Tech Titans Vie for Trump's Favor in AI-Driven Power Play

Musk, DOGE, and the AI-fueled Plot to Fire Everybody

What is DOGE? Officially, it's a "Department of Government Efficiency," intended to find and eliminate government fraud and waste; officially, it's also a joke, named after an old meme. DOGE doesn't just emit mixed signals -- the incoherent messaging is right there in the name. It's part of the plan and, for supporters, part of the fun. When necessary, the government argues that what DOGE is doing is just common sense, that Musk is spearheading a "comprehensive forensic audit of every department and agency in the federal government," and that the administration has a "commitment to an efficient and accountable federal workforce." Nearly as often, though, the mask slips, or gets pulled off and thrown on the ground: This isn't true, not that we're engaged in any meaningful sort of collective debate over facts here: According to polling, a majority of Americans think that Musk has too much power in government and share a negative view of DOGE. It's not pure MAGA polarization, either -- the public views Musk less favorably than it views Trump. You may grant, if you insist, a genuine desire to find and eliminate inefficiencies in the federal government, but Musk and his supporters are barely trying to hide other animating impulses here: The immense contempt they have for federal workers; the immense contempt they feel for courts, or anyone else that might slow their efforts to fire said federal workers; the zealous energy they bring to cutting as many people as possible; and the willingness to make things up, or to refuse to correct falsehoods, in pursuit of a nakedly ideological project, and of power more generally. Musk, who is arguably the second most powerful person in government, and the most successful capitalist in America, officially has "no actual or formal authority to make government decisions himself," according to White House lawyers, but his constant (and constantly heeded) posts betray both his functional position and the minimum bound of his ambitions. Specifically, he wants to fire as many people as possible. Punishing workers is a cause and a purpose unto itself, inseparable from a grandiose conflation of personal desires and successes with the fate of humanity. It's an ecstatic project with an accelerationist character. "I am become meme," he declares, as his team of private sector loyalists harasses federal employees with spiteful emails threatening to get rid of them. The message from the largest employer in the country to its disfavored employees could not be much clearer: You are waste, you are fraud. We want to make a spectacle of your misfortune. We cannot wait to fire you. DOGE is certainly helping to achieve, to quote the words of Project 2025 author and new director of the Office of Management and Budget Russell Vought, the goal of putting federal workers "in trauma" and casting them as "the villains." But Musk is more than a willing mercenary for someone else's project, and his stature and influence within and outside the government imbues his slash-and-burn campaign with greater significance. Within the scope of DOGE, Musk is playing the role of government auditor and contractor, which is conflict enough. But he's also a singular industrialist with multiple companies devoted to automation, a history of executing brutal and punitive firings, and a proudly hostile stance toward organized labor and regulation. Most relevant, however, is his longtime obsession with what he sees as the mother of all technologies. He's an AI guy. This might explain why DOGE's flamboyant and unprecedented performance feels both familiar and unsettling beyond the government workforce. We're committing all of our time and resources to eliminating your job, and by the way, the fate of the world hangs in the balance. We're chasing inconceivable power, and you might end up as collateral damage. We're making a new world, and this is our only chance to get it right. We're trying to fire you. This is how MAGA Republicans talk about the political moment. It's also how America's premier industry sounds when when it's talking about AI. Most other tech executives aren't openly fantasizing about detaching employee "parasites" from their hosts, but some get closer than others. At Meta, Mark Zuckerberg recently announced layoffs explicitly targeting "low performers," which many at the company saw as a pointed and unnecessarily damaging way to cut costs, adding insult to injury and damaging prospects for former employees. The layoffs came in the context of a larger attempt at an ideological pivot for the company, marked by Mark Zuckerberg's interview with Joe Rogan (and following Musk's mass layoff trial run at Twitter, about which Zuckerberg, and other corporate leaders, had publicly expressed envy.) They also came in the context of the company's massive planned investment in AI, which Zuckerberg talked about on the show. "Probably in 2025, we at Meta, as well as the other companies that are basically working on this, are going to have an AI that can effectively be a sort of midlevel engineer that you have at your company that can write code," he told Rogan. (He softened the implication by suggesting that AI would simply free people up to work on more interesting problems, but the comments reverberated, as intended, through the tech industry.) More commonly, AI executives talk about work and labor in general. In a recent blog post, OpenAI founder Sam Altman made his case for mass automation. "[I]magine AI as a real-but-relatively-junior virtual coworker. Now imagine 1,000 of them. Or 1 million of them. Now imagine such agents in every field of knowledge work," he wrote. He suggested, with an optimistic gloss, that "huge" changes are coming. "We will find new things to do, new ways to be useful to each other, and