One week ago, a new and formidable challenger for OpenAI’s throne emerged. A Chinese AI start-up, DeepSeek, launched a model that appeared to match the most powerful version of ChatGPT—but, at least according to its creator, was a fraction of the cost to build. The program, called DeepSeek-R1, has incited plenty of concern: Ultrapowerful Chinese AI models are exactly what many leaders of American AI companies feared when they, and more recently President Donald Trump, have sounded alarms about a technological race between the United States and the People’s Republic of China. This is a “wake up call for America,” Alexandr Wang, the CEO of Scale AI, commented on social media.

But at the same time, many Americans—including much of the tech industry—appear to be lauding this Chinese AI. As of this morning, DeepSeek had overtaken ChatGPT as the top free application on Apple’s mobile-app store in the U.S. Researchers, executives, and investors have been heaping on praise. The new DeepSeek model “is one of the most amazing and impressive breakthroughs I’ve ever seen,” the venture capitalist Marc Andreessen, an outspoken supporter of Trump, wrote on X. The program shows “the power of open research,” Yann LeCun, Meta’s chief AI scientist, wrote online.

Indeed, the most notable feature of DeepSeek may be not that it is Chinese, but that it is relatively open. Unlike top American AI labs—OpenAI, Anthropic, and Google DeepMind—which keep their research almost entirely under wraps, DeepSeek has made the program’s final code, as well as an in-depth technical explanation of the program, free to view, download, and modify. In other words, anybody from any country, including the U.S., can use, adapt, and even improve upon the program. That openness makes DeepSeek a boon for American start-ups and researchers—and an even bigger threat to the top U.S. companies, as well as the government’s national-security interests.

To understand what’s so impressive about DeepSeek, one has to look back to December, when OpenAI launched its own technical breakthrough: the full release of o1, a new kind of AI model that, unlike all the “GPT”-style programs before it, appears able to “reason” through challenging problems. o1 displayed leaps in performance on some of the most challenging math, coding, and other tests available, and sent the rest of the AI industry scrambling to replicate the new reasoning model—which OpenAI disclosed very few technical details about. The start-up, and thus the American AI industry, were on top. (The Atlantic recently entered into a corporate partnership with OpenAI.)

DeepSeek, less than two months later, not only exhibits those same “reasoning” capabilities apparently at much lower costs, but has spilled at least one way to match OpenAI’s more covert methods to the rest of the world. The program is not entirely open-source—its training data, for instance, and the fine details of its creation are not public—but, unlike with ChatGPT, Claude, or Gemini, researchers and start-ups can still study the DeepSearch research paper and directly work with its code. OpenAI has enormous amounts of capital, computer chips, and other resources, and has been working on AI for a decade. In comparison, DeepSeek is a smaller team formed two years ago with far less access to essential AI hardware, because of U.S. export controls on advanced AI chips, but it has relied on various software and efficiency improvements to catch up. DeepSeek has reported that the final training run of a previous iteration of the model that R1 is built from, released in December, cost less than $6 million. Meanwhile, Dario Amodei, the CEO of Anthropic, has said that U.S. companies are already spending on the order of $1 billion to train future models. Exactly how much the latest DeepSeek cost to build is uncertain—some researchers and executives, including Wang, have cast doubt on just how cheap it could have been—but the price for software developers to incorporate DeepSeek-R1 into their own products is roughly 95 percent cheaper than incorporating OpenAI’s o1, as measured by the price of every “token”—basically, every word—the model generates.

DeepSeek’s success has abruptly forced a wedge between Americans most directly invested in outcompeting China and those who benefit from any access to the best, most reliable AI models. (It’s a divide that echoes Americans’ attitudes about TikTok—China hawks versus content creators—and China’s other apps and platforms.) For the start-up and research community, DeepSeek is an enormous win. “A non-US company is keeping the original mission of OpenAI alive,” Jim Fan, a top AI researcher at the chipmaker Nvidia and former OpenAI employee, wrote on X. “Truly open, frontier research that empowers all.”

