"Is the 'Lehman Moment' approaching? A 15,000-word long article from a major short seller reveals: OpenAI is destined to collapse, and global stock markets are likely to face liquidation."

Original author: Li Dan

As OpenAI approaches its IPO, a lengthy blog post of about 15,000 words once again brings the AI bubble controversy to a climax.

Long-term bear on AI and commentator Ed Zitron, who has a large readership in the tech industry, recently made the most radical claim to date in a blog post: the real AI bubble is essentially the "OpenAI bubble”; if OpenAI ultimately fails, it will become the "Lehman Brothers" of the AI era, which would not only shatter the entire logic of AI investment but could also trigger a large-scale repricing of data centers, AI infrastructure, and even global tech stocks.

These views quickly attracted attention from financial media. To the media, Zitron's core argument is not whether AI has value, but rather whether OpenAI has a business model robust enough to support the entire AI capital cycle. If the answer is negative, then the financing, computing power investments, and capital expenditure systems built around OpenAI could face a chain reaction.

Of course, this is not a market consensus. Investors, including Oak Tree Capital co-founder Howard Marks, have recently stated that compared to their previous belief that AI might just be a bubble, they now recognize AI as a long-term value of a general-purpose technology platform, believing the industry is still in the early stages of commercialization.

Is it an AI bubble, or an OpenAI bubble?

Unlike most "AI bubble theory," Zitron presents a more shocking judgment:

What is truly concerning is not the entire AI industry, but rather one company.

In his view, since ChatGPT burst onto the scene at the end of 2022, OpenAI has effectively become the "credit anchor" of the entire generative AI era.

Investors are willing to believe: AI will change the world; large-scale data centers are worth building; GPU demand will grow at a high rate for the long term; large model companies will ultimately become profitable; AI startups can create sufficiently large end demand.

And Zitron believes that all this is predicated on the premise of OpenAI's continued high-speed growth. He argues that OpenAI not only defined the current AI boom but also shaped the capital market's valuation logic for the entire AI industry chain; therefore, once this core assumption is broken, the impact may far exceed that of a single unicorn company.

In other words, OpenAI is not just a company, but more like a "systemically important institution" for the entire AI investment cycle.

Why does he believe OpenAI's business model has fundamental flaws?

Zitron's doubts primarily focus on three aspects.

The first is that inference costs remain too high.

As the user base of ChatGPT continues to grow, every user query means increased costs for GPUs, electricity, and servers. If a large number of users remain on low-cost or even free plans for an extended period, and enterprise-level revenue growth cannot simultaneously cover costs, then scale expansion may actually lead to increased losses.

The second is that capital expenditure is growing far faster than cash flow improvement.

Currently, the largest expenditure in the AI industry is no longer model training but rather inference computing power, GPU procurement, and the construction of global data centers.

OpenAI and its partners are driving tens of billions of dollars or even larger-scale data center investments, which usually take years to recover costs. If future AI demand does not grow as expected, a lot of infrastructure may face declining utilization issues.

The third is continuously relying on external financing.

Zitron analyzes that he believes OpenAI will still need continued financing in the coming years to cover expenditures for model development, computing power procurement, and infrastructure construction; if risk appetite in the capital market declines or the financing environment tightens, its business model will face greater pressure.

These views currently remain personal judgments of Zitron and have not been recognized by OpenAI, but do reflect the market's recent debate regarding AI capital return rates (ROI).

Why are Oracle, CoreWeave, and data center operators in the spotlight?

Compared to OpenAI itself, Zitron is more worried about the industry chain's leverage effect.

Over the past two years, the U.S. tech industry has seen an unprecedented wave of data center construction.

Major cloud providers (Hyperscalers) such as Microsoft, Google, Meta, and Amazon have significantly increased capital expenditure; meanwhile, companies like Oracle and CoreWeave are taking on an increasing number of AI computing power construction tasks.

These projects heavily rely on: long-term leases, project financing, private credit, corporate bonds, and large-scale capital expenditure.

If core customers like OpenAI see demand lower than expected in the future, or if capital markets reassess AI return rates, then the utilization rates of data center assets, lease contracts, and even financing capabilities may be affected.

Media outlets point out that Zitron believes that if OpenAI encounters a major setback, companies like Oracle and CoreWeave that rely on growing AI infrastructure demand could be the first to suffer since the high valuations previously accorded to these companies are largely based on expectations of sustained AI demand surge.

Of course, major tech giants such as Microsoft, Meta, and Alphabet continue to expand AI capital expenditures and generally emphasize that AI infrastructure investments align with long-term strategies, so there have been no signs of a comprehensive capital expenditure contraction in the market so far.

Why is Anthropic and SoftBank also pulled into the discussion?

Besides OpenAI, Zitron also pointed his criticism at Anthropic.

His reasoning is that while the two companies take different paths of development, the common characteristic is that they both require continuous massive capital investment to build models, procure computing power, and rely on large tech companies for computational resources and financing support. If the speed of AI commercialization falls below expectations in the future, both companies could face profit pressure.

Another frequently mentioned entity is SoftBank.

In recent years, SoftBank has returned to the forefront of major AI investments, actively participating in financing for AI infrastructure, chips, and model companies.

If the AI industry were to enter a valuation adjustment cycle, SoftBank's massive AI asset portfolio would naturally also become a focus of the market. However, SoftBank remains firmly committed to the long-term development of AI, considering it a significant direction for the next technological revolution.

Has AI trading already overheated?

In fact, the debate over whether AI has entered a bubble phase has been ongoing on Wall Street for more than a year.

Supporters of the "bubble theory" argue that:

Investment growth in AI infrastructure is much faster than revenue growth;
The profitability model of large models has not yet been fully validated;
Capital expenditure on data centers has reached historical records;
Market valuations increasingly rely on growth expectations for the next few years.

Optimists, on the other hand, argue that AI is a typical general-purpose technology revolution, similar to the Internet and electrification, where early investments often far exceed short-term returns but can create new industries and business models in the long term.

Howard Marks recently stated that he has shifted from initially suspecting that AI may only be a bubble to more recognition of its long-term value. He believes that the reasoning, contextual understanding, and interaction capabilities demonstrated by modern AI possess unprecedented characteristics and thus cannot be simply equated with historical speculative bubbles.

Some academic research also presents a more neutral conclusion: the current AI market exhibits real technological advancement, but there are also localized overvaluation and excessive capital expenditure issues, making it closer to a "technological revolution coupled with localized bubbles," rather than purely speculative frenzy.

What is truly worth paying attention to is not whether OpenAI will fall

Regardless of whether one agrees with Zitron's judgments, the questions he raised are becoming the focus of increasing attention from investors:

When can AI investments be translated into stable cash flow?

Over the past year, the capital market has almost defaulted to the idea that the higher the capital expenditure in AI, the better.

However, recently, whether it's chip stocks, server manufacturers, or cloud computing companies, investors have begun to pay more attention to another set of indicators: corporate AI revenue growth; AI product payment rates; speed of decline in inference costs; data center utilization rates; AI investment return cycles.

If these indicators continue to improve, then the current massive capital expenditures may ultimately prove to be a forward-looking investment similar to that of the Internet era; but if the commercialization speed continues to lag behind investment expansion, the market's valuation logic regarding AI trades may also face recalibration.

Therefore, what Ed Zitron's lengthy article truly sparked discussion about is not "whether OpenAI will inevitably become the next Lehman Brothers," but rather it again places the most core question of the AI era in front of investors: after capital expenditures continue to break records, can cash flow and profitability keep pace? The answer to this question may ultimately determine the real trajectory of global AI trading in the coming years.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。