new ways to compete," he wrote, "but they may not look very much like the jobs of today." In AI, we're getting some mixed messages, too -- the people working on this stuff are excited but worried. Founders and CEOs talk about glorious abundance with the public and tease concentrated returns to investors. They muse about automation and the future of work, raise alarms, fund alignment research, and talk about existential risk. Altman, for his part, has commissioned research into the plausibility and effects of UBI for a post-AI world. A bit like DOGE, however, OpenAI's conflicted identity is embodied in its name and concept. The company was as a non-profit with the mission to "advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return." By 2019, it was teaming up with Microsoft to raise tens of billions of dollars. Now, it's in the process of converting into a for-profit company. Just as DOGE is pursuing something more than simple increases in efficiency, AI firms are pitching something more than simple increases in productivity. The gap between the industry's coy, delicate rhetoric around automation in some contexts -- oh, maybe we'll need UBI, haha, err, well, it'll be a lot of change very fast but it'll work out maybe, ehh, maybe regulate us? -- and its far less conflicted pitch to investors is not lost on people whose jobs might be defined as "knowledge work," nor is its fervor as expressed in sheer dollar amounts. "Meta, Amazon, Alphabet, and Microsoft, intend to spend as much as $320 billion combined on AI technologies and datacenter buildouts in 2025," reports CNBC, a level of investment for which observers have struggled to find comparison (although here's one from The Information: "Big Tech's Capex Gusher Tops Last Oil Spree," aka the shale boom.) Many of the largest and most visible companies in the economy, in other words, are working toward the same goal, collectively marshaling a level of resources otherwise associated with national infrastructure projects or wars. You can characterize their goals and corresponding long-term outcomes in a wide range of ways, but if you live anywhere downstream from their project, the message, again, is pretty clear, albeit with a few more apologies and caveats appended as necessary, delivered with polite contrition rather than foaming zeal: We're sorry, but we have no choice but to do everything in our power to replace you with machines. New polling from Pew suggests that this message is being heard, and the Zuckerbergian appeals to the history of automation -- that maybe it'll create more, better jobs in the long term, after periods of transitional pain -- aren't resonating. According to a Pew survey of American workers, 52% are "worried about the future impact of AI use in the workplace," while 32% "think it will lead to fewer job opportunities for them in the long run." While 36% "feel hopeful" about how AI might be used at work, 33% are "overhwelmed" while just 6% believe it'll lead to more job opportunities in the long run. The parallels between DOGE's all-or-nothing bet that it can purge the federal government and the AI industry's promise to purge humans from wide swathes of the existing private workforce are united in the figure of Musk, a former partner in OpenAI who now runs his own AI firms; to him, the projects are clearly connected, synergistic, and mutually necessary. He's making this fairly explicit! DOGE is claiming, with few details, to use AI to further its missions, however you might interpret them: AI to replace government work; AI to arbitrarily fire people with a sheen of analysis; "AI" wielded as a broader demoralizing force, or conveniently summoned threat to workers. There are intentional and unintentional aspects of his leadership at DOGE that, if nothing else, clarify the beliefs he's enacting and seems to hold. Perhaps most of the government shouldn't exist, and someone like him should run what's left of it Perhaps most jobs shouldn't exist, or at least inevitably won't, so why wait around? Perhaps employees who wish to have jobs should be reminded to do as they're told. It can be hard to call a moment or define an era without the benefit of hindsight, and tipping points tend to reveal themselves in retrospect. As top-down attempts to impose and manifest a different world go, though, this is almost parodically blunt stuff. The president is a guy who rose to national fame firing people from fake jobs on a TV show. His publicly unhinged billionaire quasi-co-president is a man who appears to believe that most government jobs are definitionally fake, and who has been deputized to fire as many of them as possible. Like many of his peers in private industry, this same guy talks in urgent and increasingly dire terms about winning the race to build human-level and perhaps superhuman AI, a project into which the tech giants are sinking the hundreds of billions of dollars in cash they've harvested from their users over the last decade. Whatever you think of its chances of succeeding, and whatever you think would come out of the other side in the long term, their fundraising resembles a bet on, at minimum, short-to-medium-term mass displacement in segments of the economy that were until recently understood as relatively safe. In fuller expression, it's not just a post-worker vision, it verges on posthuman. To anyone who isn't invested in these outcomes, either literally or ideologically, it's an unsettling wager. In Trump, Musk, and Silicon Valley's speculative CEOs -- whose musings and strategies are certainly being heard and nervously interpreted by leaders across various potentially disruptable industries -- America is living under a nightmare boss trifecta, each potentially familiar to different workers in their own way. There's one who wants to fire you for disloyalty, the one who wants to fire you because he thinks you're undeserving and