In musing about the effects of artificial intelligence on jobs James Pethokoukis remarks:
A different approach is taken in the new paper “Remote Labor Index: Measuring AI Automation of Remote Work.” The authors, from the Center for AI Safety and Scale AI, decided to treat AI systems as if they were freelance workers on real jobs. They took 240 genuine Upwork-style projects—everything from data dashboards and 3D product designs to marketing videos—and provided the same briefs, files, and deliverables to both humans and AI models such as GPT-5, Claude Sonnet 4.5, and Gemini 2.5 Pro. Human evaluators then judged whether the AI’s submissions would be acceptable to a paying client.
The result: Almost never, with a tiny 2.5 percent success rate “revealing a stark gap between progress on computer use evaluations and the ability to perform real and economically valuable work,” the paper concludes. Even the top-performing model, Chinese AI agent Manus, “earned” only about $1,700 out of $144,000 worth of human labor.
IMO he’s looking at this from the wrong perspective. The perspective he needs to consider is that of those making staffing decisions, the hiring and firing decisions. The CEOs.
In the technology sector alone in 2025 alone 100,000 layoffs have been announced. Those layoffs aren’t of minimum wage employees but of middle managers and developers at all levels (junior, senior, C-suite). Go beyond the technology sector and layoffs amounting to 1% of the total U. S. workforce have been announced in this year alone.
To take another sector the financial sector has announced roughly 50,000 layoffs just this year. I’m guessing those layoffs won’t be of either top management or those at the lowest levels of compensation but people with incomes well into six figures.
In other words in the near term it doesn’t make a smidgeon of difference. The only things that make any difference are whether the CEOs think they can use AI to trim expenses and if that will increase their stock value. There’s an old wisecrack—I don’t know who said it. “When it’s time to railroad everybody railroads.” A lot of companies are jumping on the AI bandwagon and devil take the hindmost.
I am pretty skeptical the current set of LLM’s have actually replaced many or even any jobs. There’s a lot of pre-requisites for enterprise deployment (mundane things like identity and access management) where the industry is still experimenting; so I don’t think they are deployed widely in enterprises.
As to what’s causing layoffs, its from a desire to preserve profit margins; an overvalued currency; continual bleeding from the ease of remote work and outsourcing; and adjustment from overhiring during the pandemic in the software, retail sector.
As an example, there was a report today that layoffs spiked, with a focus on the warehousing sector. Humanoid robots / automation aren’t ready for warehouse deployment, those layoffs are due to weak consumer / retail sales.
I believe a lot of angst is because the economy is rebalancing (not by choice); away from the drivers of the last 20-30 years (consumers and “software’).
So, you think that Jeff Bezos, Mark Zuckerberg, and Satya Nadella are lying when they attribute the layoffs to AI? Just preserving face in their rebalancing to maintain profits?
Yes and no on attributing to AI.
Not because AI is capable of replacing software engineers, but because of the need to increase AI capex (its getting close to eating all their free cash flow) while maintaining profit margins means they have to cut “opex” as in headcount — that’s the rebalancing that is occuring.
Look at MSFT as an example. They aren’t trying to cut / layoff their way to higher stock prices. Their capex has gone from $20 billion in FY 2021 to $64 billion FY 2025 to > $100 billion in FY 2026. In comparision, from 2021 to 2025 — their revenues have increased by 67%. That capex is spent on data centers, chips, power, cooling — so likely a lot of jobs are being created in those sectors, but its not software or retail.
Even the software engineers at MSFT is changing, another reallocation of sorts; despite the layoffs, Microsoft headcount is flat. Their AI division is hiring madly while other businesses like Windows and Xbox are laying off. Software engineering is a broad field and ML/AI require a distinct domain knowledge that is not easily transferable to game developers or system developers.