,

Palmer Luckey on Steve Jobs, Oculus and Why AI Weapons Are More Moral

Palmer Luckey tells Walter Isaacson that Steve Jobs ruined technology with the iPhone, that AI weapons making kill decisions are the only moral choice, and that Anduril exists to make war with China less likely. The 44-minute conversation on Isaacson’s series The Builders was taped on February 10, 2026, and covers Luckey’s homeschooled path to Oculus, the $2 billion sale to Facebook, his firing, the founding of Anduril, and the one area of AI he would actually regulate.

TLDW

Luckey argues that the iPhone’s “easy to learn, quick to master” design put a ceiling on what people can do with computers, and admits Oculus was part of the problem. He defends Meta’s VR spending, says he should have fought his firing harder but chose to build Anduril instead, and explains why Anduril funds products before the government pays. Most of the interview is about autonomous weapons: Luckey says letting machines make precise kill decisions saves civilian lives, that executives like him should never set foreign policy, and that the scariest use of AI is helping someone build a biological weapon.

Thoughts

Luckey’s case for autonomous weapons is stronger than its headline. He does not argue that machines should be trusted. He argues that the alternative to a smart weapon is usually a dumb one, not a human one. His example is an autonomous landmine that can tell a Russian tank from a school bus full of children, and he says some people would rather keep the dumb mine that blows up the bus “as long as it wasn’t a robot who decided.” Then he goes further back: magnetic landmines, radiation-seeking missiles used since Vietnam, and ancient pit traps built to let foot soldiers pass but catch officers on horseback. On that view, weapons have been making targeting decisions for thousands of years, and AI only changes how precise those decisions are. The weak point is the one Isaacson pressed: US doctrine keeps a human in the loop for good reasons. Luckey’s answer is that doctrine reflects the technology of the time, and that a required human approval is also a radio link an enemy can hack. That is a real argument, but it means trusting that precision claims will be “measurably and provably” true, and he never says who does the measuring.

The sharpest idea in the interview is about who decides. Asked whether Anduril’s ethics extend to refusing to arm certain countries, Luckey says no: “you don’t want to live in a world where a handful of corporate weapons executives get to have de facto control over US foreign policy.” He says Anduril cannot sell to any country the US government blocks, and must work with any ally it picks, because elected officials can be held to account and he cannot. Coming from a founder whose company is “not morally neutral,” that is a disciplined position. It also reads very differently in November 2026, when the question in Washington is whether chip executives followed the spirit of export rules or only the letter. Luckey’s rule is simple: the government sets the policy and the weapons company carries it out. That is the kind of deference policymakers now say they want from the whole AI hardware industry.

His deterrence point is the most honest part of the pitch. Luckey admits that the more Anduril succeeds, “the less credit we’ll get for it.” If there is no war with China in 20 years, people will treat peace as the natural state of things rather than “a carefully engineered beautiful garden” built by dozens of companies and countries. If there is a war, the garden was built wrong. That framing explains why defense tech is a hard sell to the public: the product works best when nothing visible happens. It also pairs with his claim, made a minute earlier, that wars are already about 90% of the way to being fought mostly by machines, so autonomy will not lower the barrier to war much. Critics will push back on that number, but the logic holds. Precision weapons have already made World War II style bombing look barbaric.

The iPhone critique that gives the episode its title is less of a hot take than it sounds. Luckey’s complaint is about skill ceilings. Products are now built to be learned in a minute and mastered in ten, with nothing for the person willing to put in “a thousand hours, 10,000 hours, 50,000 hours.” He applies it to VR headsets and even brain-computer interfaces like Neuralink, and when Isaacson points out that the Rift was just as intuitive, Luckey agrees he “was part of the problem.” The interesting part is the comparison to piano, sport and dance. Those reward years of practice with abilities nobody else has. Most software today is designed to avoid ever asking for that, and Luckey’s bet is that someone will build tools again that are hard to learn and superhuman once you have learned them.

