Postcards from the Edge of the World (Vol. 12)This is where AI actually matters... the history of the world says so...
 “The organizing principle of any society, Mr. Garrison, is for war.” Mr. X, JFK Dear Fellow Expat, Let’s do some quick math... An M1 Abrams tank... with all the bells and whistles costs at least $10 million. I’m talking about all the costs associated with technology, training, maintenance, and equipment. You’ve probably seen these tanks roll through the desert in a movie. They make the whole planet around them shake. That’s $10 million in steel, armor plating, and 70 years of military engineering. A tactical FPV drone costs about $400 in parts, give or take. I’m not talking about some fancy military drone that’s in a defense catalog. We’re talking about a first-person drone that a half-stoned teenager can build in his basement with parts he found on the internet… when he was really high... Take one of those drones, strap some explosives to it, and fly it with goggles and a controller that look like something you use on a PlayStation. How many of those drones can you produce for the cost of a tank... About 25,000. Now, here’s the question that will change the entire future of warfare and national defense. How many drones does it cost to destroy a $10 million tank? ONE. For the first 11 volumes of Postcards, I’ve talked about the extraction of time, money, and energy. Now I’m starting to focus on what it means in a changing world. As systems change... not just in war... but in careers and markets, we must discuss how power gets built, maintained, and extracted in this world. Today’s conversation isn’t really about war. It’s about the uncomfortable changes and transitions happening everywhere. It’s a story about what happens when the big, heavy, expensive, and slow thing... loses out to what is smaller, faster, cheaper, and even smarter. Which is the history of markets… and technology revolutions. So, that pattern isn’t just happening on the battlefield. This might get dense… but stick with it… Welcome back to Edge of the World. A Course ChangeWe’ve spent eleven volumes talking about how this world extracts... taxes, inflation, currency debasement, random fees, and debt. We’ve mapped the machinery that takes your time and converts it into someone else’s power.
Today’s volume is different... but just as important. We’ll discuss where they spend that extracted money and energy… to maintain the existing power systems. If I’ve learned one thing about governments, it’s that where they put their capital ultimately tells you exactly what they fear. Right now, the United States is afraid of being too slow. I don’t talk much about my time at Johns Hopkins University in Washington, D.C. When I do, it’s usually about the economics and energy courses... that was my focus in the Global Security Studies program. But the program had three security tracks: Economics, Energy, and Military. My syllabus required that I take two graduate courses on the military side. One tracked the history of warfare. The second centered on game theory and conflict. In both classes, one of the major themes was the same... Every era of warfare is defined by a technology that renders the previous era obsolete. Consider World War I. The machine gun turned cavalry charges and massed infantry into suicide. Generals spent several years still sending men into walls of bullets because they couldn’t accept that the rules had changed. Millions died because the leadership’s thinking lagged behind the technology. Airpower in World War II changed the geometry of war. You could direct force hundreds of miles without a single boot on the ground. The aircraft carrier replaced the battleship. Strategic bombing reshaped how nations thought about defense. The countries that mastered the air won. Nuclear weapons in the Cold War made total war between superpowers unthinkable... We shifted everything into proxy conflicts, intelligence operations, and deterrence theory. The atomic bomb didn’t just change tactics… It also changed the freaking logic of conflict. Fast forward to the First Gulf War. Precision-guided munitions let the U.S. hit a single building from 30,000 feet. The real innovation wasn’t precision alone, but what happened when precision merged with volume. Entire zones were saturated with “steel rain” so quickly that the space between soldiers became lethal, collapsing coordination before anyone understood what was happening. The world watched on CNN as the most powerful military in history dismantled the fourth-largest army on earth in less than 100 hours. In every case, the nation that recognized the shift first and adapted fastest gained a decisive advantage. And in every case, the nations that clung to the old