Summary
Bloomberg Technology speaks with David George, a partner at Andreessen Horowitz, in the wake of the SpaceX IPO. The conversation covers a16z’s long-term conviction in SpaceX and the firm’s view of how the company’s infrastructure advantage — particularly Starship’s path to rapid reusability — positions it to bring compute into orbit. George describes orbital data centers not in abstract terms but as “airplane-sized GPU racks in space,” each with solar arrays roughly the size of a Boeing 737, networked in low Earth orbit at scale similar to Starlink’s existing 10,000 satellite constellation.
A significant portion of the discussion centers on SpaceX’s newly named StarMind initiative, referenced in the IPO prospectus with an orbital data center target of as early as 2028. George frames this as an eventual economic inevitability: terrestrial data center construction is bottlenecked by real estate timelines, permitting, and labor — constraining an AI market moving at what he calls “blistering speed.” The $850 billion in AI data center CapEx currently planned for terrestrial deployment provides the scale context.
The interview also addresses Elon Musk’s role as the connective thesis across SpaceX, xAI, and Tesla; the strategic logic behind merging X with xAI; and the question of a potential SpaceX–Tesla combination, which George declines to dismiss. A broader point emerges: a16z’s late-stage venture thesis centers on founder-driven companies where the individual at the top is the primary value driver — a pattern George sees reflected in public markets from Apple under Jobs to Meta under Zuckerberg.
📺 Source: Bloomberg Technology · Published June 24, 2026
🏷️ Format: Interview
