B (2:55)

This was actually a really Elon blog post. There are emojis in it and jokes that are very Musk coded.

A (3:09)

Musk wrote that this new endeavor will, quote, extend the light of consciousness to the stars, enable the creation of colonies on Mars, and ensure, quote, humanity's multiplanetary future. Musk also lays out what he says is the driving cause for combining SpaceX and XAI data centers in space.

B (3:28)

See, he talks about why it makes a lot of sense to put data centers in space. He says the only logical solution is to transport these resource intensive efforts to a location with vast power and space. I mean, space is called space for a reason. Laughing EMOJI.

A (3:47)

The idea is that AI data centers on Earth take up a lot of land and power. It's a big drain. But in space, neither of those issues are problems, at least theoretically.

B (4:01)

In space there is a very big source of power that is the sun. And you need rockets to launch satellites operating those data centers into space. And if that is the case, SpaceX is the world's largest rocket launching company. And so therefore it stands to benefit a lot.

A (4:22)

The idea here is that solar will be what powers these data centers, which I suppose in space you have plenty of access to that.

B (4:28)

That's the idea. And it is fraught with a lot of engineering challenges. Like, I think if you talk to anyone, even the most optimistic investors and CEOs who are dreaming of this, they acknowledge that it'll take a few years. There are a lot of engineering challenges that come with putting them and having them operate successfully in space. But Elon Musk has. He's dreamed for the stars for his whole life. He is, you know, obsessed with moving civilization from Earth to Mars. And so if you're him, right, I mean, this is a very tantalizing prospect.

A (5:09)

This prospect is compelling enough that it's taken off across Silicon Valley.

B (5:14)

A lot of tech CEOs, including Sam Altman, Sundar Pichai at Google, Jeff Bezos, they've all talked about this idea of building data centers in space, and Google has launched some pilot programs in that regard.

A (5:30)

And here's Musk talking about orbital data centers at the World Economic Forum in Davos last month.

B (5:36)

Net effect is that the lowest cost place to put AI will be space. And that'll be true within two years, maybe three, three at the latest.

A (5:47)

SpaceX, Xai and Musk didn't respond to requests for comment. According to Wall street journal reporting, SpaceX has been studying orbital data centers behind the scenes for a while, and it had a technical breakthrough last fall. But there's something else XAI needs, and it's something you won't find in space money.

B (6:09)

In the world of AI, it costs a lot of money and requires a lot of data center power. So some investors kind of acknowled that it was very difficult for XAI to survive as an independent company in the long term. They'd either have to keep raising huge sums of money or get acquired by one of Elon's other businesses.

A (6:31)

Musk, being ultra rich and the owner of several major tech companies, has a few different chess pieces to play with here. One way to fund his AI startup is to just invest more money in it. Just a few days ago, Tesla, his electric vehicle company, did exactly that, investing $2 billion in XAI. And Musk has also sometimes combined businesses, like last year when X, formerly known as Twitter, combined with xai.

B (6:59)

For a long time, XAI and X, it was clear that their businesses were kind of intertwined, like they were sharing data center capacity and employees were moving between the two. Grok, the XAI chatbot, was put on X as a product that people could use. So he ended up merging those two businesses last spring. And a lot of people at the time saw it as a way to kind of fortify two struggling businesses in Elon's empire.

A (7:27)

This time, having SpaceX acquire Xai would give the AI startup a major financial boost. Because SpaceX has a lot of money.

B (7:36)

XAI can continue to do its AI research without having to worry that much about all these business questions around profitability and growing revenue. Because SpaceX is such a strong business and has such a strong financial profile.

A (7:57)

SpaceX is also planning to go public this year as early as this summer, according to Wall Street Journal reporting. And for Xai, being hitched to SpaceX ahead of an initial public offering gives it more financial muscle to compete with other AI companies like OpenAI and Anthropic. Both of those companies are also considering going public later in the year and through the SpaceX IPO. Now, Xai might beat them to the punch.

B (8:24)

There's an advantage to going out first if SpaceX has a huge IPO. And now, I guess since SpaceX is an AI company, right, they get to raise a lot more money. They get the benefit of the kind of halo and the attention around going public, and it definitely, I think, puts more pressure on OpenAI and Anthropic.

A (8:43)

What have XAI shareholders said about the acquisition?

B (8:47)

So XAI shareholders are very happy. I mean, I think for a long time, most XAI investors that you talk to, they know that this is the end goal, right? Like, I think none of them, even before this was announced, expected XAI to be an independent company that went public. And I mean, everyone was expecting the company to merge.

A (9:07)

But according to Wall Street Journal reporting over on the SpaceX side, they're kind.

B (9:12)

Of working through the rationale, right, and trying to buy into Elon's vision. But there's definitely more of a nervous and mixed reception amongst the SpaceX investor crowd.

