Ukraine’s Real Weapon Isn’t a Drone

I'm back from Ukraine. A week of exposure to top military leaders, defense tech founders, and people living under active siege. 

I went in with one thesis about the future of conflict and technology; I came out with ten. This is the first:

Ukraine has built the fastest-learning military organization I’ve ever seen - centralized strategy, decentralized innovation - and the architecture behind it should keep every large, slow, well-resourced institution up at night.

For most of modern military history, the threat your enemy was building took years to materialize. The Germans developed jet engines over the course of a war. The atomic bomb was a multi-year program. Even in Iraq and Afghanistan, the innovation story was mostly incremental - each side grinding through months of adaptation before the other side responded.

Centralized command organizations had time to see the threat, deliberate, fund a response, and get it there.

In Ukraine, that model is gone. The threat doesn't announce itself with a multi-year development timeline anymore. It just shows up drastically differently next week.

Any system that requires the center to identify the problem, convene the right people, approve a solution, and push it down the chain is structurally too slow for this war. The enemy has already iterated twice by the time the memo clears.

Ukraine couldn't fix that by outspending Russia.

The answer wasn’t another weapon, it was organizational. A central command sets grand strategy: the big objectives, the long-range priorities, overall budgets and major procurements, where the war is going. Everything below that, the problem-solving, equipment decisions, on-the-ground innovation, belongs to the regiments and units at the edge. 

The people in direct contact with the threat figure out how to deal with it, they don't wait for permission.

Elegant in theory. But in practice, we know that’s basically impossible to pull off. Especially for an org that loves hierarchy like a military¹.

How does the center know what's working? The hard part isn’t just empowering the edge. It’s knowing what to copy.

The first is Delta, integrated battlefield management software that looks, without exaggeration, exactly like a video game². I got to see it firsthand. A live map of the entire battlefield. Every unit on it. Click any dot, and you're watching through their camera feed in real time. Click another, and you're looking through a drone mid-operation. Everything networked, everything visible, every perspective one click away. It worked great on the big screens for the military. I was told it’s pretty slick in your pocket or on a tablet on the front lines, as well.

You'd think the new "toys" that NATO military leaders might want most are some fancy new unmanned aerial or ground vehicle. No, it's actually a piece of cloud software. I haven’t met a single American military official who had seen Delta and wasn’t envious of it.

Layered alongside Delta is the Brave1 marketplace, essentially Amazon for Ukrainian warfighting. A unit needs FPV drones? They go to the portal, filter by type, pick their configuration, and it arrives³. Remarkable enough. But the ordering mechanism is almost the least interesting part.

The latest innovation here is what is called the e-points system. After a unit deploys, they upload confirmed kill footage into Delta. Think of it like a receipt. The system scores it. 

Taking out a tank is worth 40 points. Only manage to immobilize it? That will still net you a cool 20 points. A soldier casualty? 12. Fascinatingly, you even earn points for replacing what could have been a manned action with an unmanned system instead. Those points convert directly into purchasing credit to be spent acquiring more military hardware for the unit on Brave1⁴. 

The incentive is explicit: don't waste a drone on one guy in a field when there's a tank you could be hitting. Aim for what gets you more resources to keep going.

What this creates is a feedback loop with almost no politics in it and no lag.

Central command isn't making procurement decisions based on what some general has been championing for three years. They're watching what units reorder and what they stop touching. 

The footage is the record. The purchasing behavior is the signal. When ten units independently stop ordering one drone and start stacking orders for another, you don't need a meeting. You go buy 20,000 of the one that works, this week, not next quarter. 

This didn't require the most innovative genius to sit at the top, but it did require a genius setup. Thinking about it on my way back home, I was reminded about a management philosophy we’ve seemed to have abandoned.

There was a time when some of the best companies explicitly assumed headquarters wasn't where the best ideas lived. Back in the 1940s, 3M institutionalized that with a 15% rule, it gives scientists room to chase ideas nobody asked them to work on. You can thank them for the Post-it Note. 

Google ran a louder version in Silicon Valley with 20% time. Autocomplete suggestions, Google News, and key pieces of AdSense all came out of engineers scratching an itch nobody assigned them. The point wasn't free time. The point was acknowledging that the edges of an organization often see things the center can't.

Then, somewhere along the way, we optimized the exploration out of our organizations.

Modern companies are disciplined execution machines. OKRs, quarterly sprints, dashboards, an AI copilot watching the project management software watching the project management software. All of it rewards extreme predictability. Exploration is, almost by definition, unpredictable. So we've landed on a paradox: change is happening faster than ever, breakthrough innovation matters more than ever, and organizations have become less structurally willing to tolerate the wandering that it actually requires. 

You can hit your quarterly EPS guidance to the decimal and still consistently miss the one messy thing that would have actually grown the business.

The bandaid was hackathons - sequestered, banal "extracurriculars" for the smart people at the edge. Optional, risk-averse, an after-school activity, lest it distract from the more critical job of rent-seeking optimization. I sadly even see this in too many younger tech companies. Leaders playing innovation theater keep asking for the next Post-it Note while quietly eliminating Post-it Note time.

Ukraine went the opposite direction. Headquarters doesn't try to dream up every new drone tactic or battlefield software tool. They know they can’t out-innovate the edge, so they learn from it.

Every military is studying Ukraine as an organizational petri dish, not just a technological one. Who gets to buy things. How ideas spread. What gets rewarded. 

I think we need to bring this closer to home.

For decades, business schools have assigned war books: Clausewitz, Sun Tzu (I am the dork who read both in middle school for fun, which tells you everything you need to know about me as a child). 

War is the only environment where the cost of a wrong decision is total, no second chances, which forces a quality of strategic thinking and strategic risk-taking that no peacetime organization ever had to develop. If you wanted to understand how to lead under genuine pressure, you read the people who had no other choice.

Ukraine is writing the next one, and unlike Clausewitz, it's still live.

The reason this is the case study for right now is what AI has done to the pace of innovation. New capabilities become products before the last ones have been fully adopted. New competitors appear without warning and without the courtesy of a multi-year development timeline. More organizations need to be in a wartime mentality. The organizations that survive this environment aren't the ones with the best strategy at the top. They're the ones who built a system where the edge learns and iterates faster than the ground shifts beneath it.

That system now exists in Ukraine because the alternative was losing a war. Most organizations haven't built it yet because the alternative has only recently started to feel that urgent.

Everyone wants to copy Ukraine’s drones or new software. I think they’re looking at the wrong layer of the stack. The real advantage isn’t these products alone. It’s the system that keeps producing these capabilities. Steal that instead.

Five principles worth stealing from the Ukrainian military