What this covers
Luke walks Matt through SpaceX, starting with why rockets stayed expensive for decades. The industry grew out of US-Soviet competition in the 1950s and 60s, so once the moon landings were done there was no market reason to cut costs, and the Space Shuttle ended up costing more per kilogram than Saturn V. Luke frames this as incentive misalignment: customers with billion-dollar satellites would not accept cheaper, riskier rockets. He then covers Elon Musk selling assets to fund one last Falcon 1 launch, NASA's roughly 4 billion dollar estimate for Falcon 9 against the tenth of that SpaceX spent, and the reuse numbers: about 320 of 400 flights on reflown boosters, some flying more than 20 times.
The range is wide. Luke explains why leaving Earth's surface dominates cost, so Falcon 9 charges the same for 20 tons to Low Earth Orbit as for 4 tons to Mars orbit. On Starlink, he contrasts terrestrial networks, which need road-digging for each user, with a satellite network whose marginal cost is a terminal. He notes about 11 million US households lack full broadband and that a Defense Department orbital-services program grew from 900 million to a projected 13 billion dollars. He also explains why the late-1990s ventures like Teledesic failed.
The forward-looking parts are more contested: manufacturing pharmaceuticals in orbit, a razor-and-blade model for embedded connectivity, space as the first battleground in a major conflict, and a comparison with nuclear power's boom and collapse. He closes on Starship, covering launch-pad capacity, Model 3 lessons for scaling production, and why Mars matters mainly as a forcing function.
Luke Ward argues SpaceX succeeds because it treats launch as a business problem, not just an engineering one: reusability and vertical integration drive cost down by orders of magnitude, cheaper launch unlocks satellite and orbital markets, and Starlink's revenue funds Starship, creating a self-reinforcing flywheel.
- Launch cost fell because SpaceX made boosters reusable, spreading hardware cost over many flights, which previous customers and government programs never had an incentive to do
- Starlink's per-user cost scales with Moore's Law terminals rather than civil engineering, and its revenue funds Starship development
- Starship's target of about $10 per kilogram would make orbital manufacturing, connectivity, and eventually cheap human spaceflight economically plausible
Hard-asset space businesses need ongoing business-model innovation, since technical fixes like disposable rockets do not recur.
- Business models for capital-intensive hard-asset businesses are a more difficult equation than digital and software models, but business innovation in those sectors can subsidize technology development and make them attractive.
“Once you get into capital-intensive hard asset things, it becomes a more difficult equation.”
- Making rocket sections disposable solves the physics problem that each kilogram of cargo needs about 20 kilograms of plumbing and propellant, but it is a one-time economic fix that does not recur.
“every additional kilogram of cargo requires 20 kilograms of plumbing and propellant”
Industrial demand and Starship's larger cargo capacity expand the market SpaceX can serve.
- Matt notes that he has been to drilling sites with no cell signal, so Starlink's industrial applications matter as much as consumer hotspot use.
“I've been to many drilling sites, and there was one consistent thing: I didn't have a cell phone signal.”
- Matt notes that parcel shipping prices by dimensional weight rather than actual weight, so Starship's wider cargo capacity opens more possibilities for what SpaceX can launch.
“the idea of dimensional weight rather than just traditional weight.”