Is the AI bubble bursting? Experts more inclined to call it natural cooling
With most AI projects delivering little to no revenue growth, some experts say cracks are beginning to appear in the AI balloon and are hearing echoes of the dot-com era. Others urge calm and patience.
The numbers aren’t particularly impressive. Last week, a new study from the Massachusetts Institute of Technology found that 95% of generative AI projects have delivered little to no growth in revenue.
Picture a line of ultra-modern fighter jets ready for take-off but persistently stuck at the starting line – that’s where we’re at, apparently. According to MIT’s study, most AI-based initiatives to drive rapid revenue growth are falling flat.
Less revenue, more caution
The paper concludes that there are no savings, no profits, and a lot of fear among the investors (PDF). Even OpenAI’s boss, Sam Altman, the AI industry oracle-in-chief who regularly pre-welcomes artificial general intelligence (AGI), warns that investors are overvaluing the potential returns on AI.
“Are we in a phase where investors as a whole are overexcited about AI?” Altman reportedly said at a private dinner with reporters. “My opinion is yes.”
Soon, tech stocks of companies with significant AI pushes tumbled. Palantir saw a 9% drop in shares, Oracle witnessed a 5.8% drop, and chipmaker Nvidia fell 3.5%. Even Meta, which has injected billions into its new AI division, has now paused hiring.
Sure, many urged calm late Wednesday when Nvidia said it expects revenue of $54 billion, plus or minus 2%, in Q3, compared with analysts’ average estimate of $53.14 billion.
But the chipmaker’s fiscal Q2 results came up short of some analyst expectations in its important data center segment, with some experts suggesting that cloud computing providers may indeed be more cautious about spending.
“Just beneath the breathless promotion of AI, cracks are beginning to appear,” this is how The Guardian put it this week, expressing doubt whether the hype machine touting “the supposed promise and inevitability of AI” is still working.
Others, though, are still preaching that the AI boom is to continue. To them, the current hiccup is only natural for the first stage of the adoption cycle, and chips are selling very well, after all.
“Most generative AI pilots won’t deliver revenue today – but that’s normal. It’s like the internet in the early 90s: we saw lots of experiments, most failed, but the infrastructure that mattered kept compounding,” Francis Hellyer, CEO of AI-powered travel and experiences platform tickadoo, told Cybernews.
Nvidia’s Huang is confident: Should he be?
Indeed, industry insiders such as Hellyer mostly see the current trend as an inevitable part of the AI boom and say it wouldn’t be fair to call it a bubble, at least so far.
What we are seeing is not so much a bubble set to burst, but rather a natural cooling of the initial hype, as regulation, accountability, and real-world application catch up with innovation.
Ellen Benaim.
“We’re not in an AI bubble yet, but we are in an AI hype cycle. The market is testing which applications drive real value versus which are just press releases,” said Hellyer.
“The correction is healthy. It forces companies to prove product-market fit rather than rely on investor FOMO. The next decade of AI will be won by companies solving practical, specific problems, not chasing science fiction.”
Indeed, Nvidia’s Jensen Huang seems extremely confident – even if he simply has to be in order for the business to succeed. During a call with analysts, Huang said that global spending on AI infrastructure is on course to reach $4 trillion by the end of the decade.
Boston Consulting Group also said recently that generative AI investments are projected to increase 60% in the next three years.
And even though only one in four executives says their companies are seeing significant returns from their AI investments, that only means that leaders are still searching for the critical business benefits of generative AI and need to target the highest-value initiatives that drive profits.
“In terms of the AI boom, what we are seeing is not so much a bubble set to burst, but rather a natural cooling of the initial hype, as regulation, accountability, and real-world application catch up with innovation,” Ellen Benaim, CISO at SaaS company Templafy, told Cybernews.
And even if 95% of generative AI projects have delivered little to no growth in revenue, what about the remaining 5%? Might they be creating so much value that it’s all worth it?
“People tend to focus on the projects that deliver little. Part of this is because everyone subconsciously wants AI to be less hyped because it’s intimidating,” venture capitalist Nick Davidov told Cybernews.
Even if 95% of generative AI projects have delivered little to no growth in revenue, what about the remaining 5%?
“But if you don’t implement AI now, you risk losing revenue and margins because your competitors are adopting it. AI allows companies to redistribute the existing pie before it helps to increase its overall size.”
Sustainability will take time
However, even if we’re not witnessing a bursting AI balloon, it’s still a wobble, and businesses need to react fast to the fact that markets are recalibrating expectations, Faizel Khan, lead AI engineer at Landing Point, told Cybernews.
“The market doesn’t need a hard reset. It needs a reality check,” said Khan before giving a series of advice for those in the industry.
“Dial down the AGI hype and refocus on execution. Reward teams that ship working workflows, fix data plumbing, and prove unit economics with real metrics.”
“Fund AI like any capital project: stage it, set function-level milestones, and release more budget only when those milestones are met,” he added.
Yep, the famous AGI, the moment when AI will match or surpass intelligence. Even Google’s former CEO Eric Schmidt has now warned the industry shouldn’t focus on this particular hogwash.
In an essay for The New York Times called “Silicon Valley is drifting out of touch with the rest of America,” Schmidt and technology analyst Selina Xu write that concentrating on AGI is misguided.
“We worry that Silicon Valley has grown so enamored with accomplishing this goal that it’s alienating the general public and, worse, bypassing crucial opportunities to use the technology that already exists,” the essay says.
“There’s a widening schism between the technologists who feel the AGI – a mantra for believers who see themselves on the cusp of the technology – and members of the general public who are skeptical about the hype and see AI as a nuisance in their daily lives.”
In other words, let’s maybe treat AI as infrastructure that needs improving rather than theatre, and realize that it will take time for the market to enter a healthy and sustainable space.