But for America’s top AI companies, and the nation’s government, what DeepSeek represents is unclear. The stocks of many major tech firms—including Nvidia, Alphabet, and Microsoft—dropped this morning amid the excitement around the Chinese model. And Meta, which has branded itself as a champion of open-source models in contrast to OpenAI, now seems a step behind. (The company is reportedly panicking.) To some investors, all those massive data centers, billions of dollars of investment, or even the half-a-trillion-dollar AI-infrastructure joint venture from OpenAI, Oracle, and SoftBank, which Trump recently announced from the White House, could seem far less essential. Maybe bigger AI isn’t better. For those who fear that AI will strengthen “the Chinese Communist Party’s global influence,” as OpenAI wrote in a recent lobbying document, this is legitimately concerning: The DeepSeek app refuses to answer questions about, for instance, the Tiananmen Square protests and massacre of 1989 (although the censorship may be relatively easy to circumvent).

None of that is to say the AI boom is over, or will take a radically different form going forward. The next iteration of OpenAI’s reasoning models, o3, appears far more powerful than o1 and will soon be available to the public. There are some signs that DeepSeek trained on ChatGPT outputs (outputting “I’m ChatGPT” when asked what model it is), although perhaps not intentionally—if that’s the case, it’s possible that DeepSeek could only get a head start thanks to other high-quality chatbots. America’s AI innovation is accelerating, and its major forms are beginning to take on a technical research focus other than reasoning: “agents,” or AI systems that can use computers on behalf of humans. American tech giants could, in the end, even benefit. Satya Nadella, the CEO of Microsoft, framed DeepSeek as a win: More efficient AI means that use of AI across the board will “skyrocket, turning it into a commodity we just can’t get enough of,” he wrote on X today—which, if true, would help Microsoft’s profits as well.

Still, the pressure is on OpenAI, Google, and their competitors to maintain their edge. With the release of DeepSeek, the nature of any U.S.-China AI “arms race” has shifted. Preventing AI computer chips and code from spreading to China evidently has not tamped the ability of researchers and companies located there to innovate. And the relatively transparent, publicly available version of DeepSeek, rather than leading American programs, could mean Chinese programs and approaches become global technological standards for AI—akin to how the open-source Linux operating system is now standard for major web servers and supercomputers. Being democratic—in the sense of vesting power in software developers and users—is precisely what has made DeepSeek a success. If Chinese AI maintains its transparency and accessibility, despite emerging from an authoritarian regime whose citizens can’t even freely use the web, it is moving in exactly the opposite direction of where America’s tech industry is heading.

This is Atlantic Intelligence, a newsletter in which our writers help you wrap your mind around artificial intelligence and a new machine age. Sign up here.

Earlier this week, Donald Trump unveiled perhaps the most ambitious infrastructure project in history—one that may rival the costs of the first moon missions—and all but dedicated it to Sam Altman. The project, known as Stargate, is a joint venture between OpenAI, Oracle, SoftBank, and several other corporate partners that aims to invest $500 billion over the next four years in America’s AI infrastructure: data centers, energy plants, power lines, and everything else needed to develop superintelligent computer programs. The first data center, already under construction, will soon be dedicated to training OpenAI’s next models.

The Stargate Project is a resounding victory for a start-up that was struggling at the end of last year, as Karen Hao wrote for The Atlantic yesterday. OpenAI had lost some of its most talented staff; its relationship with its most important financial backer, Microsoft, was under stress; and it was weathering any number of other public controversies. This week’s announcement, meanwhile, “reduces OpenAI’s dependence on Microsoft, grants OpenAI (rather than its competitors) a mind-boggling sum of capital for computer chips—the hottest commodity in the AI race—and ties the company to Trump’s ‘America First’ agenda,” Hao wrote.

The announcement is the capstone to a steady maneuver by Altman to align himself with the incoming administration, another “masterful display of Altman’s power” to ingratiate himself with the powerful and raise huge amounts of capital, Hao noted. Altman, along with executives from Oracle and SoftBank, stood beside Trump in the White House as he made the announcement. “I’m thrilled we get to do this in the United States,” Altman said.

Illustration by The Atlantic. Sources: Getty.

OpenAI Goes MAGA

By Karen Hao

Things were not looking great for OpenAI at the end of last year. The company had been struggling with major delays on its long-awaited GPT-5 and hemorrhaging key talent—notably, Chief Scientist Ilya Sutskever, Chief Technology Officer Mira Murati, and Alec Radford, the researcher who’d set the company on the path of developing GPTs in the first place. Several people who left either joined OpenAI competitors or launched new ones. The start-up’s relationship with Microsoft, its biggest backer and a crucial provider of the computing infrastructure needed to train and deploy its AI models, was being investigated by the Federal Trade Commission.

