Shyam Sankar

The binding constraint on US military power is increasingly the factory, not the quality of its best weapons. Sankar contrasted Ukraine consuming “10 years of production in 10 weeks” with an industrial structure that shifted from 6% of major-weapons spending going to defense specialists in 1989 to 94% today. Friedberg added the stark relative-capacity figures: a 10,000-to-one drone-production gap and 223x shipbuilding disadvantage versus China.

Anduril’s wager is that a modular, privately financed manufacturing platform can restore the volume traditional procurement lost. Its five-million-square-foot Arsenal-1 campus in Columbus is designed to pivot among Roadrunner, Barracuda, Furies and future systems like a contract manufacturer, avoiding the Ukraine-era spectacle of calling retirees back to reconstruct Stinger and Javelin assembly lines. Stephens said the model requires tremendous capital and probably cannot support “a hundred new primes.”

The required weapons mix is shifting from exquisite scarcity toward attritable mass, but not toward an all-drone force. Sankar’s simplest economic warning was that “you can’t keep shooting $2 million interceptors at $20,000 drones”; Stephens still called B-2s, Patriots and existing precision weapons exceptional, but said the high-low mix must change massively over five to 10 years. Even with unlimited cash, he estimated it would take roughly 18 months to put the country on track for a sustainable industrial base.

Defense innovation remains a monopsony fight in which products must often be validated from the battlefield backward. Palantir had to sue its customer for the right to compete, while deployed users bent rules to obtain software that worked; Anduril then reached $10 million of annual revenue in 22 months versus roughly five years for Palantir by reusing that institutional knowledge. Sankar’s governing archetype is the “heretic” founder who builds what the buyer does not yet know to request.

Defense-tech capital will follow a power law, making valuation discipline as important as access to money. Stephens warned founders that they “could raise less at a lower price” and said Anduril deliberately lowers its revenue multiple with each round, including from Series A to Series G, ahead of a medium-term IPO. Public capital can help most where supply chains bottleneck—brushless motors, critical minerals and refining—not by “peanut butter” funding every entrant equally.

Autonomy does not eliminate human responsibility; the speakers’ preferred model is accountable command rather than a universal human-in-the-loop rule. Stephens cited CIWS, which must respond to incoming threats autonomously, while leaving someone on the vessel accountable for its actions. He argued AI can improve discrimination and reduce civilian casualties, and that refusing to participate is not morally neutral: “You are making a moral decision when you decide to abstain.”

The 2040 fork is economic as much as military: dependency could produce a Chinese-led order, while reindustrialization could rebuild both deterrence and the middle class. Stephens highlighted that 80% of APIs for generic drugs are produced by China and warned that semiconductor dependence is difficult to reverse before the cited 2027 Taiwan risk window. Sankar’s downside is a Chinese century of “vassal states” playing by rules America no longer helps set; Stephens’s upside is reindustrialization, trusted institutions and a middle class that believes its children’s future will be better.

🔗 Original source & video: The State of Modern War: Palantir & Anduril Execs on Drones, AI, and the End of Traditional Warfare

Shyam Sankar’s defense thesis is that America lost deterrence by turning whole-country mobilization into a specialist industry. In 1989, only 6% of major-weapons spending went to defense-only companies; now it is 86%. Reversing that isolation means reconnecting commercial R&D and manufacturing to national security because “when a country goes to war, it’s the whole country.”

The defense opportunity is not merely adding competitors but restoring founders, heretics, and the leaders willing to protect them. The post-Cold War “Last Supper” reduced 51 prime contractors to five, but Sankar argues the deeper damage was that “consolidation bred conformity,” shifting management toward dividends, buybacks, and cash flow. The counter-model is the Higgins boat: rejected by the Navy, yet ultimately 92% of all boats in World War II.

