The AI boom is constructing a perpetual motion machine for finance through technological myths, capital loops, and media evangelism. However, this comes at the cost of draining innovation resources from other sectors and risking a repeat of historic bubbles.

本文经授权转载自动察Beating，作者：律动编辑部，版权归原作者所有。

"The only winning move is not to play."

In October, Michael Burry posted this phrase on social media. It originates from the 1983 film WarGames, where a supercomputer repeatedly simulates nuclear war and concludes that avoiding engagement is the best strategy.

A few days later, Burry revealed his Q3 holdings. Known for accurately shorting the 2008 subprime mortgage crisis, this investor allocated nearly 80% of his managed funds—about 1 billion USD—entirely to shorting Nvidia and Palantir.

For him, the strongest way to sidestep this irrational "long" frenzy is to short these companies.

His bet targets not only overvalued firms but also the dominant consensus of our time: that AI is more than a technological revolution—it's a matter of faith for capital.

How did this consensus form? How was it driven to a climax? As this belief machine keeps operating, what costs are we paying?

The Gospel

Every financial frenzy is backed by a repeatedly told story that countless people believe. In this AI wave, this narrative is textbook-level. It involves three forces working together: tech leaders crafting myths, Wall Street providing "rational" endorsements, and media spreading the gospel.

The first narrators are the evangelists of the Singularity. Figures like Sam Altman, CEO of OpenAI, and Demis Hassabis, co-founder of DeepMind, have successfully transformed the once far-fetched concept of artificial general intelligence (AGI)—once confined to science fiction and academia—into a near-at-hand, life-solving "new deity."

Altman reiterates that AGI will be "the greatest technological leap ever," bringing benefits "beyond our wildest imaginations." Hassabis describes it as a tool to help understand the universe's ultimate mysteries, with almost religious zeal.

Their language is filled with religious fervor for "the future" and "intelligence," successfully endowing this wave with a divine, almost sacred, significance beyond mere commercial interests.

If tech leaders script the myth, then Wall Street and economists serve as its rational gods, offering legitimacy.

Amid slowing economic growth and frequent geopolitical conflicts, AI is quickly chosen as a "cure" that can restore faith in the future.

Goldman Sachs forecast that by 2034, generative AI could boost global GDP by 7%, roughly 7 trillion USD. Concurrently, Morgan Stanley calls it the "core of the Fourth Industrial Revolution," comparing its productivity effects to steam and electricity.

The purpose of these figures and metaphors is to turn imagination into assets and belief into valuation.

Investors are convinced that giving Nvidia a P/E ratio of sixty isn’t madness—they're not buying a chip company but the engine that might power the future global economy.

Since ChatGPT's debut in November 2022, AI-related stocks have contributed 75% of S&P 500 returns, 80% of earnings growth, and 90% of capital expenditure increases. This techno-narrative has nearly become the sole pillar supporting U.S. equities.

Finally, media and social networks amplify the myth to its fullest extent.

From the stunning debut of text-to-image models like Sora, to every major update from Google, Meta, and others, each node is magnified, looped, and re-amplified. Algorithms push this faith into everyone's feed.

Meanwhile, discussions about "AI replacing humans" spread like shadows—engineers, teachers, designers, journalists—all unsure if they will belong in the next era.

Fears and reverence spread hand in hand, shaping a grand, nearly indisputable creation myth that prepares the stage for one of the largest capital gatherings in human history.

The Machines

As the "Gospel" spreads worldwide, a group of skilled financial engineers begins to act.

Their goal is to turn this abstract belief into an operable machine—a self-reinforcing, self-sustaining capital system. Rather than calling it a bubble, it’s better described as a meticulously designed financial engine, far more complex than the derivatives of the 2008 crisis.

This engine's core is built by a few tech giants. They weave capital, compute power, and revenue into a closed loop, where funds circulate, amplify, and recirculate—like an algorithm-driven perpetual motion system.

Leading the charge, Microsoft invests heavily in AI research organizations like OpenAI, pouring over 13 billion USD into OpenAI. Over a few years, OpenAI's valuation soared from billions to nearly a trillion USD, establishing a new market myth.

This massive funding enables more expensive training: to produce GPT-4, OpenAI utilized over 25,000 Nvidia A100 GPUs, with future models demanding exponentially increasing compute resources. These orders naturally flow toward Nvidia, the market's dominant supplier.

Nvidia's data center revenue jumped from 4 billion USD in 2022 to an estimated 20 billion USD in 2025, with profit margins exceeding 70%. Its stock price soared, making it the world’s most valuable company.

Meanwhile, major tech giants and institutional investors holding Nvidia shares track its rise, further boosting their balance sheets.

The story doesn't end at training—deployment is the next battleground. OpenAI deploys models on the cloud, partnering with Microsoft, whose Azure cloud services generate billions annually, fueling Azure's growth.

A perfect feedback loop forms: Microsoft invests in OpenAI, which purchases Nvidia GPUs and Azure services; Nvidia and Microsoft’s revenues grow, pushing their stock prices higher, reinforcing their investments.

In this cycle, capital simply shifts among a handful of tech giants, creating enormous "revenues" and "profits" seemingly out of thin air. The growth on paper validates itself, raising valuations in tandem. The machine begins to feed itself—the entire system can operate endlessly without genuine demand from the real economy.

This core engine has rapidly expanded into various industries.

The fintech and payments sectors are among the earliest to integrate AI.

