Is the AI boom finally starting to slow down? | Technology

The sudden AI cooling-off period

Hello, and welcome to TechScape.

Drive down the 280 freeway in San Francisco and you might believe AI is everywhere, and everything. Nearly every billboard advertises an AI related product: “We’ve Automated 2,412 BDRs.” “All that AI and still no ROI?” “Cheap on-demand GPU clusters.” It’s hard to know if you’re interpreting the industry jargon correctly while zooming past in your vehicle.

The signs are just one example of the tech industry’s en-masse pivot to AI, a technology that the executives who have the most to gain from it say will be universe-shifting, inevitable and unavoidable. In California’s tech heartland, every company is now an AI company, just like every company became a tech company sometime in the 2010s.

But just beneath the surface of the breathless promotion of every startup’s AI prowess, cracks are beginning to appear. Some of AI’s biggest boosters, such as OpenAI’s Sam Altman, have warned that investors are overvaluing the potential returns on AI. “Are we in a phase where investors as a whole are overexcited about AI?” Altman said at a private dinner with reporters. “My opinion is yes.”

Altman’s statements coincided with a concession that OpenAI botched the roll-out of the latest ChatGPT model – which he once promised would be a significant improvement over the current GPT 4.5 model.

Sure, Altman’s comments may have been geared at getting investors to consider not investing in his competitors. But there are other signs. Just this past week, a new study from MIT found that 95% of generative AI projects have delivered little to no growth in revenue. And tech stocks of companies with huge AI pushes tumbled this week: Palantir saw a 9% drop in shares, Oracle saw a 5.8% drop, chipmaker Nvidia fell 3.5% and chipmaker Advanced Micro Devices as well as semiconductor designer Arm each saw 5% or more drop in shares. Not even the possibility of lower interest rates, which buoyed stocks in other sectors last week, was enough to turn around tech’s market slide.

Just beneath the breathless promotion of AI, cracks are beginning to appear

Even Meta, which had been reportedly spending billions to recruit the top AI talent, announced an AI hiring freeze. Meta sought to downplay the news, with the company’s chief AI officer, Alexandr Wang, posting on X last week: “We are truly only investing more and more into Meta Superintelligence Labs as a company. Any reporting to the contrary of that is clearly mistaken.”

The sudden AI cooling-off period comes just a week after many of these companies reported largely stellar earnings despite upping their expectation for how many billions they plan to spend on building out their AI capacity over the next few months and years. At the same dinner with reporters, Altman said he expected OpenAI to spend “trillions” on datacenter expansion in the “not very distant future”, according to the Verge.

While the latest round of AI jitters is being framed as a sign that the AI hype bubble may soon burst, it’s perhaps more of a needed market correction – a reality check for investors who can’t see outside of the Silicon Valley echo chamber. Even the former Google CEO Eric Schmidt warns that some of the laser focus on achieving artificial general intelligence – or the idea that AI could at some point match or surpass human intelligence – is misguided.

“It is uncertain how soon artificial general intelligence can be achieved,” reads a column Schmidt co-wrote with China and AI policy lead Selina Xu. “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.” Schmidt and Xu spend much of the rest of the column boosting the benefits and milestones AI has already met but raising concerns about how the build-at-all-costs strategy Silicon Valley is taking in pursuit of a super-intelligent AI is not reflective of the reality for the general public.

“There’s a widening schism between the technologists who feel the A.G.I. – 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 A.I. as a nuisance in their daily lives,” they wrote.

It’s unclear if the industry will take heed of these warnings. Investors look to every quarterly earnings report for signs that each company’s billions in capex spending is somehow being justified and executives are eager to give them hope. Boosting, boasting about and hyping the supposed promise and inevitability of AI is a big part of keeping investor concerns about the extra $10bn each company adds to its spending projections every quarter at bay. Mark Zuckerberg, for instance, recently said in the future if you’re not using AI glasses you’ll be at a cognitive disadvantage much like not wearing corrective lenses. That means tech firms such as Meta and Google will probably continue making the AI features that they offer today an almost inescapable part of using their products in a play to boost their training data and user numbers.

That said, the first big test of this AI reality check will come on Wednesday when chipmaker Nvidia – one of the building blocks of most LLMs – will report its latest earnings. Analysts seem pretty optimistic but after a shaky week for its stocks, investor reactions to Nvidia’s earnings and any updates on spending will be a strong signal of whether they have a continued appetite for the AI hype machine.

Have you bonded with an AI?

Heavy ChatGPT users have formed strong emotional attachments to the AI, which means that, when it changes, they notice. In the case of OpenAI’s most recent update, the release of the AI model GPT-5, these users got upset. My colleague Dani Anguiano reports:

“It was really horrible, and it was a really tough time,” said Linn Vailt, a software developer in Sweden, about the update. “It’s like somebody just moved all of the furniture in your house.”

ChatGPTquickly made adjustments, promising an update to 5’s personality and restoring access to older models – for subscribers only – while acknowledging it had underestimated the importance of some features to its users.

***
Have you bonded with an AI? We want to hear from you. Email techscape.us@theguardian.com.

Read more:

The debate over facial recognition

TikTok believes machines will do better at removing bad things from the internet than humans

TikTok is slashing its trust and safety teams in the United Kingdom – and all over the world. My colleague Lauren Almeida reports:

TikTok has put hundreds of UK content moderators’ jobs at risk, even as tighter rules come into effect to stop the spread of harmful material online.

The viral video app said several hundred jobs in its trust and safety team could be affected in the UK, as well as south and south-east Asia, as part of a global reorganisation.

In September, the company fired its entire team of 300 content moderators in the Netherlands. In October, it then announced it would replace about 500 content moderation employees in Malaysia as part of its shift towards AI.

Last week, TikTok workers in Germany held strikes over layoffs in its trust and safety team, also the intended target of layoffs.

Read the full story: Hundreds of TikTok UK moderator jobs at risk despite new online safety rules

The cuts are part of a global push by the company towards moderating content with artificial intelligence. Automated systems already carry out some 85% of content removals on the app, according to TikTok. ByteDance, its parent company, seems to want to move the percentage upward.

The company is not hemming and hawing about the importance of human oversight and insight into thorny issues. Business is booming, with a 38% increase in revenue in the UK and Europe, per regulatory filings. American tech giants are moving in the same direction, led by Meta’s disbandment of its fact-checking initiatives and Elon Musk’s deep cuts to the trust and safety teams at X.

TikTokhas carried out smaller-scale layoffs of trust and safety teams in the United States. There have not been mass firings of content moderators. Why? Perhaps because any move that might elicit backlash would be dangerous at a moment when the official US attitude toward the app couldn’t be less clear. The White House created a TikTok account that launched last Tuesday. You may recall that the past two American presidents have tried to shut the app out of the US. The ban is still in place, held off only by flimsy executive orders extending a pause on its enforcement, which may only persist as long as Donald Trump’s goodwill does.

The wider TechScape