Stop Hunting Unicorns
Why the Department of War must design for what we can actually build.
Greg Little is a senior counselor at Palantir Technologies.
Aaron Jaffe is a deployment strategist at Palantir Technologies.
A few years ago, one of us sat in a conference room late in the afternoon as a major weapon system program team debated a marginal performance improvement. Single-digit percentages. Decimal places. Over hours of discussion, what no one talked about was how many of these systems could be built, how fast they could be replaced, or whether the parts could be sustained at scale.
That meeting captures a pattern the Department of War knows too well: we design the ideal weapon first and only later ask whether the country can actually produce it in the quantities a prolonged fight would demand.
This is why Secretary of War Pete Hegseth’s recent acquisition transformation announcement matters. His focus on speed and scale, on shifting the department to a wartime footing, and on breaking a culture that rewards compliance over execution points us in the right direction. It acknowledges something fundamental: capability delivered late—or in boutique quantities—is not capability at all. But if we really want speed and scale, acquisition reform alone isn’t enough. The deeper problem isn’t how we buy weapons. It’s how we design them and coordinate their production.
Designing for Scale
For decades, the department has treated manufacturing and the industrial base as downstream concerns. Requirements are written as if time, production capacity, and supply availability are secondary. Only after the system is defined—sometimes after it’s certified—do we focus on scaling production. This approach produces impressive prototypes. It does not produce the mass and responsiveness required for deterrence at scale. If we want to field lots of weapons fast, we need to flip the logic: design weapons systems to fit the industrial base we have and then iterate—rather than designing unicorns and hoping the supply chain catches up.
World War II is often cited as proof that America can surge production when it matters. That’s true, but the reason we surged is often misunderstood. Yes, we mobilized labor. Yes, we built factories. But the real breakthrough wasn’t just industrial effort—it was design choices that embraced industrial reality.
Consider aircraft production. In 1940, the U.S. aircraft industry was small and craft-based, producing roughly 3,600 aircraft per year. By 1944, it had become the world’s largest aircraft producer, outproducing Germany and Japan combined—and fielding over 96,000 aircraft that year alone. This transformation didn’t come from engineers pursuing optimal designs. It came from designing aircraft around manufacturing capabilities first, with performance improvements emerging organically through production experience and operational feedback.
Within four years, the U.S. designed, introduced, and built more than 30 different bombers, iterating from the workhorse B-17 to the advanced B-29 with more than twice the power, range, and load. By comparison, the latest bomber produced in the United States, the B-2, required more than a decade to move from design to the initial production of a fleet of 21 aircraft.
America’s rapid wartime evolution was spurred by a collaboration between government and manufacturers to maximize the use of their existing capabilities. At Willow Run, Ford—the automobile company—adapted aircraft designs to the logic of automotive assembly, using standardized parts, simplified subassemblies, and repeatable processes. By 1944, a B-24 bomber was rolling off the line nearly every hour. Ford optimized for throughput, not perfection—and throughput won the war.
This story was repeated across the American industrial base. At the Detroit Arsenal, Chrysler applied the same logic to tank engine production. The Sherman tank wasn’t designed to win one-on-one duels against the most advanced enemy armor. It was designed to be built quickly in enormous quantities and to be easily repaired in the field. The result was a tank that could be produced and replaced faster than adversaries could destroy it.
This design-for-scale philosophy is not a complete relic of World War II—it’s re-emerging today where speed and volume matter most. Anduril’s Barracuda cruise missile is a modern example of this same logic. Rather than designing a missile optimized solely for performance and then struggling to manufacture it, Barracuda was designed from the outset to be compatible with high-volume production, using components, processes, and tolerances that allow it to be built on current automotive manufacturing capabilities. The result is a system explicitly designed for mass production, rapid replenishment, and iterative improvement.
This is an entirely different mindset for manufacturing, and it has strategic consequences for the nations that embrace it. Production reality becomes the key design constraint, not an afterthought. Interchangeability isn’t just a maintenance convenience—it’s a warfighting advantage. Designs that can tolerate variation and substitution proliferate. Designs that can’t are redesigned or abandoned. The lesson is simple and uncomfortable: the United States doesn’t scale by asking industry to stretch to meet perfect designs. It scales by aligning designs to industrial strength.
