Interview, Fireside Chat
Shyam Sankar: The Broken Incentive Structure of How Governments Buy Defence | E1104
- US Defense Procurement Critique: Sham Sankar argues that excessive process designed to ensure "fair competition" has created a system lacking both genuine competition and speed, with the European process being even slower due to an emphasis on fairness across EU states.
- Defense Spending Context: Current US defense spending is at historic lows relative to GDP (2–3%) compared to the Cold War peak (5–6%), representing roughly half the historical "insurance premium" required to protect the Western order.
- Palantir's Business Model Shift: Palantir rejected the traditional government "labor model" (paying by the hour with cost-plus contracts) and instead pioneered a product-led model where the government pays a license/subscription fee, forcing the company to absorb R&D costs and innovate at commercial rates.
- Industrial Base Consolidation: The 1993 "Last Supper" event, where Deputy Defense Secretary Bill Perry mandated consolidation among defense primes, reduced the number of major contractors from 51 to 5, leading to a financialized industry focused on dividends and buybacks rather than creative risk-taking.
- Procurement Failure Mechanism: The government expects R&D costs to decrease over time while procurement costs remain static, a misconception that fails because software R&D costs increase annually and continuous innovation is required.
- Suggestion for DoD Leadership: If in charge of the DoD, Sankar would implement multiple competing programs for the same capability (e.g., four competing missile programs) to introduce human competition, accelerate iteration, and reduce technology risk.
- Feedback Loop Disconnection: Sankar identifies a critical disconnect between combatant commanders (who fight) and procurement officers (who buy), proposing that commanders should have direct "taste" authority to allocate resources between competing variants based on real-world performance.
- VC Engagement Recommendation: Silicon Valley Venture Capitalists should engage directly with policymakers to explain the dynamics of defense tech, bridging the cultural gap between the D.C. "no conflict" mindset and the Silicon Valley "no interest, no opinion" reality.
- Content vs. Process Philosophy: Sankar posits that "content" (creativity and substance) is the eternal substance of success, while "process" should be viewed as medicine—useful only when necessary, but becoming an "opioid" that stifles innovation when prioritized over substance.
- Talent Management Strategy: Palantir operates as an "artist colony" rather than a factory, designing roles around unique individual talents rather than slotting people into predefined boxes to avoid eviscerating their potential.
- Hiring Philosophy: Sankar advises hiring "boring" candidates who appear humble in interviews, as their unassuming nature often masks high capability, and prioritizes betting on raw talent over importing solutions with prior track records.
- AI Value Distribution: In the AI sector, value is predicted to accrue to incumbents owning the application layer or infrastructure, while model providers face commoditization; successful applications will integrate 1/3 AI, 1/3 deterministic code, and 1/3 human thought.
- AI Adoption Warning: There is a risk of an AI bubble caused by an over-focus on "proof of concept" rather than "proof of value," with enterprises requiring reliable commercial deployments rather than impressive but non-functional demos.
- Future Outlook (2028): Sankar foresees significant geopolitical instability in the next five years but views AI as a unique tool to transform and fix fundamental institutional problems by partnering with existing structures rather than tearing them down.
- Organizational Reinvention: Palantir has undergone multiple "reincarnations" where all employees effectively start from zero, a process described as a "crucible" that builds pride and capability, though it becomes harder to manage as the organization scales.
- Historical Parallel: Sankar draws parallels between the current defense industrial base and the WWII era (e.g., Kelly Johnson at Skunk Works), suggesting that the US is capable of rapid innovation when political will aligns, but suffers from "mediocrity" in steady states due to risk aversion.