OpenAI's decision to shut down Sora, its ambitious video generation tool, in March 2026 may have been a warning sign of deeper financial troubles ahead. Technology critic Ed Zitron argues that OpenAI could face a "Lehman Brothers moment" as early as the end of this decade, driven by unsustainable spending, loss-making subscriptions, and an insatiable appetite for computing resources that far exceeds revenue.

Why Did OpenAI Shut Down Sora?

The closure of Sora represents a significant retreat from OpenAI's earlier ambitions in video generation. While the company never publicly detailed the specific reasons for discontinuing the product, Zitron had predicted two years prior that Sora was not viable, a forecast that proved accurate when the tool was canned this March. The shutdown occurred as Google simultaneously moved to expand its own video generation capabilities, introducing Gemini Omni and Personal Avatars to its Google Vids platform.

What Financial Pressures Is OpenAI Facing?

According to Zitron's analysis of OpenAI's audited financials from 2024 and 2025, the company faces extraordinary financial strain. OpenAI is projected to burn over $852 billion by the end of 2030, with compute spending alone exceeding $50 billion this year, representing more than 50% of all global compute spending. This spending trajectory is unsustainable without continuous funding injections.

Zitron argues that OpenAI's business model relies on a dangerous cycle: the company raises massive amounts of capital from investors, which hyperscalers then use to fund infrastructure costs. This creates an illusion that other companies will eventually consume computing resources at similar scales, justifying the enormous capital expenditures. However, Zitron contends this assumption is flawed.

How Could OpenAI's Collapse Affect the Broader Market?

The potential consequences extend far beyond OpenAI itself. Zitron predicts that OpenAI's downfall will trigger a "violent, punishing effect on the entire stock market," with particular danger for companies like Oracle, which could be left holding "billions of wasted capex, and endless debt and leases" if the company collapses. The scale of potential damage concerns even Zitron, who stated he is "almost hoping I'm wrong" about his forecast.

Zitron, who

The collapse would likely unfold in stages. Based on his research, Zitron estimates OpenAI will need to raise funding at least three more times over the next decade, with the final reckoning coming only after "AI data center debt and venture capital funding has been almost entirely exhausted".

What Specific Factors Could Trigger OpenAI's Downfall?

• Loss-Making Subscriptions: OpenAI's subscription revenue model has not generated sufficient returns to offset massive infrastructure costs, creating a fundamental mismatch between income and expenses.
• Weak Advertising Revenue: Unlike traditional tech platforms, OpenAI has struggled to develop meaningful advertising revenue streams to diversify income sources.
• Massive Customer Costs: The expense of serving customers at scale has made it difficult for OpenAI to achieve profitability on a per-user basis.
• Reduced Compute Demand: If OpenAI collapses, the resulting loss of demand for computing resources would cascade through the entire AI infrastructure market.

Zitron also expressed concerns about Anthropic, OpenAI's closest competitor, noting that the company is "burning billions on training models, producing way too many Claude iterations and lacking focus". This suggests that financial instability may not be unique to OpenAI but rather endemic to the current AI business model.

Is There Disagreement Among Experts?

Not all prominent voices share Zitron's pessimism. Legendary investor Howard Marks, co-founder and co-chairman of Oaktree Capital Management, recently shifted his view on AI from skepticism to cautious optimism. Marks explained that he "upgraded my opinion of AI and its potential because of its ability to talk about its own strengths and weaknesses, to use humor, to put information in the context of me, to use what it knows about me. And this is really exceptional stuff". Marks noted that AI possesses a quality of autonomy that is "unprecedented" compared to past technological innovations like railroads and the internet.

Marks

This disagreement reflects a broader tension in the market: while some investors remain convinced of AI's transformative potential, others see a bubble inflated by unsustainable spending and unrealistic expectations about near-term returns.

What Does Google's Video Expansion Mean for the Market?

As OpenAI retreated from video generation, Google moved aggressively into the space. The company introduced Gemini Omni, a multimodal AI model designed for video creation that allows users to generate and edit videos using natural language prompts, along with Personal Avatars, which enable users to create AI-generated digital versions of themselves. These features are now available to Google AI Pro and Google AI Ultra subscribers, as well as eligible Google Workspace business customers.

Google's approach differs from OpenAI's in a crucial way: it is integrating video generation into existing enterprise software rather than launching it as a standalone consumer product. This strategy positions Google Vids in direct competition with specialized platforms like HeyGen, Synthesia, Captions, and D-ID, but within the protective ecosystem of Google's broader productivity suite.

The contrast between OpenAI's shutdown and Google's expansion underscores a fundamental question about the AI video generation market: can standalone AI video tools sustain themselves as independent products, or must they be embedded within larger platforms to achieve viability?