Working as Designed
Why do good people do bad things?
Preface
Palmer Luckey has a theory. It is not a complicated theory. He has been articulating it across dozens of podcast appearances for the better part of four years, and what emerges from the repetition is not so much an argument as a reflex. A way of seeing.
The theory: every dysfunctional system is producing exactly what it was designed to produce. Not what it was intended to produce. What it was designed to produce.
He stated it to Peter Diamandis like someone reading a temperature off a gauge:
And you will always get what you incentivize.
Palmer LuckeyMoonshots with Peter Diamandis
Every other argument he makes — about defense procurement, government bureaucracy, talent acquisition, geopolitical deterrence, autonomous weapons, regulation, corporate culture — is a translation of that sentence into a different domain. The world is not broken by bad people. It is broken by structures that make rational people do irrational things.
The Diagnosis
Start where Luckey starts: with cost-plus contracting, the dominant model in U.S. defense procurement since World War II. Under cost-plus, the government pays contractors for their time, materials, and labor, then adds a fixed percentage as profit. Luckey argues this inverts every incentive that matters. Contractors make more money when projects take longer. They make more money when components are expensive. They make more money, in a real and measurable sense, when things go wrong.
They get paid to do work, not to make things that work.
Palmer LuckeyAnduril Industries
The system was not always toxic. During World War II, cost-plus allowed the government to mobilize the entire industrial base without nationalizing it. It worked because everyone was already at maximum speed — multiple generations of fighter planes emerged in five years. But transplant the same structure to peacetime, where the urgency disappears but the payment mechanism remains, and the incentives reverse. Slowness is rewarded. Complexity is rewarded. Reusing building blocks you've already built means less billable work, so everything gets rebuilt from scratch.
The perversity compounds. The system doesn't just produce bad outcomes — it promotes people who perpetuate them. The people who like to run overly long test schedules naturally rise to the top. The most financially successful program managers become leaders, and their success was measured in dollars billed, not problems solved. You end up with a system where people are built to think a certain way from top to bottom. Replace the individuals and nothing changes, because the structure selects new individuals who fit it.
Even the military's own requirements are compromised. Procurement specifications are almost always deeply shaped by private sector contractors incentivized to propose only solutions compatible with their cost-plus business model. One military branch published a roadmap placing initial field trials of partially autonomous ground vehicles in 2035, with actual fielding pushed out to the 2040s and 2050s. They asked the contractors what was possible, and contractors answered within the boundaries of what their payment model would support.
Luckey draws an unexpected parallel. The only other industry with comparable cost-plus density is residential renovation construction:
Palmer LuckeyMoonshots with Peter Diamandis
Weapons systems and kitchen remodels. Same contract structure. Same dysfunction.
And here is where Luckey departs from most critics of the defense establishment:
Palmer LuckeyAnduril Industries
He extends this further than most people expect. Saving taxpayer dollars, he tells Mike Rowe, is not really the point. Money is a proxy for someone's time working in a factory, or for a building that had to be constructed that took two years to build, or for raw materials that had to be extracted from the earth. When something costs twice what it should, it means it's twice as hard to build as it should be. In peacetime, that's wasteful. In wartime, it's fatal — because even with unlimited money, that factory still takes two years to build, and that guy who was slowly building that wiring harness under a cost-plus contract only knows how to do it slowly. You cannot print productivity.
Skin in the Game
Luckey's answer is a business model. Anduril operates as a "defense product company" — a distinction he insists on with the repetitive clarity of someone who has been misunderstood too many times. Anduril uses its own money to build products, makes them work, then sells them to customers. The company bears the financial risk. When they fail, they lose their own money. When they succeed, they keep the margin.
Palmer LuckeyMoonshots with Peter Diamandis
On its most mature products, Anduril earns approximately 40% profit margins while selling at one-tenth the price of competitors. Traditional contractors operate on single-digit cost-plus margins. Anduril makes more money per unit in both percentage and absolute terms — and the customer pays a fraction of the price. Rather than distribute those profits, Luckey reinvests 100% of revenue into R&D, compared to the 1–3% typical of defense primes.
