The Brief

Perhaps nobody embodies artificial intelligence mania quite like Jensen Huang, the chief executive of chip behemoth Nvidia, which has seen its value spike 300% in the last two years.

A frothy time for Huang, to be sure, which makes it all the more understandable why his first statement to investors on a recent earnings call was an attempt to deflate bubble fears.

"There's been a lot of talk about an AI bubble," he told shareholders. "From our vantage point, we see something very different."

Take in the AI bubble discourse and something becomes clear: Those who have the most to gain from artificial intelligence spending never slowing are proclaiming that critics who fret about an over-hyped investment frenzy have it all wrong.

"I don't think this is the beginning of a bust cycle," White House AI czar and venture capitalist David Sacks said on his podcast All-In. "I think that we're in a boom. We're in an investment super-cycle."

White House AI adviser David Sacks speaks onstage during The Bitcoin Conference at The Venetian Las Vegas in January.

(

Ian Maule

/

AFP via Getty Images

)

"The idea that we're going to have a demand problem five years from now, to me, seems quite absurd," said prominent Silicon Valley investor Ben Horowitz, adding: "if you look at demand and supply and what's going on and multiples against growth, it doesn't look like a bubble at all to me."

Appearing on CNBC, JPMorgan Chase executive Mary Callahan Erdoes said calling the amount of money rushing into AI right now a bubble is "a crazy concept," declaring that "we are on the precipice of a major, major revolution in a way that companies operate."

Yet a look under the hood of what's really going on right now in the AI industry is enough to deliver serious doubt, said Paul Kedrosky, a venture capitalist who is now a research fellow at MIT's Institute for the Digital Economy.

He said there is a startling amount of capital pouring into a "revolution" that remains mostly speculative.

"The technology is very useful, but the pace at which it is improving has more or less ground to a halt," Kedrosky said. "So the notion that the revolution continues with the same drum beat playing for the next five years is sadly mistaken."

The huge infusion of cash

The gusher of money is rushing in at a rate that is stunning to financial experts.

Take OpenAI, the ChatGPT maker that set off the AI race in late 2022. Its CEO Sam Altman has said the company is making $20 billion in revenue a year, and it plans to spend $1.4 trillion on data centers over the next eight years. That growth, of course, would rely on ever-ballooning sales from more and more people and businesses purchasing its AI services.

There is reason to be skeptical. A growing body of research indicates most firms are not seeing chatbots affect their bottom lines, and just 3% of people pay for AI, according to one analysis.

"These models are being hyped up, and we're investing more than we should," said Daron Acemoglu, an economist at MIT, who was awarded the 2024 Nobel Memorial Prize in Economic Sciences.

"I have no doubt that there will be AI technologies that will come out in the next ten years that will add real value and add to productivity, but much of what we hear from the industry now is exaggeration," he said.

Nonetheless, Amazon, Google, Meta and Microsoft are set to collectively sink around $400 billion on AI this year, mostly for funding data centers. Some of the companies are set to devote about 50% of their current cash flow to data center construction.

Or to put it another way: every iPhone user on earth would have to pay more than $250 to pay for that amount of spending. "That's not going to happen," Kedrosky said.

To avoid burning up too much of its cash on hand, big Silicon Valley companies, like Meta and Oracle, are tapping private equity and debt to finance the industry's data center building spree.

Paving the AI future with debt and other risky financing

One assessment, from Goldman Sachs analysts, found that hyperscaler companies — tech firms that have massive cloud and computing capacities — have taken on $121 billion in debt over the past year, a more than 300% uptick from the industry's typical debt load.

Analyst Gil Luria of the D.A. Davidson investment firm, who has been tracking Big Tech's data center boom, said some of the financial maneuvers Silicon Valley is making are structured to keep the appearance of debt off of balance sheets, using what's known as "special purpose vehicles."

An aerial view of a 33 megawatt data center with closed-loop cooling system in Vernon, California.

(

Mario Tama

/

Getty Images

)

The tech firm makes an investment in the data center, outside investors put up most of the cash, then the special purpose vehicle borrows money to buy the chips that are inside the data centers. The tech company gets the benefit of the increased computing capacity but it doesn't weigh down the company's balance sheet with debt.

