Why the AI Bubble Is Looking Increasingly Precarious

TL;DR

Despite massive AI investment, users are cutting spending on advanced models, suggesting lower future revenues that could make the industry's massive debt burden unsustainable.

Key Takeaways

1The LLM token expenditure index has fallen nearly 50% since May, indicating users are shifting toward cheaper, less advanced AI models
2Price sensitivity suggests users view AI as a time-saving convenience rather than a transformative business tool, threatening premium pricing models
3US tech companies have borrowed over $300 billion since last year for AI investments, with debt issuance expected to hit $500 billion in 2024—exceeding dot-com bubble levels
4Long-dated AI bonds are trading at steep discounts, signaling market skepticism about long-term profitability despite near-term debt servicing ability
5Circular financing among hardware and AI companies (e.g., Nvidia investing in OpenAI) creates systemic risk if any major player defaults

Notable Quotes

“The main reason the AI looks extra bubbly at the moment is this graph, which is something called the LLM token expenditure index.”

“Rather, it means the average price that people are willing to pay for AI is falling.”

“This heightened price sensitivity and the implied skepticism towards advanced models is obviously pretty terrible news for AI companies because it suggests that they should expect tighter margins and lower revenues.”

“AI spending is ramping up faster than any previous historical episode, including the dot and railway booms.”

Chapters

1. The AI Revenue Problem

Despite hundreds of billions in AI spending, companies are struggling with profitability. The video introduces the central question: when will these investments pay off?

2. The LLM Token Expenditure Index: Price Collapse

Users are becoming price-sensitive and shifting from expensive advanced AI models to cheaper alternatives. This index has halved from its May peak, indicating falling average user willingness to pay per token.

3. What Price Sensitivity Reveals

The shift toward cheaper models suggests users see AI as a convenience tool, not a transformative technology. This undermines the business case for expensive advanced models and threatens company margins.

4. The Debt Problem

Tech companies have borrowed $300+ billion for AI, with projections of $500 billion in debt issuance this year. This debt assumes massive future revenues that may not materialize if margins tighten.

5. Market Signals of Skepticism

Long-dated bonds are trading at discounts, companies are pivoting to equity raises (which are also underperforming), and IPOs are being delayed—all indicating market doubt about long-term profitability.

6. Systemic Risk: Circular Financing

AI companies engage in circular financing (e.g., hardware firms investing in AI labs that buy their compute), creating contagion risk if any major player fails.

Key People & Entities

Anthropic: AI company noted as one of few turning a profit; reportedly considering price cuts amid competition
OpenAI: Major AI lab; reportedly considering price cuts and has delayed IPO plans
Nvidia: Hardware provider engaging in circular financing by investing in AI labs
Alphabet (Google): Tech giant planning to raise $85 billion through equity issuance for AI investments
SpaceX: Recently completed IPO; stock has fallen below original IPO price post-launch
JP Morgan: Financial institution forecasting $500 billion in tech company debt issuance for 2024

Glossary

Tokens: Basic units of text that AI reads and processes; usually a couple of characters long. Used as the standard measure of AI usage.
LLM Token Expenditure Index: Measures the average price users are willing to pay per 1 million tokens of AI usage; a key indicator of pricing trends in the AI market.
Price Sensitivity: User tendency to switch to cheaper, less advanced AI models rather than paying premium prices for cutting-edge models.
Margin Compression: Reduction in profit margins when selling prices fall faster than costs, reducing profitability despite continued sales.
Circular Financing: Financial arrangement where companies in the same ecosystem lend to or invest in each other, creating interdependency and systemic risk.

Explore