AI threatens beta SaaS far more than software that expresses a customer’s competitive advantage. Sankar’s rubric separates software that makes every company more alike from alpha-oriented platforms that help each operate differently; AI and vibe coding intensify the latter. Day-two maintenance remains “much harder” and partly unsolved, but he does not think that will protect memetically purchased, standardized software.

His AI-stack thesis places durable value at the chip and AI-infrastructure—or “ontology”—layers as models commoditize. Model companies are moving upward into software “harnesses,” while narrow applications are building downward into infrastructure to support more customers and use cases. That convergence leaves standalone models under pressure while chips and infrastructure may remain defensible, in his theory.

The preferred macro outcome is not labor replacement but AI-enabled reindustrialization that repairs the break between wage growth and GDP growth and reconnects production with innovation. Sankar points to Hadrian making people “50 to 100 times more productive” and calls AI “David’s slingshot” against China’s manufacturing scale. His warning is that separating invention from production was a strategic error: “If you don’t make the thing, you can’t innovate on how you make the thing and what the thing is.”

AI’s immediate organizational advantage belongs to domain experts who can now build instead of petitioning bureaucracies. An intel warrant officer can spend two weeks producing a working application rather than making a PowerPoint and seeking permission; sales teams similarly need an “Iron Man suit,” not replacement for its own sake. The objective, Sankar says, should be to “dominate my industry,” not satisfy an article of faith that people must be replaced.

Sankar ultimately sees America’s largest strategic danger as “suicide, not homicide”—a collapse of national will, agency, and functioning institutions. His answer spans military reserve talent, institutional competence, maximalist reindustrialization, and entertainment that makes heroism attractive again. Hard power and optimistic storytelling share one purpose: mobilizing the next generation early enough to prevent a larger war.

🔗 Original source & video: Palantir CTO on The SaaS Apocalypse & Preventing The Next World War | a16z

Sankar’s organizing thesis is that military innovation only ever comes from heretics — “they’re really founders” — with a categorical claim attached: “the only things that ever worked, the things that helped us win all the wars, were the things that the heretics actually did. Nothing that went through the machine delivered anything.” Hyman Rickover built the first nuclear submarine in 7 years from an office the Navy assigned him in the women’s restroom, and that ’50s-era sub force remains “one of our last remaining asymmetric advantages against the Chinese.”

America has lost deterrence and, arguably, only started restoring it in the last year — Crimea 2014, the militarized Spratly Islands, Ukraine, gray-zone operations in the South China Sea — but Midnight Hammer and the Maduro operation were “massively deterring,” and Sankar has seen “more change in the Department of War in the last year than the prior 19 years.” Some new defense entrants are now discussed inside the Pentagon as “plan A” — “unrecognizable from the world I started in 2006,” with hundreds of founders building in the national interest with a US capital stack and over $100B invested in their projects.

The industrial-base decay is quantified: in the Cold War era only 6% of major weapon-systems spending went to defense specialists; today it’s 86%. The 1993 “Last Supper” collapsed 51 primes to 5, and the real damage wasn’t lost competition but “financialization and conformity… you lost the crazy people.” Primes run ~9% operating margins at under 2x revenue — his fix is to make them more valuable and address cost-plus contracting, which he calls structurally “anti-heresy.”

“The biggest lie that we bought from globalization is… we will do the innovation and they will do the production.” Innovation is downstream of production: the “Attention Is All You Need” paper came from a 3% improvement to Google Translate, WuXi went from pipetting-for-hire to 50% of all clinical trials being drugs created in China, and 80% of US generic drugs come from China in one way or another. Apple has spent “the equivalent on an inflation-adjusted basis in the last 5 years of two and a half Marshall Plans” building talent and capacity in China; “how about we try to spend one Marshall Plan here?” AI is “David’s slingshot” — a 50x-more-productive American worker changes the efficient frontier of what can be made here. He is a “re-industrialization maximalist”; friend-shoring “can let us off the hook too easily.”