When competitors complained to Special Operations Command that it was unfair to compete against Anduril's years of self-funded development, Luckey's response was simple: they could have invested their own money. They chose not to. Why risk your own capital when the government pays you regardless?
The model has a second-order advantage visible only during political turbulence. Because Anduril doesn't depend on government funding to continue development, it keeps working even when programs hit political roadblocks. Traditional contractors issue stop-work orders. Anduril keeps building. Several programs survived funding breaks that would have killed any cost-plus contract.
And at the individual level, the incentive reversal reaches the cubicle:
Palmer LuckeyAnduril Industries
In large prime contractors, employees who finish work ahead of schedule are often punished — they make leadership look bad, reveal that programs were overly complex, or reduce company profits by finishing too quickly. At Anduril, speed earns promotion. Everything — career cycles, pay, equity grants — points the same direction.
Of course, the product model has its own risks that Luckey does not dwell on. A company investing its own money in the wrong technology burns through capital with no government backstop. Anduril bet and won. The model's virtue — bearing your own risk — is also its hazard, and the track record so far is one company's track record.
Why the Obvious Fix Doesn't Win
If the product model is so clearly superior, why hasn't it already replaced cost-plus? Luckey's answer is his most psychologically penetrating:
Palmer LuckeyAnduril Industries
Opposition can be argued against. Indifference cannot. And indifference has its own incentive logic:
Palmer LuckeyAnduril Industries
An official who approves a new vendor that fails gets blamed. An official who sticks with an established vendor that fails does not — because officials can easily reject new vendors by citing risk concerns without having to justify rejecting better business models. Innovation risk is punished. Status-quo failure is tolerated.
Behind the bureaucratic fog sits a structural trap Luckey traces to a single dinner. After the Cold War, the Joint Chiefs gathered defense executives for what became known as "The Last Supper" and delivered an ultimatum: "Consolidate or die". About 50 companies merged into five. The theory was efficiency through scale. The result was less competition, regulatory capture, wild inefficiencies, and dangerous too-big-to-fail dynamics. When 50 companies existed, you could let one die. Now, with only one company that can build fighter jets or submarines, the government must either prop them up indefinitely or lose the capability for decades.
The pattern repeats outside defense. The Wildfire XPRIZE — autonomous technology that could detect and extinguish wildfires instantly — faced opposition that was purely political. People whose jobs depended on large-scale firefighting operations threatened to campaign against any politician who funded it. Luckey refuses to vilify them:
Palmer LuckeyMoonshots with Peter Diamandis
That sentence contains the entire theory in miniature. Rational actors in a structure where success looks like a threat.
The Deeper Game
The cost-plus critique, the product model, and the inertia analysis form a complete argument about procurement. But Luckey extends the incentive framework further — into territory where it reveals both its power and its limits.
Consider his theory of war. Wars, in Luckey's model, are miscalculation events. They begin when one or both sides misestimate their probability of winning. When both sides agree on the likely outcome, the weaker side typically capitulates. When both believe they can win, blood follows.
The whole point of the war machine is not to win wars, it's to deter wars.
Palmer LuckeyThe Logan Bartlett Show
Deterrence, then, is an incentive problem. You must make the cost of aggression so legible that adversaries incorporate it into their calculus before they act. Smaller nations like Taiwan don't need to defeat China — they need to make themselves too prickly to step on.
But deterrence produces its own paradox. The more perfectly it works, the more invisible its value becomes:
Palmer LuckeyThe Free Press
The military that successfully prevents war becomes, over time, out of sight, out of mind. People forget it deterred the conflict. They stop enlisting. They stop funding. The capability atrophies — and the vacuum invites the very aggression it once prevented.
Post-World War I experts declared the end of history — economic interdependence would make European war impossible. Pre-Ukraine, people told Luckey the same thing about Russian gas and American-Chinese trade.
