PodcastEurodollar UniversitySteve

OpenAI's Money Problem Is WAY Worse Than You Think

The discussion weighs the scale of AI companies' spending against reported revenue and considers whether infrastructure investment could outpace demand. Speakers also point to labor-market weakness and question how much AI investment is adding to jobs and economic activity. In markets, they focus on growing AI-related credit exposure, delayed profit expectations, and the risk that tighter lending could expose investors to a sharp reversal.

Artificial IntelligenceAI InfrastructurePrivate CreditLabor MarketEconomic GrowthRisk Management

Preview · 11 more points in the app

Read the Full Summary

Key Takeaways

The reported company figures show very different financial profiles: Anthropic had less than $5 billion in revenue while losing $40 billion and making spending commitments in the hundreds of billions, while OpenAI’s annualized revenue was approaching $50 billion at the end of September. The OpenAI estimate of $70 billion included cloud-platform sales that the company excludes, so the accounting basis matters when comparing the figures.

The speakers argue that AI infrastructure could be overbuilt as companies add capacity and redundancy ahead of demand, leaving some assets unused—a risk they compare with unused railroad track from the railroad bubble. They also caution that AI may work technologically while investors still end up on the wrong side of the trade, and say the timing of a reversal is unpredictable.

One speaker says the labor market has stagnated since mid-2024, hourly earnings are declining, and payrolls could turn negative in the next month or two; the speaker also argues that AI investment is not creating jobs or additional economic activity. The same speaker says rising markets can nevertheless support activity when baby boomers spend more, while corrections prompt them to stay home.

TECHNOLOGY

Anthropic Spending — The transcript reports that Anthropic had less than $5 billion in revenue while losing $40 billion, alongside spending commitments in the hundreds of billions. Those figures point to a wide gap between reported revenue and planned spending.

“We've already seen reports of Anthropic generating less than 5 billion in revenue while losing 40 billion, having already made spending commitments measured in the hundreds of billions.”—

OpenAI Revenue — The speaker says OpenAI's annualized revenue was approaching $50 billion at the end of September, about $20 billion below the widely circulated $70 billion estimate from its backers. That higher figure included cloud-platform sales that OpenAI excludes, so the accounting basis matters when comparing estimates.

“According to the Financial Times, Open AAI told investors its annualized revenue was approaching 50 billion at the end of September…”—

ECONOMY

Labor Weakness — One speaker says the labor market has gone nowhere since mid-2024 and that payrolls could turn negative in the next month or two. The speaker also cites declining hourly earnings as evidence of labor-market weakness.

“this time, you know, as you said, if if we need it now because if we get it later, the problem is the labor market…”—

AI And Jobs — One speaker says AI investment is real, including spending on data centers, chips, hardware, and software, but argues it is not creating jobs or additional economic activity. The speaker adds that a reversal would therefore not mean the sudden loss of millions of AI-created jobs.

“The AI boom is real. There's real money being spent. There's real projects being done. you know, data centers are being built…”—

MARKETS

AI Debt Exposure — One speaker warns that credit markets are becoming heavily exposed to AI, with incoming deals increasingly involving AI loans. Some pension funds and institutional investors are screening deals to avoid excessive concentration. The speaker estimates about $1 trillion in debt is tied to AI now, potentially rising to $3 trillion or $4 trillion in a year, and warns that a reversal after such a buildup could cause much greater damage.

“It's getting to the point where the bond market is heavily exposed to AI and basically nothing else…”—

Credit-Cycle Risk — One speaker identifies dependence on the credit cycle as a major immediate risk for AI, rather than whether the technology works or is being used. Lenders could refuse to finance companies with negative cash flow even while their data centers remain unfinished.

“We're now actually realizing that your cash flow is negative, and we don't want to lend you money." And they're meanwhile, they're saying, "But the data center is half built."…”—

AI Concentration — The speaker argues that financial-market participants see little alternative to AI exposure, leaving the financial system concentrated in the AI trade. The investments, the speaker says, are based on assumptions about a future economy rather than today's economy.

“So even though people said we would never do this again with the dotcoms or after the dotcoms, you really don't have any choice…”—

AI Overcapacity Risk

The speakers compare AI spending with past bubbles, arguing that huge infrastructure commitments may exceed real demand and that reported revenue distinctions warrant skepticism.

AI Credit Concentration

They warn that lenders, institutional investors, and structured credit markets are increasingly exposed to AI, making the financial system vulnerable if expected returns fail to appear.

Boomer Wealth And Lending

The discussion links market-driven wealth to retiree spending and says banks are lending heavily to major technology firms despite growing concentration risks.

Credit Cycle Warning

The speakers explain that falling long-term Treasury yields can signal weaker growth expectations and tighter credit, potentially making it harder for AI companies to finance unfinished projects.

AI Debt Feedback Loop

They describe how rising risk aversion, widening spreads, and reduced lending could undermine delayed profit projections and trigger asset sales across concentrated markets.

Dot-Com Bubble Echoes

The speakers attribute bubble dynamics to human nature and unknowable future projections, concluding that AI investments face familiar risks even if the technology succeeds.

Related Episodes

Important Disclaimers

Not Investment Advice: For education only — not financial, tax or legal advice.

AI-Generated: Made by AI from source transcripts and may contain errors.

No Guarantees: Past performance does not predict future results; investing carries risk.

Our Commentary · Request Removal