Interview, Fireside Chat
How AI Is Changing Warfare, with Anduril CEO Brian Schimpf
- Military technology over the next 20 to 30 years will increasingly rely on vast networks of low-cost, autonomous systems equipped with numerous sensors to generate overwhelming data volumes, requiring AI to extract actionable signals for human decision-making in high-tempo scenarios.
- Defense product development is predicted to accelerate significantly, shifting from a current prototype-to-pilot timeline of three to six months and full-scale deployment of approximately two and a half years, yet new missile development remains estimated at 12 years from concept to fielding.
- Operational scaling will utilize autonomous teams of drones capable of independent routing and rebalancing, with AI augmentation deployed specifically for defensive counter-drone scenarios where human reaction times are insufficient.
- A strategic shift is anticipated where commercial investment replaces traditional defense-exclusive programs, enabling startups to produce smaller, fully autonomous assets rather than large platforms, potentially reducing the U.S. capacity to afford single shared aircraft across service branches by 2046.
- China is projected to maintain a compounding iterative advantage through hypersonic weapon testing frequencies of hundreds per year compared to the U.S. rate of four tests annually.
- U.S. readiness faces critical constraints, including an estimated stockpile depletion of key munitions within eight days during high-intensity conflict and a required two-year industrial retooling period before entering hostilities.
- To address volume and innovation gaps, the U.S. must integrate commercial supply chains, increase budget allocation for emerging capabilities from the current 0.1%–0.2% range to approximately 2%, and establish a permanent bureaucratic shift similar to "Warp Speed" initiatives.
- The U.S. defense industrial base faces risks of consolidation, capacity erosion, and persistent "cost-plus" contracting models, with a hurdle for startups rated at 30% worse than banking or oil sectors, compounded by a Pentagon culture where dissenters outnumber accountable leaders by a ratio of 10 to 1.
- Automation adoption will follow a "crawling phase" focusing on mechanical and predictable tasks before advancing to complex AI-driven decisions, while long-life assets like the Columbia-class nuclear submarine (service 2035–2085) risk outliving their technological relevance.
- Reduced troop risk resulting from autonomous force projections may alter the willingness to use force for political ends, creating a stabilizing deterrent effect while simultaneously complicating the acquisition of capabilities due to entrenched bureaucratic resistance.