Skip to main content
a16z Podcast

Inside Palantir: Building Software That Matters with Shyam Sankar

54 min episode · 2 min read
·
Inside Palantir,Shyam Sankar

Episode

54 min

Read time

2 min

Topics

Productivity, Startups, Fundraising & VC

AI-Generated Summary

Key Takeaways

  • Defense Industrial Concentration: Between 1989 and today, spending on dedicated defense contractors jumped from 6% to 86% of major weapons system budgets. This consolidation bred financial engineering over real engineering, expelling founder-type innovators into tech. Reversing this requires deliberately recruiting Silicon Valley talent back through direct commissioning programs, as the U.S. did with 100,000 people in World War II.
  • Civil-Military Fusion Model: Sankar joined the U.S. Army alongside OpenAI's former chief research officer, Meta's CTO, and OpenAI's former chief product officer as direct-commissioned senior advisors. The model mirrors Israel's post-October 7 mobilization, where 360,000 reservists with 20 years of industry experience modernized IDF technology faster in four months than in the prior decade combined.
  • SaaS Disruption Framework: Evaluate software by whether it delivers "beta" (making companies similar to competitors) versus "alpha" (expressing competitive differentiation). Beta software faces existential AI pressure. COVID exposed this: no CEO credited multibillion-dollar ERP implementations for supply chain resilience. Platforms enabling company-specific differentiation will strengthen, while standardized workflow tools face commoditization.
  • AI Value Stack: Value in AI accrues at two layers — chips and AI infrastructure (what Palantir calls ontology) — not at the model layer, which faces commoditization pressure. Model companies are expanding upward into software tooling, while vertical AI applications are building downward into infrastructure. Framing AI as a tool humans wield, not an autonomous agent, determines whether productivity gains reach workers.
  • Heretic Protection Pattern: Every major U.S. defense innovation — Higgins boats, Minuteman missiles, the nuclear navy — was institutionally opposed and survived only because a senior leader actively protected the difficult innovator. Colonel Drew Kukor, father of Project Maven, faced IG investigations and false criminal allegations while building military AI. Identifying and shielding these individuals, not just funding programs, drives breakthrough outcomes.

What It Covers

Palantir CTO Shyam Sankar discusses America's defense industrial decline, arguing that defense spending concentration shifted from 6% to 86% in dedicated contractors since 1989, expelling the "heretical" founders who drove innovation. He connects this to AI competition, SaaS disruption, and the need for civil-military fusion to prevent strategic decline.

Key Questions Answered

  • Defense Industrial Concentration: Between 1989 and today, spending on dedicated defense contractors jumped from 6% to 86% of major weapons system budgets. This consolidation bred financial engineering over real engineering, expelling founder-type innovators into tech. Reversing this requires deliberately recruiting Silicon Valley talent back through direct commissioning programs, as the U.S. did with 100,000 people in World War II.
  • Civil-Military Fusion Model: Sankar joined the U.S. Army alongside OpenAI's former chief research officer, Meta's CTO, and OpenAI's former chief product officer as direct-commissioned senior advisors. The model mirrors Israel's post-October 7 mobilization, where 360,000 reservists with 20 years of industry experience modernized IDF technology faster in four months than in the prior decade combined.
  • SaaS Disruption Framework: Evaluate software by whether it delivers "beta" (making companies similar to competitors) versus "alpha" (expressing competitive differentiation). Beta software faces existential AI pressure. COVID exposed this: no CEO credited multibillion-dollar ERP implementations for supply chain resilience. Platforms enabling company-specific differentiation will strengthen, while standardized workflow tools face commoditization.
  • AI Value Stack: Value in AI accrues at two layers — chips and AI infrastructure (what Palantir calls ontology) — not at the model layer, which faces commoditization pressure. Model companies are expanding upward into software tooling, while vertical AI applications are building downward into infrastructure. Framing AI as a tool humans wield, not an autonomous agent, determines whether productivity gains reach workers.
  • Heretic Protection Pattern: Every major U.S. defense innovation — Higgins boats, Minuteman missiles, the nuclear navy — was institutionally opposed and survived only because a senior leader actively protected the difficult innovator. Colonel Drew Kukor, father of Project Maven, faced IG investigations and false criminal allegations while building military AI. Identifying and shielding these individuals, not just funding programs, drives breakthrough outcomes.

