Last year, Elon Musk’s xAI artificial intelligence start-up bailed out investors in X, his social media platform. Then Musk’s Tesla pumped $US2 billion ($2.85 billion) into xAI, which was burning more than $US1 billion of cash a month. Now he’s caused SpaceX to merge with xAI in a deal valuing the combined businesses at $US1.25 trillion ($1.8 trillion).
See the pattern? As one of the companies he controls needs cash, he turns to another that has it. When that proves insufficient, he has looked to one of the two most valuable companies within his orbit – one without the approvals and governance complexities of Tesla – to bail them out.
There was speculation that it would be Tesla that would acquire xAI but, while Tesla has about $US44 billion of cash in its balance sheet, the switch of focus from producing electric vehicles to robotaxis and robots that Musk revealed last month and the cashflow-negative impact of the $US20 billion investment that entails probably ruled it out, at least this time.
SpaceX planned a float this year, with a mooted valuation of about $US800 billion, which would raise about $US50 billion of cash from new investors. Some of that cash will now be available to xAI.
The merged group’s valuation of $US1.2 trillion represents a big lift from the most recent valuations of its components as, just over a year ago, the merged xAI and X was deemed to have an equity value of about $US113 billion and a $US20 billion capital raising last month was struck at a valuation of $US230 billion.
For the purposes of the merger, xAI is now valued at $US250 billion, while SpaceX’s value has jumped to $US1 trillion and there’s talk of the IPO producing a $US1.5 trillion valuation.
The numbers, and their escalation, are mind-blowing, but so are artificial intelligence’s appetite for capital and Musk’s visions.
xAI is one of the minnows competing with giants in the hugely capital-intensive race to stake out a position in AI.
Where giants like Google, Amazon, Meta and Microsoft have torrents of cashflow from their legacy businesses to fund the hundreds of billions of dollars they are ploughing into AI – and even they are now having to raise external debt because of the scale of the spending on training large language models and building data centres – start-ups like xAI, OpenAI and Anthropic are reliant on seemingly endless rounds of equity raisings (and, increasingly, debt issues) at ever-greater valuations.
The sector’s insatiable appetite for capital has outstripped the capacity or willingness of equity markets to supply it, hence the increasing levels of debt in the sector and the incestuous deals between its participants.
On some estimates, the four or five heavyweight players in the sector will spend about $US500 billion over the next three years. It wouldn’t surprise if it were significantly more. This is an extraordinary period of highly focused and extraordinarily capital-intensive industry development.
At some point, perhaps quite soon, there’ll be either a consolidation of the sector or a wipeout of its weaker players.
A puncturing of the ever-inflating, and arguably already highly inflated, valuations would force rationalisation (and cause a lot of pain for the companies and their investors and lenders) if the equity spigot were turned off.
With the enlarged SpaceX not the only big AI-tinged float planned for this year – OpenAI and Anthropic have theirs on drawing boards – the market’s willingness to continue to support financial models that involve massive outlays today for uncertain returns in three, four or five years’ time will be stress-tested.
That’s why, as was the case when xAI was used to bailout investors in X (who were badly underwater on their investments), Musk has turned to SpaceX and its ability to raise a very big lump of cash – at $US50 billion, an IPO equity raising would be the biggest in history, dwarfing the $US29 billion raised by Saudi Aramco in 2019 – just to keep xAI in the game.
The industrial logic for combining Space X with xAI is thin.
There are not many synergies between launching rockets and operating a network of satellites, an AI start-up that lags its competitors and a controversial social media platform where the interest on its debt absorbs roughly half its revenues.
xAI’s Grok chatbot doesn’t have the users or financial backers (almost the entire AI sector) that OpenAI’s ChatGPT has, nor the distribution platform of Google’s Gemini or the corporate acceptance of a rival start-up, Anthropic (which moved markets with the release of new AI productivity tools on Tuesday).
SpaceX said it was acquiring xAI to “form the most ambitious, vertically-integrated innovation engine on (and off) Earth, with AI, rockets, space-based internet, direct-to-mobile communications and the world’s foremost real-time information and free speech platform.”
Musk himself said the rationale for the merger was that, within two or three years, the least expensive way to do AI computations would be in space.
He envisages putting AI data centres into space on SpaceX’s rockets and said there would be cost efficiencies that would enable innovative companies to train their models and process data at unprecedented speeds and scales.
“This marks not just the next chapter but the next book in SpaceX and xAI’s mission: scaling to make a sentient sun to understand the universe and extend the light of consciousness to the stars,” he said, whatever that means.
Maybe the future of AI and the data centres needed to support it lies in space, but Musk seems to have skated over the logistics and costs of building something in space that is already stretching, to their limits, AI companies on Earth.
The sector’s insatiable appetite for capital has outstripped the capacity or willingness of equity markets to supply it, hence the increasing levels of debt in the sector and the incestuous deals between its participants.
There is some financial logic to the combination, given that SpaceX is cashflow positive – it appears to have generated more than $US2 billion of positive cashflow last year and is on a steeply rising curve – while xAI churned through $US9.5 billion in the first nine months of last year at an annualised rate of more than $US12.5 billion.
It appears obvious, however, that the capital SpaceX raises in its IPO will be quickly devoured by xAI unless the industry trend of capital expenditures rising at a much greater rate than revenues can be reversed.
Musk does, of course, still have Tesla, valued by the market at more than $US1.5 trillion, that he could merge into a Musk Inc conglomerate if he needs more cash (and if its change in strategy from EVs to robotaxis and robots produces cashflows and earnings to validate the valuation).
On his track record, such a merger appears quite a likely eventuality, particularly if xAI’s demands for cash outstrip SpaceX’s ability to generate it.
Musk’s “vision” for SpaceX appears fanciful. He has, however, demonstrated a remarkable capacity to convince investors to capitalise his visions.
Data centres in space, transported by SpaceX rockets, fed by xAI’s models that have trained on X’s data, perhaps with some future input from Tesla’s engineers and with roles for its robots?
It all sounds fantastical but, if he gets a $US1.5 trillion valuation for SpaceX in its float, it would signal that there are plenty of investors who think his vision can be realised.
The Business Briefing newsletter delivers major stories, exclusive coverage and expert opinion. Sign up to get it every weekday morning.
Stephen Bartholomeusz is one of Australia’s most respected business journalists. He was most recently co-founder and associate editor of the Business Spectator website and an associate editor and senior columnist at The Australian.Connect via email.