For example, a special purpose vehicle was recently funded by Wall Street firm Blue Owl Capital and Meta for a data center in Louisiana.

The design of the deal is complicated but it goes something like this: Blue Owl took out a loan for $27 billion for the data center. That debt is backed up by Meta's payments for leasing the facility. Meta essentially has a mortgage on the data center. Meta owns 20% of the entity but gets all of the computing power the data center generates. Because of the financial structure of the deal, the $27 billion loan never shows up on Meta's balance sheet. If the AI bubble bursts and the data center goes dark, Meta will be on the hook to make a multi-billion-dollar payment to Blue Owl for the value of the data center.

Such financial arrangements, according to Luria, have something of a checkered past.

"The term special purpose vehicle came to consciousness about 25 years ago with a little company called Enron," said Luria, referring to the energy company that collapsed in 2001. "What's different now is companies are not hiding it. But having said that, it's not something we should be leaning on to build our future."

Enormous spending hinging on returns that could be a fantasy

Silicon Valley is taking on all this new debt with the assumption that massive new revenues from AI will cover the tab. But again, there is reason for doubt.

Morgan Stanley analysts estimate that Big Tech companies will dish out about $3 trillion on AI infrastructure through 2028, with their own cash flows covering only half of that.

"If the market for artificial intelligence were even to steady in its growth, pretty quickly we will have over-built capacity, and the debt will be worthless, and the financial institutions will lose money," Luria said.

Twenty-five years ago, the original dot-com bubble burst after, among other factors, debt financing built out fiber-optic cables for a future that had not yet arrived, said Luria, a lesson, it appears, tech companies are not worried about repeating.

"If we get to the point after spending hundreds of billions of dollars on data centers that we don't need a few years from now, then we're talking about another financial crisis," he said.

Circular deals raise even more concern

Another aspect of the over-heated AI landscape that is raising eyebrows is the circular nature of investments.

Take a recent $100 billion deal between Nvidia and OpenAI.

Nvidia will pump that amount into OpenAI to bankroll data centers. OpenAI will then fill those facilities with Nvidia's chips. Some analysts say this structure, where Nvidia is essentially subsidizing one of its biggest customers, artificially inflates actual demand for AI.

"The idea is I'm Nvidia and I want OpenAI to buy more of my chips, so I give them money to do it," Kedrosky said. "It's fairly common at a small scale, but it's unusual to see it in the tens and hundreds of billions of dollars," noting that the last time it was prevalent was during the dot-com bubble.

Open AI CEO Sam Altman speaks during Snowflake Summit 2025 at Moscone Center in June.

(

Justin Sullivan

/

Getty Images

)

Lesser-known companies are getting in on the action, too.

CoreWeave, once a crypto mining startup, pivoted to data center building to ride the AI boom. Major AI companies are turning to CoreWeave to train and run their AI models.

OpenAI has entered deals with CoreWeave worth tens of billions of dollars in which CoreWeave's chip capacity in data centers is rented out to OpenAI in exchange for stock in CoreWeave, and OpenAI, in turn, could use that stock to pay its CoreWeave renting fees.

Nvidia, meanwhile, which also owns part of CoreWeave, has a deal guaranteeing that Nvidia will gobble up any unused data center capacity through 2032.

"The danger," said the MIT economist Acemoglu,"is that these kinds of deals eventually reveal a house of cards."

Some high profile investors see bubble-popping on the horizon

Some influential investors are showing signs of bubble jitters.

Tech billionaire Peter Thiel sold off his entire stake in Nvidia worth around $100 million earlier this month. That came after SoftBank sold a nearly $6 billion stake in Nvidia.

And in recent weeks, AI bubble pessimists have rallied around Michael Burry, the hedge-fund investor who made hundreds of millions of dollars betting against the housing market in 2008. He was the subject of the 2015 film The Big Short. Since then, though, he's had a mixed reputation for market predictions, having warned about imminent collapses that never came to pass.

For what it's worth, Burry is now betting against Nvidia, accusing the AI industry of hiding behind a bunch of fancy accounting tricks. He's homed in the circular deals between companies.