model... that kept building cavalry when the machine gun existed, that kept building battleships when the carrier existed... lost. The U.S. aims to be on the cutting edge of this pattern. It’s arguably the single most consistent strategic priority in American military history... stay one technological generation ahead of everyone else. Right now, the technology that’s rewriting the rules is autonomous AI... drones, unmanned systems, software that can find, decide, and act faster than any human chain of command. The Pentagon has a term for this: Decision Dominance. It sounds like bureaucratic jargon... because it is. But what it means is that the side that can close the loop faster wins. It’s not about who has more tanks, more soldiers, or more money... It’s about who has the shorter loop between detecting a threat and acting on it. For 30 years after the Cold War, the U.S. didn’t have to worry much about this. There was no real competitor. Unipolarity... one country with unchallenged military, economic, and technological dominance. When you’re the only one at the table, speed doesn’t matter as much. You can afford to be slow because nobody’s racing you. That era is over. China is building its own autonomous weapons programs. Russia is learning why being slow is bad in real time on the battlefield in Ukraine. The BRICS nations are constructing parallel financial systems. The world is splitting into two poles... West and East. And when it splits like that, the question stops being who has the most stuff and starts being who can act first. That’s why the Pentagon created its first-ever dedicated budget line for AI and autonomous systems. It’s set for $13.4 billion in fiscal year 2026. That’s a small number… about 1% of the budget. But it’s meaningful. The Department of Defense said... and I’m quoting directly... that 2026 will be the year we emphatically raise the bar for Military AI Dominance. The total defense budget hit $1.01 trillion. And President Trump appears to be pressing for a $1.5 trillion budget. That’s not money going to more steel… It’s an investment in faster loops. We’re buying Decision Dominance… and speed in what are known as “Kill Chains.” If you want to see what that looks like in practice... look at Ukraine. How the World WorksIn June 2025, Ukraine launched Operation Spiderweb. They took 117 cheap FPV drones... first-person view, the garage-built kind I described up top... smuggled them into Russia and launched them from cargo containers strapped to the back of ordinary trucks... These were just normal trucks, parked on normal roads, with nobody looking twice. The event was seismic… and, in my view, the most important military event of the last five years… Those 117 drones hit four Russian airfields and disabled roughly one-third of Russia’s strategic bomber fleet. The estimated damage was $7 billion. Those drones might have cost about $50,000. Now... FPV drones are responsible for an estimated 60% to 70% of all vehicle losses on the Ukrainian battlefield, according to the Royal United Services Institute (RUSI). A country being invaded built an entire drone industry from scratch in two years... because a $400 flying camera with an explosive turned out to be more effective than everything that dominated warfare for the past century. Former Google CEO Eric Schmidt just wrote in the Financial Times that future wars will be defined by unmanned weapons... and that the West isn’t remotely ready to build them at scale. Tanks don’t lead attacks now... The idea that Vladimir Putin could just march across Europe with them isn’t going to happen… And the miles-wide No Man’s Land is the future of a world where boundardies will be defined by scale. Tanks now hide behind buildings and fire from distance... because if they roll into an open field, some guy with goggles and a $400 drone can find them in minutes. The most fearsome weapon of the 20th century doesn’t want attention… These drones are our generation’s version of the machine gun, airpower, or the precision-guided munition. The technology has changed, and the only question is who adapts first. And the tank isn’t just a tank. Consider it a metaphor for so many other things in our lives. Tanks are anything big, heavy, expensive, slow-moving thing that used to dominate because of its size. In business, it’s the massive corporation doing work that a small team with the right software can do in a weekend… Anyone who takes weeks to make a decision is going to lead their corporation off a cliff… In your career... the tank is anyone whose value comes from being the person who knows things or processes things in an era where knowing and processing are becoming cheap and fast. The 20th century rewarded mass. The 21st century is about speed.
