The AI backlash has already begun

Plus: Venezuela, gold price predictions, quality stocks, tokenised banking, cheap TVs, expensive RAM chips, Warhammer, woke hacking, and Taylor Swift

We’re FT Alphaville, a Financial Times blog about markets, finance and economics. Each Friday, we send a newsletter that bundles together the team’s work with some interesting stories found elsewhere, topped and tailed by commentary and charts. The newsletter’s completely free, just like Alphaville.

What are they rebelling against? Whatever you’ve got

It was fashionable about a decade ago to say that artificial intelligence is an ideology, not a technology.

Product development is engineering, which can be described in specifics, whereas AI could encompass all the storytelling and possible futures imagined by evangelists. Keeping terms separate was intended to ensure that developers couldn’t use grand theories about societal advancement to ignore social responsibility, and to encourage everyone else not to anthropomorphise technologies that are neither intelligent nor artificial. (For a reminder of how that’s going, follow those links.)

All hope of maintaining a distinction between science and sci-fi collapsed on November 30 2022, with the launch of ChatGPT. Just over three years later, AI can mean almost anything. So, to the public, it means almost nothing.

Reporting from the Consumer Electronics Show in Las Vegas, PC Gamer quotes Dell head of product Kevin Terwilliger:

We’re very focused on delivering upon the AI capabilities of a device — in fact everything that we’re announcing has an NPU [neural processing unit] in it — but what we’ve learned over the course of this year, especially from a consumer perspective, is they’re not buying based on AI. In fact I think AI probably confuses them more than it helps them understand a specific outcome.

Meanwhile, over on fantasy island …

Artificial intelligence will save time on household chores and allow more space for work or play, said OpenAI’s chief economist, arguing the technology will help millions of people in ways that do not appear in traditional economic statistics.

Aaron “Ronnie” Chatterji tells the FT that large language model AI will be as labour-saving to households as the dishwasher. One-hundred-and-seventy-six years after its invention, fewer than 50 per cent of UK households have a dishwasher, but maybe that’s because no one thought to give them away for free.

“Whether it’s caring for children or doing household chores, these are things that need to get done,” said Chatterji.

But given that it’s not counted [in GDP], a lot of those benefits will never show up. And so we have to figure out ways to measure this more carefully and figure out what the value of time savings is from AI.

It may not be obvious how chatbots will assist in caring for children rather than, for example, encouraging them towards self-harm. Neither is it obvious why a former Chief Economist of the US Department of Commerce would be ignorant of the more than 50 years of literature about how to measure the economic value of unpaid work, or of the centuries of bad predictions about how technology will create a leisure surplus. In fact, unless AI is taken to mean an ideology rather than any specific technology, it’s probably not clear on any level what on earth Chatterji is talking about.

Whatever. Now that AI is well-established as an anti-marketing slogan and a vibe religion, the challenge will be making us pay for it. Here are a couple of charts from research published this week.

That’s from Jefferies’ 2026 US utility sector outlook, which identifies affordability as an underappreciated risk.

Corporate executives worry more about reputation than investors, who are often slow to spot the significance of any event that’s hard to quantify in earnings forecasts. BP shares, for instance, were little changed in the week following news of the Deepwater Horizon disaster in 2010.

Data centres will “become the villain” of 2026, so the companies supplying their power have some hard decisions to make, Jefferies says. “We expect effectively every political candidate running for office in 2026 to be forced to take a position, whether they are pro or con data centers/AI.”

Perception matters. Rather than be seen to pass through build costs to customers, utilities will defer capex while leaning on companies to firm up their commitments. Then there’s the practicality of rushing large-scale construction, where “a single high-profile cost overrun can become a statewide affordability headline”.

Hostilities now playing out around plans to convert the Truman Brewery in East London to a data centre are a sign of things to come. Better for utilities to push the delivery of ambitious projects into the next decade and test how much of the slated demand is real, even if the delays end up killing a few highly leveraged data-farm builders on the way.

This next chart’s from Mizuho analyst Jordan Rochester.

What makes the US and UK trends unusual is the widening gap between youth unemployment and older cohorts. Rochester writes:

It could just be noise of the cycle with post pandemic data hard to read – but we suspect AI is a new factor to consider adding to the pain for the youth of today. Especially in this “low hire-low fire” environment, new recruits entering the workforce are finding it harder to land their first job.

There’s no shortage of surveys about how business leaders want to replace staff with bots, along with some evidence that hiring is on hold while trial automation projects are rolled out, but it’s much too early to talk with confidence about how it will reshape the labour market. Plenty of alternative explanations are available for the lack of entry-level jobs, such as higher taxes and persistently low growth, and it would be remiss not to note the parallel decline in US blue-collar employment.

What’s clearer is that, whichever way, AI is getting the blame. People can’t find work not because of any identifiable skills deficit or macroeconomic trend, but because of a poorly defined ideology. It’s no wonder they don’t want its name written on the Dell laptop they can no longer afford to charge.

BTW, if you missed John Plender’s Weekend FT essay on how the bubble bursts, fix that now.

A week on Alphaville

○ For all of the above, the AI spending boom as a proportion of US GDP is a bit underwhelming, at least relative to previous speculative bubbles.

○ Jay Newman et al think bondholders are underestimating the realpolitik involved in Venezuelan debt restructuring.

○ Barclays’ EM sov team had bad luck with the timing of a downgrade of Venezuela.

○ Stocks have gone up over the long term, so will probably continue to do so at approximately the same pace, say lots of strategists who make boring predictions.

○ Sovereign wealth funds are big buyers of private credit.

○ Dan Davies considers the ECB’s reverse stress-testing regime.

○ Memory prices continue to be silly.

○ Taylor Swift’s effect on the UK economy continues to be unmeasurable.

○ We can’t include the headline of this post listing senior employees of X Corp because the email would end up in a lot of spam filters.

○ Craig Coben is wary about what promises to be a very lopsided IPO calendar.

○ Pending a policy decision, the bitcoin storage facility known as Strategy has not yet been kicked out of MSCI’s indexes.

○ Tokenised bank deposits look a lot like stablecoins but are good for yield rather than crime. China’s keen.

○ The festive chart quiz asked readers to identify research houses by their typographical styles. Yes, really.

Best of Further Reading

○ The post-American internet (Pluralistic)

○ Why AI will not fix private markets (Ludovic’s Substack)

○ How did TVs get so cheap? (Construction Physics)

○ Crypto thieves move offline to terrorise investors at home (Bloomberg)

○ The Warhammer capital of the world (Dispatch)

○ Hundreds of mysterious Victorian-era shoes are washing up on a beach in Wales. Nobody knows where they came from (Smithsonian)

○ Not all heroes wear capes; some wear helmets. (Kotaku)

Promotional interlude

Ahead of the Global Insurance Summit 2026, our colleagues are running a one-hour webinar on January 21 called “Finding Opportunity In a Changing Insurance Ecosystem”. Registration is free.

Chart blast

○ Gold is expected to go up, down, or sideways this year.

○ Quality stocks (meaning those with a high return on equity, steady earnings growth and low leverage) have rarely been so out of favour, says Ruchir Sharma in his top 10 trends for 2026.

○ Having lagged behind the market rally since the pandemic, the health insurance and pharmaceutical sectors stand out among US large-caps as potentially oversold, says Unhedged in its guide to hated stocks.

What have we missed? What did we get wrong? Where can we improve next week? Let us know by email.