Wall Street’s tech titans are under duress this week, with the Magnificent Seven — Apple, Microsoft, Alphabet, Amazon, Meta, Nvidia, and Tesla — leading the Nasdaq Composite to its steepest two-day slide since August 1. The index shed 1.4% on Tuesday and another 0.7% Wednesday, with Micron plunging up to 7% and Palantir extending its retreat, while the S&P 500 recorded its fourth consecutive daily loss.
Some of the downturn stems from rate-cut uncertainty ahead of US Federal Reserve chairman Jerome Powell’s speech at the Jackson Hole, Wyoming, summit on Friday. Meanwhile, President Trump’s plan for the US government to invest in Intel — part of a push to onshore chip production — stoked fears of government overreach in private markets, prompting Intel shares to slide up to 7% and adding to the broader sector weakness, as investors contemplate the potential implications of distorted competition and politicised capital allocation.
But with major warning signals flashing in the artificial intelligence (AI) discourse recently, the dominant theme remains heightened investor anxiety that the AI boom driving tech enthusiasm is spiralling into bubble territory.
AI promises vs. business reality
A stark signal came this week from MIT’s Sloan School of Management, which published a study showing that 95% of corporate generative AI pilot programs — into which companies have already sunk billions of dollars — fail to advance beyond the testing phase. Researchers pointed to a “learning gap” inside companies — the inability to adapt workflows, culture and systems to make AI tools productive at scale.
The finding has rattled markets because it underscores how little of the massive AI investment wave is translating into immediate business results, with many projects stalling before they create measurable returns.
For investors, it raises an uncomfortable question: if the vast majority of corporate AI efforts are failing to generate value, are stock prices running far ahead of reality? This tension between promised transformation and delayed execution is increasingly feeding into the market’s bubble debate.
Data centres at the centre of the boom
Meanwhile, OpenAI CEO Sam Altman last week publicly acknowledged that the AI space may be overinflated, saying that “when bubbles happen, smart people get overexcited about a kernel of truth.”
“Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes,” he said.
Yet Altman argued that AI remains profoundly transformative, and he revealed OpenAI’s plans to spend “trillions of dollars” on data centre build-out in the near future.
Massive investment in data centres — the physical backbone of the AI boom — is also reshaping markets closer to home. In Australia, Goodman Group (ASX:GMG) has tapped into that wave with more than half its $12.9 billion development pipeline tied to data centres.
The property giant today reported a $1.67 billion swing back into profit, yet its shares slipped nearly 2.5% by the afternoon.
Howard Marks sees echoes of ‘irrational exuberance’
Backdropping all of this is Oaktree Capital co-founder Howard Marks — author of the famed “bubble.com” memo — who has recently been sounding the alarm that valuations are elevated and investor psychology is fragile. In a Bloomberg interview on Thursday, he noted that stocks are expensive relative to fundamentals in circumstances he last recalls seeing in 1997, before the dot-com bubble burst.
“Investors are by nature optimistic, and that optimism dies hard,” he said. “I think the fluctuations of the market are mostly related to psychological fluctuations — people go from neutrality to liking stocks… to liking them too much.
“That’s the continuation that creates bubbles,” he added. “We are probably in the early days of that.”
Marks said he is “not ringing the alarm bells” but noted why he’d pinpointed 1997 — when then-Fed chair Alan Greenspan issued his famous “irrational exuberance” warning — to describe how he sees markets today.
“I picked ’97 because, even though Greenspan was concerned about exuberance, the market went on to rise for another two and a half to three years,” he said. “We’re in the early days.”