their teens and kids. Complete inversion of the demographic pyramid. So a normal military structure is to take a lot of young men in their 20s and put them on the front line. That's that's just structurally not an option for the Koreans anymore, and so they've been exploring new possibilities. Drones appear at the moment to fit the bill.
Second problem, of course, is North Korea. The North Koreans, while their economy is a shriveled raisin compared to the robustness that is the South, something like a third of GDP goes to defense spending, not to mention a not insignificant nuclear arsenal. But more importantly, somewhere between 10 and 15,000 artillery emplacements within range of Seoul. So in the case of a war, by the time the first shell hits Seoul, there will be tens of thousands in the air already, and there simply is no conventional or even nuclear solution to that. But maybe with drones, there might be. Uh, we'll see.
Third problem is the United States. Uh, the United States is in the process of ripping up all of its alliances, and for countries like South Korea that have faced artificial restrictions put on them by the Americans on what sort of military technologies they can develop, that really puts the Koreans in a bind. So, for example, one of the ways you might be able to counter the North Korean nuclear program and all that artillery is develop your own ballistic missile program. The Americans flat out have barred that since 1955. Uh, and so the Koreans are now coming into this period where the Americans have become unreliable to absent, and they just don't have the tools of statecraft or military craft that are necessary for the strategic environment in which they find themselves. Drones offer a cheap and fast way to move into something new. Maybe. We'll see.
These technologies are still being developed, and that brings us to the fourth thing. Uh, the industrial plant. Most countries who are moving into drone warfare are discovering that almost all of the parts come from China, and since China is the primary sponsor, uh, with Russia, of North Korea, the South Koreans definitely don't want to become dependent upon that. But unlike other countries that have been starting to move into the drone space, the Koreans are already a premier technological manufacturing power, and you don't have to do a lot of creative thinking in order to convert what the Koreans already have in things like aerospace and weapons and automotive and turn it into a drone industry. Everything right down to the microchips and the uh, the batteries. So, when the Koreans say they're going to be producing millions of drones a year within a few years, I tend to believe them because they have a lot of the base industrial plant already in place. It just requires modification.
Now, as the plan stands, and it is very early in development, the idea is you make everybody a drone controller. And with the technologies of literally yesterday, that makes a lot of sense. But that's not the technologies that are likely to be used on the battlefield in two, three, four, five, ten years. You see, at this point, most of the drones that have been used by the United States, the Russians, and the Ukrainians have been what are called first-person view drones. Basically, you have a live telemetry feed to your drone. You can see through its optics, and you can direct it manually. And that works in any number of circumstances. But it faces two problems when you're dealing with, say, North Korea. Number one, that link can be jammed. And while there are fiber optic drones with ranges upwards of 50 km now, which is nuts, uh, that's not enough to get all of the potential artillery that could target Seoul. So, some decision-making has to be put into that drone, so it's not simply unmanned, but it's to at least a degree autonomous. Uh, the way you've been doing that to this point is with something called NAND, which is flash memory, cold memory, think a thumb drive. Uh, NAND memory can hold a short decision tree, so you can tell your drone to look for four or five things and select a target among that list. And once the target has been selected, there's nothing to be jammed because the decision was made on board. You then have to kinetically intercept the drone, which is a lot harder.
But there's another generation of memory beyond NAND. It's called DRAM, dynamic random access, hot memory. Uh, much faster. Can hold a much larger decision tree. Instead of four or five steps, maybe several thousand steps. You train that on pilot data, which the Ukrainians have been generating in mass, and the Koreans, South Koreans tend to like the Ukrainians because the Ukrainians are fighting North Koreans now. The North Koreans have sent troops to assist the Russians in the war. So, the South Koreans have become one step removed. Anyway, the the DRAM approach is a different sort of drone with a lot more autonomy, and it's better at, say, following a roadway until it finds a train and then targeting specific parts of the train, for example.
Now, this is new technology. It is not being used in Ukraine yet. It's in the process of being invented, quite honestly, but the prototype, the first-generation drone, Xaden, is a South Korean product. It's not militarized yet. They're, it's still in the kind of advanced testing stages of the prototypes, but it already exists. And once you modify those drone brains, for lack of a better term, to a more of a combat chassis, then you really are opening up a new phase of warfare. Uh, it's not swarm intelligence. It's not AI where you can throw anything at them and they'll be able to interpret. We don't have that hardware, but it is a significantly large enough decision tree to, in many ways, mimic full decision-making capabilities.
And guess what? Remember I said that the South Koreans are already a technological economy? About 90% of the world's high-end DRAM comes specifically from South Korea. So, what's likely to be the next phase of the drone revolution is something that the Koreans are already the world leader at.
You put it all together, and by 2030, the Korean military is going to look very, very different. The only question is to the degree to which they decide to go all in on drones. Because if you take one of the world's top 10 manufacturing economies and apply these technologies and the need, which is very, very real and very present, and the Koreans have proven time and time again that they can change direction on a dime and build huge amounts of functional product in a matter of months. So, we're going to see a significant shift not just in the way that drone warfare happens, but in a number of strategic environments around the world that appear to have been locked in stasis for decades. The DMZ in the Korean Peninsula probably being at the top of that list.