18.06.2026

Who wants to be a trillionaire?

Amid an increasingly obvious tech bubble, Elon Musk has added a 13th digit to his net worth. Paul Demarty asks what this tells us about American capitalism today

So, finally, humanity has minted its first trillionaire.

The lucky man is Elon Musk, who made his initial fortune in the flurry of mergers that produced PayPal, and has since concentrated his efforts on the hardware end of high technology, assuming control of Tesla and turning it into America’s first mass-market manufacturer of electric cars, creating The Boring Company as a way to create car-based, high-speed transit in underground tunnels (so far a failure), and recently purchasing Twitter and turning it into an even more miserable social network called X.

Commenting back in 2003 on the scandalous corporate bankruptcies of the dot-com bust era, comedian Lewis Black expressed astonishment that the owners of the Tyco IT conglomerate had extracted $500 million from the company. “What were they going to do?” he asked. “Start their own space programme?” If only he had known what was coming! Musk broke the 13-digit net-worth barrier by floating his personal space programme on the stock market. The initial public offering (IPO) of SpaceX, as it is called, went off fairly cleanly (though the powers that be had to relax their standards to allow this entirely unprofitable enterprise into public markets). He is not even the only member of the ultra-rich to have such an endeavour - Jeff Bezos’s Blue Origin continues its work, albeit still under Bezos’s personal tutelage.

What even is a trillion dollars? What could one do with it? How much land would you need to do the old Keynes trick and pay people to bury it and dig it up again? How many trees would you need to fell to print the banknotes to represent it? What, more to the point, could you spend it on? Obviously not enough luxury goods exist in the world; and there seems little point in owning 100% of all Lamborghinis, Swiss watches and so forth. It could be invested, and at that point Musk could set himself up as his own ‘institutional investor’. But all he could get from that is … more money.

Fortunately for him, of course, it is not precisely the case that he has a bank account with a 13-digit balance. His worth is largely in asset holdings in his own and presumably other companies; the trillion dollars is obtained from adding up the market capitalisation of this portfolio and assessing his relative ownership stake. No doubt Musk has plenty of ready cash to be going on with, and - so long as the stock price of his various ventures remains buoyant - he can liquidate some of his possessions. But a lot of this money exists, at best, only in theory.

Price and value

What is the ‘theory’? The stock price of a company - and the price of other associated assets, like corporate debt - is supposed to be a function of its present and historical financial performance, on the one hand, and expected lifetime value, on the other. There are standard, off-the-peg measurements of this - take this year’s profit and multiply it by some factor, for example. Yet there has always been the possibility of selling at a premium, if there are plausible reasons to expect far better performance than in an average company.

This little get-out clause, as one might call it, has been exploited mightily by the tech industry, broadly conceived, in the past 15 years or so. Venture capitalists have been broadly successful in convincing institutional investors that economies of scale are so easy to achieve in verticals like software-as-a-service that valuations which look, prima facie, ludicrously inflated can be a good deal. When I started programming as a career, it was still rare for startups to achieve billion-dollar valuations, and so such companies were called “unicorns”. By the end of the zero-interest-rate era in 2022, it seemed that there were more unicorns in northern California than there were buffalos on the great plains prior to the coming of the Pilgrim Fathers.

This period was not without its scandals. WeWork - a short-term office-space subletter - caused a stir when its S-1 (the formal IPO filing to the US Securities and Exchange Commission) bombed completely, and proved to be full of guff about “elevating the world’s consciousness”, including plans for a network of woo-woo schools run by the CEO’s wife (a relative of woo-woo god-empress Gwyneth Paltrow). Nonetheless, a lot of idiotic stuff got waved through, particularly in the region of cryptocurrencies and related technologies.

The Federal Reserve put an end to that, and the startup scene is now at best stagnant. However, a much more concentrated bubble has taken form since, particularly focused on artificial intelligence (and, specifically, large language models). A third of the value of the American stock market is in a small number of tech firms heavily involved in the AI business. A decent chunk of US economic growth is in data centre expansion, specifically for these purposes (LLMs are enormously hungry for very specific kinds of computer hardware). Anthropic, one of the two major AI model vendors, quietly began preparing for IPO last week. That will certainly be an S-1 worth reading.

SpaceX is not obviously part of the AI bubble per se (a certain amount of bandwagon-jumping guff about AI was added to the various investors’ prospectuses). It nonetheless has something of a similar smell to it. Its present-day financials are pretty awful, with annual losses in the order of $20 billion. Amazingly, the various bankers involved in this sale blithely admitted that they expect losses to increase over the next 15 years. This is not the ‘conventional’ prospectus of the tech premium: investors buy in with the expectation that the stock price will go up regardless.

Similar dynamics swirl around Anthropic, OpenAI and friends; the vast capital expenditures required to make their products work at all do not seem to unlock superlinear revenue growth of the sort required to justify exposing the world’s pensions to their shares. The promise is of a new industrial revolution forever just around the corner.

