Before the spectacular implosion of his $45 billion AI hedge fund late last month, twenty-four-year-old Leopold Aschenbrenner had laid out a road map for a new Cold War with China.
In June 2024, shortly after being fired from OpenAI, Aschenbrenner published a treatise titled “Situational Awareness,” analogizing his work in Silicon Valley to the race to build the atomic bomb. The nuclear secrets that had permanently changed the world’s balance of power and won World War II for the United States, he wrote, had been upstaged. Advancements being made across Silicon Valley augured the imminent arrival of artificial superintelligence, a technology upon which the future of American economic and military supremacy rested.
It was a race, said Aschenbrenner, that would “determine the future of the free world.” The only question was whether the United States or its authoritarian adversary China would get there first.
Aschenbrenner’s manifesto skyrocketed his profile in Silicon Valley, earning him the trust of major tech investors who poured assets into a new hedge fund under his control. He named it, of course, Situational Awareness.
The ideas articulated in his essay, however, had not always been so well received in the industry. During the first Trump administration, Silicon Valley reacted with widespread concern as the White House sanctioned and blacklisted top Chinese tech firms, worried that decoupling American tech firms from the enormous Chinese market could harm the tech industry’s growth and revenue. After the Biden administration signaled bipartisan political support for the new Cold War by implementing sweeping semiconductor export controls on China, the industry rankled, with some US chip firms lobbying furiously against them.
Today, however, Aschenbrenner’s manifesto is a foundational text in the growing Silicon Valley consensus around the new Cold War. In his own 6,500 word essay this month, Meta’s Mark Zuckerberg wrote that obstructions to the AI build-out’s necessary energy and data center infrastructure represented a national security risk to US dominance in the AI race with China. Others, from liberals like OpenAI CEO Sam Altman and Anthropic chief Dario Amodei to right-wing venture capital heavyweights Peter Thiel and Marc Andreessen, agree.
In an address at the conservative Hudson Institute that quoted Samuel Huntington’s Clash of Civilizations
, Palantir CEO Alex Karp put it bluntly. When it comes to the AI race, he said, “There are only two cultures that are going to win in the next year. It’s going to be us or China.”
Ironically for Aschenbrenner, it was Chinese tech competition that brought his hedge fund crashing down. On July 30, in a seeming repeat of January 2025’s DeepSeek moment, a Chinese start-up named Moonshot AI released a cheaper, open-weight AI model that bested its closed-Aschenbrenner’s highly leveraged portfolio — tumbling
The Moonshot moment, however, also highlighted material differences between the US and Chinese tech sectors, much of which can be traced to the state policy that shaped them. The Chinese government has kept a tight rein on its tech industry. After Jack Ma, the billionaire head of the Chinese e-commerce giant Alibaba, criticized the country’s financial regulators in 2020, the state torpedoed his firm’s historic $37 billion public stock offering, removing him from control of the company and splitting Alibaba into six different firms.
It was the start of a sweeping regulatory crackdown that disciplined the nation’s technology industry and wiped out $2 trillion in value from the sector. Today the Chinese state has ushered in AI development plans focused more on practicality than dreams of superintelligence, a strategy that prioritizes efficiently diffusing AI tools and services throughout the economy and has been described by Xi Jinping as an approach “strongly oriented towards applications.”
Even the rise of a new technology like AI highlighted the centralization of power in Silicon Valley. Microsoft entered into exclusive deals and profit-sharing partnerships with OpenAI, while Google and Amazon took major stakes in OpenAI’s main rival, Anthropic. Nvidia, which controls 90 percent of the AI semiconductor market, has similar arrangements with much of the AI industry. Much of the trillions in capital flowing into the AI investment boom represents gains extracted from state-sanctioned monopolies that span cloud services, online advertising, social media, and e-commerce, as well as hardware monopolies like Nvidia.
