#AIS: Palmer Luckey on Anduril

A host opens by saying this is the 15th and final episode from the All-In Summit 2022. He thanks his team for working more than 100 days on the event, thanks the audience, and says next year they will try to have twice as many people there. He then runs through a long list of production staff and speakers, a roll call that the transcript garbles in places.

He warns that some listeners have called this a controversial episode and hints at a twist. He says he will come back after the talk to give context, so that nothing is confusing or spoiled. Before the talk starts, he also addresses questions about the show's future, joking that they will never be answered and insisting he is not leaving: "The show goes on."

After a muddled stray fragment in the transcript, the speaker introduces himself as Palmer Luckey and says he has founded two companies. The first was Oculus VR, which he started at 19 while living in a camper trailer. By his account he sold it to Facebook for a few billion dollars and was fired a few years later.

He then started Anduril, saying he wanted to work in the national security space "for a variety of reasons," which he promises to explain over the course of the talk.

Luckey argues the technology industry prides itself on seeing where things are heading, yet was among the last to recognize the rise of strategic adversaries. In his telling the cause was ideological as well as commercial, and Silicon Valley took "the exact wrong position." He points to big technology companies that refused to work with the Department of Defense, with Google as one example, but says the worse cost is the startups that were never founded because people feared ruining their careers.

He says that even after selling a company for billions, investors still would not back him. According to him, the venture capital meetings were not about hiring, execution or product, since everyone believed he could do those things. They were about whether it was ethically acceptable to build a company that makes weapons.

He claims the investors who passed thought the team and product-market fit were good. In his account the sticking point was a belief that building tools capable of violence is inherently wrong, and that deterring violence with a strong arsenal is obsolete. These are his recollections of private conversations and cannot be checked from the transcript.

Luckey cites a 2019 Bloomberg story that, as he recalls it, called Anduril tech's most controversial startup. He contrasts that label with what he describes as that same year's other controversies: TikTok banning users who drew attention to a genocide in China (the transcript's wording is garbled) and who posted homosexual content, Adam Newman paying himself tens of millions of dollars for the right to use the word "we," Uber facing a federal investigation over workplace culture after a board coup, and Facebook being hauled before Congress. These are his characterizations, not verified here.

His point is ironic: a tiny defense company making a handful of purely defensive base security systems won "the belt" as the world's most controversial technology company, simply for building technology for the military.

Luckey says the war in Europe has shattered the idea that we live at the end of history, a notion he calls popular in DC. By his account, every few decades people come to believe economic ties have ended the prospect of war, and every few decades they are reminded it is not true.

He offers a historical parallel: in 1909 an English economist and politician, Norman Angel, published The Great Illusion, arguing that war in Europe was impossible and that money spent on deterrent militaries would be better spent building utopia. Luckey says the book was the number one best seller that year, and that Angel compared one European country annexing another to London annexing Hartford.

He thinks many people are now waking up, but not through reasoned principle: supporting the military, defense and Ukraine in particular has become "the current thing." In his view, defense stakes are too high and timelines too long for people to start caring only when they need to. He then says the talk will explain why he started Anduril and why the audience should think as he does.

Luckey says he expected to work on virtual reality for his whole life and had no plans to leave Oculus. He says he said as much less than 30 days before he was fired, and that the decision was made for him: he gave $9,000 to what he calls the wrong political candidate, which was unpopular in Silicon Valley. He says there are other reasons, to be covered later.

Before Oculus, he says, he worked at an Army-affiliated research center on a program called Brave Mind, an Army project that used virtual reality exposure therapy for veterans with post-traumatic stress disorder. Patients entered simulated environments that triggered symptoms and practiced coping skills with a licensed therapist, with the aim of reducing dependence on medication. He describes his own role as lab technician, a "cable monkey."

The job, he says, exposed him to how broken defense procurement was: slow, built on old technology, with misaligned incentives. He had always wanted to contribute to national security and says it took a few years for the right circumstances to arrive.

Luckey says the United States has the strongest commercial artificial intelligence industry in the world, followed closely by China. Yet he argues the US military and the prime contractors that dominate the military industrial complex lack the tools, talent and incentives to apply autonomy to their systems.

His incentive argument is that contractors are paid for doing work, not for making things that work, and gain prestige and money from having more people on bigger projects, so there is no reason to use autonomy to cut costs. He also claims the US military is well behind China's People's Liberation Army in implementing AI, and offers two comparisons: there is more and better AI in John Deere tractors than in any US military vehicle, and better computer vision in the Snapchat app than in any system the Department of Defense has deployed. These are rhetorical claims, not measured ones.

Luckey says other countries have noticed this weakness. Russia and China, he argues, do not want to compete with the US toe-to-toe. He rejects the idea that China could never fight the US because it lacks a blue water navy and has only one aircraft carrier coming online: in his view the contest will not be fought there.