AI, work and the next downturn
Machines are beginning to take over tasks that people are paid to do. If that continues, income may move from wages to the owners of what remains scarce, chiefly land, location and power, and the next downturn could cost jobs that do not return. This site sets out that argument, follows the evidence as it arrives, and publishes the model behind it for anyone to run.
The current assessment
- Now
- Long-term borrowing costs are rising across rich countries faster than their economies are weakening, while the labour market is quiet on the surface but hollow underneath.
- The coming months
- The most likely course over the next six to twelve months is higher rates for longer, with fiscal tightening in Europe and a growing risk that the AI investment boom stalls on its financing costs.
- The longer run
- To about 2030-32, the evidence is consistent with a slow but persistent shift of income from wages toward the owners of capital and sites, with public budgets under growing strain.
Written by Claude, an AI model, on 7 Oct 2026, 15:52 Stockholm time, from the 100 items flagged on this site since 24 September 2026, the recent ones weighted most. It is revised as the evidence changes, and it may depart from our own view below. The reasoning, and how it has changed
Key facts
- 107%of net US job growth from Aug 2024 to Aug 2026 came from health care and social assistance. Outside care, private employers added 1 thousand jobs and government shed 92 thousand, as federal cuts outweighed state and local hiring.
- 128basis points is the extra yield France paid over Germany on ten-year debt on 6 Oct 2026, against a high of 143 since 2019.
- 93.4is labour's share of US business output (2017 = 100), the lowest since records began in 1947, down 3.3 points in a year.
- 0.6 pointsis how far the share of Americans aged 25–54 in the workforce, working or looking for work, fell in June 2026, after holding between 83.0% and 84.0% since 2023. People who stop looking are not counted as unemployed.
- 24%of the US workforce is out of work by Q2 2032 in the model after a dot-com-sized bust, with today's policies and AI taking jobs at the pace the model assumes; 16% at the fastest pace the trend in the jobs data allows, 18% at the pace measured to their latest months.
- 4.8×is how far US wages measured in durable goods and measured in shelter diverged between 1964 and 2024.
What we think is happening
AI has, on our reading, been displacing work for two to four years, and the pace may now be increasing. A stock market bust larger and faster than the dot-com crash seems to us plausible, and concern over government debt and a round of tariffs could turn it into a recession in which the lost jobs do not return. If that reading is right, the model puts 14–16% of the US workforce out of work by the end of 2028, against 10% in its central case. Worse is not hard to conceive of: if firms automate twice the usual share of the jobs they cut, the figure is about 46% by mid-2032, and still rising. Our reading, point by point
What the indicators show
8 of 29 indicators have triggered: Jobs outside care, Labour's share of income, People leaving the workforce, The premium on US government debt, France's borrowing costs, Oil prices, US mortgage rates, Japan's bond yields. Each indicator is checked against a fixed trigger. About the indicators
The combined signal
Jobs
- US unemployment4.2%Normal
- US jobless claims200kNormal
- Permanent job losses25%Normal
- US hiring3.3%Normal
- Office jobs−0.0%Normal
- Factory and warehouse jobs+0.5%Normal
- Jobs outside care−24kTriggered
- Labour's share of income93.4Triggered
- People leaving the workforce−0.33 ptsTriggered
- Swedish unemployment8.9%Normal
Markets and credit
- Technology stocks+0%Normal
- Corporate credit3.1%Normal
- Financial stress−0.81Normal
- Long-run inflation expectations2.35%Normal
Government bonds and interest rates
- The premium on US government debt1.08Triggered
- France's borrowing costs128 basis pointsTriggered
- Other euro-area borrowing costsClose to triggering
- Sudden moves in bond yieldsClose to triggering
- Expected interest ratesNormal
- Riksbank policy rate1.75%Normal
Energy, housing and credit
- Oil prices$114Triggered
- US mortgage rates7.28%Triggered
- US mortgage arrears1.86%Normal
- US house prices+1.9%Normal
- Bank lending to non-bank lenders$2.07tnNormal
Japan and Korea
Latest
- US 30-year yield hits a fresh 24-year high near 5.71% on 7 Oct as Brent returns above $100; 30-year gilt 6.02%
- Sweden's CPIF inflation 1.5% in Sep (flash, 7 Oct), up from 0.7%; excluding energy 0.5%, unchanged
- HSBC to cut ~70% of UK wealth advisers and ~half of management and specialist roles in AI push (FT, 7 Oct)
- OpenAI (6 Oct): internal model produced 722 maths manuscripts on open problems; 185 main results checked in Lean
- US trade deficit widens 13.7% to $105.6bn in August (6 Oct); capital-goods imports a record $146.4bn
The guide
The argument and the evidence in nine short sections.
Our view
What we think may be happening, what the model shows if we are right, and what would show that we are wrong.
The model
Choose the shock and the policies, and see what follows in four economies.
The indicators
The data that would show it early, each against a fixed trigger, back to 1990.