A (9:23)

So what does SpaceX get from this merger? That's after the. Foreign. Out of all of Musk's private companies, SpaceX is by far the most valuable. The company develops cutting edge space technology like starlink, a satellite network that provides Internet connectivity. And of course, there's SpaceX rockets. They're some of the most advanced in the world, and backed by billions of.

B (10:04)

Dollars in government contracts, they control the rocket launch business. I mean, it's almost impossible to break into that market. That's why SpaceX is so valuable, and it's kind of the crown jewel of Elon's empire.

A (10:18)

Even before it acquired Xai, SpaceX was already a customer and an investor in the AI startup, according to people familiar with the matter. SpaceX uses a custom version of the XAI chatbot Grok, which is known internally as Spock. SpaceX also invested $2 billion in XAI last summer. But despite the rocket company's involvement with xai, actually combining the two companies has been the subject of intense discussion. For people close to SpaceX, the rationale.

B (10:50)

For the acquisition there isn't really like an immediate tangible benefit to SpaceX's bottom line. Right. The case you can make is that they could work together to figure out the technology needed to build data centers that are able to operate successfully in space. Right. They're able to work on the design together. SpaceX would have more of a sense of the research direction that XAI is going in. You know, how to design systems that kind of work very well with the models that xai's training. So I guess that's the kind of the benefit that SpaceX would get in theory.

A (11:28)

It was late last year that work on the deal picked up steam. According to recent Wall Street Journal reporting, at that point, the idea of combining SpaceX and XAI became a priority for people in Musk's inner circle. Soon, SpaceX executives were selling the company's new Focus on AI satellites to banks and investors, people familiar with the discussion said. On January 31, SpaceX and Xai signed a merger agreement that created a $1.25 trillion company. The deal closed two days later. Now, with SpaceX reportedly set to debut on the stock market in the coming months, executives at the mega company and Elon Musk have to sell their vision to the public. Is this risky for the valuation of SpaceX? Will it change anything about its plan to go public, if at all?

B (12:24)

I think it definitely complicates the plan to go public for two reasons. I mean, first, they have to actually integrate and unify the cultures. Right. So there's a lot of legwork to do there. And that takes time. And usually when a company goes public, they already have that stuff mostly sorted out. Out. The second thing goes to this question of how are they going to sell this new business? The way Wall street works is people. I mean, it's all about narrative. The question is, are they buying the narrative that the company is selling?

A (12:56)

It's not unusual for a more established, profitable company to buy a small startup with a lot of potential but not a lot of revenue. But Berber says this acquisition is different in part because Musk owns both companies. So investors might see it as a conflicted transaction. Also, the synergies Musk is promising, specifically these data centers in space, are an unproven technology.

B (13:20)

Even if you're going to give Elon buy into his vision, it is a bet that is very much ambitious. Some people would even say, you know.

A (13:29)

Out of this world, out of this world.

B (13:32)

Overly idealistic. And if you're a cynic, Right. And I've spoken to a lot of investors who are a little bit more cynical of Elon's motivations. They would say it's just a way for him to save Xai.

A (13:48)

The mega merger shows how committed Musk is to his grand ambitions for a multi planetary human species. But it also makes clear that his focus on AI is paramount.

B (13:59)

He is really willing to do whatever it takes to win in AI. And it's become clearer than ever that he is willing to leverage the power of other parts of his empire to shore up Xai.

A (14:13)

Yeah, it's kind of interesting. It's almost like we're once again asking the question, is Elon Musk reckless or is he a brilliant genius playing 4D chess while the rest of us are in 3D?

B (14:25)

Yes, I think that's exactly the way to look at it. And I think the IPO will really be an interesting moment in time where we'll see, right, whether he gets rewarded or punished for this acquisition. And we'll potentially even be able to get a look under the hood of just how much it costs to train AI models. Right. It'll depend on how they break it down the financials. But I mean, we could learn a lot. I think we will learn a lot from that process.

A (15:03)

That's all for today. Friday, february 6 additional reporting in this episode by megan bobrowski, corey driebush, micah madenberg, becky peterson and alexander saidi. The journal is a co production of Spotify and the wall street journal. The show is made by kathryn brewer, victoria dominguez, pia gadkari, isabella japal, sophie codner, ryan knudson, matt kwong, colin mcnulty, annie minoff, laura morris, enrique perez de la rosa, sarah platt, allen rodriguez espinosa, heather rogers, pier singhe, jeevika verma, lisa wang, katherine whalen, tatiana zemis and me, jessica mendoza. Our engineers are griffin tanner, nathan singapak and peter leonard. Our theme music is by so wiley. Additional music this week from katherine anderson, peter leonard, bobby lord, emma munger, nathan singapok, griffin tanner and blue dot sessions. Fact checking this week by mary mathis. Thanks for listening. See you on Monday.