Human nature is somewhat fixed... People don't change their nature.
Palmer LuckeyIMPAULSIVE
When pressed on whether eliminating scarcity would end conflict, he answers no — because humans will simply value whatever remains scarce. Eliminate food scarcity, and competition migrates. Eliminate material scarcity, and status becomes the contested resource.
Palmer LuckeySpecial Competitive Studies Project
He calls pacifism a luxury belief — one that people can hold only because they are protected by others who don't share it. If that conviction became pervasive, the moment adversaries believed NATO genuinely held it, that's when they move on Europe.
And nations can lose without anyone firing a shot. The Dutch navy was once arguably the world's most powerful. Nobody destroyed it. The world simply grew faster. South Korea's birth rate of 0.67 means North Korea doesn't need to fight — they just need to keep having children and making artillery shells. Within four generations — achievable within a single lifespan — there could be 16 to 20 times as many North Koreans as South Koreans. At that point, Luckey asks, will the United States justify fighting World War III over a city with fewer people than a small town?
Here the incentive framework reaches a wall it acknowledges. No weapons system, no procurement reform, no structural redesign can make a civilization want to reproduce.
Filtering for What Incentives Cannot Create
If all of Luckey's thinking were strictly mechanistic — redesign the payment structure, watch outcomes change — he would be a systems consultant with a good theory. What complicates the picture is what happens when he talks about people.
His hiring philosophy begins as incentive design. The "Don't Work at Anduril" campaign tripled recruitment volume while attracting more qualified applicants, because it raised the psychological cost of applying — repelling careerists while attracting the mission-driven. The ads show the demanding reality: field assignments, getting dirty, no remote work. Repelling wrong candidates is more important than attracting right ones, because a hire who passes interviews but lacks genuine commitment is more dangerous than an unfilled seat.
So far, so structural. But listen to what Luckey actually looks for. His key interview question is "What have you done that you didn't need to do?" He wants candidates who have created projects in their own time with their own money that weren't mandated by work or school — people who build things because they want them to exist. Not because anyone rewarded them. Because something inside them demanded it.
He is one of these people. When Rick Rubin asks why he still works despite having billions:
It isn't the money... I love to build things.
Palmer LuckeyTetragrammaton with Rick Rubin
He describes himself as obsessive, distinguishing between a "creator mentality" consumed by the work itself and a "consumer mentality" that works for external rewards. Both are valid. But the people he hires, the people he trusts, the people he builds companies with — they come from the first category. The Mod Retro forum he started as a teenager, where hobbyists modified game consoles purely to show each other what they could do, produced the founding teams of both Oculus and Anduril. When Facebook fired Luckey, those same people quit and followed him. Not because of an equity package. Because of who he was and what they wanted to build.
The man whose diagnostic framework treats motivation as incentive-responsive builds his companies around people whose most important quality is that they are not incentive-responsive — people who would build the thing anyway, money or none. Incentive design can identify these people. It cannot produce them.
Don't Let Me
The layer that separates Luckey from the typical Silicon Valley founder who reads the world through systems thinking is the one where he argues against his own power.
Should defense companies decide how their technology is used? Luckey's answer is no, and the reasoning goes beyond deference.
When AI companies impose ethical restrictions on military use — prohibiting "offensive autonomous weapons" or "targeting innocent civilians" — Luckey notes that every key term is ambiguous. What defines "offensive"? What counts as "human in the loop"? Who determines innocence — the courts, The Hague, Brussels, or the company CEO?
If the government accepts these restrictions, it effectively grants private executives more unilateral power over military operations than the president and commander in chief. A patchwork of corporate moralities would make consistent military planning impossible.
Then he turns the analysis on himself:
Palmer LuckeyWBNS 10TV
When critics invoke Dick Cheney — arguing that democracy sometimes fails to hold leaders accountable and therefore corporate executives should exercise independent moral judgment — Luckey acknowledges the premise and rejects the conclusion. That failure describes a flaw in democracy, not a justification for corporate override. If the American people chose not to punish a leader, that reflects the democratic will, however imperfect. A corporate executive appointing himself the arbiter of accountability is, in Luckey's word, "spookier".
