Oil prices are tumbling now that the Strait of Hormuz is open, potentially averting a continued rise that could have led to broader, more persistent inflation.
But as energy prices plunge, the Federal Reserve may have a new inflation hurdle: artificial intelligence.
Technology giants from Amazon to Google are racing to build the foundation of AI, buying and making millions of specialized AI computing chips and building data centers packed with liquid-cooling systems to keep computers from overheating.
TD Cowen expects the major hyperscalers to spend $745 billion this year, with another $1 billion-plus in 2027 and 2028. According to their calculations, spending by these mega players is set to rise to approximately 3% of GDP next year, up markedly from under 0.5% in 2020.
The impacts are beginning to emerge in inflation data and in the rising demand for construction workers.
The components used to make the chips and build the infrastructure are in such high demand that they're causing ripples across other sectors and affecting prices. Memory and storage chips that are in high demand for AI are also used in consumer electronics such as video games, cars, and smartphones. Apple announced this past week that it's raising prices on iPads and MacBooks in response to higher costs for memory and storage. And the ramifications are likely to continue to grow.
"The initial phase of AI infrastructure will be inflationary as demand exerts pressure on a fixed supply side of the economy before it grows the pie through higher productivity growth, which may well end up as a disinflation driver in years to come," said Oscar Muñoz, head of economics for TD Securities.
This means that the neutral rate — the level of the Fed's benchmark interest rate designed to neither boost nor slow economic growth — is higher now, Muñoz said, and the Fed will have to decide whether to raise rates as a result.
"This poses a challenge for the Federal Reserve," Muñoz said. "AI's large capital needs, even if only lasting during the build-out phase, are lifting the shorter-run [neutral rate], regardless of if it proves to be disinflationary in the medium term."
What about the supposed productivity boost?
The developments run counter to what Fed Chairman Kevin Warsh and others have said: that AI will boost productivity, push down inflation, and allow the Fed to cut rates. In his debut press conference on June 17, Warsh didn't affirm that he still believes that, only noting the Fed has work to do on price stability, and that is the focus.
But Greg Daco, chief economist for EY, argued that there's a misunderstanding about technological revolutions — that they raise productivity and ease prices. He says that does happen, but not before prices rise first.
Phase one of any technological revolution comes with an increase in inflation because, Daco noted, it's highly capital-intensive. That initial investment wave leads to pressures on resources with a lot of demand chasing a limited supply.
"So long as we're in phase one, which is the investment phase, we're going to see these pressures," Daco said in an interview. "Over the next one to two years, you're going to see these inflationary pressures gradually pass through onto consumers and then abate, once we've reached the peak investment growth rates."
Then enter phase two, when stronger productivity kicks in. Daco said once the infrastructure and talent are in place, that will drive productivity growth.
"It takes time for companies to adopt the technology, incorporate it into the existing processes, create new processes, have the right talent on hand. All of those ingredients are necessary to see, then the productivity," Daco said. "This is a multiyear process."
TD Securities' Muñoz also noted that both Warsh's take and the current increase in prices can be true. He says an eventual second phase could see inflation drop, and as AI adoption across the economy could lead to an increase in output and production.
"Those savings can be passed on to consumers in the form of lower prices as the economy increases productivity growth," he said. "So those two stories can be true. Just different timelines for it."
AI 'is unlikely to be a reason for lowering policy rates'
Minneapolis Fed president Neel Kashkari said in a Q&A this past week that he's gone from penciling in one rate cut this year to one hike, based in part on massive investment in data centers and the expected resulting inflation.
"The data center build-out … may have disinflationary effects over the long term, but it's certainly putting pressure on prices in some sectors of the economy today," he said. "Anything that touches those sectors, the prices are skyrocketing on those parts of the economy."
New York Fed president John Williams said this week that strong demand for some technology components used in the AI investment boom is one of the three major factors driving inflation now. And Fed governor Michael Barr has disputed the idea of AI as a productivity accelerator that puts the Fed on a rate-cutting path. He warned that AI could be inflationary, offering the example of electricity supply constraints on the power grid colliding with booming energy demand from data centers.
"For all of these reasons, I expect that the AI boom is unlikely to be a reason for lowering policy rates," Barr said in a February speech.
The emerging picture is of a Fed board with little choice but to confront a new force that could make its mandate of price stability harder to achieve, even if AI ultimately delivers on its revolutionary promises.
"That dream world is going to take a lot longer to get there than the forecasters are telling us," Kashkari said.
"It's probably pushing up interest rates across the economy now and for the next several years, and then where does it lead long term, you know, we'll have to see."
Jennifer Schonberger is a veteran financial journalist covering markets, the economy, and investing. At Yahoo Finance she covers the Federal Reserve, Congress, the White House, the Treasury, the SEC, the economy, cryptocurrencies, and the intersection of Washington policy with finance. Follow her on X @Jenniferisms and on Instagram.