Teahose.
SIGN IN
NEW HERE — WHAT TEAHOSE DOES
We read the entire AI & tech firehose — so you don't have to.
PODPodcastsAll-In, No Priors, Acquired…
NEWNewslettersStratechery, Newcomer…
PAPPapersPhysical AI research
PHProduct Huntdaily launches
VCInvestor ScoutSequoia, a16z, Benchmark…
CLAUDE DISTILLS →
7 reads, 30 sec each — free, 6 AM ET.
+ a live graph of the companies, people & themes underneath.
HOME/AXIOS AI+/🫧 Bubble trouble
NEWS
// NEWSLETTER ISSUE
AXIOS AI+

🫧 Bubble trouble

DATE November 16, 2026SOURCE AXIOS AI+PARTICIPANTS AXIOS AI+
// SUMMARY

1. Key Themes

Theme 1: AI is real, which is precisely what makes a financial bubble possible

Technology success and financial bubbles are not mutually exclusive

The core thesis is that the AI boom's legitimacy doesn't protect investors from a market reckoning. Multiple experts frame the dot-com analogy.

  • "No sane person questions whether AI works. And that 'working' status is precisely what makes a bubble possible," veteran technology investor Paul Kedrosky told Axios.
  • "I have no doubt that AI will be transformative. That does not preclude a financial bubble," MIT economics professor Ricardo Caballero told Axios.
  • "History shows that world-changing technology and devastating financial bubbles often go hand in hand. A market reckoning won't stop the AI revolution. But as with the dot-com crash, it could vaporize trillions in investor wealth as expectations come crashing back to earth."

Value creation for users does not guarantee returns for infrastructure investors

  • "AI can deliver enormous value to users while the companies building the infrastructure earn far less than their investors expected," Columbia Business School professor Stijn Van Nieuwerburgh told Axios.

Theme 2: Capex is outrunning revenue, and the financing structure is fragile

The spending-to-revenue gap is large

  • "AI providers are projected to spend $7 trillion on infrastructure through 2030, while Gartner forecasts no more than $3.6 trillion in related revenue over the same period."

Circular and debt-based financing amplifies risk

  • "The risks have grown as technology companies commit trillions to infrastructure, increasingly relying on debt, investments in one another and complicated financing arrangements to keep the expansion going."
  • "Rising borrowing costs and an avalanche of new bond offerings have made lenders wary."

Reflexivity could turn a correction into a crisis

  • "A retreat in AI investment could become self-reinforcing, turning a market correction into a much broader financial reckoning."

Theme 3: Valuation and sentiment are already resetting

Private-market expectations are being marked down

  • "Anthropic's newly announced IPO terms imply a valuation well below earlier hopes of as much as $2 trillion. OpenAI also reportedly secured less cash and a lower valuation than it had hoped for in a recent funding round."

Retail "buy the dip" behavior has flipped

  • "Retail traders, who have routinely bought previous AI-related market dips, instead became net sellers of U.S. stocks during the sharp Nov. 5 sell-off, according to Vanda Research."

Theme 4: Political and regulatory backlash is becoming a material risk

Data center and AI skepticism is winning elections

  • "A wave of election victories for AI-skeptical candidates, along with widespread voter support for data center restrictions, threatens to delay construction, increase costs and undermine the assumptions behind some projects."

Chip export restrictions are gaining bipartisan momentum

  • "Huang's downfall has intensified bipartisan demands for tighter restrictions on advanced chip exports to China, which increasingly appear likely to pass."

The right is trying to split Trump from Big Tech

  • "Right-wing players are hoping the bombshell revelations about Huang's affairs will chip away at the president's embrace of the tech industry and offer an opening for anti-tech voices in his administration."
  • "They view the midterm elections as proof that people across the political spectrum hate AI and data centers and view Silicon Valley as too powerful."