"True end demand is ridiculously small. Almost all customers are funded by their dealers," Burry wrote on X. He later wrote: "OpenAI is the linchpin here. Can anyone name their auditor?"

As tech companies sink billions into data centers, some executives themselves are freely admitting there looks to be some over exuberance.

OpenAI CEO Sam Altman told reporters in August: "Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes. Is AI the most important thing to happen in a very long time? My opinion is also yes."

And Google chief executive Sundar Pichai told the BBC recently that "there are elements of irrationality" in the AI market right now.

Asked how Google would fare if the bubble burst, Pichai responded: "I think no company is going to be immune, including us."

Copyright 2025 NPR

L.A. County Sheriff Robert Luna confirmed Monday that ICE will play a role in World Cup security, but said he's been told federal agents won't conduct immigration enforcement.

He made the comments at a news conference on law enforcement's plans for the tournament, and said he'd been speaking directly with the head of Homeland Security in the Los Angeles area.

"There will be federal agents," Luna said. " Because it's gonna take all of us to make sure that all the venues, the scoped and unscoped events, are secure."

SoFi Stadium is set to host eight tournament matches, including the U.S. team opener against Paraguay on June 12. Los Angeles will also host a historic match three days later when Iran is set to take the field in Inglewood, making the U.S. the first host nation in World Cup history to be at war with a participating country.

Luna said the federal government had said that civil immigration enforcement would not occur at the tournament. But he made no guarantees.

" They told us that specifically would not be occurring at any of the games. Any of that's subject to change," he said. "But I have trust that they're giving me the appropriate information because if that starts occurring, we're gonna have a whole new host of problems."

In a statement to LAist, Assistant Secretary Lauren Bis wrote that Department of Homeland Security is working with federal, state, local and international partners.

“The safety and security of the American people and the millions of visitors attending these events remain our highest priority," Bis wrote in an email. "DHS will continue leveraging every available authority, technology, and partnership to protect the Homeland while ensuring the World Cup remains safe, secure, and successful for everyone involved.”

Luna is the latest official to confirm that U.S. Immigration and Customs Enforcement will play a role in the tournament. Kathryn Schloessman, who leads L.A.'s World Cup host committee, told reporters last month that ICE would be at the World Cup, and that its presence was typical at these types of major events.

ICE has two main branches: Enforcement and Removal Operations, which detains and deports people, and Homeland Security Investigations, which conducts international criminal investigations.

Todd Lyons, the former head of ICE, said at a congressional hearing earlier this year that it would be ICE’s investigatory branch, not its enforcement division, playing a key role in World Cup security.

Still, some in L.A. aren't satisfied. The World Cup is coming to Los Angeles at exactly the year mark since immigration agents ramped up arrests in the region. Masked agents in neighborhoods across the county sparked protests and widespread fear, and ICE arrests in the L.A. area last year tripled.

SoFi Stadium workers represented by Unite Here Local 11 have threatened to strike over ICE's role in the tournament. They'll vote on whether or not to authorize a strike later this week.

California air regulators on Friday approved a contentious overhaul of the state’s carbon market, creating a program that could steer billions of dollars in free pollution permits to oil refineries and other major polluters over the objections of environmental groups, key lawmakers and three of the board’s own members.

Ten members of the California Air Resources Board voted to adopt the changes to its cap-and-invest program after two days of lengthy hearings, including a full day dedicated to hundreds of public comments.

The overhaul followed intensive lobbying by the oil industry as well as pressure from Gov. Gavin Newsom’s administration to help keep refineries operating in the state amid rising gas prices.

The approval sets up a potential budget fight in Sacramento. The Legislative Analyst’s Office projects that quarterly auction revenue for state climate programs will drop from roughly $4 billion a year to about $2 billion under the new overhaul.

Such a shortfall would effectively zero out programs lawmakers spent last year fighting to fund: affordable housing, public transit, drinking water in low-income communities and pollution monitoring in California’s most polluted neighborhoods.

The governor’s office praised the measure as a compromise that balanced economic uncertainty with the state’s climate goals. Refinery closures and the Iran-Israel war have driven average California gas prices above $6 a gallon.

Newsom, in a statement, used the moment to draw a contrast with President Donald Trump.