On AI value capture: “the value is going to accrue at the chips layer and at the ontology layer.” Models are commoditized — model companies are running up the stack while AI application pure-plays run down and reinvent AIP from first principles. Palantir’s delivery metric compounded from “inside of 8 weeks” two years ago to “it feels like a week.”

The forward deployed engineering model he pioneered is “build software through back propagation” — validating outcomes “on the factory floor, in the foxhole,” not at the point of sale. It only makes sense when customers sit on a power law (solving the future-dweller’s problem puts you “5 years ahead of everyone else”) and when you can capture value downstream of customer economics; if you fit a Gartner box, “doing forward deployed engineering would be crazy.”

Palantir’s talent machine: hire people who can turn “an inch into a mile,” irradiate them Hulk-style with problems they’re unqualified for (“your maximal rate of learning will be coincident with your maximum ability to tolerate pain”), and enable superpower/kryptonite discovery — because career ladders are “perfectly designed to make you feel comfortable that you are growing… but actually it’s all fake.” Sankar calls the company the Founder’s Factory.

🔗 Original source & video: How AI Is Changing Warfare | Palantir CTO

Washington’s 90-action AI plan treats innovation, physical infrastructure and global ecosystem reach as one national-security strategy. Jacob Helberg laid out the plan’s pillars: America must out-innovate competitors, accelerate data centers, energy and domestic manufacturing, then create the AI stack for the world. The plan targets actions achievable within six to nine months because “you can’t regulate your way to winning the AI race.”

The investable bottleneck is shifting from model intelligence to power, permitting, machine tools and skilled labor. Michael Kratsios wants federal scientific data made usable, not merely open, while warning that AI will enter regulated products from drones to medical diagnostics. Jacob Helberg separately flagged more than 1,000 proposed or enacted state measures as a potential path toward “a patchwork of 50 different state regulatory regimes.”

Hadrian’s factory results support the episode’s central labor thesis: AI can create industrial capacity where qualified workers no longer exist. Chris Power reported 4x manufacturing productivity, 10x workforce productivity and 30-day training for recruits entirely from non-factory backgrounds; the company’s Arizona expansion is planned at four times the Los Angeles facility’s size with 350-plus jobs. “Hadrian’s advanced factories look and operate more like a data center.”

Gecko Robotics reframed energy as both AI’s constraint and one of its highest-return applications. Jake Loosararian’s example began with a 620-megawatt plant producing only 580; robotic inspection and AI reportedly unlocked 1% efficiency, which he extrapolated to 11.9 gigawatts across the US thermal fleet “without putting a shovel in the ground.” His call: “AI shouldn’t just consume, it should create energy.”

Palantir’s Shyam Sankar sees the strongest adoption where frontline workers author the applications and institutional leadership releases their agency. He wants workers made 50 times, not merely 50%, more productive; cited factory training falling from three years to three months; and described a four-week fellowship for mechanically intuitive, often self-taught workers. His categorical conclusion was that “the traditional college degree is dead.”

Paul Buchheit’s abundance case is that natural language expands the pool of builders, while capital intensity preserves scarcity at the model layer. With only 2%-3% of Americans able to code—and perhaps half that number capable of building a startup—“English is the new programming language” could produce 10x or 100x more startups, robotics companies and local applications. Buchheit expects the number of foundation-model providers to remain relatively stable, with open source constraining censorship and lock-in among closed vendors.

Small businesses are positioned as AI’s distribution channel, but the resulting market may be a barbell rather than a universal uplift. Kelly Loeffler said 60% of the SBA’s $21 billion lent this year went to companies with one to five employees; Keith Rabois expects those firms to gain incumbent-grade information, products and administration, then take share from the mid-market while compute leaders such as NVIDIA also benefit. The counterweights are energy and materials costs, industrial supply-chain exposure, rigorous underwriting and Power’s warning that offshore competition remains “companies versus the CCP.”

🔗 Original source & video: Winning the AI Race Part 1: Michael Kratsios, Kelly Loeffler, Shyam Sankar, Chris Power