Silicon Valley is moving sharply and openly to the right. In a presidential inauguration where the best seats were essentially sold at auction, tech billionaires and CEOs from Apple, Amazon, Google, and Meta paid premiums to sit close to Donald Trump. The world’s richest man, Elon Musk — once a darling of those hoping for a greener capitalism — even managed a fascist salute.

What a stark change from the 2016 election. Then Facebook’s (now Meta) Mark Zuckerberg felt the need to publicly affirm that he “care[s] deeply about the democratic process,” and that he wants to “make sure that Facebook is a force for good in democracy.” Now he’s going on Joe Rogan’s podcast, worrying, “I do think a lot of our society has become very, like, I don’t know, I don’t even know the right word for it, but it’s, like, kinda like neutered or, like, emasculated.” The corporate world, he goes on, needs more “masculine energy” — energy that is perhaps best exemplified by Facebook’s recent rule changes that now explicitly allow posts that dehumanize LGTBQ people. After Trump’s 2017 inauguration, the Washington Post, owned by Amazon’s Jeff Bezos, changed their slogan to “Democracy Dies in Darkness.” In this election cycle, Bezos personally blocked the newspaper’s editorial board from endorsing Kamala Harris.

There has long been a part of Silicon Valley that was, to put it mildly, fascism-adjacent. But until quite recently, most of Big Tech (Musk included) was aligned with the neoliberal centrists of the Democratic Party, a bond that was perhaps strongest during the Obama years. But today, Trump can proudly proclaim that “they did desert [Biden]. They were all with him, every one of them, and now they are all with me.”

There are many reasons for Big Tech to cuddle up to the far right, from hopes for more protectionism — whether it concerns Chinese competition like TikTok, or regulatory issues related to the European Union — to fears of antitrust regulation, or even Republican reprisals against businesses that don’t fall in line. But there is another important, and largely overlooked factor pushing Silicon Valley to the right: artificial intelligence.

To understand this shift, we need to understand class antagonism and the role that AI is playing in it. Arguably, the wealthy, the venture capitalists, the self-declared “entrepreneurs,” and the managers have always had a penchant for chauvinistic elitism. And the top of Silicon Valley was no different — there has always been a right-wing cesspool. As Becca Lewis has recently laid out in an article about the industry in the 1990s, tech was then already full of people who lamented the “pussification” of Silicon Valley, who wanted to fight against “political correctness,” and who were already accused of “technofascism.” The same thinking echoes through in Zuckerberg’s podcast appearances.

To the bosses’ misfortune, however, Big Tech depends on a workforce to make their profits. And for two essential reasons this workforce has, until this point, often been an effective counterweight to tech’s most reactionary tendencies.

First, tech workers tend to lean toward left and liberal political positions — just as the working class in general is, in the words of United Auto Workers president Shawn Fain, “the arsenal of democracy,” and just as organizations of the working class have generally been the major historic driver of democratization. At Alphabet/Google, for instance, workers have loudly raised their collective voice against everything from sexual harassment to Google’s involvement in the defense industry. Tech workers are certainly pushing back against their industry’s rightward shift even now.

Second, tech workers at large have generally been in advantageous bargaining positions: their work is highly skilled, in high demand, and in relatively low supply. Hence companies had to woo workers — and workers have demanded not just higher wages or office perks, but quite often they have demanded for their work to be meaningful, or at least for their workplaces not to be forces of evil in the world. In other words, they used their individual and collective bargaining power to countermand the elitist and antidemocratic tendencies of the capitalists in their industry.

When any commodity, including skills, is scarce and thus expensive, capitalists who depend on that commodity will try to increase its supply. In this case, the desire to expand the supply of skilled labor also pushed Big Tech toward an uneasy alliance with liberals. Under Barack Obama, for instance, Democrats heavily promoted expansion of STEM education. Big Tech and the government worked hand in hand to increase the supply of tech’s desired labor power, from funding coding boot camps, to offering special funds for low-income students, to creating organizations like Girls Who Code (a nonprofit funded by various tech companies and started under the White House Science & Technology Initiative). From the perspective of tech, these measures were simultaneously PR tactics that could be touted as increasing diversity and equity, and means to increase their labor supply.

For a while, it seemed like the workers’ demands for equity and progressive change could be squared with capital’s desire to increase its supply of capable workers. Back in the Obama years, Google’s corporate motto was still “Don’t Be Evil,” reflecting this sense that the drive for profit and progressive values could coexist. We can see remnants of the forces that made this coexistence possible even today. Far-right billionaires like Elon Musk who support calls for mass deportations still threaten to “go to war” with far-right allies like Steve Bannon over the issue of H-1B visas. These visas, generally reserved for individuals with special qualifications, increase the supply of the very same skilled labor that Silicon Valley depends on. Or in the words of Musk himself: “There is a permanent shortage of excellent engineering talent. It is the fundamental limiting factor in Silicon Valley.” Alas, the support for H-1B visas may just be the swan song of that older unstable and uneasy class compromise, this constraining force on Silicon Valley’s most reactionary tendencies, that is now fracturing.