And then there was Elon Musk. He’d co-founded OpenAI with Sam Altman and others, but the two had become fierce rivals. As “first buddy” to Donald Trump, Musk was suing OpenAI while rapidly building up his own AI venture, xAI, whose chatbot, Grok, has become a central feature on X. Amid all of this drama, Altman was notified by his sister, Annie, that she intended to sue him; she alleges that he sexually abused her when she was a child. (That lawsuit was filed at the start of this month; Altman and members of his family strongly denied the allegations through a statement posted on X.)

Read the full article.

What to Read Next

Sam Altman doesn’t actually need Trump: As I noted on Wednesday, the Stargate Project felt more like a display of weakness from Trump. These companies could have gone elsewhere; AI’s rapid development would have continued with or without Stargate, and under Trump or a President Kamala Harris. “Only a day into his presidency, Stargate showed Trump taking cues from China, Microsoft, OpenAI, and Biden all at once—from a foreign adversary, the tech giants he vilified in 2020, and a political rival he has ruthlessly vilified,” I wrote. OpenAI takes off its mask: “For the first time, OpenAI’s public structure and leadership are simply honest reflections of what the company has been—in effect, the will of a single person,” Hao wrote last fall.

P.S.

Of course, Altman wasn’t the only one cozying up to Trump this week. At his inauguration, tech titans whose tools collectively touch billions of lives—including Mark Zuckerberg, Jeff Bezos, Tim Cook, and Elon Musk—stood right beside Trump’s family. “The tech industry has officially placed itself in the palm of Trump’s hand,” Atlantic senior editor Damon Beres wrote on Monday.

— Matteo

Elon Musk is already casting doubt on OpenAI’s new, up to $500 billion investment deal with SoftBank (SFTBY) and Oracle (ORCL), despite backing from his allies — including President Donald Trump.

Read more...

Things were not looking great for OpenAI at the end of last year. The company had been struggling with major delays on its long-awaited GPT-5 and hemorrhaging key talent—notably, Chief Scientist Ilya Sutskever, Chief Technology Officer Mira Murati, and Alec Radford, the researcher who’d set the company on the path of developing GPTs in the first place. Several people who left either joined OpenAI competitors or launched new ones. The start-up’s relationship with Microsoft, its biggest backer and a crucial provider of the computing infrastructure needed to train and deploy its AI models, was being investigated by the Federal Trade Commission.

And then there was Elon Musk. He’d co-founded OpenAI with Sam Altman and others, but the two had become fierce rivals. As “first buddy” to Donald Trump, Musk was suing OpenAI while rapidly building up his own AI venture, xAI, whose chatbot, Grok, has become a central feature on X. Amid all of this drama, Altman was notified by his sister, Annie, that she intended to sue him; she alleges that he sexually abused her when she was a child. (That lawsuit was filed at the start of this month; Altman and members of his family strongly denied the allegations through a statement posted on X.)

It’s remarkable, then, that with its latest maneuver, OpenAI has once again reestablished its dominance. On Tuesday, President Donald Trump announced the Stargate Project, a joint venture between SoftBank, Oracle, and OpenAI to pump $500 billion of private-sector investment over four years into building out U.S. AI infrastructure, with the intent of securing America’s leadership in AI development against China. Very little is known about how any of this will work in practice, but OpenAI is speaking as though it will reap most of the rewards: In its blog post announcing the partnership, it said that all of the infrastructure will be “for OpenAI.” The company’s president, Greg Brockman, underscored the point on X: “$500B for AI data centers for OpenAI.”

In one fell swoop, the project reduces OpenAI’s dependence on Microsoft, grants OpenAI (rather than its competitors) a mind-boggling sum of capital for computer chips—the hottest commodity in the AI race—and ties the company to Trump’s “America First” agenda, providing the best possible protective shield against Musk. (Musk blasted the project yesterday, alleging that it doesn’t “actually have the money,” which Altman then denied.) OpenAI (which entered into a corporate partnership with The Atlantic last year) did not respond to a request for comment.

