Oculus VR – PJFP.com

Palmer Luckey sat down with AIAA CEO Clay Mowry and flight test engineer Jessica “Sting” Peterson at Anduril’s Costa Mesa headquarters for an episode of Up Next, the American Institute of Aeronautics and Astronautics interview series. Over nearly an hour he covers why he left consumer tech for defense, how the Fury became the first production fighter with a proper FQ designation, why America has to design weapons for the factories it still has, why patents help adversaries, how Thunder extends the loyal wingman idea to attack helicopters, why touchscreens set a low skill ceiling, and why he thinks the crust of the Earth is the next warfighting domain.

TLDW

Luckey explains that after being fired from Facebook he chose between three problems (obesity, prison reform, and national security) and picked defense because the other two were political rather than technical. He frames Anduril as a product company that spends its own money rather than a cost-plus contractor. He calls patents “Chinese instruction manuals” and says interoperability standards should be owned and enforced by the government. His core industrial argument is that the US has to design missiles that can be built in car factories and aircraft that can be built in tractor factories, as it did in World War II, and that Arsenal-1 in Ohio is deliberately built like an auto plant so the government can nationalize the designs and farm them out in wartime. He describes Lattice as an open system with roughly 700 partner companies and over 100 integrated DoD platforms. Thunder, a hybrid-electric tiltrotor built with Archer, is pitched as a loyal wingman for attack helicopters. He argues deterrence only counts for force you are politically willing to risk, that drone threats to helicopters will be solved with close-in countermeasures like the Trophy system, and that Ukraine is a snapshot rather than the permanent future of war. He admits a lot of Anduril’s gear fails in truly adverse exercises, says the iPhone set a skill ceiling that too much software now copies, and lays out his case for subterranean warfare using narrow, autonomous boring vehicles. He closes with Heinlein, Jules Verne, and his dream of a 727 re-engined with afterburning Volvo RM8s.

Thoughts

The most interesting design idea in the first stretch is small but telling. Luckey points out (around the ten minute mark) that Robert Heinlein imagined computer-flown fighters in the 1940s, before anyone had a graphical display, so the pilots in those stories simply talked to the machines. He says that is how Anduril approaches Fury: don’t give the human pilot another computer in the cockpit, let them talk to the drone the way they would talk to a wingman. This matters because the hard part of collaborative combat aircraft may not be the airframe or the autonomy. It may be how much extra work the human has to take on. A wingman you have to manage through a tablet adds to the pilot’s workload. A wingman you can brief by voice is closer to the actual promise. Later in the interview he ties this to testing. Find out early whether the tablet is unusable under real workload, because after five years of development nobody will rip it out.

The industrial argument in the middle of the conversation (roughly 18:00 to 24:00) is the part defense readers should keep. The usual story is that Detroit’s car plants were converted into tank and bomber plants. Luckey’s correction is that the US designed tanks and bombers around the welding, fasteners, bend radii, and workforce that car plants already had. That flips the question from “how do we build more aerospace capacity” to “what can we design that the capacity we still have can build.” He pairs it with a position that sounds strange coming from a founder: he expects the government to nationalize his designs in a real war and hand them to other manufacturers, he wants the government to own the IP on critical weapons, and he criticizes competitors who build capacity that nobody else could copy within ten years. Read next to his “patents are Chinese instruction manuals” line, the logic holds together. Protection only makes sense against allies, because adversaries ignore it anyway, so the better strategy is trade secrets, speed, and designs that can be copied at home.

The deterrence point at 27:45 deserves more attention than it will get. “You only get credit for deterrence for strength that you are politically able to deploy.” China knows Congress will not park a carrier with 6,000 sailors inside anti-ship missile range, so the carrier deters less than its price suggests. Autonomous systems change that because an adversary can believe you will actually use them. The argument is uncomfortable because it implies part of the value of an unmanned fleet is that losing it is politically cheap. It is also a better argument for autonomy than the usual “take the human out of harm’s way” framing. It pairs with his drone point a few minutes later. People are overindexing on Ukraine, he says, where drones dominate because the countermeasures haven’t been fielded yet. A quadcopter can be killed with a shotgun, and a small gimballed gun on an Apache would change that math. Both arguments look at the adversary’s calculation, not the current headlines.

