Companies traded people for tokens. The returns haven’t shown up

Jensen Huang has a check for whether or not an engineer is value preserving, and it comes with a token price range connected. On the All-In Podcast on the shut of GTC 2026, the Nvidia chief government mentioned that if a US$500,000 engineer’s annual AI token consumption got here in underneath US$250,000, half their wage, “I’m going to be deeply alarmed.” Nvidia, he confirmed, is working towards a US$2 billion yearly token invoice for its engineering pressure.

It’s a memorable provocation from the person who sells the compute. It’s additionally a tidy description of the trade-off now being made in company budgets all over the place, normally with much less candour: cash that when paid people is more and more being paid for tokens. The query the trade has been slower to ask is whether or not that commerce is definitely working, and the trustworthy solutions arriving from the businesses that moved first recommend it usually isn’t.

Where the cash went

The reallocation itself shouldn’t be in dispute. The 4 largest hyperscalers have guided roughly US$700 billion in mixed 2026 capital expenditure, practically double final 12 months, whereas Gartner projects AI agent software program spending will attain US$207 billion, up 139%. On the opposite aspect of the ledger, Challenger, Gray & Christmas data exhibits AI because the most-cited cause for US job cuts for a report fourth straight month, with tech accounting for 31% of first-half layoffs. 

An inner Meta memo obtained by Reuters described May’s cuts of 8,000 roles as offsetting the firm’s substantial investments, whilst income grew 33% that quarter. Oracle’s filings present headcount down 21,000 as financial savings feed its information centre buildout. These are extremely worthwhile firms. The layoffs aren’t survival measures. They’re financing.

Andy Challenger’s abstract of his agency’s information is the plainest out there: “Companies are shifting budgets towards AI investments on the expense of jobs.” The job a employee carried out might not have been automated in any respect. The price range that paid for it has merely moved.

What the cash purchased

Here, the report turns awkward. Gartner surveyed 350 executives at firms with over US$1 billion in income, all deploying AI brokers or automation, and located roughly 80% had reduce headcount, with no correlation between the cuts and improved returns. Analyst Helen Poitevin’s verdict: “Workforce reductions might create price range room, however they don’t create return.” 

Her analysis discovered the organisations that did enhance ROI had been these utilizing AI to amplify their people moderately than take away them. The token aspect of the ledger has its personal reckoning underway. 

Uber gave 5,000 engineers AI coding instruments in December and exhausted its total 2026 AI price range by April. Chief working officer Andrew Macdonald conceded that regardless of 70 per cent of dedicated code being AI-generated, the connection to something clients expertise is lacking: “That hyperlink shouldn’t be there but.” Uber’s engineers are actually capped at US$1,500 a month in AI spend.

Walmart imposed related token rationing on its inner assistant after demand blew previous projections, Bloomberg reported. Something is evident in that element. When tokens exceed the price range, they get capped. When people exceed price range, they get severance.

The admission

No firm has travelled the total circle extra publicly than Klarna. The fintech big changed roughly 700 customer support roles with an OpenAI-powered assistant, froze human hiring for greater than a 12 months, and made its AI-first mannequin a part of its pitch to public market traders. 

Then buyer satisfaction fell, complaints rose, and chief government Sebastian Siemiatkowski went on Bloomberg to say what few executives have mentioned aloud: “We targeted an excessive amount of on effectivity and value. The outcome was decrease high quality, and that’s not sustainable.” Klarna is hiring people once more, and its CEO now argues that investing within the high quality of human help is the firm’s future.

Gartner expects the Klarna sample to generalise, predicting that by 2027, half of the businesses that reduce customer support workers for AI will rehire, usually underneath new job titles. Its separate survey of 321 customer support leaders discovered solely 20% had genuinely decreased staffing due to AI within the first place, which suggests a lot of the slicing was unusual value self-discipline sporting an AI costume. 

OpenAI’s Sam Altman has acknowledged as a lot, conceding some “AI washing” in company layoff bulletins, and enterprise capitalist Marc Andreessen, co-founder of Andreessen Horowitz, calls AI the “silver bullet excuse”. The displacement narrative, in different phrases, is partly theatre. The price range shift beneath it’s actual, and so is the injury.

Who absorbs the experiment?

The verified hurt lands on the people least capable of soak up it. Stanford HAI’s 2026 AI Index discovered that employment for software program builders aged 22 to 25 fell practically 20% from 2024, whilst older cohorts saved rising. Companies are successfully eradicating the underside rung of the ladder whereas nonetheless anticipating senior engineers, those directing all these tokens, to exist in 5 years.

The world arithmetic is harsher nonetheless. Huang’s thought experiment assumes a US$500,000 engineer, a compensation bracket that covers maybe 2 to five% of American software program engineers and vanishingly few wherever else. Apply his half-salary token ratio to a typical engineer in Kuala Lumpur, Manila or Jakarta, and the token price range prices greater than the individual. 

In the markets the place a lot of the world’s software program work and buyer help truly occurs, the trade-off he describes doesn’t amplify staff a lot as value them in opposition to a machine, utilizing ratios set in Santa Clara.

What Klarna realized at the price of 700 jobs and a dented model is roughly what Gartner’s information exhibits in mixture: the returns observe firms that spend on people who use AI, not on AI that replaces people. The CFOs now capping token budgets after burning by them in 1 / 4 have began to rediscover one thing the trade spent two years unlearning. Talent was by no means the road merchandise holding the enterprise again.

(Image by kate.sade)

See additionally: Per-token AI charges come to GitHub Copilot

Want to be taught extra about AI and large information from trade leaders? Check out AI & Big Data Expo happening in Amsterdam, California, and London. The complete occasion is a part of TechEx and is co-located with different main know-how occasions together with the Cyber Security & Cloud Expo. Click here for extra data.

AI News is powered by TechForge Media. Explore different upcoming enterprise know-how occasions and webinars here.

The publish Companies traded people for tokens. The returns haven’t shown up appeared first on AI News.