less-than, and the one who grimaces through the PIPs and the RIFs but lights up during the earnings calls. The news can, at times, resemble a zoom call from hell, with Americans on mute as management talks through plans for getting rid of them. It's layoff morning in America. These parallels also extend into an unknown future. It's increasingly clear that Musk -- and perhaps Trump -- have ambitions for DOGE that extend beyond cuts and into meta-governmental control. It represents, in other words, an all-or-nothing bet on remaking, destroying, or perhaps controlling the government, with massive potential upside for those involved and correspondingly large risk if they fail. Big tech's push into AI is similarly high-stakes, with firms both excited about possible returns but also fearful of getting stuck on the wrong side of a massive reallocation of power. A superpower freed of its opposition and a superintelligence freed of its labor force are versions of the same fantasy in which absolute power is seized, or obtained, through different means. These expressions of absolute self-belief shouldn't be confused with inevitability, of course, nor, should they come up short of counterrevolution and post-human superintelligence, respectively, will their almost unbelievable nakedness be forgotten. (Perhaps the most pressing question about both is whether they're accelerating toward a lasting new status quo, ramping like a manic episode, or a little bit of both. Either they have the potential to be widely galvanizing.) Given overlapping once-in-a-generation opportunities, America's most empowered and resentful bosses have converged on the same simple plan: laying everyone off.

Techno-Fascism Comes to America

When a phalanx of the top Silicon Valley executives -- Mark Zuckerberg, Jeff Bezos, Elon Musk, and Google's Sundar Pichai -- aligned behind President Trump during the Inauguration in January, many observers saw an allegiance based on corporate interests. The ultra-wealthy C.E.O.s were turning out to support a fellow-magnate, hoping perhaps for an era of deregulation, tax breaks, and anti-"woke" cultural shifts. The historian Janis Mimura saw something more ominous: a new, proactive union of industry and governmental power, wherein the state would drive aggressive industrial policy at the expense of liberal norms. In the second Trump Administration, a class of Silicon Valley leaders was insinuating itself into politics in a way that recalled one of Mimura's primary subjects of study: the élite bureaucrats who seized political power and drove Japan into the Second World War. "These are experts with a technological mind-set and background, often engineers, who now have a special role in the government," Mimura told me. The result is what, in her book "Planning for Empire" (2011), she labelled "techno-fascism": authoritarianism driven by technocrats. Technology "is considered the driving force" of such a regime, Mimura said. "There's a sort of technicization of all aspects of government and society." In the nineteen-thirties, Japan colonized Manchuria, in northeastern China, and the region became a test ground for techno-fascism. Nobusuke Kishi, a Japanese commerce-ministry bureaucrat, was appointed to head the industrial program in Manchuria, in 1936, and, with the collaboration of a new crop of the Japanese conglomerates known as zaibatsu, he instituted a policy of forced industrial development based on the exploitation of the local population. When Kishi returned to national politics in Japan, in 1939, along with a clique of other Japanese technocrats who had worked in Manchuria, he pursued similar strategies of state-dictated industrialization, at the expense of private interests and labor rights. This fascistic regime would not be structured the same way as Mussolini's or Hitler's, with power concentrated in the hands of a single charismatic leader, although Kishi had travelled to Germany in the nineteen-twenties, as the Nazi movement expanded, and drew inspiration from German industrialization for his Manchurian project. Instead, Mimura said, Japan "kind of slid into fascism" as bureaucrats exercised their authority behind the scenes, under the aegis of the Japanese emperor. As she explained, techno-fascist officials "acquire power by creating these supra-ministerial organs and agencies, subgroups within the bureaucracy that are unaccountable." Today, Elon Musk's DOGE is the Trumpian equivalent. American corporations of the twentieth century flirted with a merging of state and industrial power. The entrepreneur Henry Ford promoted a system of industrial organization that came to be known as "Fordism," whereby the state would intervene in the economy to guarantee mass production and consumption. In the nineteen-thirties, I.B.M. did business with the Nazi government through a German subsidiary, lending its technology to projects like the 1933 census, which helped identify Jews in the country. As a recent feature in the Guardian by Becca Lewis laid out, Silicon Valley itself has exhibited right-wing tendencies for decades, embracing misogynist and hierarchical attitudes about achievement. The journalist Michael S. Malone was issuing warnings about emerging "technofascism" way back in the late nineties, when he warned about "IQ bigotry" in the tech industry and the willingness of people to push forward digital revolution while "tossing out the weak and wounded along the way." But our current moment marks a new conjunction of Internet entrepreneurs and day-to-day government operations. American techno-fascism is no longer a philosophical abstraction for Silicon Valley to tinker with, in the vein of intermittent fasting or therapeutic ketamine doses. It is a policy program whose constitutional limits are being tested right now as DOGE, staffed with inexperienced engineers linked to Musk's own companies, rampages through