America's About to Go Off: Why Orbital Infrastructure Cements the Next Century

You've heard the story. America is finished. China is about to overtake us. Debt, deficits, hollowed-out factories — doom on a loop. Those problems are real. Farzad has covered them. They still do not decide the next century.

What decides it is already flying. Literally right above your head. America is about to own space infrastructure — launch, satellites, compute, lunar jump-off, deep-space logistics — in a way no peer can copy on any useful timeline. Control that layer and you shape what happens on Earth too: bandwidth, intelligence, materials, manufacturing, deterrence. Most people are still scoring the Cold War with twentieth-century scoreboards. The game moved.

The stack already in orbit

Start with what is live today, not a white paper.

Starlink is operational with more than 9,000 satellites delivering global high-speed internet. That constellation is American. AI-satellite efforts like StarCloud are aiming the next wave of orbital compute at the same problem every terrestrial data center is slamming into: power, cooling, land, and permits. Reusable lift that makes dense constellations affordable sits under it — Falcon 9, Falcon Heavy, Starship, Blue Origin's New Glenn. All American.

China is racing. It recently flew a first reusable test vehicle roughly in the Falcon 9 class. Fine. Falcon 9 is already legacy relative to Starship. Catching yesterday's workhorse while the United States iterates the vehicle meant to obsolete it is not parity. It is trailing a curve that compounds with every landing, every reflown booster, every mass-to-orbit drop.

Launch cadence is the scarce asset. Whoever can put mass in orbit cheaply and often owns the rest of the stack: internet, sensing, compute, depots, landers. Right now that cadence is American private industry, not a ministry five-year plan.

Why the next AI wave wants vacuum

Superintelligence talk usually stays on Earth: GPUs, substations, water fights, NIMBY lawsuits. The better place to put a huge fraction of that compute is space.

Unlimited solar. No night in the right orbits. Heat rejects into the void — you radiate instead of pumping rivers through chillers. No land cost. No county permit. No protest over cooling water. Launch the hardware, point the arrays at the sun, and run.

Who flies the freighters? SpaceX. Who builds the models that want that compute? OpenAI, xAI, Google, Nvidia — American companies with American capital markets behind them. Rockets plus models plus finance is the full loop. Rivals can copy slides. They cannot copy that industrial and capital machine overnight.

This is not a side quest for "space enthusiasts." It is the escape hatch from Earth's power wall. Every year terrestrial AI fights the grid harder. Orbital AI skips the grid.

Asteroids, rare earths, and a monopoly that dies

Asteroid mining still sounds like comic-book cover art. The economics changed when launch costs collapsed.

There are rocks with more platinum than Earth has ever mined, rare earths beyond what China currently controls, and iron past any industrial appetite we have. Rare earths go into phones, computers, batteries, and basically every advanced electronic. Beijing's choke point on those minerals is a geopolitical weapon today. Mine the belt and that weapon goes dull. Bottlenecks for advanced manufacturing stop being a map of Chinese processing plants and start being a question of who can rendezvous, refine, and return mass.

You do not need full sci-fi utopia for this to matter. You need launch cheap enough that prospecting and return missions close a spreadsheet. That is the curve Starship is built to drive.

The Moon as the cheap on-ramp

NASA's Artemis program is already laying track for the early 2030s. SpaceX is lined up as the human lander. A Moon base is not a flag-planting photo op. It is physics arbitrage.

Lunar gravity is about one-sixth of Earth's. Launching from the Moon into deep space is dramatically cheaper than climbing out of Earth's well every time. Mine regolith. Build manufacturing in situ. Use the surface as the jump-off for Mars, the asteroid belt, and everywhere else. Whoever owns that depot infrastructure owns the cheap ticket to the rest of the solar system.

America is positioned to own it: lander, heavy lift, satellite backbone, and a private sector that actually ships hardware. That is not a press release. It is a supply chain.

Mars is the insurance policy — and only one country is trying

Elon Musk has talked about a Mars city for years and caught endless memes for it. Starship is flying. The cargo math is brutal and clear: on the order of a million tons to build a self-sustaining city. Hit the cost targets and you can imagine that in roughly twenty years; miss and it takes thirty or forty. Duration is not the strategic point.

The point is America is the only country seriously attempting the economics. China talks Mars without the reusable heavy-lift machine or the private ecosystem iterating it weekly. Russia's program is a shadow of what it was. Europe is trying and still not in the same weight class.

Why does a Mars colony matter for American dominance? Catastrophe insurance for civilization — nuclear war, asteroid impact, supervolcano, a pandemic worse than COVID — and the colony that survives flies American tech, American capital, American institutions. Values follow the stack that built the life support. Farzad's bet is blunt: American-led, American-built, American-financed. The flag debate can wait. The industrial lead cannot.

The feedback loop nobody prices

Put AI satellites in sunlight. Those models design better rockets, better mining gear, better habitats, better factories. Better designs mean more space development. More development means more orbital compute. Smarter AI again.

That loop compounds. America is the only player holding all the pieces at once: frontier AI labs, launch providers, deep capital markets, and a regulatory culture that still lets private companies move without total state capture. Ask whether Beijing lets an entrepreneur loft thousands of satellites or fund a Mars city the way SpaceX does. Look at what happened to Jack Ma. That system optimizes for control. This stack optimizes for iteration speed. In a compounding domain, iteration speed is the whole game.

And the tech does not stay upstairs. Starlink already lights up remote regions. Orbital AI that solves materials, medicine, and energy feeds back to Earth. Asteroid extraction techniques clean up terrestrial mining. Mars life-support work improves sustainability here. Spillover is not charity. It is how the lead hardens into living standards people can feel.

Wrong scoreboard

People still argue great-power competition as GDP growth rates, military line items, and manufacturing tons. Those metrics still matter. They are not where the decisive game is being played.

The game is being decided a few hundred miles up, roughly 240,000 miles away at the Moon, and on the order of 140 million miles away at Mars. America is winning that contest before most cable panels notice it exists. Why does the Space Force exist? Not for parade optics. Because orbit is already contested infrastructure.

None of this claims America is healthy across the board. Aging infrastructure. A healthcare mess. Schools that should be better. Doom narratives that treat those failures as proof of terminal decline are looking at the wrong century. We are refighting the last Cold War on trade balances while the United States builds the plumbing for the next phase of civilization: space-based intelligence, asteroid resources, lunar access, multiplanetary backup.

The gap likely widens. More flights make the next flight cheaper. Cheaper flights unlock more constellation and depot work. More orbital industry funds the next leap. China could drop a trillion dollars on space tomorrow and still sit a decade back — and by the time it reaches today's American position, America is another decade out. That is what real dominance looks like: not a slight lead, an unbridgeable compounding lead.

Worry about debt and deficits. You should. Then watch the weekly launches, the growing constellations, the AI labs training larger models, and the Starship test cadence. The future that is forming is not America clinging to status. It is America extending it into a domain that is effectively infinite — as long as the gains compound for Americans and, practically, for a world that rides the same networks and spillover tech.

Check the video here.