Tech stocks have taken a hit in recent weeks and questions are mounting over whether the frenzy around artificial intelligence has finally peaked.
The talk of a coming crash is uncomfortably reminiscent of the dotcom bust two decades ago, when hyped-up internet firms collapsed and dragged markets down with them.
At Jackson Hole, Federal Reserve chair Jerome Powell acknowledged the risks of stagflation but hinted that interest rate cuts could be on the horizon to ease pressure on indebted firms. As The Guardian noted, the health of the stock market is now tied closely to AI-driven giants like Microsoft, Google, Amazon and Meta, all of which have poured billions into development with little in the way of profit to show for it.
A recent MIT study found that 95% of companies investing in generative AI have yet to see returns, while OpenAI boss Sam Altman has warned that valuations are “insane,” remarks that, according to the Financial Times, triggered pullbacks across the sector. Palantir slid nearly 10% in a single session, Nvidia fell more than 3%, and other AI-linked stocks like AMD and Oracle also dipped.
Not everyone agrees that panic is warranted. Investopedia notes that mega-cap tech remains under-owned by institutions and continues to produce strong earnings. Others, like fund manager Simon Edelsten writing in the FT, say investors should return to basics: look for firms that demonstrate real productivity gains and avoid chasing overvalued names. Meanwhile, SoftBank’s Masayoshi Son has doubled down with a $500 billion AI data centre plan, a bet covered by Reuters that shows long-term confidence in AI infrastructure despite short-term jitters.
History provides some comfort. The dotcom crash wiped out 75% of the Nasdaq, but survivors like Amazon and Google went on to dominate. Scholars like Carlota Perez argue that such cycles of boom and bust are part of the natural “installation phase” of new technologies, which eventually give way to a “golden age” once the excess is burned off, as explained in the FT.
My View: A Correction, Not an End
From where I stand, this looks more like a necessary cooling than an outright implosion. Valuations such as Palantir’s price-to-earnings ratio above 500 were clearly unsustainable. A pullback allows stronger firms to consolidate, acquire innovation cheaply, and re-set investor expectations to something more realistic.
At the same time, it would be naïve to dismiss the sector. AI tools like Microsoft’s Copilot are already embedded in offices, cutting down on tasks and shifting workflows in ways that are unlikely to reverse. Even if weaker firms fade, the giants have the cash, scale, and infrastructure to carry the technology forward.
In short, the bubble may deflate, but AI is not disappearing. Investors should be cautious, diversify, and avoid stocks priced on hype alone. Panic selling now would mean missing out on the long-term transformation still underway. Prudence, not fear, is the smarter strategy.
