Elon Musk’s Starlink satellite company has been getting a fair amount of attention here in Norway lately. Last month, a newspaper editor on the small island of Stord on Norway’s west coast told public broadcaster NRK that he’d replaced his local broadband provider with Starlink: it was both cheaper and faster, he said. Previously, he’d been paying 849 NOK (around $91) a month; Musk’s company only charged 589 NOK ($63). “The hardest part,” he said, “was making a hole in the wall” to connect the satellite kit.
He’s not alone; in fact, Starlink is doing well in Norway. The number of subscribers has nearly quadrupled, to around 15,000 subscribers by the end of last year. The Norwegian government is increasingly opting for Starlink as well: earlier this month, the country’s defense procurement agency Forsvarsmateriell signed a framework agreement worth up to 450 million NOK ($48.5 million) with Marlink, which resells Starlink services, covering the military, police, and domestic intelligence. The armed forces are already paying for 180 Starlink subscriptions.
This despite growing concerns, both in Norway and across Europe—indeed much of the world—about Musk’s increasingly unchecked power, and in particular about Starlink’s restrictions earlier this fall on Ukraine. A Norwegian Green Party politician, Ingrid Liland, criticized the Norwegian government for being “frighteningly naive” in selecting Starlink: “The Norwegian Armed Forces’ operational capability cannot depend on the mood of a right-wing radical and unpredictable tech oligarch,” Liland said, “or on an American president who grovels for Putin and who will not shy away from threatening allies to get what he wants.” Earlier this year, in Foreign Policy, Robert Muggah and Misha Glenny’s assessment was stark: “Starlink Has Privatized Geopolitics.” They argued that Starlink is “far more than a commercial connectivity service. It is strategic infrastructure that increasingly shapes how wars are fought.” Meanwhile, the EU is planning a “Starlink rival” under its IRIS² program, an “answer to pressing challenges ahead,” as the union diplomatically puts it.
Musk’s power is now wide-ranging—from launch capabilities and satellites, to AI and EVs, social media, election campaigns, and brain-computer interfaces. After some corporate reshuffling, he has built a trillion-dollar conglomerate covering space, tech, communications, and manufacturing: SpaceX now controls Starlink and the military Starshield satellites (it has filed for up to a million AI satellites under Starmind), and has absorbed the social media platform X, the AI model family Grok, xAI’s gigawatt-level Colossus data center cluster, and the AI coding company Cursor (bought for $60 billion, and paid for in stock, not cash, around the time SpaceX went public). Musk also wants in on telecoms: since September last year, SpaceX has committed $25 billion to buying mid- and low-band spectrum blocks in the US, both for satellites that beam directly to phones and for cell phone towers themselves, positioning itself to become a “serious competitor in the US mobile market,” in Reuters’s words. Musk is preparing to dominate not only low-Earth orbit but an ever-widening, frontier-level tech stack. The boards of Tesla and SpaceX have even formed special committees to assess a possible merger of the two, which would further cement the conglomeratization of the Muskverse.
Structurally, Musk is asserting himself as a dominant member of both a manufacturing-oriented capitalist class and what cultural theorist McKenzie Wark calls the “vectoralist class,” which controls the essential pathways, or “vectors,” of information in late-modern, digitized societies, and is therefore in prime position to harvest economic value. Musk is fusing power across these two camps, controlling both physical and digital infrastructural layers, often through complex financial acrobatics and narrative phantasmagoria. This also means he is well positioned to exercise social and political power: what makes Musk specifically an oligarch, in addition to being the world’s wealthiest man with a net worth of around a trillion dollars, is his ambition to reshape politics and wider society, both in the US and globally—even if the recent US midterms show that, in the face of popular mobilization, spending nearly half a billion dollars doesn’t guarantee a particular outcome.
Some of Musk’s business proposals are still pie-in-the-sky, like SpaceX’s fabulistic notions of a $28.5 trillion potential market, the “largest actionable total addressable market in human history,” as the company claimed in its S-1 filing with the SEC earlier this year. More than 90% of that market was said to be in AI, not space—a hard, unexpected pivot for a successful space launch company. Musk’s bigger bet on “orbital AI”—putting AI chips in space to do heavy processing—seems similarly fanciful: it faces serious technical and business limitations, not least because the vacuum of space doesn’t allow GPUs to be cooled by convection as on Earth.
But to their defenders, the criticisms of Musk or the wider Big Tech world don’t touch the core issue that sees both consumers and governments—like the editor on Norway’s island of Stord, or various countries’ defense forces—coming back to them: the quality and sheer usefulness of their services.
