Nov 9, 2026 · 41 min
Eisman warns AI spending could outrun returns
Big Short Investor Explains How the AI Bubble Could Burst
The episode tests whether AI’s economic importance can justify the concentrated, debt-heavy investment cycle powering its expansion.
- 1AI’s biggest risks may be excessive capital spending, financing complexity, and weak returns rather than technological failure.
- 2Power constraints, supply-chain concentration, and dependence on OpenAI and Anthropic expose the broader economy to a narrow group of companies.
- 3Eisman favors AI infrastructure suppliers while separating confidence in the technology from confidence in richly valued individual stocks.
Don't miss
Eisman connects OpenAI’s financial vulnerability to a much wider chain of exposure involving Oracle, hyperscalers, banks, industrial companies, and the U.S. economy.
The brief
Steve Eisman argues that AI has become the dominant force in markets and the U.S. economy, but extraordinary spending now rests on a small group of companies delivering extraordinary results.
The bull case sees AI as the largest technology buildout in history; the bear case centers on excessive capital spending, weak returns, debt, and a financing structure that magnifies disappointment.
The infrastructure is harder to build than the software suggests: data centers require chips, power, permits, and grid capacity, while community opposition and generation timelines can slow expansion.
OpenAI and Anthropic sit near the center of the risk because hyperscaler spending, projected economic growth, and several suppliers depend on their ability to sustain demand and revenue.
Eisman’s standout investment distinction is simple: belief in AI does not require belief in every AI stock, especially where valuations assume profits without durable customer lock-in.
What was said on this episode
39 statements · 6 positive · 28 negative · 5 neutral
The AI investment boom increasingly depends on a few companies succeeding.
“This entire AI investment boom has become increasingly dependent on the success of just a few companies.”
Listen at 1:12
The U.S. could enter recession and experience a large market correction.
“could easily enter a recession, and we could almost certainly see a large market correction.”
Listen at 1:28
A gigawatt-scale AI data center could cost about $30 billion.
“A new large AI data center capable of supporting a gigawatt of power demand could cost around, get this, $30 billion to build and equip.”
Listen at 4:34
Hyperscalers are expected to spend roughly $850 billion on AI infrastructure in 2026.
“In 2026 alone, hyperscalers, just the hyperscalers, are expected to spend roughly $850 billion on AI infrastructure”
Listen at 5:18
AI will change the world.
“AI will change the world.”
Listen at 7:06
The bullish AI scenario has a 30% probability.
“I give this narrative a 30% chance.”
Listen at 7:28
Chinese open-weight models will pressure prices and weaken AI investment returns.
“The Chinese open-weight models will put pressure on pricing, a price war will ensue, moats vanish, and the returns on investment don't measure up.”
Listen at 7:50
The bearish AI scenario has a 30% probability.
“I give this narrative also a 30% chance.”
Listen at 8:25
The intermediate AI scenario has a 40% probability.
“I give this narrative a 40% chance.”
Listen at 9:50
Machines taking over Earth is not foreseeable.
“Machines taking over the earth, like in Terminator, is not in the foreseeable future.”
Listen at 10:09
The AI doomsday scenario has approximately zero probability.
“I give this narrative about a 0% chance.”
Listen at 10:51
Market direction depends on AI’s future health and success.
“the direction of the market hinges on the future health and success of AI.”
Listen at 12:43
AI will transform software, medicine, and other sectors.
“AI will change it.”
Listen at 13:37
ServiceNow and Salesforce likely have prohibitive enterprise switching costs.
“some software companies are so embedded in the operations of enterprises that the switching costs are too prohibitive. I think that probably applies to companies like ServiceNow and Salesforce.”
Listen at 15:27
Private-equity investors are likely to suffer substantial losses.
“private equity investors are likely going to get hurt and hurt a lot.”
Listen at 17:21
Google raised $85 billion in equity capital for AI needs in 2026.
“Google raised $85 billion in equity capital in 2026 to help fund its AI needs.”
Listen at 19:07
Oracle’s total debt rose from $94 billion to $155 billion in two years.
“In the past two years, Oracle's total debt has increased from $94 billion to $155 billion.”
Listen at 19:37
AI capital spending has made hyperscaler business models capital-heavy.
“AI CapEx requirements have altered the business models of the hyperscalers from capital light, capital heavy.”
Listen at 19:52
Meta’s Beignet data-center debt is kept off its balance sheet.
“the debt will not sit on Meta's balance sheet.”
Listen at 21:45
More off-balance-sheet AI financing transactions are likely.
“you should expect many more such transactions in the future, and not just at Meta.”
Listen at 23:02
The AI investment thesis requires construction of AI data centers.
“The AI story cannot work unless AI data centers are built.”
Listen at 23:48
Permitting for AI data centers is delayed in many U.S. areas.
“the process of permitting ai data centers is being delayed in many parts of the country”
Listen at 24:23
Five customers accounted for 70% of NVIDIA’s accounts receivable.
“70% of NVIDIA's account receivable, which is a proxy for revenue, was from five customers.”
Listen at 26:09
OpenAI accounts for about half of Oracle’s backlog.
“around 50% of that backlog is from OpenAI alone.”
Listen at 26:57
Hyperscaler AI spending accounts for about half of projected 2026 GDP growth.
“that spending accounts for around half of the 2% GDP growth projected for 2026.”
Listen at 27:12
U.S. economic health is highly dependent on hyperscaler capital spending.
“the health of the U.S. economy is extremely dependent on hyperscaler CapEx.”
Listen at 27:20
Steve Eisman is not yet predicting OpenAI will fail.
“I'm not predicting, at least not yet, that OpenAI will fail.”
Listen at 30:50
OpenAI’s failure would push the U.S. into immediate recession.
“the demise of open AI would push the U.S. into an immediate recession.”
Listen at 32:37
Failure of Anthropic or OpenAI within a year would have severe consequences.
“If either company fails within, I believe, a year, the ramifications would be severe.”
Listen at 32:52
OpenAI failure would reduce hyperscaler CapEx and slow cloud revenue growth.
“Failure of OpenAI would mean they would pull back on CapEx. Cloud revenue growth would slow.”
Listen at 33:15
Investors will reallocate toward defensive sectors after AI weakness.
“Investors will reallocate to safety sectors and subsectors.”
Listen at 34:10
AI providers currently lack durable competitive moats.
“There don't seem to be any moats around AI.”
Listen at 35:08
Switching costs between AI models are likely very low.
“the switching cost between models seems to be perhaps zero, but certainly very low.”
Listen at 35:14
Chinese open-weight models cost substantially less than Anthropic and OpenAI models.
“open weight models are much cheaper than the models of anthropic and open ai”
Listen at 36:44
Anthropic and OpenAI cannot withstand an AI price war financially.
“Anthropic and OpenAI do not have the balance sheets to withstand a price war.”
Listen at 37:20
Trouble at Anthropic or OpenAI within a year would signal serious AI risks.
“if anthropic or open AI get into trouble in the next year, that would be telling.”
Listen at 37:53
Eisman prefers AI infrastructure suppliers over model providers.
“I'm much more interested in the picks and shovels.”
Listen at 38:29
Investors should be cautious with low-switching-cost businesses priced for huge future profits.
“Where I'd be much more careful is in businesses where investors are assuming enormous future profits despite very low switching costs.”
Listen at 38:53
AI stocks can disappoint if results fall short of expectations, even without technology failure.
“you don't necessarily need the technology to fail for the stocks to disappoint you just need reality to fall short of expectations”
Listen at 39:41
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.
Weekly email
State of the Pod, every Monday
The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.