Descriptions:
Bloomberg Technology sits down with Sequoia Capital’s Michael Maguire for a deep dive into his investment thesis on SpaceX across what he describes as five distinct business layers: launch (the foundational moat), connectivity via Starlink, terrestrial and eventual orbital compute, the AI model layer, and long-term moonshots. His most contrarian argument is that SpaceX’s terrestrial compute business — validated by $2 billion per month contracts from Anthropic and Google — is being heavily underappreciated by public market investors focused on launch and Starlink.
On the $60 billion Cursor acquisition, Maguire pushes back hard on the “acquihire” framing. He argues the Cursor team, which required a roughly two-month intensive trial period before the deal was finalized, represents one of the best AI engineering teams ever assembled — capable of going “bare metal” from compute hardware all the way through model architecture. He assigns 99.999% probability the deal will be accretive, citing the rigor of the vetting process.
Maguire maps out why he believes xAI can close the gap with Anthropic and OpenAI in coding AI, pointing to the Cursor team’s culture and technical depth. Looking five years out, he predicts orbital compute will dominate net new AI inference workloads — making launch SpaceX’s moat again — but argues that for the next five years, the company’s 20-year talent distillation process is the competitive edge in AI cloud infrastructure. He says he intends to hold his SpaceX shares indefinitely.
📺 Source: Bloomberg Technology · Published June 16, 2026
🏷️ Format: Interview