“While Trump sows ongoing chaos and uncertainty, California is staying focused by protecting our economy, safeguarding public health, and doubling down on the clean energy future all Californians deserve,” he said.

Environmentalists warned the changes to the program amount to a giveaway to the fossil fuel industry that weakens California’s only program setting a firm cap on greenhouse gas emissions.

Katelyn Roedner Sutter, California senior director for the Environmental Defense Fund, called the decision “deeply misguided” for prioritizing polluters over communities.

“Newsom’s air regulators are handing billions to oil executives at the expense of our climate, health, and affordability for working families in a rushed process that has shortchanged meaningful public participation,” said Bahram Fazeli, policy director at Communities for a Better Environment.

How the program works — and what changes

California’s 13-year-old carbon market forces major polluters to buy permits while the state lowers the overall cap each year. Friday’s vote will reduce those permits – and creates a new subsidy program carved out of the market.

The program, which may still see changes, could make available a new pool of free pollution permits available to industry valued at as much as $4 billion. Companies that pledge to invest in clean energy and efficiency may qualify for the permits in exchange for investments in clean energy.

The pool will be capped at 118.3 million permits — the same number the air board has said must come off the market for California to hit its 2030 climate target. Environmentalists say the proposal risks wiping out those reductions.

Half are reserved for the fossil fuel sector. A recent Berkeley analysis, by the chair of an independent committee that oversees the carbon market, found refineries could end up with more free permits than they need to cover their emissions.

The air board has defended the design. Officials say the credits will go only to companies undertaking decarbonization projects, will be limited and temporary and can be clawed back if companies misuse them. The plan, they say, is meant to keep California refineries operating at a time of mounting closures and global market pressure. According to air regulators, the amended program will spur clean-energy investment as Trump cuts federal support.

Here's what you should know about the vote totals currently released:

Keep in mind that June 9 will be the final day for votes postmarked by June 2 to arrive at county elections offices, so the bottom line on the vote totals won't be known until then.

In L.A. County, the combined tallied votes as of Sunday add up to about 10% of registered voters.

In Orange County, the current tallies represent about 22% of registered voters.

How vote counts will be released

I thought it was an election NIGHT?

That hasn't been true in quite a while. It takes a while to get results because after the initial tallies on election night, there are still many, many votes to count and more mail-in ballots are usually arriving.

Here’s what we know so far:

L.A. County turnout

Los Angeles County has more than 5.8 million registered voters. As of Sunday, May 31:

• 580,720 ballots have been processed
• 95% voted by mail
• 5% voted in person

What's next:

• We'll know more on election night and the following days how many ballot remain to be counted.
• Has your ballot been tallied yet? You can track the status of your ballot here.

Orange County turnout

Orange County has more than 1.8 million registered voters. As of Sunday, May 31:

• 401,865 ballots have been processed
• 95% voted by mail
• 5% voted in person

What's next:

• We'll know more on election night and the following days how many ballot remain to be counted.
• Has your ballot been tallied yet? You can track the status of your ballot here.

Expected total turnout

Political Data Inc. is tracking ballot returns across California and in some high-profile races.

As of midday Monday, turnout statewide was at 16%. While Democrats outnumber Republicans statewide by almost double, Republicans have returned more ballots pre-election (21% of their voters compared to 16% for Democrats).

See the latest totals ▶

Why election day has turned into ballot-counting month

Because of the increasing use of vote-by-mail ballots, the vote count has gotten longer, according to the California Voter Foundation. In an analysis, the organization found:

• In November 2004, more than 80% of votes were counted within two days of Election Day, with 32.6% voting by mail. 
• In June 2022, about 50% of ballots were counted within two days of Election Day, with more than 90% of people voting by mail. 
• In November 2024, 66% of votes were counted within the first two days of Election Day, with 81% of the vote by mail.

A closer look at ballot counting times in California where an increasing number of vote-by-mail ballots has slowed ballot counts.

(

Courtesy California Voter Foundation

)

Election officials must physically open mail-in ballots and verify signatures.