Both the US and Chinese technology industries are products of their respective states’ policies, neither of which have been particularly democratic. The very first words of Aschenbrenner’s Situational Awareness manifesto describe the results of the long-standing US approach to industrial policy and technology regulation: “You can see the future first in San Francisco,” he wrote. “Right now, there are perhaps a few hundred people, most of them in San Francisco and the AI labs, that have situational awareness.” His proximity to the concentrated power and capital of America’s inventor-investor-policymakers facilitated his vision. “Let me tell you what we see.”
What he sees is what the CEOs of OpenAI, Anthropic, Google, Meta, and Amazon see, along with many of their twenty-four-year-old engineers: opportunity. The idea of looming conflict with China has its conveniences, offering an existential AI arms race that entices funders and increases firm valuations while providing built-in marketing for products that are still searching for consumer desire.
Some benefits are more personal. The Situational Awareness treatise enabled Aschenbrenner, who lacked investing experience entirely, to raise an initial $100 million for his hedge fund — which, in the span of under two years, reached a scale that its peers took decades to achieve — and cushioned his fall from grace when it came crashing down around him. Today investors are clamoring to reinvest.
But the ultimate return on this vision is even more lucrative. At its core, the new Cold War with China is a framework that provides Silicon Valley crucial insulation from democracy. At a congressional hearing in 2018 where he was grilled about an infamous Facebook data breach, Mark Zuckerberg’s notes laid out the strategy as simply as it gets: “Break up FB? US tech companies key asset for America; break up strengthens Chinese companies.” A photograph of the notes, taken by the Associated Press, quickly went viral.
The national security refrain developed into a commonly used antitrust and tech regulation efforts during the Biden administration. In 2021, faced with the possible passage of a bipartisan slate of antitrust bills aimed at their industry, top Silicon Valley firms organized a letter from a dozen senior Trump, Obama, and George W. Bush administration national security officials, including former Defense Secretary and CIA Director Leon Panetta, later a key speaker at the 2024 Democratic National Convention. Legislation to rein in American tech giants would degrade America’s military and intelligence dominance, they wrote, a regulatory mistake that would come “at the expense of U.S. economic and national security” and “cede U.S. tech leadership to China.” It was a useful argument: despite bipartisan backing in Congress and growing public support, none of the antitrust bills were passed.
Today the Cold War with China that Leopold Aschenbrenner so breathlessly relayed is a potent shield both for Silicon Valley and the national security state. A bipartisan backlash against the growing power of Silicon Valley to dictate the terms of American life is underway, with Americans across the political spectrum finding common ground in opposing the automation of their jobs, data centers and their associated energy costs, the increasing prevalence of surveillance tools like Flock cameras, and the spread of extractive rent-seeking tools like algorithmic price-fixing of groceries and apartment rents. Against this backdrop, arguments about national security and great-power rivalry offer an attractive way for tech firms to short-circuit growing public demands for control over the technologies shaping their lives.
The new Cold War with China offers the national security establishment a similar opportunity to safeguard its own power from democratic control. Recent disastrous choices, including Israel’s genocide in Gaza and the US war on Iran, are significantly weakening the foreign policy establishment’s hold on both parties and igniting support for antiwar politics. A new existential threat abroad that must be managed by experienced operatives with the commensurate security clearances, however, could provide fertile ground for re-entrenchment.
Neither trillions of dollars in AI investment nor a worldwide program of gunboat coercion can conceal the slipping grasp of Silicon Valley and the national security state over American public opinion. How many people desire to live out Leopold Aschenbrenner’s vision — a world in which a few hundred people scattered across Silicon Valley AI labs and Washington national security posts are afforded “situational awareness” through their proximity to unassailable power and can act to shape the world accordingly? Few to none. The shadow of the new Cold War with China points to the unanswered question straining at the heart of Leopold Aschenbrenner’s Silicon Valley vision: How much longer will Americans be willing to surrender democratic control of the world-structuring institutions that shape their lives?