The most honest moment is at about 36:40, when he says a lot of Anduril’s equipment “totally fails to work” in worst-case scenarios because “my guys are computer kids.” Right after that comes his long rant about touchscreens, and the two belong together. His complaint about the iPhone is not nostalgia. It is that the interface that made the phone learnable in five minutes also capped how good anyone could get with it, and then every app copied that trade-off. For an F-35 pilot at their 30,000th hour, or a helicopter pilot flying with the hydraulics out, at night, in weather they didn’t expect, the right interface is the one with the highest ceiling, not the gentlest learning curve. Defense software built by people raised on consumer apps will lean the wrong way unless someone forces the other question.

Then there is the subterranean domain (42:00 to 47:00), which he knows sounds ridiculous and says anyway. The reasoning is more concrete than the laughter suggests. Crewed boring machines are huge because people are huge. Take the people out and the diameter can shrink a lot. In rock, diameter is the expensive part, while length is almost free, because anything that follows through an existing bore travels at no extra cost. His claim is that the energy needed to move that much earth fits within batteries, tethered power, or nuclear sources. Whether or not subterrines show up in his lifetime, his larger point is fair. Air power at sea was mocked, and careers ended over it. His freedom to say this comes from controlling Anduril’s voting shares, which is itself a quiet argument for founder control in defense tech. The line to remember is his last one: most of these things “are not waiting to be invented, they’re waiting to be implemented.”