Had Silicon Valley permanently refused to work with the military, tech executives would have wielded more foreign policy power than the President of the United States. Anyone who argues that a defense company should go beyond what elected leaders authorize is effectively arguing for corporatocracy.
Palmer LuckeyNew York Post
It is worth noting, without cynicism, that this argument for democratic control aligns neatly with Anduril's competitive position. Luckey becomes the compliant supplier; competitors who impose ethical restrictions look anti-democratic. The principle may be entirely genuine — the consistency of his argument across years and contexts suggests it is — but the convenient overlap of conviction and commercial interest is the kind of thing an honest accounting should mention.
The Corrosive Weight
Palmer LuckeyfOx Hsiao
Corrosive. He chose that word. He has worked on entertainment, games, high-tech toys. The stakes in defense are different in kind. If he can't push a georectified target coordinate to a system that needs to respond for close air support, people die. What he tells himself and his employees:
Palmer LuckeyInternational Spy Museum
No incentive structure made him choose this work. He chose it because he believes there is no moral high ground in leaving defense work to less ethical, less competent, or less reticent people. The weapons will be built regardless. The only question is whether they will be built by people who care about the weight of what they are doing.
Trust as Precondition
The deepest challenge to Luckey's framework comes from his own analysis of why the American right is turning isolationist.
He recounts his mother's childhood lesson — a playground analogy about strong people having a moral duty to protect those they care about. This used to be the Republican position. It isn't anymore. The reason, in Luckey's telling, is that people have lost trust in the system's ability to act in the national interest. They see grift. They see poorly planned interventions. They see foreign aid that never built a single house in Haiti. And they conclude that international engagement is just another scam.
This corrodes the preconditions of everything Luckey is trying to build. You can design the most efficient defense procurement system in history, but if the public does not trust that the system serves them, they will not fund it. Trust is not an output of incentive design. It is a precondition. And the people who destroyed it — through corruption, incompetence, and self-dealing — did something Luckey considers unforgivable, because they corroded the public's willingness to support the very institutions that keep them safe.
Given this reality, Luckey proposes that America become the "world's gun store" — providing allies with everything they need to defend themselves while requiring them to do the fighting. Ukraine, Poland, Japan, the EU: give them everything, but don't send American sons and daughters to die for another country's sovereignty. Not because it is the ideal. Because it is the only position the current state of American trust can sustain.
What Holds It Together
Luckey's worldview is layered, and the layers do different work.
The surface is incentive analysis — the lens through which he identifies why procurement is slow, why bureaucrats resist change, why contractors gold-plate, why regulation stifles, why consolidation corrupts. At this level, the framework is elegant and powerful. You will always get what you incentivize.
Beneath that sits a conviction about human nature. Scarcity will always migrate to new domains. Competition backed by violence will persist as long as humans value anything others also value. Peace is a maintenance problem, and every generation that forgets this invites the catastrophe it thought it had outgrown.
Beneath that sits a moral layer the incentive framework cannot generate. Working on weapons is corrosive and necessary. Abdication is a choice to let worse people do worse work. Democratic legitimacy matters more than optimal outcomes. Trust is a precondition of collective action, and those who destroy it through graft commit something close to treason. Strong people have a duty to protect the weak — not because it is incentivized, but because Luckey's mother explained it on a playground when he was six.
And beneath all of it, the personal drive that powers the enterprise. He builds because he must. Not for money, not for recognition. Because something inside him — obsessive, relentless — demands that the thing exist.
You will always get what you incentivize. But the choice of what to incentivize, and the willingness to constrain your own power over that choice, comes from moral conviction, democratic principle, and the obsessive need to build — none of which any incentive structure can manufacture.
The next issue
Find out when the next portrait lands.
One email per portrait. We won't share your address.