It is increasingly clear that, beneath all the talk of industrial revolutions, the marquee names of AI - and the cloud hyperscalers like Amazon and Microsoft, for that matter, who effectively resell this functionality - will in fact have to cash out in deals with governments, and specifically in the military. It is suspected that Anthropic’s Claude was used in the targeting of the girls’ school on day one of the Iran war; but, of course, this is already a routine matter, and it will get more so.

That is radically more true of SpaceX, which - when you get down to it - has very little potential for profitable civilian business at all. Its Starlink network is already a real factor in various hot wars around the world, and a telling military asset for the US. It is a successful contractor for government rocket launches, including the recent Artemis II mission; for all the Star Trek utopianism, in reality this work will largely turn into line items at the Department of Defense - sorry, War. The US military budget, after all, also hovers around $1 trillion at present.

Deferral

So there are a couple of reasons why a bursting of the bubble may be deferred. The first is simply that, since deindustrialisation began to hit hard, the United States has managed its social fabric effectively by asset price inflation, backed by ballooning public and private debt. This periodically goes into crisis, of course; credit dries up, individuals and corporations must sell assets to meet liabilities; the asset prices begin to shrink, making more debt go bad, and so on. Sooner or later, the cost must fall on someone. In 2008, the cost fell on ordinary mortgage-holders; the banks were able to sell gilts under quantitative easing and use that to pay off their underwater liabilities.

This came at an enormous cost to the political coherence of the American state. It is not surprising, then, that - despite the fact that everyone from JP Morgan to the Financial Times recognises that we are in a bubble - the only actions being taken are to protect the bubble. The political choice between hammering ‘savers’ and imposing big haircuts on the banking sector is singularly unattractive, so it is deferred.

The second reason is that we are moving into a period of great-power competition, and it is in the interests of the state to prop up industries it considers to be of military and strategic importance. Plainly this includes SpaceX. It also to some extent includes AI, although much of the big-name products in the recent AI boom are tailored either to consumer-grade trivialities or automating business processes substantially irrelevant to global conflict. However, military and surveillance uses of the fundamental technologies of AI are plentiful, and America has an interest in the ‘research and development’ taking place there instead of in China.

The enormous increase in the share of the surplus going to capital (and, within that, specifically to the tech industry, broadly conceived) rather than to labour has, of course, provoked a certain amount of theoretical reflection on the left. Many well-known intellectuals, notably Yanis Varoufakis, Jodi Dean and Cédric Durand, have begun to advocate a theory of ‘techno-feudalism’ - the business model of the great tech platforms is, after all, centred on rent extraction, typically from other capitalists, and fuelled by mass, unpaid labour in the form of platform engagement (which generates the data on which the platforms are trained).

Alternatively there is the theory of Robert Brenner and Dylan Riley that what we have is a new regime of accumulation they call ‘political capitalism’: we are going through a long period of secular stagnation, which depresses the rate of profit and tends to make the system more zero-sum than positive-sum in its dynamics. The result is a move towards profiting by direct control of the political system, with rival camps of the working class absorbed into competing class-collaborationist alliances and, in the end, a substantial upward redistribution of wealth.

The techno-feudal thesis seems insupportable. It radically overstates the novelty of large-scale rent-seeking in capitalist society, as well as the use of surveillance as a labour-disciplining system. It has the advantage, from its advocates’ point of view, of recommending a politics of class collaboration on essentially populist grounds: workers and petty proprietors alike are united against a tiny elite of stupefyingly wealthy people.

Brenner and Riley are more interesting, since their objective is plainly to reject class collaboration, though their account is at best radically incomplete. They introduce a distinction between “working class politics” defined through participation in struggle for class-level demands per se, and an individualist competition for sectional advantage. Correct in principle, but they draw so sharp a distinction that it risks caricature: “Working-class politics in this sense has been a highly unusual occurrence in US history. There were only two brief spells of it in the 20th century.”¹ It is better, I think, to consider sectional politics a degraded form of class politics, in that it is likewise oriented to a situation of real class antagonism, albeit in a strategically hopeless way.

Related to this is the near-total absence both of the intermediate classes (effectively folded into the proletariat in their account) and the bourgeoisie, and how the latter might fight for their interests. Nonetheless, their account does throw an interesting light on the bourgeoisie, since presumably they, also, fight for sectional advantage against the background of stagnation and feeble profit rates. The growth of fortunes like those of Musk and Bezos, in short, inescapably flows from parasitic relations with the state and alliances with fractions of the state core. At this stage of capitalist decline, that is where the action is, and perhaps this justifies a term of art like ‘political capitalism’.

Regardless of that, nobody, surely, could be so stupid as to see this news as evidence of the system’s vitality.

1. ‘Seven theses on American politics’ New Left Review November-December 2022.↩︎