Key Takeaways

Luckey started Oculus because VR was the logical end state of PC gaming. After six monitors and multi-GPU rigs he saw a dead end and concluded the next step was presence, not more screens.
He did Oculus because he liked it. He started Anduril because he explicitly wanted his next act to be chosen for impact rather than fun.
After Facebook fired him, he considered three missions: zero-calorie foods to fight obesity, a nonprofit private prison chain paid only when people stayed out of prison, and national security.
He dropped obesity and prison reform because they were more political than technical problems.
His core worry was that the US tech industry had stopped working with the national security establishment, largely to stay in China’s good graces.
Before Oculus he worked at the USC Institute for Creative Technologies mixed reality lab on Bravemind, which used VR exposure therapy to treat veterans with PTSD.
He keeps a letter from the Secretary of the Air Force thanking his grandfather for flying as a civilian pilot in support of Desert Storm. He says that letter helped him choose Anduril.
Anduril has a public showroom and a separate one only the government can see, and products regularly move from the classified room to the public one.
The YFQ-44A Fury prototype has moved to serial production as the FQ-44. Luckey admits it is a vanity metric but wanted it to be the first production fighter with the F (fighter) and Q (unmanned) designation.
He argues much of today’s world was invented by older science fiction written by engineers who understood their craft. He calls most modern science fiction “space themed fantasy.”
Heinlein described computer-flown fighters and bombers in Astounding Science Fiction in the early 1940s. Anduril’s voice-driven approach to Fury echoes that, because pilots should talk to a drone wingman the way they talk to any other pilot.
Voice control only recently became good enough, not just at transcription but at understanding intent and turning it into something a computer can act on.
Jules Verne’s submarine, which rebuilt its batteries from minerals in seawater, points to what Luckey calls perhaps the most promising non-nuclear undersea propulsion idea today: using seawater as a reactant, the way an air-breathing turbine uses atmospheric oxygen.
Anduril sees itself as a product company. It funds development with its own money and sells finished products, instead of billing time, materials, and a fixed profit on top.
Under cost-plus contracting, the engineer who cuts a million dollars from production is penalized. In a product model, that engineer is rewarded.
Luckey concedes that some national capital assets, such as aircraft carriers, will probably stay cost-plus because there is only one buyer.
He believes everything should talk to everything, that no one should be allowed to build a proprietary silo, and that the government must own and enforce interoperability standards.
Oculus DK1 and DK2 were fully open-source hardware and software. His side company ModRetro has open-sourced its Game Boy and Nintendo 64 clones.
Anduril files very few patents. Luckey calls them “Chinese instruction manuals,” because they block Western allies from building on the technology while adversaries ignore them. Anduril relies on trade secrets instead.
The Anduril edition of the ModRetro Chromatic uses a sapphire screen lens, the same aluminum-magnesium alloy as Anduril’s attack drones, the low-IR Cerakote from the Ghost X helicopter drone, and titanium nitride on its connectors.
The consensus fix for US battlefield dominance is unified command and control, where every sensor serves every shooter across services and allies.
The real competitor is China and its partners. Chinese manufacturing equipment supports Iran’s attack drone supply chain and Russia’s weapons factories.
The US has to design missiles that can be built in car factories and aircraft that can be built in tractor factories, because automotive, agricultural, and some industrial plants are most of what is left.
In World War II the US did not simply convert car plants. It designed tanks and aircraft around the welding, fasteners, and metal forming those plants could already do.
Designing for common factories matters for two reasons: wartime scale-up, and deterrence, since adversaries weigh America’s total industrial capacity before acting.
Luckey credits organized labor with preserving most of the manufacturing that remains in the US.
Arsenal-1, Anduril’s roughly 5 million square foot plant in Ohio, deliberately looks more like an automotive factory than an aerospace one.
He expects the government to nationalize Anduril’s designs in a real war and farm them out. He supports government ownership of IP on critical weapons, despite leaning libertarian.
Some fielded systems that were contractually required to be interoperable were never actually tested. The documented calls simply don’t work.
Lattice is an open system with about 700 partner companies and integrations with over 100 existing DoD platforms. Government customers have integrated with it without talking to Anduril.
Thunder is a hybrid-electric, long-range, high-speed tiltrotor that acts as a loyal wingman for attack helicopters. It carries heavy munitions loads and vertical launch tubes for countermeasures and launched effects.
Luckey prefers jet fuel to batteries for now. Fuel burns off and can be dumped, which keeps emergency landing weights far lower than a battery aircraft that must carry its full mass into a crash.
Archer is building composite structures and drivetrain systems for Thunder, reusing components developed to FAA crewed-aviation standards for its civilian eVTOL.
You only get deterrence credit for strength you are politically willing to deploy. Adversaries increasingly believe only unmanned systems will actually be put at risk.
The Thunder launch video illustrates the concept, not the real concept of operations. Missile interception would happen miles out, not 100 yards ahead of the lead aircraft.
He expects radar-guided systems that shoot bullets out of the air, and more Trophy-style active protection on aircraft.
Drones are deadly to vehicles and helicopters today mostly because countermeasures haven’t been fielded yet. A small gimballed gun could protect an Apache.
Much of the resistance to automatic safety systems disappears when there is no human on board to be thrown around.
Exercises should be unscripted, overloaded, and degraded (damaged systems, weeks without maintenance, bad weather at night). Luckey admits a lot of Anduril’s gear fails under those conditions.
The iPhone made computing easy to learn but set a very low skill ceiling. Too many interfaces now optimize for the first five minutes instead of the 5,000th hour.
Anduril works in every domain, including space. It is working on space-based interceptors and has had AI on orbit since 2022.
Luckey believes subterranean warfare, with vehicles, people, and supplies moving through the Earth’s crust, is inevitable, and that autonomy makes narrow-diameter boring vehicles workable.
He points to the Soviet nuclear subterrine program, which he says lost its prototype underground, as evidence the problem is workable.
He can say radical things publicly because he is not in government and controls most of Anduril’s voting shares.
His dream aircraft is a Boeing 727, ideally a Valsan Super 27, re-engined with afterburning Volvo RM8 engines from the Saab Viggen, so he can do unlimited vertical climbs, including at Oshkosh.