What’s Beneath the NoiseLet me explain the “kill chain.” The entire premise is based on a simple loop. Find. Decide. Act. That’s the loop. You find the military target, decide what to do, and act on that decision. For most of human history, humans lived inside the delay between those steps. Generals, intelligence officers, and layers of bureaucracy… Someone makes a phone call because they see the target… then it takes way too much time to take action. That delay used to be hours, days, sometimes weeks. AI is compressing it toward zero. A drone swarm doesn’t need a committee meeting. The sensors detect, the AI decides, the system acts... all in seconds. BOOM! That compression is Decision Dominance in action. But here’s what most people miss: that Find-Decide-Act loop isn’t just for the military. It’s the structure of almost everything. Your doctor finds symptoms, decides on treatment, and acts by prescribing. Your insurance company finds a claim, decides its validity, and acts by paying or denying. Your bank identifies a transaction pattern, determines whether it is fraudulent, and acts by flagging or approving it. Every single one of those loops is being compressed by AI right now. The battlefield just gets there first because a slow loop gets people killed. So the funding and the urgency come first. And here’s the historical pattern that I was taught… What gets built for war eventually gets used for everything else. GPS... the thing your phone uses to find the nearest Chipotle or so that guy can deliver you pizza... was built to guide missiles. The internet was a military project. Seriously… Ask Al Gore… Jet engines, radar, nuclear energy, microwave ovens... all developed first for the military. Look up DARPA. It’s been around since 1958. The AI coordination systems being built right now for drone swarms will become the AI systems that run logistics networks, hospital triage, financial markets, and everything in between. It’s inevitable.. And financial trading? Come on… there’s no way half the theory hasn’t already been employed. I talked about Druckenmiller’s 2018 statement around algorithms… AI will take over trading and outperform human managers. Full stop. In Volume 9, I wrote about AI threatening personal identity. In Volume 10, I showed how the financial system extracts your time. In Volume 11, we explored how currency debasement should forces you into scarce assets. This volume connects those threads. The kill chain is the beneficiary of the extraction chain. Every time a loop compresses, the humans living inside it get squeezed out. How Power WorksAs I’ve showed, the loop is getting tighter. The real question that matters is… “Who owns the loop?” I’ll just highlight a key example… A private company called Anduril Industries is building something called Lattice... This acts like air traffic control for drone swarms. It fuses data from every sensor on the battlefield into a single decision layer. The Pentagon doesn’t buy it like hardware. It subscribes to the product. This is software-as-a-service (SaaS) for war. Peter Thiel and Eric Schmidt are both putting serious personal money into this space... and they’re not doing it for fun. I bring up Anduril not because you can buy it... you can’t… It’s now a private company. The more impoirtant thing is that the business model tells you where everything’s heading. The drones are designed to be expendable. You use them, they get destroyed, you order more. It’s not a tank you maintain for 30 years. It’s consumable warfare... it’s a recurring revenue model from the battlefield. The process never stops. It just finds new loops to compress. The companies that will define the next 20 years aren’t building things. They’re building operating systems for things. The value isn’t in the steel. It’s in the software that tells the steel what to do. Venice got rich not by making anything, but by controlling the passage. Whoever owns the coordination layer collects the toll. What to Do at HomeAs always, I don’t want you to think of this as just an investment letter. I want you to think strategically about life and what you can do to ensure you succeed amid the major shift in technology and society…
If your job involves a Find-Decide-Act loop... and most knowledge work does... start learning the AI tools that compress that loop. Don’t wait for your employer to train you. The people who learned to use GPS early didn’t wait for permission either. The person who sits atop the coordination layer survives… and can get rich. The person inside the loop gets squeezed out.
Most of you already track inflation data, interest rates, and central bank moves… I hope that’s because I’ve spent 11 volumes showing you that’s where the extraction signals are. This volume gives you a new signal to track... defense procurement and autonomous systems contracts. It’s all out there. When the Pentagon awards a drone production contract or increases the AI budget line, that’s a leading indicator of where commercial technology is headed over the next three to five years. In the same way, Cantillon effects tell you where money flows first… Defense spending tells you where technology flows...
This one hits the identity thread from Vol. 9. If the value of “knowing things” is collapsing as AI learns faster, then the skill that matters is understanding how systems work... You must know how loops connect, where the chokepoints are, and how to sit at the coordination layer rather than inside the process. That’s a different education priority than “get a credential.”
I don’t mean this literally. But the same process that lets 117 cheap drones take out $7 billion in bombers applies to business and career. As always, small, fast, and coordinated will crush “big and slow.” If you’re building a business, build it light. You don’t need more people. You need tools… If you’re building a career, build it around speed and adaptability, not mass and credentials. Where you went to school will matter incredibly less in the future… Seriously… the fact that I went to the schools I attended… doesn’t matter anymore. It’s important for networking… but it means nothing if I can’t scale, build… and run FAST now (after years of crawling and then walking.) The person with five AI tools and a clear thesis will outperform the team of 50 doing it the old way. An introvert with ADHD has just been handed a jet… A person stuck in the past… they’re done. CEOs who don’t create horizontal cultures will lead to failure. People who take 8 months to make a basic decision, will lead their companies to ruin. The next 24 months are going to create a sprint… and you’d better take the time to master prompting skills and understand what platforms stand to overhaul existing systems. Now then… as always, I have a stock recommendation… This one will be a little different… It’s not necessarily the traditional chokepoint that I recommend. It’s something that you could take the existing yield from those other chokepoints and pour into the speculation and speed of this trend… Be fearless… because Schmidt is correct. The future of warfare is in the No Man’s Land… And the world is about to change significantly in the next five years… Invest that way…... Continue reading this post for free in the Substack app
|
Post a Comment
Post a Comment