The bioweapons answer stands out because Luckey calls himself “not a pro censorship guy.” The one place he would accept limits on AI is help with biological and chemical weapons. He describes AI-run labs where the model chooses the experiments and lab technicians simply follow instructions, “using the people as meat robots,” and says a person could soon build a dangerous pathogen without anyone in the lab knowing. His fix is blunt: remove or deliberately corrupt certain data outside approved institutions, even if that means a teenager in a garage cannot design a new drug. Compared with his relaxed view of autonomous weapons, the line he draws is clear. Things that “self-replicate and live in the real world outside of that data center” get regulated, and machines that can be held to account do not.

Key Takeaways

  • Luckey says Oculus would not exist if he had not been homeschooled, because finishing schoolwork efficiently left him time for VR. Its early hardware engineers came from ModRetro, the online electronics forum he started.
  • He argues the iPhone set technology on the wrong path by optimizing for products learned in a minute and mastered in ten, with no reward for deep skill, and says Oculus was part of the problem.
  • Facebook won Oculus over Google and Microsoft by promising $1 billion a year in R&D for 10 years. Luckey says it ended up spending about $70 billion.
  • He calls Meta Horizons “a terrible product” but says the recent Reality Labs layoff of about 10% leaves a team still ten times bigger than any other VR team.
  • Luckey says he probably should have fought his firing harder, but suing would have made him a “professional victim,” so he started Anduril instead.
  • Anduril began after Trae Stephens at Founders Fund met about 2,000 defense startups and found none worth backing to take on Lockheed, Northrop and Boeing.
  • Luckey points to Anduril’s Fury aircraft and the Air Force’s Collaborative Combat Aircraft program as proof it can beat the big primes.
  • Anduril is a product company, not a contractor. It has spent billions on systems, including aircraft developed for five years, before receiving government money.
  • Lattice is the software brain behind Anduril’s 25-plus products, and as many people work on it as on all of the company’s hardware combined.
  • Luckey says AI weapons that can make kill decisions are “the only rational moral” choice when they reduce civilian deaths, provided people stay accountable for their use.
  • He argues weapons executives must not set foreign policy, so Anduril follows whatever the US government decides about who it can sell to.
  • His biggest AI fear is AI helping people make biological and chemical weapons, and it is the one area where he supports restricting data.
  • He predicts automation and vertical integration could bring a $1,000 car within our lifetime, and that megastructures at sea will make settling the oceans practical.

Chapters

02:16 Homeschooling, ModRetro and Jules Verne

Luckey says homeschooling let him finish high school and college credits efficiently and spend the rest of his time on VR. His closest friends were in ModRetro, an online community for electronics enthusiasts, and its moderators became Oculus’s early hardware engineers, starting with Chris Dycus, whom he talked out of a pizza job. He loved science fiction that focused on technology, like Jules Verne’s Twenty Thousand Leagues Under the Seas, and says he was born at the right time to turn VR from fiction into fact.

06:45 Why Steve Jobs and the iPhone Changed Technology

Asked about human-computer interfaces, Luckey says the whole industry went the wrong way. Before the iPhone, computers were Jobs’s “bicycle for the mind.” After it, everything was built to be easy to learn with a low skill ceiling, from apps to VR headsets to brain implants. He admits Oculus followed the same model and sold well because of it, but wishes people invested in computer skills the way they invest in sport or piano.

10:15 Why Facebook Bought Oculus and Meta’s VR Future

In 2014, Google and Microsoft had plans that did not need VR, while Facebook had no operating system or hardware of its own and needed a new platform. Its pitch was $1 billion a year in R&D for 10 years, and Luckey says it spent about $70 billion. He thinks Google would have shut Oculus down, citing its own cancelled VR efforts. He defends Meta’s recent cut of about 10% of Reality Labs as a decision to stop competing with game developers on its own store.

