Banking giant UBS dismisses AI bubble fears, noting that current capital spending levels are nowhere close to historical innovation booms.
In a new CNBC interview, UBS managing director Alli McCartney pushes back against mounting bubble-watch commentary, saying long-term infrastructure waves have historically required sustained spending far above today’s levels.
She points to past industrial revolutions as the correct benchmark for judging AI CapEx, not short-term stock volatility or near-term investor anxiety.
“If we look back at everything from the steam engine to telecom to electrification of the grid for an extended period of time, so over years, not over months, somewhere between 2% and 5% of GDP was spent on infrastructure to keep those things going.”
McCartney notes that today’s AI investment pace remains modest in comparison, despite rapid data-center buildouts and hyperscaler capex guidance accelerating across the sector.
“Right now, we’re at about a year, maybe two, of slightly less than 1% being spent.”
Goldman Sachs echoes McCartney’s call to reject AI bubble calls. Goldman Sachs senior global economist Joseph Briggs recently said that when compared to investment cycles of the past, the current AI CapEx boom is still small.
“And when we do that, we see that the total amount of AI spending in the US today is actually a little bit below 1%. If we look back at IT spend, the buildout of infrastructure, railroads, canals in the UK in the 1600s, generally, these infrastructure investment impulses rose to 2 to 5% of GDP. And so we’re not at levels that are unprecedented or even that large.”
Disclaimer: Opinions expressed at CapitalAI Daily are not investment advice. Investors should do their own due diligence before making any decisions involving securities, cryptocurrencies, or digital assets. Your transfers and trades are at your own risk, and any losses you may incur are your responsibility. CapitalAI Daily does not recommend the buying or selling of any assets, nor is CapitalAI Daily an investment advisor. See our Editorial Standards and Terms of Use.

