What’s the point of Elon Musk’s AI company?

Look, I’ve been following the adventures of xAI, Elon Musk’s AI company, and I’ve come to one conclusion: the only real idea is, “What if AI, but this time with Elon Musk?”

What is publicly available about xAI makes it seem like Musk showed up late and beerless to the generative AI party. This is 2024. The party is busy now. xAI doesn’t seem to have anything that makes it stand out outside of Musk.

However, that won’t stop Musk from selling his idea to investors! Last December, xAI said it was seeking to raise $1 billion through a filing with the Securities and Exchange Commission. (This is not the same company as the X formerly known as Twitter.) There is also reporting from the Financial times say Musk is seeking up to $6 billion in funding.

xAI (not Twitter) has one product so far, a supposedly brutal LLM called Grok

To be sure, Musk tweeted that “xAI does not raise capital and I have not had discussions with anyone in this regard.” Musk says a lot of things in public and only some of them are true, so I’m going to rock with the filing, which I saw with my own eyes.

xAI (not Twitter) is kind of a strange entity. Other than its entanglement with X (Twitter), it doesn’t really seem to have any particular purpose. The xAI pitch deck obtained by Bloomberg depends on two things:

  • OpenAI had a very successful 2023. Musk was of course one of the founders of OpenAI and stormed out in a panic in 2018. Lately, he says OpenAI has offered him shares and he turned them down.
  • An investment in xAI gives you access to the ‘Muskonomy’. Note: That word isn’t my fault. It is true that Musk is involved in many companies and is the alleged CEO of some of them. In addition to xAI (not Twitter) and X (Twitter), there is Neuralink, SpaceX, the Boring Company and Tesla. We’ll be at Tesla in no time.

xAI (not Twitter) has one product so far, a supposedly Brutal LLM called Grok, which users can access by paying $16 per month to to go. xAI (not Twitter) does not have a standalone interface for Grok. My colleague Emilia David has characterized it as having “no raison d’être” because it isn’t meaningfully better than its competitors’ free chatbot offerings. Its most obvious distinguishing feature is that it uses X-data (Twitter) as real-time input, allowing it to serve as a kind of operatic spectacles for platform drama. The Discover/Trends section of the

Grok was developed terribly fast. One possible explanation is that Musk has hired an in-demand team of the very best in the field. Another is that it is a refined version of an open-source LLM like Meta’s Llama. Maybe there’s even a secret third thing that explains its rapid development.

Besides X (Twitter), the other data source for xAI (not Twitter) is Tesla. Bloomberg‘s reporting. That’s curious! In January, Musk said, “I would rather build products outside of Tesla” unless he gets “~25 percent voting control.” Musk has also said that he does not feel he has enough control over Tesla to feel comfortable “growing Tesla into a leader in AI” and that he would definitely prefer Tesla stock “to products would want to build outside Tesla’.

Tesla has been working on AI in the context of self-driving cars for some time and has encountered some of the same obstacles as other self-driving car companies. There’s also the Optimus robot, I think. These seem like specific use cases that are significantly less general than building any other LLM. That Tesla data is valuable and goes back years. If xAI transfers it, I wonder how Tesla shareholders will feel about that.

Who wants to fund a very generic AI company in a crowded space?

There are certainly real applications for AI. Databricks exists! It’s not consumer-facing, but it does seem to have a specific purpose: data storage and analysis. There are smaller, more specialized companies that deal with industry-specific types of data. Take Fabric AI: its goal is to streamline patient intake data for telemedicine. (It also makes a chatbot that threatens to replace WebMD as the scariest place to ask about symptoms.) Or Abnormal Security, an AI approach to blocking malware, ransomware, and other threats. I don’t know if these companies will achieve their goals, but at least they have a compelling reason to exist.

So I’m wondering who else wants to fund a very generic AI company in a crowded space. And I wonder if the reason Musk denies raising money at all is because there isn’t much interest in xAI, and he’s trying to minimize his embarrassment. Why does one of the richest men in the world even need outside financing for this?

Silicon Valley’s assessment of Musk has been remarkably resilient, probably because he has made a lot of people a lot of money in the past. But the debacle at X (Twitter) was disastrous for its investors. And Musk has been distracted at a crucial time for Tesla, which is facing increasing competition. Tesla’s latest product, the Cybertruck, comes without clear coat; some owners say it is rusting. (A Tesla engineer claims the orange dots are surface contamination.) And in its most recent earnings, Tesla warned that growth is slowing. Meanwhile, Rivian’s CEO has been open about his attempts to directly undermine Tesla.

A perhaps underappreciated development over the past twenty years has seen Elon Musk go from ahead of the investment curve to a top signal. Take, for example, the “GameStonk” movement, when Musk’s tweet was the perfect sell signal – not just for retail investors, but also for sophisticated hedge funds. Or the Dogecoin crash that happened when he called himself the Dogefather on SNL. Or even Twitter, which certainly wasn’t worth what Musk ultimately paid for it and has since rapidly declined in value, to the point where the debt from the deal is being labeled “non-investable” by a firm that specializes in distressed debt.

I don’t see a compelling case for xAI. It has no specialized purpose; Grok is an LLM that is also managed and aims to strengthen an existing product:

Okay, I looked into it: ChatGPT isn’t growing anymore.

Musk is trying to pitch a new AI startup without a clear focus as the generative AI hype begins to wane. It’s not just ChatGPT: Microsoft’s Copilot is experiencing a sharp drop in usage after a month. There is now an open question as to whether the productivity gains from AI are enough to justify its costs. So here’s what I’m wondering: How many investors think just adding Elon will solve it?

With reporting by Alex Heath.

Leave a Reply

Your email address will not be published. Required fields are marked *