Kim Alexander, president of the California Voter Foundation, recently wrote about the ripple effect of turning in mail-in ballots by hand or in drop boxes on election day. She wrote for our partner newsroom CalMatters:

"We turn in ballots in envelopes on Election Day that take time and care to process and cannot be processed until after Election Day. Processing these ballots — which account for as much as a quarter of all ballots cast — creates a bottleneck I like to call 'the pig in the python effect'. It prevents counties from doing other tasks they need to do to certify the results."

The buzz on college campuses is that AI is disrupting the job market for young college graduates.

But new research from the Federal Reserve Bank of New York finds that the culprit may be something else: remote work.

An analysis of federal employment data, paired with a deep dive into the flexible work arrangements at one unnamed Fortune 500 tech company, reveals that companies are less likely to hire recent college grads into occupations that can be done remotely.

Researchers speculate that employers are reluctant to put such workers in a setting where it's harder to absorb lessons from coworkers.

The researchers found the unemployment rate among younger college grads — those under the age of 29 — rose 20% after the pandemic, while unemployment among older college grads fell slightly.

The study compares unemployment rates pre-pandemic, from 2017 to 2019, with unemployment rates after the pandemic, from 2022 to 2024.

Unemployment rose as remote work grew fourfold, the researchers write. "Our analysis suggests that these trends are related, with remote work making it more difficult for managers to train and mentor new employees."

Remote work leads to less feedback on the job

The research began with a look at how much feedback software engineers at a Fortune 500 tech company were getting, says Emma Harrington, an assistant professor of economics at the University of Virginia and one of the authors of the report.

"What we saw was this pretty striking pattern that software engineers got about 20% more feedback if they were sitting near their colleagues than if they were distant from them," she says, adding that that was true even before the pandemic.

But after the pandemic, feedback plummeted.

"And that really hit young workers much harder," says Harrington. "It was these people who had the most to learn that really saw this deficit in feedback."

The researchers then looked deeper into who was getting hired at the tech firm. Turns out, as the company embraced remote work, they switched away from hiring younger people.

"So they used to hire a bunch of new grads for their software engineering jobs," Harrington says. "Then they shifted really towards hiring much older people, like a decade older on average."

Later, the company pivoted again, implementing what Harrington calls a "pretty aggressive" return-to-office policy. At that point, the company resumed hiring new graduates.

"So [there was] some sense that these problems with mentorship were translating into whom this firm was deciding to hire," she says.

A look at the broader economy

The researchers then wanted to see if what was happening at that single tech company was playing out in the broader economy.

Using a widely-used index that measures how feasible it is to do a job from home, the team divided all occupations into two categories: "remotable," which included software engineering, and "non-remotable," which included mechanical engineering.

They found the gap in unemployment between recent graduates and older workers was significantly higher in "remotable" jobs than in jobs that have to be done in person.

The unemployment rate for younger grads in "remotable" jobs jumped by almost a full percentage point after the pandemic, while the unemployment rate among older grads fell marginally.

They concluded that remote work explained nearly two-thirds of the rise in unemployment among young graduates during this period.

"This relative increase in young people's unemployment coincided with the pandemic and has remained elevated since then, as have rates of remote work," the researchers write.

AI isn't disrupting so many jobs for recent college grads — yet

To see how the rise of AI chatbots may have contributed to rising unemployment among the younger set, the researchers used another index that divides occupations into those more exposed to AI, such as engineering and accounting, and those less exposed, such as teaching and nursing.

They found exposure to AI didn't explain the divergence in unemployment rates in the 2022-24 time period. Remote workflows were much more of a driving force, Harrington says, while emphasizing that this could change.

"It's always hard to make guesses about what's going to happen with generative AI," she says. "It's certainly possible that this story could really change over the next few years."

Researchers at the London School of Economics have reached a similar conclusion — that remote work is having a clearer impact on early-career hiring than AI — in a working paper examining new hires in the U.S., the U.K., Canada and Australia.

Regardless of the cause, the New York Fed report warns that a high unemployment rate among young college grads is concerning.

"Early-career experiences can have lasting consequences," the researchers write. "Research finds that individuals who began looking for jobs in slacker labor markets tend to have lower earnings and slower career progression relative to comparable peers who began their job search in better market conditions."

Copyright 2026 NPR