SpaceX, Palantir, Anduril & Onodrim | The Return of Conglomerates

He says SpaceX has been 137 Ventures’ largest position after investing for 16 years, and he frames the company as dominant across multiple critical industries with rapid recent growth. He also emphasizes that, in his view, the business is "just getting started" despite its current scale.

He argues the underappreciated shift is SpaceX moving toward a hyperscaler model by building software and services around its compute stack and monetizing at multiple layers (applications/models/compute). He adds that vertical integration lets it benefit across different model outcomes and shift capacity internally if external demand changes.

He highlights management’s stated sub-12-month payback periods as a demand signal and says speed, payback period, and cost of capital drive who can build capacity. He also notes that debt financing is central to the AI build-out and that market volatility has raised questions about underwriting revenue and access to capital.

He says the most important metric is not any single-year growth rate but the duration of compounding, citing SpaceX compounding ~40% for 20 years. He notes 137’s concentrated bets (e.g., SpaceX, Anduril) reflect conviction that these companies can compound for decades, which matters because they are minority shareholders who can’t control holding duration.

He argues today’s market can be both real business growth and capital-driven price inflation, because there’s a lot of capital chasing few growth-stage opportunities. He says the venture/growth “power law” means a handful of winners can pay for widespread overpayment and failures, pointing to fintech as a category with many “dead” companies but also huge outcomes.