Elon Musk has spouted his fair share of bullshit—but his latest claims about... – Chronik des laufenden Wahnsinns

7. Juni 2026Elon Musk has spouted his fair share of bullshit—but his latest claims about the SpaceX IPO are on a whole new level.

Let me break it down. Musk is notorious for pie-in-the-sky promises that either never materialize or show up years late. We’re talking about fully autonomous driving “in five years” (a claim he's been recycling since 2014), the infamous “funding secured” tweet during his botched attempt to take Tesla private, and the elusive $35,000 Model 3 that was practically impossible to actually buy at that price point.

But the narrative he’s spinning for the SpaceX IPO is in a league of its own: moving AI data centers into space because Earth supposedly lacks the power grid to handle them.

The argument goes like this: On Earth, solar energy is severely handicapped by weather, cloud cover, atmospheric absorption, and the simple day-night cycle. In space, you get near 24/7 unfiltered sunlight. True. BUT.

Here on Earth, AI data centers are cooled using massive amounts of circulating water and air. In the vacuum of space, there’s no matter to carry that heat away. Cooling up there relies entirely on thermal radiation dissipating into the void.

To radiate just 1 megawatt of waste heat at 80 to 85 degrees Celsius (the tolerable limit for silicon chips), you need nearly 1,600 square meters of radiator surface area. Scale that up to a single gigawatt of computing power, and you’re looking at over 800,000 square meters of nothing but cooling panels—that's more than 110 football fields.

The International Space Station (ISS) uses a highly complex cooling system just to shed 70 kilowatts of waste heat. Those radiators alone weigh 7 metric tons and span roughly 422 to 475 square meters. Yet, SpaceX is planning to cram 100 kW of computing power into a satellite with a total weight of just 1 ton. Physically impossible.

Even if they could magically solve the thermal management problem, hardware in space degrades vastly faster than on Earth due to cosmic radiation—and swapping out fried components is virtually out of the question because of the sheer cost.

Assuming a realistic hardware lifespan of two to three years, you’d have to replace hundreds of thousands of tons of equipment annually. That means thousands of Starship launches every single year—just to maintain the status quo, let alone scale up a single chip.

And then there's latency and bandwidth: hard physical limits that you simply can't argue away.

Elon Musk’s Wall Street stories have always thrived on pure fantasy. But if a valuation actually holds up based on this fairy tale, we’re looking at straight-up mass psychosis.