Detailed Summary

From Oculus to Anduril: choosing impact over fun

Asked, in a nod to the Mandalorian, how he knew defense was “the way,” Luckey traces Oculus back to a gamer’s question about what the final platform looks like. After building a six-monitor, dual-GPU setup, he concluded that more displays and more graphics cards led nowhere and that virtual reality, which tricks the subconscious into believing you are present, was next. Oculus made its investors hundreds of millions of dollars each, but he did it because he wanted to. When Facebook fired him about three years after the acquisition, he decided his next project would be chosen for impact. He weighed zero-calorie foods built on long-chain hydrocarbons to fight obesity, a nonprofit prison operator paid only for keeping people out of prison, and national security. He picked defense because the other two were political problems. He wanted to pull engineers away from building “augmented reality mustache emojis” and toward autonomous fighter jets and robotic submarines. His early work on the Army-affiliated Bravemind PTSD therapy project, and his grandfather’s letter from the Secretary of the Air Force, both fed into that decision.

Fury, the FQ-44, and science fiction written by engineers

After a tour of Anduril’s public showroom (the other showroom is government-only), co-host Sting Peterson asks about Fury’s path through the Collaborative Combat Aircraft program. Luckey notes the YFQ-44A prototype is now in serial production as the FQ-44, and that he wanted it to be the first production fighter with a proper F and Q designation. Both bond over the 2005 box office flop Stealth, which Peterson saw as a kid and which made her want to work in aviation. Luckey argues the world we live in was invented by older science fiction, written by NASA and aerospace engineers for whom the science mattered as much as the fiction. He jokes that his wife finds these novels unreadable because the characters only exist to deliver technical ideas. Heinlein wrote about computer-flown fighters and bombers in the early 1940s and imagined voice or punch-card commands with no displays at all. Luckey says that is effectively how Anduril is building Fury: you want to talk to it like any other pilot. Clay Mowry adds that AIAA’s forerunner, the American Rocket Society, was founded in the 1930s by science fiction writers and rocket enthusiasts whose work led to the engine on the Bell X-1.

Jules Verne and seawater batteries

One of Luckey’s favorite childhood books was Twenty Thousand Leagues Under the Sea, which he describes as a thin story wrapped around maritime technology. Captain Nemo doesn’t recharge his batteries. He rebuilds them from zinc and magnesium pulled from seawater, which Luckey calls a continuous underwater battery manufacturing system. Anduril isn’t building this, but he has looked at it. He notes that L3Harris bought the company doing the best work on saltwater-reactive lithium fuel cells, and that oxide buildup on the plates is the practical problem. Using seawater as a reactant is like an air-breathing turbine, which is why turbines beat rockets. The segment ends with Luckey singing “A Whale of a Tale” from the Disney film.

Product company, open standards, and why patents help adversaries

Luckey describes Anduril as a product company that picks what to build with its own money and sells finished products. Cost-plus contractors, by contrast, get paid for time and materials plus a fixed margin, which penalizes cost cutting. Some assets like aircraft carriers will probably stay cost-plus because they have only one buyer. He says everything should talk to everything and that the government should own interoperability standards. He is a longtime open-source advocate: early Oculus development kits were fully open, and ModRetro has open-sourced its Game Boy and N64 clones. On defense work, open-sourcing usually isn’t allowed, but Anduril files few patents because patents publish the design for adversaries who ignore IP law while blocking allies for the life of the patent. Anduril keeps its work as trade secrets instead, and if someone copies it and executes better, “they deserve to win.” He also describes the Anduril edition of the ModRetro Chromatic, which uses drone-grade alloy and Cerakote.

