Technology consultant Jeffrey Funk warned here Wednesday that the AI financial bubble is finally popping.
Yesterday, a massive tech stock selloff wiped $2.5 trillion from market value (CNN). There’s been an explosion of news on the topic in the last week and here are some highlights:
Why is the bubble bursting?
At Gizmodo, Bruce Gil traces the “Wall Street bloodbath” to growing doubt about the huge investment in AI relative to foreseeable returns:
Alphabet [Google], for example, had its worst day ever on the stock market Thursday, with shares plunging more than 12% and wiping roughly $500 billion off the company’s market capitalization.
The dramatic selloff came a day after Alphabet reported disappointing third-quarter earnings. The tech giant reported Wednesday after the market closed that it generated $127.6 billion in revenue, falling short of Wall Street’s expectations. (October 30, 2026)
The question comes up more and more: Looking beyond the hoopla, does AI really increase productivity? The Bank of England says no and generally, employees are not using it.
In many regions, anyone who has phoned a big company recently about a complex problem must allow considerable extra time to get through the wall of “AI assistants” in order to connect with someone who knows how the business works and has access to the history of transactions. That adds greatly to consumer frustration. The AI as such does not create this problem; the attempt to use AI to reduce the expense of providing responsive customer service is the culprit here.
Not too surprisingly, chatbot makers like Anthropic are also coming under fire. How much are most people really willing to pay for artificial conversation with a bot? The question for the financial markets is not whether tech writers, celebs, or anyone else find the bots amusing (or amazing). No, the question is whether the bots will ever provide a worthwhile return on an enormous investment (ROI).
Opposition to data centers ramps up
Political changes, particularly the rejection of data centers for high speed transmission play a big role too. Last Tuesday has been called the worst post-midterms day for the market since 1896 and many think the centers are a key cause. They were rebuffed on all but one ballot.
We learn from CNN,
Key to the sell-off is Texas’ gubernatorial Democratic victory, given the state has played a crucial role in accelerating AI advancements, with the likes of Google, Meta, Oracle and OpenAI making massive investments in data centers there. Texas Governor-elect Gina Hinojosa campaigned on a temporary data center moratorium and eliminating certain tax incentives for businesses building data centers in the state. (November 5, 2026)
At Fortune, tech editor Alexei Oreskovic predicts that the decisive rejection of data centers during the U.S. midterms “will bring the AI boom to a premature halt, or at least, a significant slowdown”:
Meta President Dina Powell McCormick, speaking at a conference organized by the Wall Street Journal on Wednesday, acknowledged the new hurdles. Building data centers in the U.S. is “going to be harder,” she said. “I can also tell you that every hyperscaler is now thinking about ‘Do we need to build outside the United States?’ Because they’re enormous investments, hedging those investments is a wise thing to do.” (November 5, 2026)
That’s been bad news for chip makers like Intel (“shares fell 29% on Friday, posting the stock’s worst day since the 1970s.”) Oreskovic notes that, in general, “Some of the biggest names in tech took it on the chin, with Facebook parent company Meta down 3.4%, Microsoft down 2.8%, and Nvidia down 8%.”
Some politicians have knives out for big AI figures like Elon Musk. At Forbes, Alison Durkee reports
Democrats have particularly singled out DOGE as a clear target for any initial investigations, as Musk’s cost-cutting initiative in the early days of Trump’s second term led to the gutting of the U.S. Agency for International Development and other government agencies, widespread government layoffs and numerous privacy concerns over the sharing of Americans’ data. ( November 5, 2026)
All that is early days yet
Going forward, the big question is, are people willing to live with, for example, slower internet in order to avoid data centers? Or will they consider a responsible environmental approach that incorporates them so they can have the benefits of AI? Perhaps there are alternatives to data centers that do not require this choice. But they will require research and development (R&D).
The controversy currently generates enough heat that it isn’t yet clear how many people understand the underlying issue: Everybody seems to want powerful internet when it helps them but no one wants data centers.