the federal government. Musk has slashed the ranks of federal employees, shut down agencies whose authority challenges his own, and leveraged artificial intelligence to decide where to cut, promising a government executed by chatbots such as Grok, from Musk's own A.I. company. DOGE has gained access to Americans' private data and developed tools to e-mail the entire federal government at once, a digital megaphone that Musk recently used to demand that employees send in a list of their weekly accomplishments. As Mimura put it, "You try to apply technical concepts and rationality to human beings and human society, and then you're getting into something almost totalitarian." The techno-fascist opportunism goes beyond Musk; one can sense other tech entrepreneurs and investors slavering to exploit the alliance between Trumpism and Silicon Valley capitalism, building infrastructure on a national scale. Sam Altman, the C.E.O. of OpenAI, has arranged his own deals with Trump's government, including Stargate, a heavily hyped data-center project worth a potential five hundred billion dollars. Apple recently announced its own five-hundred-billion-dollar investment campaign in the U.S. over the next four years, including a plan to begin building A.I. servers in Texas. However nebulous, these extravagant plans signal a spirit of collaboration. On Truth Social, Trump posted approvingly that Apple's plans demonstrated "FAITH IN WHAT WE ARE DOING." Erin McElroy, a geographer at the University of Washington who studies Silicon Valley, has used the term "siliconization" to describe the way that places such as San Francisco or Cluj-Napoca, Romania, to which many western tech companies have outsourced I.T. services, have been remade in the image and ideology of Silicon Valley. According to McElroy, the first signs of Washington's current siliconization can be traced back, in part, to the Administration of Barack Obama, who embraced social-media platforms such as Facebook as a vector of government communication. For a time, digital platforms seemed to support democratic government as a kind of communal megaphone; but now, a decade later, technology seems to be supplanting the established authority of the government. "There is a crisis of the state," McElroy said, and Silicon Valley may be "trying to corrode state power" in order to more quickly replace it. Silicon Valley is premised on the idea that its founders and engineers know better than anyone else: they can do better at disseminating information, at designing an office, at developing satellites and advancing space travel. By the same logic, they must be able to govern better than politicians and federal employees. Voguish concepts in Silicon Valley such as seasteading and "network states" feature independent, self-contained societies running on tech principles. Efforts to create such entities have either failed or remained confined to the realm of brand-building, as in the startup Praxis, a hypothetical plan for a new tech-driven city on the Mediterranean. Under the new Trump White House, though, the U.S. government is being offered up as a guinea pig, McElroy said. "Now that we've got Musk running the state, I don't know if they need their little offshore bubbles as much as they thought they did before." Such visions of a technologized society represent a break from the Make America Great Again populism that drove the first Trump Administration. MAGA reactionaries such as Steve Bannon tend to be skeptical of technological progress; as the journalist James Pogue has explained, their goal is to reclaim an American culture "thought to be lost after decades of what they see as globalist technocracy." Bannon has denounced Silicon Valley's ideology as "technofeudalism" and declared war on Musk. He sees it as antihuman, with U.S. citizens turned into "digital serfs" whose freedom is delimited by tech companies. In a January interview with Ross Douthat, of the Times, Bannon said, "They have to be stopped. If we don't stop it, and we don't stop it now, it's going to destroy not just this country, it's going to destroy the world." Whereas the MAGA right wants to restore things as they were (or as they imagine things were), the tech right wants to, in Mark Zuckerberg's phrasing, break things. In the Times interview, Bannon called Musk "one of the top accelerationists," referencing another technology-inflected political ideology that treats chaos as an inevitability. Accelerationism has been popularized in the past decade by the British philosopher Nick Land, who is part of the so-called neo-reactionary or Dark Enlightenment movement populated by figures including Curtis Yarvin, a former programmer and blogger whose proposals for an American monarchy have enjoyed renewed relevance during Trump 2.0. The accelerationist attitude is, as Andrea Molle, a professor of political science at Chapman University who studies accelerationism, put it to me, "This collapse is going to come anyway -- let's rip the Band-Aid." Accelerationism emerged from Karl Marx's idea that, if the contradictions of capitalism become exaggerated enough, they will inspire proletarian revolution and a more egalitarian society will emerge. But Molle identifies what he calls Muskian "techno-accelerationism" as having a different end: destroying the existing order to create a technologized, hierarchical one with engineers at the top. Musk "has to completely break any kind of preëxisting government architecture to impose his own," Molle said. He added that a government thoroughly overhauled by Musk might run a bit like the wireless system that operates Teslas, enabling the company to theoretically update how your car works at any moment: "You're allowed some agency, but they are still in control, and they can still intervene if the course is not going in the direction that it is supposed to go to maximize efficiency."

