Been watching this Musk-xAI situation unfold and there's honestly a lot to unpack here. So we've got co-founders jumping ship—Tony Wu bailed Monday, Jimmy Ba followed Tuesday—and now we're at six out of twelve founding members who've walked. But instead of addressing the exodus directly, Musk calls an all-hands to pitch... a lunar manufacturing facility. Yeah, you read that right. A factory on the moon.

According to reporting, he's talking about building AI satellites up there and launching them into space via some kind of giant catapult setup. The pitch: it gives xAI access to computing power no competitor can touch. Look, I get the vision—more power, less interference from Earth's limitations. But the timing is wild. You've got people leaving, a potential SpaceX IPO targeting $1.5 trillion valuation supposedly coming this summer, and suddenly the conversation shifts to the moon.

Here's where it gets interesting though. Back in January, Musk posted that SpaceX was pivoting from Mars to the Moon, saying a lunar city gets there in half the time compared to a 20+ year Mars timeline. For a company that's never actually sent a mission to the moon, that's a pretty dramatic direction change. Investors seem way more into orbital data centers than interplanetary colonies anyway.

But there's a theory floating around that ties this all together. The idea is that Musk has been building toward one thing the whole time: the world's most powerful world model. Tesla brings energy systems and road data. Neuralink gives brain insights. SpaceX provides physics and orbital mechanics. The Boring Company adds subsurface information. Now add lunar manufacturing to that stack and suddenly you see something genuinely powerful taking shape.

The legal side is where it gets tricky. The 1967 space treaty says no one can claim ownership of the moon itself. But here's the loophole: a 2015 U.S. law lets you own whatever you extract from it. So technically, the resources you pull out of the lunar surface are yours. Some legal experts point out this distinction is pretty fuzzy in practice—you can't own the house but you can own the floorboards, except the floorboards basically are the house. And not everyone's playing by these rules anyway. China and Russia certainly aren't.

So we've got a shrinking team, an upcoming IPO, and increasingly ambitious space plans. The question isn't really whether the moon factory concept is cool—it obviously is. The question is whether any of this actually gets built, and more pressingly, who's going to be around to build it. The all-hands meeting raised more questions than answers.