Executive Summary
The Trump administration's AI strategy represents a fundamental shift from regulatory constraint to infrastructure acceleration. David Sacks and Michael Kratzios outlined a three-pillar approach: innovation through federal preemption of state regulations, massive infrastructure buildout with self-funded power generation, and global AI technology export. The administration rescinded 300 pages of Biden-era AI regulations in its first week, creating what Sacks calls 'permissionless innovation' environment. Most critically, they revealed Microsoft has pledged that data centers won't increase residential electricity rates, with other hyperscalers expected to follow. This addresses the primary political constraint on data center expansion. The infrastructure thesis is validated by zero 'dark GPU' capacity—every chip deployed is generating revenue through token production. Sacks noted this buildout added 2% to GDP growth in 2025, with similar impact expected in 2026. The regulatory patchwork problem—1,200 state AI bills currently in process—creates competitive moats for large incumbents while constraining startups. Federal preemption legislation targeting summer 2026 could unlock significant venture capital deployment in AI applications. However, heavy insider selling at NVIDIA ($343M) and Microsoft ($28M) over 90 days suggests management teams are taking profits despite public optimism. The China competition dynamic has shifted from complacency to active rivalry following DeepSeek's emergence, with export controls now working both directions as China blocks NVIDIA chips to protect domestic champion Huawei.
Key Insights
what David Sacks and Michael Kratzios said“There's no such thing as a dark GPU right now. Every GPU that's being put in a data center is getting used. And it's being used to generate tokens.”
This is a preview. Log in to see the full analysis including investment opportunities, risks, catalysts, and detailed insights.