Designing weapons for the factories America still has

Asked what it will take to regain battlefield dominance, Luckey starts with the consensus answer, unified command and control and information sharing across services and allies, then moves to his more debated point. China supplies manufacturing equipment and support to Iran’s drone programs and Russia’s weapons factories, so the US has to design weapons its remaining industrial base can build: car plants, agricultural equipment plants, and some industrial plants. He says the World War II story of converting car factories is not quite right. The US designed tanks and aircraft around the welding, fasteners, bend radii, and heat-treatment processes car makers already used. That matters for wartime scale-up, since anything that needs hand-laid composites in a bespoke aerospace facility won’t scale, and for deterrence, since China should know GM could produce cruise missiles by the hundreds of thousands. Arsenal-1 in Ohio is built to look like a car factory on purpose. Luckey expects the government to nationalize his designs in a major war, criticizes companies that build capacity no one else can copy, and supports government ownership of IP for critical weapons, even though he leans libertarian.

Interoperability that actually works, and Lattice

On connectivity, Luckey says the government must actively enforce the standards it owns. Anduril has run into fielded systems whose contracts required interoperability, yet the documented interfaces were never tested and don’t work. On paper they are open. In practice they are silos. He pushes back on the idea that Lattice is closed: about 700 companies are in its partner program, it integrates with more than 100 existing DoD platforms and every messaging system and radio Anduril can get, and some government customers have integrated with it without involving Anduril at all.

Thunder, deterrence, and the drone countermeasure gap

Thunder is a hybrid-electric tiltrotor, not a pure electric aircraft. Luckey loves jet fuel because it burns off and can be dumped, while batteries force every emergency landing to carry their full weight, and the landing gear and crash structures that requires get heavy fast. Archer supplies composite structures and drivetrain components built to FAA crewed standards, which Anduril chose to reuse rather than redesign. If the CCA is a loyal wingman for fighters, Thunder is one for attack helicopters, a forward sensor and shooter that goes in before people do. Luckey argues deterrence only counts for force you will actually use, and no one believes Congress will risk a carrier and its 6,000 sailors inside Chinese missile range. He says the Thunder launch video illustrates ideas rather than tactics: interceptions would happen miles out, and the countermeasures would be canister, electronic, and kinetic rather than literal Anvil drones. He expects radar-guided systems that shoot down bullets, points to the Trophy active protection system on armored vehicles as a model for aircraft, and says drones threaten helicopters mainly because cheap close-in defenses aren’t fielded yet. People overindex on Ukraine, he says, which is “a reflection of a moment in time.”

Trusting autonomy and testing in the worst case

Peterson, who has worked on ground and air collision avoidance, asks how to build trust in AI and collaborative aircraft. Luckey, a helicopter pilot himself, notes that pilots dislike automatic systems that yank them around, and that problem disappears when nobody is on board. Commercial pilots also work within chauffeur-like constraints and avoid abrupt maneuvers, while a robot will take the most evasive action at the first sign of trouble. Simulators help, because a GPS-jamming scenario that kills most pilots can become one where everyone lives once safety systems are integrated. When Peterson points out that things that work in the sim often fail in flight, Luckey agrees that exercises are too scripted. He wants overloaded, degraded scenarios: systems shot out, three weeks without maintenance, hydraulics out at night in unexpected weather. He admits a lot of Anduril’s equipment fails in those conditions because its engineers are “computer kids,” which is why testing has to happen early, before a bad interface choice becomes five years of sunk cost.

The iPhone skill ceiling rant

Touchscreens set Luckey off. He respects Steve Jobs’s “bicycle for the mind” goal but argues the iPhone made computing so easy to learn that it capped how skilled anyone could become. A keyboard and mouse take thousands of hours to master but become a superhuman interface, like an Excel power user running macros at 150 actions per minute. The problem is not the iPhone itself. It is that everything became an iPhone, optimized for the first five minutes. He praises chorded keyboards and vector swipe keyboards as ideas that never caught on because nobody wants to invest hundreds of hours anymore. He wants technology designed for what an F-35 pilot can do on their 30,000th flight hour, not constrained by what Jobs showed on stage in 2007.