How Sam Altman Could Break Up Elon Musk and Donald Trump

Two of the most powerful tech executives in the world are desperate for the president's approval. The rivalry between Sam Altman and Elon Musk is entering its Apprentice era. Both men have the ambition to redefine how the modern world works -- and both are jockeying for President Donald Trump's blessing to accelerate their plans. Altman's company, OpenAI, as well as Musk's ventures -- which include SpaceX, Tesla, and xAI -- all depend to some degree on federal dollars, permits, and regulatory support. The president could influence whether OpenAI or xAI produces the next major AI breakthrough, whether Musk can succeed in sending a human to Mars, and whether Altman's big bet on nuclear energy, and fusion reactors in particular, pans out. Understanding the competition between these two men helps illuminate Trump's particular style of governing -- one defined by patronage and dealmaking. And the rivalry highlights the tech giants' broader capitulation to the new administration. Executives who have sold a vision of the future defined by ultra-intelligent computer programs, interplanetary travel, and boundless clean energy have bowed to a commander in chief who has already stifled free expression, scientific research, and the mere mention of climate change in government work. Why? Simply because doing so will advance their interests. (And, in some cases, because tech leaders are true believers -- ideological adherents to the MAGA worldview.) Altman's MAGA turn is best understood as a search for a lifeline. In 2017, as Trump's first term was just beginning, Altman tweeted, "I think Trump is terrible and few things would make me happier than him not being president." This time around: "I think he will be incredible for the country in many ways!" In the months before the election, Altman and OpenAI leaned on connections to Trump allies to curry favor, according to The New York Times. In June, two of the start-up's executives met with Trump in Las Vegas, showcasing their technology and emphasizing its land and energy needs. Meanwhile, OpenAI's technological lead over xAI, Google, Anthropic, and other firms has dwindled. The company's relationship with its main financial backer, Microsoft, has also frayed so much that OpenAI is actively courting other corporate partners. (Microsoft, despite approving OpenAI's ability to find other data-center partners, maintains that it will remain a key partner going forward.) Over the past year, a number of senior researchers have departed, and the start-up faces several lawsuits and investigations. A new and friendly administration, then, could provide Altman with a much-needed boost to maintain his firm's shrinking edge in the AI race. (The Atlantic recently entered into a corporate partnership with OpenAI.) And Musk, for all his criticism of federal bloat, is plenty dependent on the government. Over the past decade, his companies have been awarded at least $18 billion in federal contracts. SpaceX relies heavily on NASA for its rocket business and as of Monday is reportedly testing its Starlink technology to improve the Federal Aviation Administration's national airspace system, despite an existing $2 billion contract that the FAA has with Verizon. Tesla, with shrinking sales and a relatively stagnant lineup of models, could benefit mightily from friendly regulation of self-driving cars. Musk also appears jealous of Altman's it-boy reputation in Silicon Valley and beyond: He started xAI within months of ChatGPT's launch, has taken to calling his rival "Scam Altman," and recently led an unsolicited $97.4 billion bid for OpenAI (which the start-up's board refused). "Probably his whole life is from a position of insecurity," Altman told Bloomberg Television the next day. "I feel for the guy." Anything that OpenAI might gain from Trump, xAI could reap as well. Altman and Musk both hope to build data centers that use a tremendous amount of electricity -- each one potentially requiring as much as would be provided by a large nuclear reactor or even several, a demand equal to millions of American homes. The government can open federal lands to data-center and power-plant construction, and it can expedite the construction of natural-gas or nuclear plants (or the now-less-likely renewable-energy sources). Trump could attempt to cut down the sometimes interminable permitting process for the transmission lines that carry that electricity to data centers. He might intervene in or make it difficult to enforce the outcomes of AI copyright litigation, and generally make the regulatory environment as friendly as possible for the industry and its investors. Read: For now, there's only one good way to power AI Musk, of course, has cemented his place in the president's