It’s unclear whether Stargate will even be able to spend $500 billion in four years. But consider just how astounding that goal is. In late 2023, as Microsoft started spending roughly $50 billion a year on expanding cloud-computing capacity, one semiconductor analyst had already declared that that was “the largest infrastructure buildout that humanity has ever seen.” Rene Haas, the CEO of the semiconductor company Arm Holdings, said that even this pace of expansion across the industry would put global computing on track to consume more energy than India by 2030.

[Read: Microsoft’s hypocrisy on AI]

The move is a masterful display of Altman’s power at work. Altman has shown an uncanny ability throughout his career to get himself out of the toughest binds by leaning on his influential network, ingratiating himself with the powerful, and fundraising extraordinary amounts of capital. It was for these reasons that Altman successfully orchestrated his return to OpenAI as CEO in late 2023, after the board briefly ousted him. And it is why so many people have expressed alarm about his leadership in recent years. This week, he was at it again, standing next to Trump during the Stargate announcement in a symbol of solidarity and praising him later on X: “watching @potus more carefully recently has really changed my perspective on him … i’m not going to agree with him on everything, but i think he will be incredible for the country in many ways!”

Although OpenAI has led the pack, many AI companies have worked over the past two years to influence policy and grow without government interference. Silicon Valley has always operated like this, and many other major tech CEOs took their place next to Trump this week. But the demands of generative AI are meaningfully different from, say, those of a traditional search engine or a social-media platform: Its development requires far more crucial physical infrastructure. Generative-AI models are of a size that necessitate the build-out of data centers at unprecedented scale. This, in turn, will give Silicon Valley outsize influence over the placement of power plants and even water lines across the country. Already, the past few years of dramatic data-center expansion have affected power reliability for millions of Americans and threatened to raise the cost of drinking water.

[Read: Billions of people in the palm of Trump’s hand]

The tech industry expertly laid the groundwork for this outcome: It made big promises about the wondrous potential of its technologies while creating a sense of peril by evoking China’s own technological advancement. During the Stargate announcement, Trump said that he would do what he could to strip away any regulatory barriers. “China is a competitor, and others are competitors,” he said. “I’m going to help a lot through emergency declarations, because we have an emergency. We have to get this stuff built.”

Standing at the same podium, Altman emphasized America’s leadership. “I’m thrilled we get to do this in the United States of America,” he said. And then, in recognition of his new benefactor: “We wouldn’t be able to do this without you, Mr. President.”

Late yesterday afternoon, the president of the United States transformed, very briefly, into the comms guy for a new tech company. At a press conference capping his first full day back in the White House, Donald Trump stood beside three of the most influential executives in the world—Sam Altman of OpenAI, Larry Ellison of Oracle, and Masayoshi Son of SoftBank—and announced the Stargate Project, “the largest AI infrastructure project, by far, in history.”

Although Trump’s rhetoric may seem to suggest otherwise, Stargate is not a new federal program but rather a private venture uniting these three companies with other leaders in the AI race, such as Microsoft and Nvidia. The new company—for which Son will serve as chairman and OpenAI will be in charge of operations—will spend a planned $500 billion over the next four years to build data centers, power plants, and other such digital infrastructure in the United States, all in hopes of developing ever more advanced AI models. Trump presented Stargate as a victory for his “America First” agenda, saying that it may “lead to something that could be the biggest of all”—an apparent reference to superintelligent machines. The executives concurred, speaking of AI’s potential to generate cures for cancer and heart disease. “It’s all taking place right here in America,” Trump said.

Although the project will likely produce many jobs and generate some value for the companies involved, it is hard to ignore the feeling that Trump needs this more than any of the men he was standing beside. “It’s an honor that they want to come to our country” for their AI-infrastructure build-out, Trump said of these “three great people, great CEOs, and great geniuses.” Over the course of roughly 45 minutes, he said seven separate times that it was an honor to host them, adding, “For Larry to be here and do this is very unusual, because he doesn’t do this stuff; he doesn’t need it.”

