Behind the Curtain: The AI super-stimulant

Why it matters: A wild cycle is unfolding. The biggest companies in history are spending a stunning amount of money to fuel the AI revolution, driving demand for more chips, more energy and more capital.

• This dynamic is basically powerful enough to cover for an economy that otherwise looks like it's weakening by the day.

The big (Nvidia, Microsoft, Apple, Alphabet/Google, Amazon and Meta) are getting bigger, and burrowing deeper into direct investment and ownership of the products feeding into their AI.

• They're buying up land, building data centers, sucking up chips, investing in energy sources. Microsoft and Nvidia together are worth about $2.5 trillion more today than they were a year ago.

Alphabet, Google, Amazon and Meta together will spend nearly $400 billion this year on capital expenditures, largely to build AI infrastructure — more than the EU spent on defense last year, The Wall Street Journal notes.

• Investors are getting comfortable with the mammoth spending on this "capex war," as shown by Big Tech's runaway results last week, the Financial Times reports.

The big picture: The Wall Street Journal's Christopher Mims captured the tectonic impact of the AI arms race in a must-read column, "Silicon Valley's New Strategy: Move Slow and Build Things."

• "Call it an 'age of infrastructure,' in which companies spend vast sums on actual stuff," Mims wrote. "It's reminiscent of the age of business titans and 'robber barons' who dominated railroads, steel and other enterprises."
• Paul Kedrosky — a Ph.D. and trained engineer who became a tech founder, investor and pundit — found that capital expenditures on AI data centers have passed the peak of telecom spending during the dot-com bubble (1995-2000) — and are second only to railroads in the 1880s. "In a sense," Kedrosky writes, "there is a massive private sector stimulus program underway in the U.S."

Think about that: A small number of companies — some nearly the size of Germany's economy — are spending more than the 340 million American consumers, whose status as the engine of the U.S. economy has always been all but gospel.

• As Derek Thompson, co-author of the bestselling book "Abundance," put it after Friday's disappointing jobs report: "GDP is only growing because of AI capex."

Zoom in: Investment in information processing equipment and in software increased at a 25% annual rate in the first half of the year — while overall GDP rose at a paltry 1.2%, Axios chief economic correspondent Neil Irwin notes.

• Neil Dutta of Renaissance Macro notes on X that this measure of AI capital spending has contributed more to growth this year than consumer spending.

The big questions: Will U.S. workers — not just tech giants and big shareholders — benefit with better jobs and higher salaries? Will this boom ever bust?

• The Wall Street Journal's Greg Ip points to "hidden risks from the AI boom": "No one doubts its potential to raise growth and productivity in the long run. But financing that boom is straining the companies and capital markets."

• Econ blogger Noah Smith has an even more dire warning: With a "large and increasing amount of debt being used to fund one single sector of the economy (data centers)," a data-center bust could trigger a financial crisis.

Between the lines: Theoretically, heavy investment backed by generous government incentives should, over time, create a new class of higher-end work to build, operate and maintain data, chips and energy centers. But these jobs need to more than offset the jobs AI wipes out.

• "The AI layer is just the newest, the latest layer," Jay Timmons, president and CEO of the National Association of Manufacturers, tells Axios' managing editor for business Ben Berkowitz. "We have to be able to make sure we're ready for today with those skills."

Reality check: Tech in general, and semiconductors in particular, have always been cyclical businesses. First, you overbuild, then you suffer through a glut, then there's a shortage — and then you overbuild again, Axios managing editor for tech Scott Rosenberg notes.

• It happened with memory chips at the start of the '90s. It happened with internet connectivity during the dot-com boom and bust. Plenty of people in the AI industry — many of whom never lived through those cycles — believe this time is different.
• And data centers generate more direct jobs in their construction phase than once they're built.

The bottom line: No one disputes that the vast majority of America's growth right now flows from AI investment competition. And it's only accelerating.

• It's a steroid injected at a level and speed we've never seen.

Go deeper ... "Behind the Curtain — Jensen vs. Dario: "There will be more jobs."