Space, and the case for subterranean warfare

Anduril works in every domain. It is publicly working on space-based interceptors and has had AI on orbit since 2022. The “weird one” is the subterranean domain. Luckey doesn’t mean tunnels or bunkers. He means vehicles, people, and supplies moving through the Earth’s crust as a three-dimensional battlespace. The US and Soviets both pursued subterrines. Autonomy removes the need for people-sized bores, and since diameter is expensive and length is cheap, the optimal design is very narrow and very long. The energy to displace or compact that earth, he says, fits within batteries, tethered power, or nuclear sources. He compares the ridicule to what early naval air power advocates faced, notes that his voting control of Anduril means no one can fire him for saying it, and cites the Soviet nuclear subterrine that was reportedly lost underground as proof the problem is workable. He retells the scene from The Core where a general shows a scientist a check and asks, “Would this be enough?” and says he wants the government to ask him that question. Mowry adds Journey to the Center of the Earth, and Luckey closes the thread by saying these ideas are waiting to be implemented, not invented.

Jetson ONE, a Black Hawk, and the afterburning 727 dream

Luckey was the first owner of a Jetson ONE, which he calls the Polaris of the sky: a short-range thrill ride with a redundant architecture that can lose about half its rotors and still land. He owns a UH-60 Black Hawk assembled from surplus parts on an FAA restricted certificate, bought before the Army began surplusing them cheaply, and a 1985 ex-Marine Corps Humvee bought when real ones were rare. His daily flyer is a Eurocopter EC120. His dream aircraft is a Boeing 727, his late grandfather’s favorite in 45 years at United Airlines, ideally a Valsan Super 27 conversion. His secret plan is to fit it with Volvo RM8 engines, the licensed, afterburning, thrust-reversing version of the Pratt and Whitney JT8D built for the highway-capable Saab Viggen, so he can request unlimited vertical climbs from the tower and fly it to Oshkosh with his grandfather’s 727 paperwork on board.

Notable Quotes

“I wanted to try to get people out of big tech and into work on national security problems with the same rigor and vigor that they were working on consumer electronics products and social media products.”
Palmer Luckey, on why he founded Anduril after leaving Facebook

“I often call patents Chinese instruction manuals. You’re just putting everything out there for an adversary to rip off.”
Palmer Luckey, on why Anduril relies on trade secrets instead of patents

“We need to design missiles that can be made in car factories. We need to design aircraft that can be made in tractor factories. And we’ve done this before. We did this in World War II.”
Palmer Luckey, on rebuilding US defense production around the industrial base that remains

“I fully anticipate that the government is going to nationalize my designs, farm them out to a whole bunch of other people. This is what we did during World War II as well. But we need to be building for that assumption.”
Palmer Luckey, on why Arsenal-1 is built to look like a car factory

“You only get credit for deterrence for strength that you are politically able to deploy.”
Palmer Luckey, on why autonomous systems carry deterrent weight that crewed ones increasingly lack

“People are overindexing on what warfare looks like in Ukraine. They’re saying, oh, this is the future of warfare. And I think it’s actually a reflection of a moment in time.”
Palmer Luckey, on why drone dominance over vehicles and helicopters will fade as countermeasures arrive

“A lot of our stuff totally fails to work when you get in those scenarios because my guys are computer kids.”
Palmer Luckey, on the gap between scripted exercises and real combat conditions

“The same interface that made it easy to learn to use also put a maximum skill ceiling on it that was very, very, very low.”
Palmer Luckey, on the iPhone and the touchscreen-ification of everything

“It is inevitable at this point. The only thing stopping us is that it sounds so crazy.”
Palmer Luckey, on subterranean warfare as the next warfighting domain

“They’re not waiting to be invented. They’re waiting to be implemented. And I’m less of an inventor and more of an implementer.”
Palmer Luckey, on living in an age of unprecedented possibility

Watch the full Up Next conversation with Palmer Luckey here.

Tag: Oculus VR

Palmer Luckey on AIAA Up Next: Anduril’s Fury FQ-44A, Designing Missiles for Car Factories, Patents as Chinese Instruction Manuals, the iPhone Skill Ceiling, and Why Subterranean Warfare Is the Next Domain