The AI Bubble is Worse Than You Think

Unprecedented Spending Commitments

• 🚀 OpenAI has committed to an astounding $1.15 trillion in infrastructure spending over five years through deals with major tech players like Nvidia, AMD, Broadcom, Oracle, Microsoft, Amazon, and CoreWeave.
• 📊 This massive spending is vastly disproportionate to OpenAI's projected $20 billion in revenue by the end of 2025.
• 📈 To meet its obligations, OpenAI would need to achieve an 85x revenue growth in five years, effectively becoming the largest company globally, without accounting for operating costs.
• 💡 The scale of these commitments is nearly equivalent to the annual capital expenditure (CapEx) of all US companies combined.

The Circular Economy of AI Deals

• 🔄 Many of these deals involve interconnected, circular investments, where one company's investment in OpenAI is often tied to OpenAI purchasing services or products from that investor (e.g., Microsoft investing in Azure credits).
• 🧩 This creates an "ouroboros-like" flow of money, where revenue for one company is a cost for another, making it incredibly difficult to assess real cash flows and true company valuations.
• ⚠️ The complexity of these arrangements, including various contingencies and financial instruments, makes it impossible to accurately track and measure the financial health of the involved entities.

Parallels to Past Bubbles

• 🔍 The current AI situation is compared to the 2008 Great Financial Crisis (GFC) and the dot-com bubble of the late 1990s.
• ✅ Unlike the GFC, the AI bubble currently lacks widespread securitization and derivatives, and major tech players generally possess significant cash reserves and low debt.
• 📉 Compared to the dot-com bubble, today's major tech companies are largely profitable, but the market exhibits high valuations and an underlying economy that cannot sustain low interest rates.

The "AI Bubble" Assessment

• 💬 Despite some differences from past crises, the AI landscape is characterized by trillions of inter-company dealings, questionable earnings, and relatively high valuations.
• 🗑️ The speaker highlights the presence of "turds" (e.g., fake revenue, unsustainable spending, discussions of government backstops) alongside "raisins" (companies with strong cash flows and real business models).
• 🚨 The overall setup is described as insane and fragile, raising significant concerns about potential ripple effects if one interconnected company were to fail.