The message to investors in the software, wealth management, legal services and logistics industries this month was clear: AI is coming for your business.
The release of new ones, ever more powerful AI tools coincided with a plunge in stock markets, which swept through industries as diverse as drug distribution, commercial real estate and price comparison sites. Technological advances are giving increasing credence to predictions that they could make millions of white-collar jobs obsolete — or, at least, eat into the profits of established companies.
Carl Benedikt Frey, author of How Progress Ends and associate professor of AI and work at the University of Oxford, says investors are reassessing the value of companies that rely heavily on the sale of software or specialized knowledge.
“AI is transforming once-scarce expertise into cheaper, faster, and increasingly comparable production, squeezing margins long before entire jobs disappear. »
Fears of widespread job losses were amplified this week by a viral testwritten by AI entrepreneur Matt Shumer, titled: Something Big is Happening. In it, Shumer claims to explain to the world outside Silicon Valley that new models will come for coding tasks and then “everything else”, comparing the present moment with February just before the Covid pandemic.
The post was viewed 80 million times on X, sparking fear and fury – including from people pointing out that Shumer has a history of hyping AI. (He previously excited the Internet by announcing the release of the “best open source model” in the world, that it was not.)
Shumer and the markets were reacting to the capabilities of recently released models such as Anthropic’s Claude Opus 4.6 and OpenAI’s GPT-5.3-Codex, both improvements over previous powerful AI products.
But there are other reasons for the excitement of the moment, notably the companies which build these models. AI “hyperscalers” – the term for big US tech players in the field – collectively plan to spend $660bn (£484bn) this year. It follows a year of colossal, often circular deals between the world’s biggest tech companies.
However, cracks have appeared in these numbers, along with questions about what they really mean. Nvidia and OpenAI recently appeared to abandon a $100 billion deal, replacing it with a smaller, as-yet-unknown commitment.
Meanwhile, none of the AI model builders – not OpenAI, xAI or Anthropic – have a clear path to the huge revenues that would justify these expenses; revenues for the entire global software industry this year are expected to be just $780 billion.
It emerged this week that both arguments about AI – that it is an unsustainable boom or that it heralds a destructive revolution in white-collar work – have been picked up by some investors, after shares of Google parent Alphabet and Mark Zuckerberg’s Meta were hit by apparent concerns about a spending bubble.
Frankly, investors expect these companies to recoup their investment through hordes of individuals and businesses paying for their tools because they allow certain tasks and jobs to be done by fewer people or for fewer hours. Or, in economic jargon, a productivity boom.
“The two themes are intrinsically linked but not necessarily contradictory,” says Jason Borbora-Sheen, portfolio manager at investment management firm Ninety One.
Initially, investors supported spending by “hyperscalers” during the initial phase of the AI gold rush. Those concerns have now turned to cash burn and the scale of investment needed to remain competitive, says Borbora-Sheen, while at the same time the stock prices of wealth managers and others have been affected by the perception that AI is “now here, will evolve and can move around.”
Companies have cited AI as an influence on job cut plans, including British American Tobacco this week, but there has yet to be a wave of massive disruption. Greg Thwaites, research director at the UK think tank Resolution Foundation and associate professor at the University of Nottingham, says the evidence for a tangible jobs impact from AI on large Western economies is “quite ambiguous so far”.
Not all white-collar jobs will be affected, he says, although AI could test the axioms around the old capitalist concept of “creative destruction,” which involves entirely new jobs replacing obsolete jobs, like auto mechanic replacing farrier. Will AI be a different case because change has come so quickly or because it will be good at absolutely everything?
He adds: “Some jobs are going to be very different very quickly. But the idea that there will be gangs of unemployed lawyers and accountants wandering around London within a few years seems far-fetched to me.”
Forrester analyst Alvin Nguyen says the fears that have roiled the stock market are based on sentiment, not evidence: No one has had time to evaluate the performance of a wealth manager with Opus 4.6.
“It’s a knee-jerk reaction,” he said. “Is that true? Look, there are a lot of leaders who thought, ‘I can replace people with AI at the beginning. And a lot of people acted on it. And I think one of the things that was discovered is that in many cases, no, it didn’t work.’
Aaron Rosenberg, a partner at venture capital firm Radical Ventures – whose investments include leading AI company Cohere – and former head of strategy and operations at Google’s AI unit DeepMind, says the impact of AI is underappreciated in the long term, but the adoption of breakthrough models will not be uniform.
“History shows a repeating pattern: there is a significant gap between a technology working in a laboratory and its diffusion into the broader economy, as well as a chasm between early adopters and the majority of users,” he says.
Other new models will come; other huge AI deals could also fail. Meanwhile, this month, faint rumblings of discontent have emanated from high-level tech workers; a series of departures from AI companies for reasons as diverse as boredomAI doom and gloom and concerns about the prospect of adult content in ChatGPT.
There’s a nervous, fuzzy energy at work. As Borbora-Sheen says: “There is a strong dynamic between winners and losers. »