Stripe epitomizes this. Valued at over 100 billion USD, it handled a total payment volume of 1.4 trillion USD in 2024—about 1.3% of global GDP. By 2025, it partnered with OpenAI to launch "instant checkout" in ChatGPT, embedding payment functions into conversational AI.

Stripe's role is multifaceted: it is both an infrastructure buyer—continuously acquiring compute to improve fraud detection and recommendations—and a beneficiary, monetizing AI-powered features to boost its valuation.

PayPal swiftly follows, becoming the first major digital wallet integrated fully with ChatGPT in October 2025.

The ripple effect extends beyond finance. Traditional industries like manufacturing are feeling the tremors. Once relying on automation hardware, they now pay for algorithms. In 2025, a German automaker announced a €5 billion investment over three years to AI-enable its production and supply chain, mainly buying cloud and GPU services. Similar efforts are happening across automotive, steel, electronics, and other sectors—treating compute power as the new fuel.

Retail, logistics, advertising—all sectors are undergoing similar transformations, purchasing AI compute, partnering with model providers, repeatedly emphasizing their "AI strategies" in reports and investor calls—belief in the narrative boosts their valuation and financing.

All these streams ultimately flow into a few core companies— Nvidia, Microsoft, OpenAI—circulating among GPUs, cloud, and models. As their revenues soar, stock prices climb, reinforcing the AI narrative's trust.

The Costs

However, this machine isn’t without costs. Its fuel comes from real economic and social resources, which are gradually extracted, transformed, and combusted into growth noise. Although often obscured by hype, these costs are real and subtly reshape the global economic structure.

The first is opportunity cost of capital. In venture capital, funds always chase the highest returns. The AI gold rush has created an unprecedented black hole of capital. According to PitchBook, in 2024, about one-third of global venture funding flowed into AI; in the first half of 2025, this rose to nearly two-thirds in the US.

This means that capital which could otherwise support climate tech, biotech, renewable energy—key areas for humanity's future—are disproportionately diverted into the same story.

When all the smartest funding goes into one narrative, the soil for innovation is drained. Focus does not always mean efficiency; it often erodes diversity.

In 2024, global venture capital in clean energy was only one-fifth of that in AI. Climate change remains one of humanity's greatest threats, yet funding flows into compute and models instead. The biotech sector faces similar neglect, with many entrepreneurs noting that investors find AI stories more glamorous and quicker to yield returns.

This capital frenzy approaches a dangerous tipping point.

The US tech industry's capital expenditures nearly match the peaks of the 1999-2000 dot-com bubble, when “new paradigms” were the hype, companies expanded before profit, and investors poured into visions of changing the world. When the bubble burst, Nasdaq lost two-thirds of its value, and Silicon Valley faced a prolonged winter.

25 years later, similar sentiment reignites—yet the protagonists are AI giants. Capital spending curves steepen again, with hundreds of billions invested in data centers and compute clusters, as if just spending guarantees the future.

Such historical comparable raises concerns: the outcome might not mirror the past exactly, but the highly concentrated capital momentum means society will bear the costs when the inevitable correction arrives.

The second cost involves talent and intellect. An unprecedented brain drain is underway. Top engineers, mathematicians, and physicists are pulled from solving fundamental problems to focus on large-scale models. In Silicon Valley, top AI scientists command salaries over a million USD per year, dwarfing salaries in academia’s physics labs.

Behind this salary gap lies a shift in focus: the brightest minds are moving away from basic science, energy innovation, and biotech—long-term fields—toward highly commercialized AI. Knowledge flows rapidly but into increasingly narrow channels.

The third cost is strategic for industries. Under AI's pressure, nearly every traditional sector faces a sense of urgency—either to join the costly AI arms race or risk falling behind. Many companies invest heavily without clear return paths, driven by fear of obsolescence rather than strategic necessity.

By 2025, global data center capital expenses are projected at around 500 billion USD, mostly AI-related; giants like Amazon, Meta, Google, and Microsoft plan to invest over 200 billion USD. Many retail and manufacturing firms announce multimillion-dollar AI investments, but according to MIT research, most returns are insufficient to justify the costs. For these firms, AI is more a show of modernity than a practical necessity.

The Turning Point

However, viewing this AI surge solely as a financial bubble and resource misallocation is overly simplistic. Deep structural changes are underway—whether markets rise or fall.

"Intelligence" and the compute power driving it are replacing traditional capital and labor—becoming new fundamental factors of production.

Similar to electricity in the 19th century and the internet in the 20th, AI’s importance is irreversible and indispensable. It is quietly infiltrating every industry, rewriting cost structures and competitive norms.

Control over compute resources has become a new oil race. Dominance in semiconductors and data centers is no longer just industrial; it’s a matter of national security.

The U.S. Chips Act, EU technology export bans, and Asian government subsidies form a new geopolitical battleground—an accelerating global race for compute sovereignty.

Meanwhile, AI is setting benchmarks across industries. Possessing a clear AI strategy is now key for companies to gain investor trust and survive future competition. Whether willing or not, learning AI language and thinking along its lines is the new business grammar and survival rule.

Michael Burry’s foresight isn’t infallible; he’s misjudged directions before. This gamble may once again prove his prescience or turn him into a tragic figure swept away by the times.

Regardless of outcome, AI has already reshaped our world forever. Compute power has become the new oil, AI strategy the survival necessity, and global capital, talent, and innovation resources are converging on this path.

Even if the bubble bursts and the wave recedes, these fundamental shifts will persist—continuing to shape our environment, becoming the irreversible backdrop of this era.

Recommended reading: Chinese low-cost AI robots defeating ChatGPT in crypto trading battles.