The drone revolution

The history of war is, among other things, a history of cost-exchange ratios. The defender's per-kill cost vs. the attacker's per-kill cost determines what is sustainable. When that ratio inverts dramatically — as it has in the last decade with small drones — entire strategies and weapon systems become obsolete almost overnight.

This isn't speculation; it is documentable from the Nagorno-Karabakh war of 2020, the Ukraine war from 2022 forward, the Red Sea attacks of 2024, and the Iran-Israel exchanges of 2024-25. In all of them, cheap unmanned systems repeatedly destroyed expensive manned platforms at ratios that nobody's Cold War-era doctrine assumed possible.

The math of asymmetry

An Abrams main battle tank costs roughly $10 million. A modern manned fighter, $80-100 million. A guided-missile destroyer, $2 billion. These platforms were designed in a world where the things that could kill them — other tanks, other planes, other ships — cost comparable amounts. Each side was spending high-value chips on high-value chips.

A commercial FPV drone costs $400-1,000. With a small shaped charge or anti-armor warhead, it becomes a precision-guided munition that an 18-year-old can fly into a tank's engine deck. Ukrainian and Russian forces are now producing these in the millions per year. The cost of killing a tank has fallen by three orders of magnitude.

What this changes

The naval problem. The Houthis, with weaponry that costs in the tens of thousands of dollars per unit, have forced the US Navy to fire million-dollar interceptors. The exchange ratio is unsustainable for the defender even when the defender wins every engagement. A $20,000 Shahed-style drone defeated by a $2 million SM-2 missile is, in pure economics, a Houthi victory. A US carrier with a $13 billion price tag now faces a credible threat from swarms whose unit cost is the price of a used car.

The armor problem. The "tank is obsolete" thesis is older and has been wrong before. But the Ukraine evidence is real: heavy armor without comprehensive drone defense (which barely exists) is uniquely vulnerable. The response is not "no more tanks" but "tanks need top-attack defense, electronic warfare bubble, and infantry-drone integration." Most of the world's armored forces are not built that way.

The infantry problem. Drones give a foot soldier ISR (intelligence, surveillance, reconnaissance) and precision strike capabilities that used to require an artillery battery and an air force. The kill chain has compressed from hours to minutes. A small unit with quadcopters, FPV drones, and a Starlink terminal has more lethal reach than a 1980s infantry company.

The production problem. Russia and Ukraine are each producing millions of drones per year. China's commercial drone industry (DJI alone) could pivot to military production at a scale Western primes literally cannot match. The premium-priced, low-volume Western weapons procurement model — built for an era when the platform mattered more than the round count — has not yet caught up to a world where rounds are the bottleneck.

What this doesn't change

To keep perspective: drones haven't yet won a war. Russia hasn't collapsed Ukraine; Ukraine hasn't pushed Russia out. Israel's air force, with conventional munitions, still does what it does. Nuclear deterrence still anchors great-power restraint. The drone is a tactical and operational revolution; the strategic order it remakes is still being remade.

And: drone proliferation cuts both ways. A weapon cheap enough to be everywhere is also a weapon every adversary, every cartel, every militia can have. The same logic that empowers the small state empowers the criminal organization and the lone actor. We are at the beginning, not the end, of figuring out what that means for civilian life.

The broader frame

The drone case is one instance of a larger pattern: commoditization of capability. The same logic applies to:

Cyber offense. Capabilities that were state-only in 1995 are state-or-criminal in 2010 and state-or-criminal-or-bored-teenager in 2025. Ransomware-as-a-service offerings sell for what a used car costs. The exchange ratio of attack vs. defense has shifted hard toward attack.

Biological capabilities. Synthesis costs for DNA have collapsed 10,000-fold in 25 years. Designer biology that required a state lab now requires a competent grad student. The biosecurity implications are not theoretical.

Disinformation. Generative AI has commoditized the production of plausible text, images, voice, and video. A 2020 disinformation operation cost a state-sized budget. A 2025 version of the same campaign costs subscription fees.

The pattern: capabilities that were the exclusive domain of well-funded actors are progressively democratizing. This is sometimes empowering — and sometimes the worst kind of empowering.