Theme 5: Key-person and governance risk at the center of the ecosystem

Nvidia's leadership crisis is a systemic risk

  • RSM chief economist Joe Brusuelas told Axios that Huang's departure could trigger a "wholesale bursting" of the AI bubble.
  • "Treasury Secretary Scott Bessent was having none of it yesterday, urging Nvidia to hire an outside CEO."
  • "Allies of Huang are plotting for him to return to the top spot, apparently seeing an opening for a Sam Altman-style comeback."

2. Contrarian Perspectives

Tighter export controls could help U.S. AI developers even as they hurt Nvidia

The consensus assumes chip restrictions are purely negative for the AI trade. The article notes a winner within the damage.

  • "Tighter U.S. export controls could also benefit American AI developers by limiting Chinese rivals' access to advanced computing power, even as they complicate Nvidia's business prospects."
  • Supporting evidence: Anthropic's run rate rose to $81 billion from $71 billion in a month, and OpenAI's quarterly revenue grew roughly 50% to $10 billion, suggesting model developers have strong demand independent of Nvidia's China sales.

Revenue growth and a bubble can coexist

Strong fundamentals are often read as an all-clear. The article argues the opposite: demand evidence is exactly what sustains over-investment.

  • "Anthropic reported an annualized revenue run rate of $81 billion at the end of October, up from $71 billion a month earlier." Yet "OpenAI... grew quarterly revenue roughly 50% to $10 billion in the third quarter... even as its adjusted operating losses reached $10.9 billion."
  • Reasoning: the issue is capital intensity versus returns, not usage. "Spending is outpacing expected returns."

Trump's China posture may be naive, not hawkish

Conservative thinkers argue that Trump's "transactional" view of China is the vulnerability that Nvidia's ties exposed.

  • Trump's perception of China "has always been entirely commercial and transactional. His view was, 'We're getting screwed. I'm going to get a better deal,'" said Oren Cass.
  • Cass argued China "is not a commercial partner but an adversary that is going to 'subvert the American political process in ways that end up humiliating the president.'"

3. Companies Identified

Nvidia

  • Description: Leading AI chip maker, now in leadership crisis.
  • Why mentioned: Huang's downfall is the proximate catalyst for bubble fears and for the export-control push.
  • Quotes: "The downfall of Nvidia CEO Jensen Huang adds an explosive new risk to an already-strained AI boom, threatening to burst a financial bubble even if the technology ultimately succeeds."

Anthropic

  • Description: AI lab preparing an IPO.
  • Why mentioned: Evidence of both real demand and valuation reset.
  • Quotes: "Anthropic reported an annualized revenue run rate of $81 billion at the end of October, up from $71 billion a month earlier." Also: "Anthropic's newly announced IPO terms imply a valuation well below earlier hopes of as much as $2 trillion."

OpenAI

  • Description: Leading AI lab, still deeply loss-making.
  • Why mentioned: Case study in rapid growth alongside heavy losses and a disappointing fundraise.
  • Quotes: "OpenAI also reportedly secured less cash and a lower valuation than it had hoped for in a recent funding round." Also: "grew quarterly revenue roughly 50% to $10 billion in the third quarter... even as its adjusted operating losses reached $10.9 billion."

Gartner

  • Description: Research and advisory firm.
  • Why mentioned: Source of the revenue forecast that contrasts with projected spend.
  • Quotes: "Gartner forecasts no more than $3.6 trillion in related revenue over the same period."

Vanda Research

  • Description: Investment research firm tracking retail flows.
  • Why mentioned: Data point on retail sentiment turning.
  • Quotes: "Retail traders... instead became net sellers of U.S. stocks during the sharp Nov. 5 sell-off, according to Vanda Research."

Megaspeed

  • Description: Entity referenced in the New York Times' reporting on the Nvidia saga.
  • Why mentioned: Origin point of the NYT's narrative on the scandal.
  • Quotes: The NYT offered "a six-byline, comprehensive narrative of the entire saga, going back to last year when the paper asked about Megaspeed."

American Compass

  • Description: Conservative think tank.
  • Why mentioned: Source of anti-Nvidia, China-hawk framing.
  • Quotes: "Oren Cass, founder of conservative think tank American Compass."

