X + xAI = a16z, Fidelity, Kingdom Holdings, Sequoia Capital

And: Final notes on CoreWeave

Welcome to Cautious Optimism, a newsletter on tech, business, and power.

Happy Monday! The economic calendar for the week is here. The big upcoming news is the package of tariffs that POTUS will announce on Wednesday (April 2nd). Today is the final day of the first quarter — so make sure to close out those KPIs, yeah?

The stock market is off around the world today, with especially sharp losses Germany and Japan (tariff concerns). Recall that to close out last week, domestic tech stocks sold off sharply. Currently another 1.5% selloff is in the cards (premarket trading). With all that said, get your chin up and let’s get to work! — Alex

• 📈 Trending Up: Perplexity’s financial health … karma … humanoid robots … Tesla protests … tariffs … Japanese domestic chip production … Chinese AI … État de droit … economic meddling … Apple in India … Isomorphic Labs …

  • AI costs? While the cost to run released AI models falls quickly, upper-tier price points for some AI tech is going up. OpenAI’s $200 monthly subscription tier is a good example. As is Manus AI’s new pricing. The company recently secured more compute capacity and access to Anthropic’s models, mind.

• 📉 Trending Down: GDP growth … GDP growth … Apple-Starlink … peace? … EU fines …

• Grift Watch: The Trump family is investing more heavily in crypto operations. Past defi and memecoins and stablecoins, the family is expanding its footprint in bitcoin mining. Given that the Trump patriarch is busy ending investigations into Web3 rogues and cheats while stumping for national investment in the same assets that his sons are boosting their exposure to, well, it all feels more than a little gross.

X + xAI

As last week came to a close, Elon Musk’s social network X (previously Twitter), and his AI company xAI agreed to merge. The official verbiage went like this:

@xAI has acquired @X in an all-stock transaction. The combination values xAI at $80 billion and X at $33 billion ($45B less $12B debt).

The idea of one of the two companies acquiring the other is a bit rich. Both are owned by Musk, and share an investor pool. To wit, here’s an abbreviated rundown:

• Select X investors: a16z, Fidelity, Kingdom Holdings, Sequoia Capital

• Select xAI investors: a16z, Fidelity, Kingdom Holdings, Sequoia Capital

I could find more I suspect, but I think that sampling suffices.

Making the “acquisition” even more humorous, the two companies already had strong financial ties. And by that I mean xAI was paying X’s bills. Recall this bit of reporting from the WSJ in February:

Financial documents reviewed by investors showed that the artificial-intelligence company transferred hundreds of millions of dollars to the social-media company, the people said. That money has helped X pay its bills and stay current on its obligations, the people said. Growing advertising revenue at X should mean fewer transfers in the coming months and years, the people said.

xAI was also reportedly paying X a $65 million fee to license its data, and X used its own money to purchase chips for xAI at one point. To cap off the pre-deal ties, X owned a big chunk of xAI and buyers of X debt were offered that equity as collateral.

To say that the two companies were distinct before would have been a stretch. So, to see them come together is not a shock. (Musk’s personal history includes other, similar deals of course.)

Snark aside about the optics of merging two sister companies as if they were born from entirely different roots, the deal makes good sense. I have no view — at this point — of the valuation set to the combined entity. If you would put your capital into the X+xAI combo is a personal perspective.

The deal’s potential financial fecundity, however, rests on the fact that without X, xAI would be without an edge, and X has found that xAI’s AI products are great boosters of its own business. Therefore, with each company already leaning on one another, it is reasonable to formalize the pairing.

Let me explain:

• xAI needs X because it offers a unique dataset and a platform for both distribution and revenue generation. Valor Equity Partners’ Antonio Gracias spoke about the importance of X’s data to xAI last April. His argument was that AI foundation models sans proprietary data will depreciate to zero, while AI models that have a unique and high quality data source to learn from will succeed. Therefore, xAI sans X is just a big compute cluster with products that, thus far, have yet to slow down domestic rivals OpenAI and Anthropic.

  • X’s integration of xAI’s Grok is — seemingly — going well. I see folks chatting with it frequently, and it became so popular in India that it ‘kicked up a storm’ in the nation.

  • Finally, X has an existing set of paying subscribers that were partially enticed with more, and greater access to xAI’s AI technology.

• X needs xAI because its debts are huge, its business desultory, and consumers are willing to pay for AI tools. How many major social media funding rounds have you seen in the last few years? None? Apart from that Farcaster deal (data on the Web3 social network’s performance here), zero, right? Musk told staffers at X that the company’s “user growth is stagnant, revenue is unimpressive, and we’re barely breaking even,” which is not a huge endorsement of the social platform’s health.

  • But xAI can raise as much money as it wants, given market enthusiasm for AI foundation model companies. X had big debts it needed to manage and xAI can raise lots of money. Ta-da!

  • Most importantly, modern consumers will pay for AI tools. More willingly I’d wager than they will cough up subscription fees for social networks. X was therefore a good vehicle to sell xAI access, but a poor reciptical for those revenues which it would have to mostly pass to its sibling entity. Why bother? Just merge!

I do not know if the combined X+xAI company will succeed as its owner and backers hope. But certainly, the two are an obvious pairing. One that could offer cost savings through the reduction in duplicate staff and better allow for the sale of customer value without the need to pass revenue between corporate entities.

Final notes on CoreWeave

Pricing an IPO is no small task. If you go public, and your shares rise from their set IPO price, private-market investors will berate you for leaving money on the table. If you go public and your shares fall, public-market investors will bemoan a mispriced IPO. And if you go public and trade nowhere, it casts a bit of a pall on your newly tradable equity.

CoreWeave finds itself stuck between options 2 and 3. CoreWeave traded closely to its IPO price of $40 per share Friday. Today, like the rest of the market, the company is going to lose some value, pushing it under its IPO price at least at the open.

For company that had to price under its proposed IPO range, it’s a all a bit grey. But, CoreWeave now has a huge slug of new capital to manage its debts and keep growing. We leave it in the hands of Wall Street, even as its final acts as a private entity failed to kick off fireworks. Or, perhaps, entice a wave of other IPOs.