He may be correct, and not just about Ellison. Altman has reportedly proposed similarly massive AI-infrastructure projects to investors in the Middle East and computer-chip makers in Asia. Just this week, Jensen Huang, the CEO of the computer-chip giant Nvidia, visited China—America’s biggest geopolitical foe—apparently thanking local staff and lauding his company’s contributions to “one of the greatest markets, the greatest countries in the world.” SoftBank is a Japanese corporation. Oracle has substantial investments and AI infrastructure in the Middle East. A United Arab Emirates firm, MGX, is Stargate’s fourth initial financial backer, and the British chip manufacturer Arm is a technical partner alongside Nvidia. In other words, AI development is proceeding within, but also outside of, the U.S., Stargate or not. (The Atlantic recently entered into a corporate partnership with OpenAI.)

As such, the project may be less a vote of confidence in Trump’s vision for America so much as the latest sign of the country’s capitulation to the AI industry, which has repeatedly pushed for lenient regulations and invoked the specter of China to clear a path for rapid development. (Although, to be clear, tech giants have done plenty of capitulating to Trump too.) Trump emphasized that his role is to welcome these companies and get out of the way: “We’re going to make it as easy as it can be,” he said. He also referenced China more than once. “China is a competitor; others are competitors. We want [AI] to be in this country,” he said, later adding, “This is money that normally would have gone to China.”

[Read: A virtual cell is a ‘holy grail’ of science. It’s getting closer.]

AI may well change the world, but the announcement provided little in terms of specifics of how it would get there. Despite promises of AI-enabled cancer vaccines and personalized medicine, exactly how the technology will revolutionize the military, biology, or any other industry is unclear, and the path to “superintelligence” is hazier still. Even if generative AI yields productivity gains and speeds up medical research, there will be trade-offs: The technology and its infrastructure are as likely to displace millions of jobs, require massive natural-gas and nuclear power plants to meet tremendous electricity demands, raise consumer energy prices, and take up substantial public land. Even some AI enthusiasts expressed skepticism: Elon Musk broke with Trump by publicly bashing the announcement, posting on X that SoftBank doesn’t “actually have the money” to support Stargate. (Altman called this characterization “wrong” in a post of his own.)

To hear these companies tell it, however, the path forward is all but inevitable. Put together, major American tech companies are already spending perhaps hundreds of billions of dollars a year developing their technology with a questionable path to profit. Instead of acting as a deterrent, those costs have been spun into a selling point. Executives at OpenAI, Anthropic, Microsoft, Nvidia, and their competitors are fond of touting the lucrative sums—$100 billion, or perhaps $7 trillion—their technology will require, as if to say: This will be big. Don’t miss out. They have seemingly willed demand into existence.

In an interview after the press conference, Altman said that Stargate “means we can create AI and AGI in the USA. It wouldn’t have been obvious this was possible—I think with a different president, it might not have been possible—but we are thrilled to get to do this. I think it will be great for Americans.” Now the White House is fully embracing tech executives’ messaging. But all of this started well before Trump’s inauguration. Ellison himself said that Stargate had been in the works for “a long time,” and the nationwide build-out of data centers, power plants, and transmission lines is well under way. Days before his term ended, Joe Biden signed an executive order for “advancing United States leadership in artificial intelligence infrastructure,” which would open up federal lands for data-center construction. (Trump, when asked if he would rescind the order, responded, “No, I wouldn’t do that. That sounds to me like something I would like.”)

[Read: Microsoft’s hypocrisy on AI]

Winning the generative-AI race would, in Trump’s telling, be a display of his geopolitical and economic might. But only a day into his presidency, Stargate showed Trump taking cues from China, Microsoft, OpenAI, and Biden all at once—from a foreign adversary, the tech giants he vilified in 2020, and a political rival he has ruthlessly vilified. During yesterday’s briefing, Trump read a statement that the tech executives had apparently prepared. “This monumental undertaking is a resounding declaration of confidence in America’s potential under a new president,” he said, looking up from the dais and grinning at the final two words. “New president. I didn’t say it; they did. So I appreciate that, fellas.” Altman and the others knew exactly how to play this. Trump—and the rest of the nation—is merely tagging along.

Oracle (ORCL) stock closed up by more than 7% on Tuesday amid reports of a joint venture with OpenAI and Softbank (SFTBY).

Read more...

Elon Musk is already casting doubt on OpenAI’s new, up to $500 billion investment deal with SoftBank (SFTBY) and Oracle (ORCL), despite backing from his allies — including President Donald Trump.

Read more...

OpenAI, Oracle and Softbank join forces to build artificial intelligence infrastructure.