We don’t actually need to be afraid to concede that sometimes their technology can be highly useful. I’m the first to acknowledge that the “artificial unintelligence” of a few years ago has morphed into sometimes highly capable tools for academic summarizing and research discovery. (I still find it deeply unethical that billions of people were subjected to experimental gen AI well before it was ready for prime time.) Similarly, Musk’s rockets really have pushed down the price of launch, allowing more companies and countries to access low-Earth orbit for all kinds of communications purposes, including rural broadband. On a good day, Starlink has allowed Ukraine to defend itself. Nor, of course, should we hold as axiomatic that Big Tech’s products are a net good: they might be low-quality, troubling, unnecessary, wasteful, or dangerous. But our political choices about technology, and about the companies that currently produce it, should not hinge first and foremost on the quality of their services. Our politics of tech should be quality-agnostic, if you will. Ownership and concentrations of power are not completely distinct issues from the question of what those companies are making: Big Tech might build big, bad tech products, just as I do believe that a socialized producer of tech could make better, even utopian, technologies that could solve real social problems at scale.
What if we acknowledged that Big Tech’s products were never the personalistic creations of billionaire founders or owners, or the result of some unique genius of “nerd capitalists,” but the fruits of the labor of thousands of engineers, designers, coders, annotators, testers, and others? What if we pierced the ideological veil of Silicon Valley’s myth-making machine and started to recognize that frontier technologies are not the work of an oligarchic class—but rather the product of extremely skilled, creative human labor, in combination with vast quantities of economic capital, which need not reside in private hands?
Combining thousands of talented, world-class engineers and other professionals with hundreds of billions of dollars in economic capital is likely to produce sophisticated technological results. Now, that tech can certainly be put to ill use, of which there are plenty of examples, from Facebook’s addictive algorithms to Palantir’s surveillance systems. But it is precisely because these products are so advanced that they are crucial stakes in a progressive politics of the twenty-first century. Ironically, the concentrated nature of these combinations makes the case for nationalization clearer—what Marx called a situation in which the “expropriators are expropriated.” One of the odd benefits of Musk’s in-gathering of power under SpaceX is that he has, unintentionally, made the target of nationalization simpler to identify. The idea has gained a lot of traction lately: Nobel laureate Paul Krugman has advocated nationalizing Starlink and Starshield, while Clinton’s former Labor secretary Robert Reich wants to nationalize both “Starlink and SpaceX.”
Our societies are inextricably digital. There’s no getting around the need for increasingly complex technology to run things. We might even say that socialism in today’s world would mean something like people-power plus digitalization. While some technologies should be scaled back—space launch is becoming a significant source of harmful emissions, for instance, and satellite de-orbiting could pose a threat to the ozone layer—tech’s proper ends and limits can only be settled if societies wrest back control over the tech stack. That will require something like public ownership or nationalization.
Nationalization isn’t a guarantee of the democratization of tech, of course. The Trump administration has already shown a willingness to take an equity stake in Intel, and days after Senator Bernie Sanders’s proposal for an AI sovereign wealth fund in June, Trump suggested the government might take a stake in AI companies. Couldn’t nationalization place a great deal of tech power in the hands of an authoritarian, far-right leader? Nationalization does raise the stakes of who wields executive authority. But private ownership under a bad executive is hardly some ideal alternative. The absence of public ownership won’t necessarily prevent authoritarians from coercing firms into obeying their will. Apple was pressured by the Justice Department to remove ICE-tracking apps from its App Store. When Anthropic insisted that Claude not be used for mass surveillance of Americans and that no autonomous weapons systems use its tech without a human in the loop, the Department of Defense retaliated by branding it a “supply-chain risk.”
The problem of authoritarianism doesn’t mean we should abandon the idea of public ownership, but that we need to sharpen how we design it with democratic resilience in mind. One solution might be to organize public tech companies as public utilities under an independent commission or with various arm’s-length procedures in place. Another solution could be to internationalize public control, so that clusters of democratic societies keep each other in check, preventing any one government from capturing and corrupting the stack.
It isn’t founders, entrepreneurs, or billionaires who are advancing the technological frontier, but human labor. Socializing the combinations within which that human labor works could be a net benefit. And with the unprecedented “blue tsunami” of the US midterms, a unique political opportunity has arisen for democratizing tech and ensuring that it is put to good, sane, prosocial, and environmentally aligned uses. For now, that may mean tighter regulation and increased democratic oversight, with more far-reaching action awaiting a change in the White House. Whatever the outcome, technology is too important to leave to the billionaires and techno-oligarchs.