Trae Stephens, Shyam Sankar: The Companies Changing Warfare Forever: Palantir & Anduril Execs on Drones, AI & the Future of War

The conversation, recorded at the Hill and Valley Forum, opens with the host welcoming Trey Stevens and Sean Sankar and noting that the two go back a long time. Asked how Palantir and Anduril are connected, the guests begin with an origin story from Palantir's early years.

In the account given, Palantir staff were then giving demos to anyone willing to watch. Trey was working at an intelligence agency, saw one of these demos, and was frustrated with bureaucracy. He applied, travelled to Palo Alto, and turned up in a full suit and tie with CIA cufflinks to interview at a company of roughly 20 people who wore T-shirts and hand-me-down clothes. A receptionist caught him in the lobby and told him to lose the tie. He was hired and helped build the government business. The host describes him as employee thirteen; he says he joined in early 2008, when the company had about 25 to 30 people.

Asked about Palantir's early setup and Peter Thiel's role, a guest describes a slow start. The five cofounders, including Peter, thought the post-9/11 argument over whether privacy or security mattered more was a false choice. Their goal was to push out the "efficient frontier": more privacy for any given level of security, or more security for any given level of privacy. The guest says the business at first was "pretty myopically focused" on a handful of counterterrorism problems for a handful of institutions.

The host asks the guests to set out their philosophy, given claims that a military-industrial complex has an incentive to promote war. The answer is that anyone who has been to war knows it is awful. That does not make it always avoidable: some actors will use might to set the rules, and a nation must be able to protect its people and interests.

War is bad, categorically bad.

The guests frame their work around deterrence. The aim is to be ready enough that, if war comes, it is won decisively and quickly. A guest says they have never met a general who wants to call parents to tell them their children died in combat. The goal at both Palantir and Anduril is described as making it "unthinkable to your adversaries that they should ever challenge you."

Asked why defense work was looked down on in Silicon Valley for so long, one answer calls the taboo a "beautiful consequence" of a peaceful world, and the speaker says this is meant unironically. Silicon Valley's origins lie in defense: Lockheed was the region's largest employer in the 1950s, and the Corona spy satellites were built there. After the Cold War, with talk of the end of history and an optimistic globalism, threats felt unreal and defense work was seen more cynically.

The turning point cited is Russian tanks crossing into Ukraine. Many people in Silicon Valley with Ukrainian or Eastern European ties were affected, and some began to see a more nuanced issue than Eisenhower's warning about the military-industrial complex. A guest points out that many who protested tech's national security work were the same people with a Ukraine flag in their bio, and calls this a mismatch in policy or understanding.

A second explanation is structural. Large tech companies were global and did not necessarily see themselves as American, and the guest says many signatures on the employee protests did not come from US citizens. That global character was probably absent during the Cold War. Asked whether things have changed, a guest says awareness of geopolitical complexity has grown and the work is much less controversial than in 2017–2018.

The host lays out figures behind the opposite worry, that Silicon Valley is taking over defense. Anduril is reportedly raising money at a $60 billion valuation and has just landed a $20 billion Army contract. Palantir is worth $400 billion. The host calls this a "twenty year overnight success" and asks how today's position compares with the defense industry after World War II.

The reply is that World War II and the early Cold War were won by an American industrial base, not a separate defense industry. Chrysler was prime contractor on the Minuteman ICBM. General Mills had a mechanics division that applied its grain-processing R&D to torpedoes and inertial guidance systems. Ford built satellites until 1990. On this account, the whole economy supported both prosperity and the security that made it possible.

Spending figures cited on major weapon systems: - In 1989, while the Berlin Wall still stood, only 6% went to pure-play defense specialists and 94% to what the speaker calls dual-purpose companies. - Today, 86% goes to defense specialists.

The speaker notes that a missile is single-use, not a dual-use product you could buy at Walmart. The point is that the companies building missiles were active in both commercial and military markets. The speaker rejects the idea that, in a crisis, auto factories could simply "flip a switch" to wartime production. History shows it took eighteen months, starting with supplies to the British and Soviets in World War II, and the speaker argues similar warning signs are being missed today.

Ukraine is the main example. A guest says that going through "ten years of production and ten weeks of fighting" should have been a "five alarm fire" showing that the US had misjudged deterrence. The stockpile was assumed to be the deterrent.

It was always the factory. It was the ability to generate and regenerate the stockpile.

The host lists readiness gaps: a 10,000-to-one drone production gap with China, a 223x disadvantage in shipbuilding capacity, and a 2027 "window of danger" for Taiwan, which the host attributes to one guest. The host then asks how ready the US is, given that it outspends every other nation on defense.

The answer begins with a caveat: the US joint force is the best in the world, and alarmist commentary should be read with that in mind. The concern is how fast adversaries are moving. The guest lists what they call an empirical loss of deterrence: the annexation of Crimea in 2014, the militarization of the Spratly Islands in 2015, Iran's breakout capability toward a bomb in 2017, a pogrom in Israel, the Houthis holding Red Sea trade hostage, and the present conflict.

The guest says US high-end capability is "still unquestionably amazing," but adversaries are investing heavily in cheap systems, and the cost gap cannot be sustained.

you can't keep shooting $2,000,000 interceptors at $20,000 drones and have that math work very long.

Losing the broad American industrial base matters, the guest argues, because it costs volume and the R&D stimulus that produces creative ideas. The example given is using the methodology of building a bathtub to make a next-generation low-cost cruise missile. Without that base, the military gets stuck on platforms that are excellent but cost about $2 million each and cannot be produced at the scale or speed needed.

Turning to an article titled "no solvency, no security," a guest says thirty years of globalization sent these capabilities abroad and gutted whole communities. Their own grandparents, aunts and uncles worked at GM, Ford, Frigidaire, National Cash Register and Armco Steel in Ohio, and every one of those factories has closed. Anduril is building a factory campus in Columbus, Ohio, transcribed as "5,000,000,000 square foot." The guest says the site can draw on local manufacturing know-how that is now underemployed.

The guest argues the US has built little new manufacturing muscle. The only at-scale manufacturer founded this century that comes to mind is Tesla. Because the capacity has atrophied, it cannot be ramped up quickly, so investment must come before a conflict rather than when it is urgently needed.