inner circle, acting as a Trump surrogate during the campaign and now leading his efforts to remake the civil service. He has fused his political ideology -- reactionary, authoritarian, nativist -- with Trump's. But Altman, too, has quietly gained the president's confidence, albeit with a much narrower appeal to American-AI dominance. His company has ramped up its public messaging and lobbying about the importance of America's AI leadership over China -- a goal that Trump has repeatedly emphasized as a priority. The maneuvering is already starting to pay off. The day after the inauguration, Altman stood beside Trump in the White House as the president announced Stargate, a new company planning to spend $500 billion on AI infrastructure, and in which OpenAI is a principal investor. According to the Times, Altman had struggled to raise money for Stargate for months -- potential investors worried that government approval for the necessary, extensive construction would be too slow -- until Trump's victory, when sentiment flipped. During the press conference, Trump said the government's job would be "to make it as easy as it can be" to build. Altman was sure to signal gratitude, saying that "with a different president, [Stargate] might not have been possible." Within hours of the announcement, Musk, not to be excluded or outdone, chimed in on X. "They don't actually have the money," he wrote, suggesting that Stargate's main investors could not fund the project. Altman denied this, writing on X, "I realize what is great for the country isn't always what's optimal for your companies, but in your new role i hope you'll mostly put [America] first." Read: OpenAI goes MAGA For Musk to break ranks with his newfound presidential ally suggests that the world's richest man is still focused on an old grudge and affront to his ego. After being one of OpenAI's initial investors, Musk left its board in 2018, at the time citing potential conflicts of interest with future AI projects at Tesla. Four years later, when OpenAI released ChatGPT and kicked off the generative-AI boom, Musk was caught off guard -- not just behind in the race, but not even an entrant. Within weeks, he was suggesting that the chatbot was too "woke." Soon after, Musk formed his own AI start-up, xAI, and last year, he sued OpenAI for betraying its original nonprofit mission. In response to the lawsuit, OpenAI released old emails from Musk suggesting he had departed because he thought that without merging with Tesla or otherwise securing substantially more funding, OpenAI's chance of "being relevant" was "0%. Not 1%. I wish it were otherwise." (Oops.) Ever since he left, Musk has been playing catch-up. The first and second iterations of xAI's model, Grok, lagged behind the most powerful versions of ChatGPT. Musk's latest, Grok 3, appears to be in the same ballpark as OpenAI's new, state-of-the-art "reasoning" models -- but xAI accomplished this months later and likely with far more computing resources. Despite, or perhaps because of, repeatedly coming up short, Musk has evinced a willingness to use any tactic to maintain his own relevance, or at least slow down his competitors. In late March 2023, Musk signed a widely circulated letter calling for at least six months' pause on training AI models more powerful than OpenAI's then-just-released GPT-4 -- even though he had incorporated xAI weeks earlier and was actively recruiting staff. Musk's lawsuit denounces OpenAI as profit-hungry and secretive, and he has dubbed the start-up "ClosedAI," but the code and training data underlying Grok 3 are as opaque as that of ChatGPT. And despite Musk's claims that Grok 3 is the "smartest AI on Earth," OpenAI researchers have accused his start-up of misrepresenting the chatbot's performance to make it appear on par with their own top model, o3-mini (a sort of manipulation common in the generative-AI industry, and one that OpenAI itself has been accused of as well). Still, he is now closer than ever to catching Altman, and his position at the helm of the Department of Government Efficiency and his perch on Trump's shoulder could push him over the edge. The president, for now, seems content to keep both relationships open; certainly, his association with two tech executives considered visionaries has its perks. Altman, with his tunnel vision on AI, seems unlikely to affect or sour Trump and Musk's ideological bond and attempt to reshape the federal government. Perhaps the greater risk to xAI is that Musk overstays his MAGA welcome and attracts the president's ire. And Musk does not appear to have turned Trump against OpenAI. When asked about Musk's criticism of Stargate, the president shrugged it off: "He hates one of the people," Trump told reporters. "But I have certain hatreds of people, too."