AI’s Abrupt Cooling Period

Welcome to TechScape.

As you cruise down the 280 highway in San Francisco, it may seem like AI is omnipresent. Billboards are filled with AI-related advertisements. “We’ve automated 2,412 BDRs.” “Is all that AI and ROI really there?” “Affordable on-demand GPU clusters.” It’s tough to decipher industry jargon while speeding by.

These billboards symbolize the tech sector’s mass shift toward AI. Executives are rapidly rebranding their companies as AI-focused. In California’s tech epicenter, just as every business turned high-tech in the 2010s, they are now redefining themselves as AI enterprises.

Yet beneath the dazzling promotions of AI capabilities, troubling signs are emerging. Prominent AI advocates like OpenAI’s Sam Altman caution that investors are misjudging AI’s potential returns. “Are we in a phase where investors are overly enthusiastic about AI?” Altman remarked during a private dinner with a journalist. “In my view, yes.“

Altman’s words align with OpenAI’s acknowledgment of struggles in launching its latest ChatGPT model, which he promised would be a substantial upgrade over the existing GPT 4.5 version.

Of course, Altman’s comments might be aimed at dissuading investors from financing rivals. But other indicators have emerged. A recent MIT study found that 95% of generative AI projects reported little to no revenue growth. Major tech stocks incorporating AI have suffered as well: Palantir’s shares dropped by 9%, Oracle’s by 5.8%, Nvidia’s by 3.5%, and Arm’s by over 5%. A slump in tech stock support from other sectors contributed to this downturn.

Cracks are starting to show beneath the dazzling AI promotion.

Moreover, Meta has reportedly invested billions in securing top AI talent but has announced an AI hiring freeze. Last week, AI executive Alexandr Wang stated on X that Meta is investing in its Superintelligence Labs, asserting, “The reports are grossly inaccurate.”

This abrupt cooling of AI interest comes just as many companies announced sizable investments in building AI capabilities while reporting less-than-stellar revenue. Altman noted during that same dinner that he aims to invest “trillions” in data center expansion in the near future, according to The Verge.

The current wave of trepidation surrounding AI might signal a necessary market correction rather than an outright bust of the AI hype bubble. Even Eric Schmidt, the former Google CEO, cautions against the notion that achieving artificial general intelligence (AGI) is imminent or that AI will eventually surpass human intelligence.

“The speed at which AGI can be reached is uncertain,” he stated in a column co-written with AI policy expert Selina Xu. “There is a worry that Silicon Valley is fixated on this goal.” Schmidt and Xu also emphasized the achievements AI has already delivered, expressing concern over Silicon Valley’s preoccupation with the horizon.

“There exists a significant divide between engineers who believe AGI is just around the corner and the general public, who often view AI through a skeptical lens and see it as an inconvenience in daily life,” they wrote.

It remains to be seen if the industry heeds these warnings. Investors are eagerly awaiting quarterly revenue reports for signs that each company’s multibillion-dollar investments are warranted while management aims to keep morale high. The ongoing promotion and allure of AI play a crucial role in alleviating investor anxiety, particularly amid a quarterly rise in projected spending across the board. For instance, Mark Zuckerberg recently suggested that those not engaging with AI tools could be at a cognitive disadvantage, indicating that companies like Meta and Google may continue integrating AI into their essential products, leveraging them to enhance training data and user populations.

The first major test of this AI reality check will occur on Wednesday, as chipmaker Nvidia, a key player in developing large language models, releases its latest revenue figures. While analysts are optimistic, the volatile stock week poses a challenge, making investors’ reactions to Nvidia’s earnings and spending updates a critical indicator of their future enthusiasm for the AI hype.

Have You Bonded with an AI?

Frequent CHATGPT users often develop a strong emotional connection with AI. When changes occur, they take notice. Many users were dismayed by the introduction of OpenAI’s latest update to the GPT-5 model. My colleague Dani Anjano reports:

“It felt truly unsettling; it was a challenging time,” remarked Swedish software developer Linn Vailt about the update. “It seemed like someone shifted all the furniture in your home.”

ChatGPT quickly adapted, promising updates to the personality features and allowing access to older models for subscribers, recognizing the significance of these features for users.

***
Have you formed a bond with AI? We want to hear from you. Please reach out at techscape.us@theguardian.com.

Read More:

The Debate on Facial Recognition

TikTok Believes Machines Outperform Humans in Content Moderation

TikTok is downsizing its UK trust and safety team. My colleague Lauren Almeida reported:

TikTok’s decision puts hundreds of roles from the UK content moderation team at risk, even amid stricter measures aimed at curbing harmful content online.

The popular video app has revealed that hundreds of jobs in its trust and safety teams could be impacted in the UK, along with South and Southeast Asia, as part of a global restructuring.

In September, the company let go of an entire team of 300 content moderators in the Netherlands, and in October, it announced the replacement of around 500 content moderation roles in Malaysia as part of its transition towards AI.

Recently, German TikTok employees protested against the layoffs within its trust and safety teams, which was the driving force behind this restructuring.

Read All Episodes: Despite new online safety regulations, hundreds of TikTok UK moderators face uncertainty in their roles.

These layoffs are part of a larger global initiative to moderate content using AI. According to TikTok, 85% of content removals on the platform are currently handled by automated systems. The parent company, ByteDance, appears eager to increase this percentage.

The company is not downplaying the significance of human oversight in tackling sensitive issues. It’s generating considerable revenue, with reports indicating a 38% increase in the UK and European markets. This strategy mirrors similar moves by other tech giants, like Meta, which have dismantled fact-checking initiatives and made significant cuts to their trust and safety teams.

TikTok has also conducted minor layoffs within its US Trust and Safety Team. The absence of mass terminations of content moderators raises questions: Is this move too risky amid unclear US policies towards the app? The White House’s recent TikTok account launch may signal a shift. Recall previous administrations’ attempts to limit the app’s presence in the US. The ban remains in limbo, upheld by a fragile executive order.

Broader Technology Landscape

Source: www.theguardian.com