Elon Musk’s acquisition of Twitter was, in a purely financial sense, a bad deal, and everyone (including Musk!) knew it was a bad deal.
The primary issue was that Musk agreed to pay $54.20 per share, or $44 billion, to acquire the company right as the entire tech sector sold off. Musk made the initial bid of $44 billion on April 14, 2022, and, after trying to bail, he finally closed the deal on October 27, 2022. For context: in that timeframe, Meta’s stock price slid from $210 to $99 per share.
Say what you want about the value of X as a beacon of free speech or a “digital town square” or whatever, but the price he paid at the time that he paid it was way too high — hence why Musk tried and failed to bail. To make matters worse, the post-acquisition business continued to deteriorate, with advertisers like Apple, Coca-Cola, and Uber leaving (though, since the election, many of these companies are looking to come back to the site). Between the steep acquisition price and decline in advertising revenue, investors have since marked down their internal valuations of X big-time, with Fidelity pricing its stake in the company 79% lower than its acquisition price, effectively valuing X at ~$9.4 billion.
However! A pending fundraise for one of Musk’s other ventures, AI startup xAI, has introduced an interesting wrinkle for X shareholders.
The most visible connection between X and xAI is X Premium users’ access to xAI’s chatbot, Grok. However, it’s X investors, not users, that have the most to gain from Musk’s AI business. Musk, an OpenAI co-founder, launched xAI in March 2023 to compete against his former project, and since then he’s had no issues attracting investors. Yesterday, I wrote about xAI closing in on a $6 billion fundraise that will value the company at $50 billion. This 11-figure price tag is huge for shareholders in his struggling social media company, thanks to a Musk tweet from November 2023: “X Corp investors will own 25% of xAI.”
To do some quick math here, per Fidelity’s estimate, X was worth $9.4 billion in September, and per Musk’s tweet (assuming this figure is still accurate), X Corp investors will own a $12.5 billion stake in xAI after the fundraise. Even if we account for some dilution as a result by outside fundraising, it’s quite possible that X’s stake in xAI is worth at least as much as the equity value of X itself: an 18.8% ownership stake in xAI would equal the total equity value of X.
And, unlike X, which has seen its business deteriorate over the last two years, xAI is booming. Last week, The Information published a good piece covering xAI’s newest datacenter in Memphis, Tennessee:
Two things about Musk’s supercomputer have jolted competitors: its size and the speed with which xAI built it. The supercomputer, fittingly known as Colossus, consists of 100,000 graphics processing units, the chips best suited to training and running AI software. That is several times bigger than similar supercomputers built in the past by Meta and other tech giants.
Stringing together so many GPUs into a single supercomputer isn’t as simple as it sounds because of how much power the servers consume and bottlenecks in the networking equipment used to connect the chips to each other. And completing the project as quickly as xAI did is unheard of.
Musk and Nvidia, the AI chip powerhouse that supplied the GPUs for Colossus, said the data center and supercomputer were built in just 122 days. On a recent podcast, Nvidia CEO Jensen Huang said a GPU cluster of that size would normally take three years to plan and design and an additional year to get working.
The speed with which xAI built “the world’s largest supercomputer,” 122 days, has put big tech companies, AI competitors, and investors on notice. AI has largely been seen as a compute arms race: the companies with the most computing power can build the most powerful models, and the most powerful models will, eventually, come out on top. While OpenAI and Anthropic’s revenue figures still dwarf xAI’s, the latter now has the most powerful supercomputer on the market, and he wants to triple its size to 300,000 GPUs by next summer.
While Musk hasn’t shown much success running a social media company profitably, you can’t deny his track record when it comes to infrastructure and manufacturing-related products. Tesla and SpaceX revolutionized their industries, and they’re now worth $1.1 trillion and $255 billion, respectively. AI warehouses feel more like Tesla and SpaceX than they do social media, and investors have been more than willing to bet on another Musk success story.
Thanks to xAI’s ballooning valuation, X shareholders are in an interesting position. On one hand, they paid way too high of a price for a social media site that is, by any conventional measure, on the decline. And yet, that same investment proved to be a call option for Musk’s AI venture that could, quite possibly, more than compensate for paper losses on the social media company.
Think about it like this: if you invested $4.4 billion in X for 10% of the company, and it’s down 75%, you’re left with a $1.1 billion stake in the social media company. But let’s say that xAI becomes the big winner of the AI model companies, hitting, I don’t know, a $200 billion valuation (it sounds crazy, but Musk is about to raise at a $50 billion valuation, and OpenAI is worth almost $160 billion), and X shareholders maintain a 20% stake in xAI which would, at that point, be worth $40 billion. Your 10% stake in X would now control a $4 billion stake in xAI.
Investors who joined Musk’s bid to acquire Twitter were buying a social media company (and allegiance with Musk, which, presumably, is worth a premium), but they may or may not have an interest in xAI. On the other hand, there are definitely investors who would like to invest in xAI, but can’t get allocation in a competitive round.
I feel like there is a trade to make here where an investor in X, who isn’t interested in xAI and wants to offload their stake, could sell to another party who isn’t all that interested in X, but wants to backdoor their way to having a stake in xAI, no? If you’re bullish on xAI at a $50 billion valuation, wouldn’t you be willing to invest in the X stake that controls 25% of xAI?
For Musk, the biggest value prop of buying X was probably some mixture of enforcing “free speech” and ensuring that your tweets end up at the top of everyone’s timeline. For other investors, the biggest benefit might be its investment in Musk’s better performing AI startup.