4. People Identified

Jensen Huang

  • Description: Former Nvidia CEO, central figure in the scandal.
  • Why mentioned: His downfall is the catalyst for market and political risk.
  • Quotes: "Huang's affair with an alleged Chinese spy." Also: "Allies of Huang are plotting for him to return to the top spot, apparently seeing an opening for a Sam Altman-style comeback."

Paul Kedrosky

  • Description: Veteran technology investor.
  • Why mentioned: Articulated the "working technology enables a bubble" thesis.
  • Quotes: "No sane person questions whether AI works. And that 'working' status is precisely what makes a bubble possible."

Joe Brusuelas

  • Description: RSM chief economist.
  • Why mentioned: Warned of the macro impact of Huang's exit.
  • Quotes: Huang's departure could trigger a "wholesale bursting" of the AI bubble.

Ricardo Caballero

  • Description: MIT economics professor.
  • Why mentioned: Historical framing of transformative tech and bubbles.
  • Quotes: "I have no doubt that AI will be transformative. That does not preclude a financial bubble."

Stijn Van Nieuwerburgh

  • Description: Columbia Business School professor.
  • Why mentioned: Distinguished user value from investor returns.
  • Quotes: "AI can deliver enormous value to users while the companies building the infrastructure earn far less than their investors expected."

Oren Cass

  • Description: Founder of American Compass.
  • Why mentioned: Leading conservative critic of Nvidia and Trump's China approach.
  • Quotes: "Cass, who has warned about Nvidia risks for more than a year, said the allegations of espionage are the strongest evidence yet that China is not a commercial partner but an adversary."

Steve Bannon

  • Description: MAGA figure.
  • Why mentioned: Pushing for a chip cutoff and tying Huang to Chinese influence.
  • Quotes: "We need to actually cut OFF ALL THE CHIPS NOW."

Scott Bessent

  • Description: Treasury Secretary.
  • Why mentioned: Publicly urged Nvidia to bring in an outside CEO.
  • Quotes: "Treasury Secretary Scott Bessent was having none of it yesterday, urging Nvidia to hire an outside CEO."

Kanishka Narayan

  • Description: U.K. AI minister.
  • Why mentioned: Leading Britain's pitch for AI hardware and manufacturing.
  • Quotes: "Britain is back in business when it comes to AI."

Sam Altman

  • Description: OpenAI CEO.
  • Why mentioned: Cited as the precedent for a comeback after ouster.
  • Quotes: "apparently seeing an opening for a Sam Altman-style comeback."

5. Operating Insights

Stress-test your business against a funding and sentiment downturn

Capital markets are tightening for AI-adjacent companies, so assume less favorable terms.

  • "Financing is getting harder... Rising borrowing costs and an avalanche of new bond offerings have made lenders wary."
  • "OpenAI also reportedly secured less cash and a lower valuation than it had hoped for in a recent funding round."

Build for demand that is real, and watch your dependence on political-risk infrastructure

Customer demand is validated, but data center and chip supply chains face regulatory and electoral friction.

  • "Businesses are already using AI to write software, automate tasks and perform work that previously required significant human effort."
  • Data center restrictions "threaten to delay construction, increase costs and undermine the assumptions behind some projects."

Diversify supply chains and allies rather than relying on one vendor or one country

Britain's approach models strategic leverage without total dependence.

  • "Rather than attempting to recreate the entire AI industry domestically, Britain wants to build strength in areas where it can gain leverage while relying on a diversified group of allies for other technologies."

6. Overlooked Insights

Hardware and manufacturing as a new geopolitical opening

The U.K. is explicitly targeting AI hardware as a supply-chain play, a niche that could create opportunities for allied manufacturers if export controls tighten.

  • He said Britain wants "a larger share of AI hardware and manufacturing, both to create British jobs and to strengthen the resilience of the U.S. and U.K. technology supply chains."

The U.K. is positioning a third regulatory path

A middle-ground regime between EU legislation and U.S. laissez-faire could attract AI companies seeking predictability.

  • Its approach "offers a contrast to the European Union's sweeping AI legislation and the largely hands-off approach favored by the Trump administration."