16:18 Getting Fired From Facebook and Starting Over

Luckey says he only “kind of” wishes he were still at Meta, because most of the Bay Area was against him politically, though he grants the mood has shifted. Isaacson raises his firing over a donation to a pro-Trump group, which Facebook denied was political. Luckey says at 24 he should have fought harder, but a lawsuit could not make him CEO again and would have turned him into a “professional victim.” He calls Elon Musk a singular person for running several companies at once and mentions his own side projects, ModRetro’s premium Game Boy tribute and a stake in an ice cream burrito company.

21:18 Why Palmer Luckey Founded Anduril and Took On the Primes

Luckey says the US was falling behind in defense technology, the big contractors had no incentive to innovate, and big tech refused to work with the military for the first time in US history. He first wanted to invest in defense startups, which led him to Trae Stephens at Founders Fund, who had met about 2,000 companies and backed none. After the firing they decided to build one themselves. He cites Fury and the Air Force’s Collaborative Combat Aircraft program as a win over Northrop, Boeing and Lockheed, and says Anduril funds products itself and sells them when they are ready.

26:18 The Ethics of Building Weapons, and Lattice

Anduril, named after a sword in The Lord of the Rings, is “not a politically neutral company,” Luckey says, and works with nations that protect free expression and self-government. But he says the US government, not executives, must decide who Anduril sells to. Lattice is the software that links Anduril’s missiles, submarines, fighter jets and sensor towers, so every sensor serves every weapon. Every dollar spent on Lattice improves every product, which is why it gets as many engineers as all the hardware combined.

29:29 Should AI Be Allowed to Kill Without Human Approval?

Luckey says yes, if the system is measurably more precise and causes fewer civilian deaths, and calls for regulating algorithm quality and holding people accountable instead of banning lethal autonomous weapons, as some at the United Nations propose. Isaacson notes that US drone operations keep humans in the loop. Luckey says that doctrine reflects past technology, that a human approval link can be hacked, and that landmines, radiation-seeking missiles and even ancient pit traps already make targeting decisions on their own.

35:06 Robot Wars, Deterrence and China

Luckey says we already live in an early version of machine-fought war. Precision-guided weapons cut deaths so much that World War II tactics would now be war crimes, and he puts the shift at about 90% complete. He hopes Anduril deters more wars than it enables, and says that if there is no war with China in 20 years, it will be because many companies and countries built a world where China felt it could not move.

38:51 The Biggest Danger of Artificial Intelligence

Responding to a line from Fei-Fei Li about AI responsibility, Luckey names biological and chemical weapons as his top fear. AI is opening up knowledge that used to require advanced training, and AI-directed labs can run experiments with technicians who never know what they are making. He supports removing or corrupting certain data outside approved institutions, and accepts the cost to legitimate drug discovery.

41:48 $1,000 Cars, Cities at Sea and Science Fiction

Luckey’s two picks for the future are a vertically integrated chain of power, manufacturing, processing and recycling that makes goods like cars and TVs as cheap as fabric, with a $1,000 car within our lifetime, and settling the oceans with megastructures that last thousands of years. He says people tend to build the future that science fiction writers put in front of them, and that seasteading will come back into fashion.

Notable Quotes

“Steve Jobs ruined everything with the iPhone. We’re still suffering the impacts of his incredible product design.”

Palmer Luckey, on skill ceilings in modern technology

“Was I going to be a professional victim? Or was I going to start a new company and prove that I’m not a one-hit wonder.”

Palmer Luckey, on choosing not to sue Facebook after his firing

“You don’t want to live in a world where a handful of corporate weapons executives get to have de facto control over US foreign policy.”

Palmer Luckey, on why Anduril defers to the US government

“If I put a dollar into making lattice better, everything that uses lattice gets made better.”

Palmer Luckey, on Anduril’s Lattice software

“I think it is the only rational moral outcome or decision to make.”

Palmer Luckey, on autonomous weapons making kill decisions

“The key is that a person has to be accountable for the use of force.”

Palmer Luckey, comparing AI weapons to landmines and pit traps

“The only reason that we will avoid it is because we carefully build a world where China doesn’t feel that they can move.”

Palmer Luckey, on deterrence and the risk of war with China

Watch the full conversation here.

Related Reading