The AI Bubble is Worse Than You Think

Nvidia's $100 Billion Investment in OpenAI

Overview of Major Investments

• In September 2025, Nvidia announced a significant investment of up to $100 billion in OpenAI aimed at funding a data center build-out of at least ten gigawatts.

• Shortly after, OpenAI and AMD entered into a deal where AMD would provide warrants for 160 million shares at $0.01 each, while OpenAI committed to purchasing six gigawatts of AMD GPUs worth approximately $90 billion.

Financial Implications

• The deal with AMD implies that while they receive $18 billion in EBITDA, the cost of stock is valued at nearly $40 billion.

• Another partnership was formed with Broadcom to develop ten gigawatts of custom AI chips over four years, estimated at a value of $350 billion.

OpenAI's Massive Spending Commitments

Total Infrastructure Commitments

• OpenAI has made spending commitments totaling approximately $1.15 trillion over five years across various partnerships including Oracle ($300 billion), Microsoft ($250 billion), and Amazon ($38 billion).

Revenue vs. Spending Discrepancy

• Despite having only $13 billion in revenues, questions arise about how OpenAI can sustain such massive spending commitments.

• Sam Altman projects an annualized revenue run rate of $20 billion by December 2025, which starkly contrasts with their spending obligations.

Understanding CapEx and Growth Projections

CapEx Contextualization

• U.S. companies collectively spend around $1.2 trillion on capital expenditures (CapEx) annually; OpenAI’s commitments approach this total within five years.

Industry Comparisons

• Major companies like Amazon and Microsoft have combined CapEx spending of only $225 billion over the last year, highlighting the enormity of OpenAI's financial commitments.

Revenue Growth Requirements for Sustainability

Projected Revenue Needs

• To meet its projected spending from 2025 to 2030—growing from $6 billion to an astonishing $295 billion—OpenAI would need to increase its revenue from $12 billion to nearly $983 billion.

Feasibility Concerns

• This growth requirement suggests that OpenAI must become the largest and most profitable company globally within five years—a highly improbable scenario given current market conditions.

Potential Solutions and Broader Implications

Operating Costs Consideration

• Current projections do not account for operating costs such as energy expenses or debt servicing, complicating the financial outlook for OpenAI.

Government Backstop Discussion

• There are discussions about seeking government support if obligations cannot be met, indicating potential taxpayer involvement in covering losses.

The Wider AI Landscape: Competition and Spending Trends

Other Major Players' Investments

• Competitors like Anthropic have raised substantial funds (e.g., $27 billion), while Meta plans significant investments including a projected spend of $70 billion on data centers.

Market Dynamics

• The overall trend indicates unsustainable levels of investment across the AI sector without clear paths to profitability or revenue generation.

Understanding the Complex Financial Interconnections in AI Investments

The Intricate Web of Investments

• Altman and Jensen Huang, despite not being finance experts, are involved in complex financial dealings that resemble those of seasoned financiers. Their recent deals highlight interconnected spending commitments among major tech companies.

• Microsoft has invested over $13 billion into OpenAI's for-profit arm, primarily through Azure Cloud Computing credits rather than cash. This investment is designed to ensure OpenAI purchases Microsoft products.

• As a result of this arrangement, Microsoft must expand its Azure data centers to accommodate the increased demand from OpenAI, creating a cycle where investments lead to further purchases across companies.

• Nvidia enters the picture by committing up to $100 billion in investments towards OpenAI, which will be used for purchasing Nvidia chips. This creates a feedback loop where funds circulate between these entities.

• The cyclical nature of these transactions complicates tracking and measuring revenue flows, as one company's revenue often becomes another's cost or investment, leading to confusion about financial realities.

Challenges in Valuation and Market Stability

• The complexity of inter-company dealings raises questions about how to accurately value these companies when distinguishing between real and artificial earnings becomes challenging.

• Concerns arise regarding potential ripple effects if one company within this interconnected network fails. Comparisons are drawn with past financial crises like the 2008 Great Financial Crisis (GFC).

Historical Context: Comparing AI Investments with Past Crises

The Great Financial Crisis (GFC)

• The GFC resulted in significant job losses and a dramatic decline in household net worth due to subprime mortgages that represented around 20% of all US mortgages.

• A systemic loss exceeding $10 trillion occurred as defaults on low-quality loans triggered widespread economic collapse, raising questions about how such losses could stem from relatively small amounts of bad debt.

Differences Between AI Investment Dynamics and Past Crises

• Unlike during the GFC, current obligations related to AI investments are not being securitized into risky derivatives that amplify exposure; major tech players today possess substantial cash reserves with minimal debt.

Insights on the Dotcom Bubble

Characteristics of the Dotcom Bubble

• The dotcom bubble was characterized by rapid growth driven by excitement over internet capabilities but ultimately led to market panic when profitability was questioned.

• By March 2000, the Nasdaq had surged significantly before crashing by October 2002, wiping out trillions in market value as investors realized many companies were unprofitable.

Current Market Conditions Compared to the Dotcom Era

• Today’s market shows lower valuations compared to peak dotcom levels; however, many tech firms are now profitable unlike during the previous bubble period where few survived without profits.

Conclusion: Navigating Current Market Realities

Evaluating Risks and Opportunities

• While today's AI landscape lacks some risks seen during past crises (like excessive debt), it still faces challenges such as questionable earnings reports and inflated valuations amidst an economy struggling with interest rates.

• There exists a mix of strong businesses alongside weaker ones within this ecosystem—illustrated metaphorically by mixing "raisins" (good companies with solid cash flows) with "turds" (poorly performing entities).

Final Thoughts

• Despite some promising elements within AI investments, there remains skepticism regarding unsustainable spending patterns that may require extraordinary growth from firms like OpenAI.