SpaceX Buys More Spectrum, Musk and Market Structure, Distributed Cellular

Thompson revises his earlier skepticism about SpaceX’s spectrum buying. Previous purchases seemed poorly matched to existing handsets and satellite-to-phone approvals, while major carriers chose other partners. He argues that declining Starlink might eventually hurt airlines whose passengers expect its Wi-Fi, but does not claim the same customer preference has been shown for mobile carriers.

A reported purchase of spectrum formerly used by Nextel, together with SpaceX executives’ description of ground equipment at broadband-dish sites, gives him a different explanation of the mobile plan. Satellites could supply backhaul for many small terrestrial access points; lower-frequency spectrum could help cover buildings and places beyond an individual node’s reach. He thinks the architecture could become hard to match if deployed at scale.

SpaceX’s own deployment description is conditional on regulatory approval, and its executives did not give the terrestrial build-out cost. Thompson notes that incumbents already have strong networks and consumers, not airline executives, choose mobile service. His advantage claim is a forecast, with no observed coverage, customer switching or unit economics here. The opening podcast reference to AI does not turn this into evidence about coding agents or software-team productivity.

Read at stratechery.com · 10 min