“Most People Think SpaceX Is a Rocket Company”

The most memorable line in James Altucher’s Musk Millionaire presentation is also the core of the thesis: “SpaceX isn’t a rocket company… It’s a trojan horse hiding the most powerful AI blueprint ever conceived by mankind. It’s Elon’s plan to control AI.”

It’s a bold reframing of a company that just completed the largest IPO in history at $1.75 trillion. The market prices SpaceX as a launch and satellite internet business. Altucher’s argument is that the rockets were always the means, and the end is something much bigger: total control of the artificial intelligence industry, run from orbit.

The Pattern: Three Takeovers, Then a Fourth

The presentation builds its case on Elon’s history of not creating companies but “transforming industries,” each time seizing dominant share of a market he essentially birthed:

  • Digital payments. PayPal took online payments from basically zero to over $100 billion, with Elon walking away from the 2001 deal controlling 70% of transactions on the internet’s largest marketplaces.
  • EVs. Tesla launched the EV market from about $2 billion to over $1 trillion, at one point controlling nearly 80% of the entire US EV market.
  • Space. SpaceX took a government-owned industry to a $613 billion commercial sector, and Elon now controls more than 80% of everything the U.S. launches into orbit.

The through-line is control of the entire stack, not just a product in it. That’s what makes the AI move familiar: “Elon doesn’t want to just ‘compete’ in the AI industry. He wants to control the AI industry from end to end.”

The Stack, Piece by Piece

What does end-to-end control of AI actually look like? The presentation assembles the pieces like a checklist:

The rockets. SpaceX is, per Altucher, “the only man on earth who can reliably send satellites to space.” Launch capacity is the moat nobody else has crossed — Blue Origin’s program is FAA-grounded after failures, and Bezos admits his version could take twenty years.

The AI. The xAI-SpaceX merger, the largest private merger in history, means Elon now has “the most advanced AI on earth to put on” his satellites. Days after filing the million-satellite plan, the gap between Elon and Bezos, Zuckerberg, and Altman widened further.

The data centers. The orbital plan itself — one million solar-powered AI satellites, starting with the 230-foot-wingspan AI1. We cover the filing in our FCC filing breakdown.

The chips. Terafab, the largest chip factory ever attempted, could cost up to $119 billion. Musk says its chips will be “two to three times more powerful than Nvidia’s best” at a tenth of the cost — with Intel confirmed as a manufacturing partner, up over 100% since the April announcement.

The transmission layer. The one piece Elon didn’t build: the “nervous system” that beams data from orbital data centers back to Earth. That’s the unnamed “AI MasterKey” supplier — publicly traded, under $100, already in all 10,000 existing satellites — that the presentation’s investment case is built around.

Altucher names the playbook explicitly: John D. Rockefeller’s “vertical integration,” controlling all aspects of the supply chain in the fastest-growing sector of the economy. It’s the same strategy, a century later, aimed at a $25 trillion-per-year industry.

The Trojan Horse Logic

Why “trojan horse”? Because the rocket business funded and legitimized the buildout of everything else. Every Starlink launch perfected the satellite production line, the reuse economics, and the orbital real estate strategy. The commercial launch business — the thing the market values SpaceX on — quietly assembled every component of an AI empire: rockets, satellites, launch sites, spectrum rights, and now, post-merger, the models themselves.

The presentation adds a geopolitical flourish: rivals “Bezos, Zuckerberg, and Altman… will be on their knees, hands out, begging Elon for mercy,” because they now need what only he has — cheap, reliable, at-scale access to orbit for their own space-AI ambitions. NVIDIA’s Starcloud launched one refrigerator-sized satellite. Google’s Project Suncatcher is two test satellites, maybe, by 2027. Sam Altman looked into buying a rocket company and quietly gave up.

How Solid Is the Argument?

The vertical integration read is genuinely one of the cleaner analytical frames we’ve seen applied to the Musk empire. The pieces are real: the launch dominance, the merger, the Terafab announcement, the orbital filings. Where it stretches is in the certainty. “Two decades ahead of them all” is a presenter’s superlative. Terafab’s cost and performance claims are Musk’s own. And the trillion-dollar question — whether orbital data centers beat terrestrial ones on economics — remains open. Luke Lango’s XPanse research makes a related integration argument, and our Terafab article digs into the chip side.

Even discounted for hype, though, the trojan horse framing changes how you look at SpaceX. The rocket company valuation may be the floor, not the ceiling — and the suppliers plugged into every layer of the stack, from the confirmed (Intel) to the speculative (the Elon Eleven), are how public-market investors get exposure.

For the full presentation breakdown including the offer details and our take, see our Musk Millionaire review, and for the energy constraint driving the whole orbital strategy, our AI energy bottleneck article.

NewsletterVetter is an independent publication. We receive compensation from some of the services we review through affiliate links. Nothing on this site is investment advice. Always do your own research.