We’ve all heard in ever so many ways that Silicon Valley is betting big on AI: Meta is planning to spend $65 billion dollars this year, Microsoft has $80 billion set aside for AI, and most recently, Trump announced that Big Tech players in the Stargate Initiative would spend half a trillion dollars on AI over the next five years. Regardless of the PR that is supposed to sell AI to the public, their bet is not on postscarcity worlds “so unbelievably good that it’s hard for me to even imagine” (OpenAI CEO Sam Altman), but on a massive push to undermine the bargaining position of tech workers. They are, in other words, hoping that AI will help them to massively deskill their own workforce and myriad workers beyond that.

In our book, Why We Fear AI, we describe AI as a kind of Techno-Taylorism. The sociologist Harry Braverman described Taylorism as the explicit verbalization of the capitalist mode of production and argued that its core tenet is to centralize as much knowledge as possible among management. The purpose is to control labor processes, and to divvy up work such that as much of it as possible can be performed by people without expensive training. In other words, management is about gathering knowledge in order to deskill. Today AI represents an attempt to achieve the same aims not by managerial means, but technological ones, by embedding knowledge and skills directly into AI models. The aim is much the same: to replace skilled and expensive labor with unskilled labor, in this case by reducing the number of skilled programmers, and supplanting them, as much as possible, with code-generating language models like Microsoft’s Copilot.

Certainly, there is a lot of hype in the sales pitches for these AI models and their alleged capabilities, and we should not simply believe what the most ardent AI-mongers tell us about the future. But even if these models will forever remain crappy alternatives to competent humans, they may still prove their worth in lowering wages. After all, we have seen cheap goods of lower quality crowd out the mid-level-quality market before. From Ikea furniture to clothing, we have gotten used to things that are not made to last in the way they used to, but which are so much cheaper than whatever normal quality existed in the past, that they manage to undermine the competition anyways. The same may very well happen to a lot of code.

Certainly, Big Tech seems dead set on making this bet on wage suppression through AI. Marc Andreessen, Silicon Valley venture capitalist, billionaire, and author of the fascist-quoting “Techno-Optimist Manifesto,” is already publicly speculating about “human wages crashing from AI.” Mark Zuckerberg is expecting 2025 to be the year in which AI’s code-generating abilities reach a level that allows them to “effectively be a sort of mid-level engineer” — in other words, a tool to crash the wages of the engineers and tech workers of today. This interview, in which he talks of “AI engineers instead of people engineers” is, perhaps not incidentally, the very same Joe Rogan interview in which he expressed his worries about his emasculation. What he really meant by that is now becoming clear: it’s his workers’ power that he resents, and that he is hoping now to undermine with AI. It’s the fact that hiring skilled workers required compromising on Silicon Valley’s most elitist and reactionary ambitions. It’s the power of workers to demand dignity for everyone that is experienced as emasculating by the ruling class. And in their bet on AI’s deskilling power, they are ready to break free from the class compromise that they clearly experienced not as peace but as a shackle.

None of this means that the rich will inevitably win or that their bet will pay off. But if we want to stop them, we need to be clear-eyed about what it is they’re doing. If tech workers act collectively, they have far more power than the venture capitalists do — they can shape what technologies we develop and how we use them. But they have to face the fact that this is no longer possible in a world of class compromise; they have to face the fact that the tech capitalists have thrown down the gauntlet.

From AI-fueled dreams about fully automated misery capitalism to fantasies of escaping to Mars or surviving climate change in bunkers, the ruling class is moving on from lies about trickle-down economics and “rising tides that lift all boats.” Instead, their politics are increasingly driven by a desire to make people superfluous, and even disposable. The billionaire investor Warren Buffett famously said, “There is class warfare, all right, but it’s my class that’s making war, and we’re winning.” As Big Tech and the ruling class is moving sharply to the right, that rings increasingly ominous. And yet the paucity and misery of their dreams show that they themselves have no plan for the future. Their anxious desire to make themselves independent from us is simultaneous proof that they do not even have the slightest thing to offer to the world. As they dream about making us superfluous, we should be emboldened, not frightened. As they strategize at the level of class, we should do the same. If they think they no longer need us, we should seriously ask if we need their class at all.

If we no longer get to choose whether we engage in class war or class compromise, then we only get to choose whether we fight back. We should pick up the gauntlet that they have thrown down — we have a world to win.