Posts tagged AI
Patrick Boyle on the repercussions of the U.S. government’s shutdown of Anthropic’ flagship AI models on June 12:
For years, the theory was that building artificial intelligence required Silicon Valley genius, billions in venture capital, and an exclusive relationship with a cloud computing giant […] It turns out that there is no monopoly on math. The open-source models are good enough. They’re 60 times cheaper, and nobody in Washington can switch them off on a Friday afternoon. The productivity gains everyone was promised will still happen; they’ll just flow to the businesses running the code on their own machines rather than to the investors who paid almost a trillion dollars for a global monopoly on math.
It is hard to grasp how absurd and evil the SpaceX IPO is when you pick out the facts from Patrick Boyle’s hilarious analysis.
[…] According to the [IPO prospectus], 93% of [SpaceX’s] total addressable market is attributed to AI. And based on the capital expenditure breakdown, roughly 60% of all spending is now going into artificial intelligence infrastructure.
[…] When Musk says that this is an AI company, what he means is that the one business that actually works [Starlink] is subsidizing a money-losing AI venture while he tries to convince public market investors that the money-losing AI venture is the entire reason the company’s worth $1.75 trillion.
[…] In 2025, the company generated $18.7 billion in revenue. […] at a proposed valuation of $1.75 trillion, SpaceX is asking investors to value it at almost 100 times its revenue.[…] If SpaceX were dropped into the S&P 500 today, it would be the seventh largest company in the index by market cap. But if you ranked it by actual revenue, it wouldn’t even meet 200th. It would be down near Kelloggs […] Unlike Kelloggs, SpaceX is not profitable.
[…] Capital expenditure reached $20.7 billion last year, quintupling from 2023, driven by Starlink expansion, Starship development, and AI infrastructure spending. To fund all of this, the company is carrying $29 billion in debt, including a $20 billion bridge loan taken out in March 2026 — eight weeks ago. So when investors hand over $75 billion in this IPO, a meaningful portion of it […] goes straight back to the banks who lent SpaceX money eight weeks ago.
[…] SpaceX claims its total addressable market is $28.5 trillion. This is larger than the entire GDP of the United States and almost entirely built on businesses that do not yet exist. According to the prospectus, enterprise applications of AI make up to 80% of SpaceX’s total addressable market. […] The filing effectively concedes that the enterprise AI market is dominated by OpenAI, Anthropic, and Google. SpaceX’s own AI product, Grok, holds a market share of roughly 3.4%.
[…] SpaceX’s near-term AI revenue story rests substantially on a $15 billion a year deal to rent compute capacity to Anthropic, one of the very competitors whose existence proves that Grok is losing the AI race. […] The more pressing problem is that this $15 billion a year contract — roughly 40% of projected near-term revenues — can be cancelled by Anthropic with 90 days’ notice. SpaceX’s entire AI business model at the moment is renting GPUs to a competitor on terms that can vanish in a fiscal quarter. […]
Now the related-party concerns:
Antonio Gracias is a long-term Musk ally and a sitting member of the SpaceX board of directors. He also runs a private investment firm called Valor Equity Partners. According to the IPO prospectus, SpaceX and XAI are carrying more than $20 billion in AI infrastructure lease obligations tied to agreements with entities connected to Valor. Some of these deals are structured as what accountants call „failed sale-leasebacks.” […] The practical effect is that SpaceX is carrying billions in obligations to a firm run by one of its own board members on its balance sheet as debt.
[…] Back in 2016, Elon Musk used Tesla’s shareholder money to acquire SolarCity, a struggling solar company that he also chaired and partly owned.[…] The court reviewed the entire transaction […] describing the process as deeply flawed […] [Musk] has now completed a structurally similar acquisition, this time of XAI, in a state with far more lax securities rules. […] The lesson that Musk appears to have drawn from SolarCity is not „don’t do conflicted acquisitions”; it’s „do them somewhere with friendlier courts.”
Who is profiting?
The clearest beneficiaries of this acquisition are not the SpaceX engineers trying to get to Mars. They’re the venture capitalists and banks, including Morgan Stanley, who were at first stuck holding billions in underwater debt from Musk’s chaotic buyout of Twitter. Thanks to the XAI-SpaceX merger, those lenders have been made whole in SpaceX equity, conveniently timed just before the largest IPO in history.[…] 23 banks have crammed their names onto the cover of this prospectus. The fees constitute the most lucrative Wall Street IPO payday ever.
and more:
All of this brings us to the corporate governance structure of SpaceX […] While Musk only owns around 41% of SpaceX, he will control more than 85% of the votes, meaning that he can’t ever be fired by shareholders. The company has also placed limits on how shareholders can file legal challenges. As SpaceX writes in the prospectus, „this will limit or preclude your ability to influence corporate matters and the election of our directors.”
[…] On top of that, because Elon Musk convinced NASDAQ to change their index inclusion methodology in order to win this listing, SpaceX will be fast-tracked for inclusion in the NASDAQ 100 index. […] countless pension funds and passive investors will become forced buyers of the stock almost immediately, long before there’s any serious price discovery, and they’ll own shares in a heavily cash-burning company where they have no voting power and no meaningful ability to sue management.
[…] Musk is getting 1 billion performance-based restricted shares and how these shares only vest when SpaceX establishes a permanent human colony on Mars and hits a $7.5 trillion market cap. […] But there’s a trick. As law professor Ann Lipton points out, Musk doesn’t actually need to achieve any of those hard-to-reach goals. […] He gets all the benefits of owning the shares immediately based on a Mars colony that the company admits he will probably never build.
[…] nothing stops Elon Musk from setting up a separate private company tomorrow to pursue whatever the next big tech trend is rather than doing it inside SpaceX […] As Ann Lipton summarized it: public shareholders in SpaceX will have no votes that matter, no ability to sell at a fair price discovered through normal market mechanisms, and no ability to sue the company for wrongdoing. No votes, no sales, no suits.
This is a transcript from an interview of David Sirota with Ed Zitron on Youtube, published on Dec 15, 2025.
Nvidia claimed […] that they shipped 6 million Blackwell GPUs. Now, I’ve confirmed with sources that this was specifically AI GPUs, 6 million of these. So, doing a little napkin maths, that’s about 10 to 12 GW of required power IT load. So, when you turn them on… but we’ve not built anywhere near that many data centers in the last year […] there are likely millions of GPUs sitting in warehouses waiting to be turned on. […] Satya Nadella of Microsoft said in an interview a couple of weeks ago that he has chips sitting in inventory that do not have power to turn on.
I think this is one of the most insane things that’s ever happened in history, ’cause I think that Nvidia is pushing billions of dollars of GPUs on companies that cannot install them.
When Will The AI Bubble Pop? Full Interview With Ed Zitron (10:38 onwards)
I am now collecting quotes about the AI craze, just for my records.
LLMs are dangerous for many, many reasons, but the under-discussed one is how well they play to a certain kind of executive imbecile. Generative AI is — to quote Mo Bitar — really good at doing an impression of work, much like most managers and c‑suite executives, and even if it’s completely incapable of doing something, it’ll absolutely say it can and tell you you’re amazing for suggesting it. And that’s why Business Idiots love it.
26 May 2026, Ed Zitron: Revenge of the Business Idiot