The Promise Was Specific

This isn't about vague green marketing language. Musk put it in writing.

In Tesla's original Master Plan, Musk declared the company's "overarching purpose" was to "help expedite the move from a mine-and-burn hydrocarbon economy towards a solar electric economy." That's a direct quote. That's a mission statement, not a press release.

Four Master Plans later: xAI, Musk's AI startup, is running data centers on unregulated natural gas turbines and has committed to buying $2.8 billion more in gas infrastructure, according to TechCrunch's reporting on the SpaceX IPO filing released the week of May 23, 2026.

What the SpaceX Filing Actually Shows

The IPO document provides specifics. According to TechCrunch, SpaceX spent $131 million on 1,279 Cybertrucks. xAI has spent $697 million over the last two years on Tesla Megapacks — the grid-scale battery systems Tesla sells for managing peak loads.

So Musk's companies ARE buying from each other. Just not solar panels.

xAI has purchased ZERO materially significant solar capacity from Tesla, according to TechCrunch's analysis of the filing. The Megapacks are there to manage grid peaks, not to store solar energy. They're propping up a natural gas operation.

The filing does mention solar — but only as a setup to argue that space-based solar is superior. SpaceX claims orbital arrays can generate more than five times the energy of terrestrial solar panels, thanks to around-the-clock illumination unaffected by weather or day-night cycles.

The Space Solar Pitch Is Real — and Genuinely Difficult

The physics argument is sound. Space-based solar doesn't go dark at night. It doesn't get cloudy. The energy density is legitimate.

But TechCrunch rightly points out the economics are brutal. Power costs for Starlink satellites run multiples higher than what a ground-based data center pays. Radiation-hardening chips for space isn't cheap or easy. And it's not even clear whether AI training workloads can be distributed across multiple satellites in orbit — which means a massive chunk of AI computing stays on the ground regardless.

It's not one problem SpaceX needs to solve. It's a stack of hard problems that compound each other.

Why Natural Gas? The Honest Answer

Mainstream coverage keeps avoiding a central fact: AI has a power problem that solar genuinely cannot solve right now.

Training large language models requires continuous, uninterrupted gigawatts of electricity. Solar is intermittent. Wind is intermittent. Battery storage at the scale needed is expensive — the kind of expensive that kills data center economics.

As techbuzz.ai noted, future AI systems could require terawatt-scale computing power. Current global data center consumption sits around 40 gigawatts. That gap is not something you close with rooftop panels.

Natural gas plants can spin up and run 24/7. That's why every major AI company — not just Musk — is turning back to fossil fuels. Microsoft, Google, and Amazon are all reopening or recommitting to gas and nuclear capacity. Musk isn't uniquely hypocritical here. He's doing what the industry is doing.

The difference is that Musk specifically and repeatedly promised he wouldn't.

Tesla's Solar Business Was Already Struggling

Fortune reported in November 2023 that Tesla's solar ambitions had already collapsed well before the AI boom. The flagship Tesla Solar Roof product had achieved a cumulative 3,000 total U.S. installations — against Musk's stated goal of 1,000 per week. Tesla's Gigafactory 2 in Buffalo, a 1.2-million-square-foot facility built on a former steel mill site that Tesla leases from taxpayers, is barely operational. Solar deployments of all Tesla products were tracking toward a nearly one-third annual decline as of 2023, according to Fortune.

The solar business wasn't a thriving operation that Musk pivoted away from. It was already a quiet failure before AI energy demand became the story.

What Mainstream Coverage Gets Wrong

Left-leaning outlets like TechCrunch frame this primarily as hypocrisy — Musk the climate evangelist betraying his own principles. That framing isn't wrong, but it's incomplete.

Every major tech company is doing the same thing. The AI power crisis isn't a Musk-specific moral failure. It's an industry-wide collision between climate commitments and engineering reality.

Tesla's solar division was already deeply troubled before xAI existed. The pivot to natural gas didn't kill Tesla's solar ambitions. Those were dying on their own.

SpaceX making ambitious claims in an IPO filing has an obvious incentive. Investors need a story. "We're going to power AI from orbit" is a compelling narrative. Whether it's a viable business in this decade is a separate question.

The Record

Musk made a specific, documented promise: a solar-electric economy. His AI company is now running on natural gas, his space company is pitching orbital solar as a years-away solution, and his terrestrial solar business has been underperforming for years.

The real engineering constraints of AI power demand are genuine. So is the gap between what Musk sold and what he's building.

Regular people paying for Tesla solar products deserve to know the guy pitching them clean energy is betting his most powerful new company on the opposite. That's just the number on the invoice: $2.8 billion in natural gas.