Notable Moment

Sankar recounts that Hyman Rickover, who built the nuclear navy despite opposition from Oppenheimer and the Navy itself, was given a women's restroom as his first office — a deliberate humiliation tactic. Rickover documented every slight but channeled them into results, producing submarines with zero radiation deaths across decades of operation.

Know someone who'd find this useful?

Episode Transcript

World events remind us that there is actually evil out there. Just horrendous barbarism is still possible. When a country goes to war, it's not enough to just have the Department of War fight these wars. It is actually the whole country. The idea that somehow the American people are not capable is the beggar's belief. I think our biggest risk as a country is suicide, not homicide. How do we win the AI race, particularly as it moves towards more physical AI and robotics, etcetera? The things that we did to to win in the past, we accidentally turned our back on, and there's an opportunity to reclaim that with vigor. In the moment right now, you could say, we need to build more web. We need to do this. We need yes. But the most important thing we need to do is The companies that won World War two weren't built to win wars. Chrysler made minivans. Every cereal box an American consumer bought was quietly subsidizing national security. In 1989, only 6% of spending on major weapon systems went to dedicated defense contractors. Today, it's 86%. The heretics who built the arsenal of democracy left for Silicon Valley, and the Pentagon lost its front door. Now some of them want to come back. Shyam Sankar has spent more than two decades at Palantir building technology for national security. He was commissioned into the US army and published Mobilize, arguing that America's real risk isn't China. It's losing the will to compete. I speak with Shyam Sankar, chief technology officer at Palantir Technologies, alongside a 16 z general partner, Catherine Boyle. So, Catherine, when we were talking about guests that we had to add on, Sean was at the top of your list. Why was that? You know, after the Jeremy Stern profile in Colossus and a lot of, I think, stories that have come out recently or podcasts that have come out about Shyam. He's one of these people that if you were in the know several years ago, you knew. He was the OG, like, fixer for everyone. You know, I think Trey Stevens, who's the cofounder of Anduril, came out on Twitter and said he's single handedly responsible for my career. John Doyle, so many of our founders have pointed to Shyam as the person who made their career and introduced them to Palantir, supported them in Palantir, but also sort of gave them wings to fly away from Palantir and to start something new. You hear that story time and time and time again. And Eric, you and I were talking about this. It wasn't until a couple years ago that I think Shom actually became more of a public figure. He was sort of the behind the scenes guy, the behind the scenes fixture. And I think the thing that really changed it was, Shom, and I'd love to to talk to you about kind of what was the inspiration for this. You …

Get the full transcript (11,412 words) + summary by email — free

One-time email with the complete transcript and AI summary of this episode. No account needed.

One email, no spam. We’ll also show you what SignalCast does.

Browse all a16z Podcast transcripts →

You just read a 3-minute summary of a 51-minute episode.

Get a16z Podcast summarized like this every Monday — plus up to 2 more podcasts, free.

Pick Your Podcasts — Free

Keep Reading

More from a16z Podcast

We summarize every new episode. Want them in your inbox?

Similar Episodes

Related episodes from other podcasts

Explore Related Topics

This podcast is featured in Best Business Podcasts (2026) — ranked and reviewed with AI summaries.

Read this week's Startups & Product Podcast Insights — cross-podcast analysis updated weekly.

You're clearly into a16z Podcast.

Every Monday, we deliver AI summaries of the latest episodes from a16z Podcast and 192+ other podcasts. Free for one show.

Start My Monday Digest

No credit card · Unsubscribe anytime