The Collapse of OpenAI | Exposing the Fraud.

Artificial Intelligence Bubble & OpenAI’s Financial Concerns

Key Concepts:

• AI Bubble: The potential for inflated valuations and unsustainable investment in Artificial Intelligence companies.
• Fine-tuning as a Service: A product offering (like Thinking Machines Lab’s Tinker) that allows users to customize AI algorithms for specific research purposes.
• Greater Fool Theory: An investment strategy relying on finding someone else to buy an asset at a higher price, regardless of its intrinsic value.
• AGI (Artificial General Intelligence): Hypothetical AI with human-level cognitive abilities.
• Burn Rate: The rate at which a company is spending its cash reserves.
• Credit Default Swaps (CDS): Financial contracts used to transfer the credit risk of a borrower to another party.
• Stargate: OpenAI’s ambitious, and costly, infrastructure project for future AI development.

---

I. The Emerging AI Bubble & Questionable Valuations

The speaker asserts a strong belief that a significant bubble is forming within the Artificial Intelligence sector, predicting a correction and consolidation. This assessment is driven by the sheer volume of investment – $500 billion in checks being written – fueled by the hype surrounding AI. The core argument is that current valuations are detached from fundamental realities, potentially bordering on fraud. The speaker frames this as potentially the “biggest bubble in the history of mankind,” contrasting it with government waste, but emphasizing the global economic implications.

II. Thinking Machines Lab: A Case Study in Suspect Funding

A central example used to illustrate this bubble is Thinking Machines Lab, founded by former OpenAI CTO Meera Veratti, along with Barrett Zoff and John Schulman (OpenAI co-founder). Despite being only 11 months old, the company secured a $2 billion seed round at a $10 billion valuation from Andre Horowitz, followed by further funding at $12 billion, and is now seeking investment at a $50 billion valuation.

The speaker questions the justification for these valuations, focusing on the company’s core product, “Tinker,” a “fine-tuning as a service” API for AI research. The concern arises from the discovery that positive reviews for Tinker on the company website were allegedly authored by PhD students (including a first-year student at Princeton) and, critically, by an employee of Thinking Machines Lab itself. This is characterized as “borderline fraudulent” and indicative of a manufactured perception of value. The speaker highlights the role of Andre Horowitz’s backing in attracting further investment, suggesting a “greater fool theory” is at play – investors relying on finding someone else to pay an even higher price.

The situation is further complicated by the departure of co-founder Barrett Zoff due to disagreements with leadership, followed by OpenAI rehiring Zoff, Schulman, and others. This is presented as evidence of instability and a lack of genuine progress.

III. Humans: Raising Funds with Minimal Substance

The speaker then introduces Humans, a three-month-old company with no revenue and, seemingly, no product, which recently raised $480 million at a $4.5 billion valuation. The company’s stated mission – focusing on “human-centric AI” and “making connections” – is ridiculed as a justification for the massive investment. The speaker emphasizes the absurdity of this valuation, again pointing to the involvement of OpenAI alumni and the influence of firms like Andre Horowitz. The company’s website copy is dissected, highlighting the vague and aspirational language used to justify the investment.

IV. OpenAI’s Financial Struggles & Shifting Strategies

The analysis shifts to OpenAI itself, detailing a series of concerning developments. Despite previously stating a reluctance to use advertising, OpenAI is now implementing ads. This follows an internal “Code Red” memo initiated by Sam Altman to counter the growing competition from Google’s Gemini, which is gaining market share, particularly among major clients like Salesforce.

OpenAI is also introducing a cheaper subscription tier (GPT Go at $8/month) in an attempt to convert free users into paying customers, acknowledging the difficulty in monetizing its user base. Statistics reveal that 95% of GPT users do not pay for the service. The speaker predicts that the commoditization of basic chatbot functionality, with the integration of similar features into platforms like Apple’s Siri and Amazon’s TVs, will further exacerbate OpenAI’s revenue challenges.

A particularly alarming point is the rumor that OpenAI plans to take a cut of revenue generated from ideas developed using its AI tools, which the speaker condemns as “theft” and a potential driver of customer alienation.

V. Financial Data & Concerns of Insolvency

The speaker presents data suggesting OpenAI is facing a severe financial crisis. HSBC estimates the company will be cash flow negative until at least 2030 and require an additional $27 billion to remain solvent. The speaker highlights the company’s high burn rate, estimating losses of $9 billion in 2025, but then presents leaked Microsoft financial data suggesting losses could be as high as $50 billion per year.

This is contrasted with Google’s profitability, which generated $34 billion in net profit in the last quarter, or $136 billion annually. The speaker points out that Google, Meta, and Microsoft have sustainable business models, while OpenAI relies heavily on fundraising and hype.

The speaker references Elon Musk’s statement that OpenAI will bankrupt Microsoft and notes that Microsoft is now allowing OpenAI to seek funding from other sources due to its inability to cover OpenAI’s expenses.

VI. Enron-Like Transparency Issues & Investor Concerns

The speaker draws a parallel between OpenAI’s financial opacity and the scandal surrounding Enron, citing a Wall Street Journal quote describing OpenAI’s losses as comparable to national government deficits and noting that investors are “not even allowed to ask about the amount of cash it’s burning” during fundraising meetings. Sam Altman’s response to an investor questioning funding commitments – suggesting they sell their shares – is presented as further evidence of a lack of transparency.

VII. Conclusion: A Looming Correction

The speaker concludes that the current situation is unsustainable and predicts a significant correction in the AI market. The speaker’s personal experience building a real-world AI application in the real estate sector, with a focus on tangible value creation, fuels skepticism towards the hype-driven valuations of companies like OpenAI and Thinking Machines Lab. The speaker emphasizes the potential for these inflated valuations to collapse, leaving investors with substantial losses and setting the stage for future case studies in financial mismanagement. The speaker’s bias is acknowledged, stating a preference for investing in a company with a real product and sustainable business model.