Foundation models are a new technological substrate fundamentally changing how software is built and distributed. This shift enables companies to scale at a dramatically accelerated pace, as exemplified by the hyper-growth of AI-native companies like Cursor.
A new wave of companies is combining AI with physical hardware to solve complex, real-world problems in sectors like defense (Anduril), aerospace (SpaceX), and neurotech (Nudge). While capital-intensive initially, these businesses can build highly durable competitive moats and exhibit accelerating growth at multi-billion dollar scale.
AI is being applied to core scientific disciplines, turning complex research challenges into solvable engineering problems. Companies like Isomorphic Labs (drug discovery) and Nudge (neuroscience) are using AI to accelerate breakthroughs that could have a profound impact on human health and materials science.
There is a growing strategic imperative for nations to control critical technology supply chains, from semiconductor manufacturing to space launch capabilities. The decline of US semiconductor manufacturing and its previous reliance on Russia for astronaut transport highlight the national security risks of technological dependency.
The speaker advocates for a venture capital strategy centered on making concentrated, high-conviction bets on generational founders and companies, rather than broad diversification. This is exemplified by Thrive Capital's early and decisive investment in OpenAI when few other firms were willing to commit at that valuation.
Keep pulling the thread on Philip.