Here’s where this gets uncomfortable. If a modern weapon system depends on a bespoke material, a limited supply base, or a certification regime that freezes design for a decade, then it may be technologically impressive—but it is strategically brittle. In a contested industrial war, elegance is not resilience. A system that cannot tolerate supplier substitution, component swaps, or rapid redesign is not “high-end.” It is pre-failed the moment production is disrupted or losses exceed peacetime assumptions.
We often describe these bad outcomes as manufacturing problems or supply-chain shocks. And yes, we need investment to expand industrial capacity and deepen critical supply chains. But at root these are design decisions—made early, reinforced by incentives, and rarely revisited. And then we act surprised when scale never materializes.
Orchestrating Production
Too many in the Department of War still operate on the assumption that if we simply fix acquisition policy, streamline contracts, and empower program offices, the rest of the system will somehow align. It won’t, and Secretary Hegseth’s acquisition transformation recognizes this. Even if every acquisition reform memo is executed perfectly, the department is still missing something fundamental: a way to operate the war machine as a single, connected system.
Today, weapon design lives in one world. Manufacturing capacity lives in another. Supply chains live in spreadsheets. Sustainment data arrives late. Readiness has little connection to its industrial inputs. Learning happens—but slower than the fight evolves. This is a coordination problem.
What’s missing is a war machine operating system—not a product, but a capability. A digital layer that connects weapon design decisions, factory throughput, supplier health, substitution options, and sustainment realities into one continuously updating picture.
There is precedent for this idea, and it comes from an era long before software existed. During World War II, the United States effectively created a human version of a war machine operating system. Through institutions like the War Production Board—and figures like William S. Knudsen, a former auto executive elevated to general officer and placed in charge of industrial production—the government unified design, manufacturing, supply chains, and battlefield demand under a single coordinating authority. Knudsen didn’t just manage contracts. He managed tradeoffs. He could simplify designs, redirect factories, force standardization, and prioritize throughput over elegance. He had visibility into what industry could produce and the authority to shape what was being designed accordingly. Knudsen wasn’t merely a bureaucrat. He was the conductor of a vast industrial orchestra.
What’s different today is that this coordinating function no longer has to live in a handful of individuals or emergency boards. It can live in software. In limited cases, it already does. During recent Navy ShipOS pilot deployments, AI-powered manufacturing and sustainment capabilities demonstrated what a war footing looks like in practice. At General Dynamics Electric Boat, submarine schedule planning was slashed from roughly 160 manual hours to fewer than 10 minutes. At Portsmouth Naval Shipyard, material review timelines that once took weeks were compressed to less than an hour. These changes show how a wartime industrial system moves when information, authority, and execution align.
The same model is already visible in commercial manufacturing enterprises that must operate continuously, at scale, and under real-world stress. Airbus’s Skywise platform connects design data, production systems, maintenance records, and operational telemetry across thousands of aircraft—creating continuous feedback between engineering, production, and operations. This is how modern industrial systems function when downtime, delay, and uncertainty are unacceptable.
War Footing
The key takeaway for the Pentagon is that operating a complex industrial machine at scale on a war footing is already possible. The limiting factor is adoption, integration, and leadership. Today, those feedback loops are fractured in the defense industrial base. Requirements, engineering, production, logistics, and sustainment are treated as sequential phases instead of one living system. By the time a problem surfaces, the opportunity to adapt cheaply has already passed. Now that the Pentagon is taking a war footing mindset and approach, it needs an operational backbone that treats design, production, and sustainment as one system, not as disconnected handoffs. Without this connective tissue, the department is forced to manage by lagging indicators and intuition—exactly the opposite of what a wartime footing demands.
Secretary Hegseth’s push for speed and performance is necessary—but it will only deliver results if the department changes what it rewards upstream. Designing for substitution, modularity, and common components is a strategic choice. It’s how you stay in the fight when conditions change—and they always do.
So here’s the rule leaders should apply before approving the next major weapon system: can this weapon be built, repaired, and replenished faster than it will be destroyed? If the answer is no, no amount of acquisition reform, contracting flexibility, or urgency memos will save it. In a fight defined by attrition, adaptation, and industrial endurance, the winning systems will not be the perfect ones on paper but the ones that can be produced, replaced, and improved the fastest.
America didn’t win past wars by hunting unicorns. We won by designing for scale—and then scaling relentlessly.
 |
|