🔍 Read the full analysis: A Closer Look At AI And Three Forms Of Unemployment on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Thorsten Meyer’s November 16, 2026, article explains how official unemployment is measured and distinguishes frictional, structural and cyclical joblessness. It classifies technological unemployment as generally structural, while noting that the source presents no evidence that AI has caused mass unemployment; it points instead to a higher share of long-term unemployed Americans.
In a November 16, 2026, article, Thorsten Meyer explains how economists distinguish frictional, structural and cyclical unemployment and why job losses linked to artificial intelligence would usually be counted as structural. The piece reports no current mass unemployment attributed to AI, but highlights a rise in the share of unemployed Americans searching for work for at least six months.
Meyer’s article, published by ThorstenMeyerAI.com as Episode 14 of the series “Work After Work,” sets out the US Bureau of Labor Statistics’ basic test for being counted as unemployed: a person must have had no job in the reference week, been available to start work and actively looked for a job during the previous four weeks. People who want work but have stopped searching are not included in the headline unemployment count.
The article cites an US unemployment rate of 4.1% in October 2026. It says 27.6% of unemployed Americans had been looking for work for at least 27 weeks, compared with about one in five in 2019. The median spell of unemployment that month was 11.9 weeks. These figures describe duration and the overall rate; the article does not attribute the changes to AI.
Meyer distinguishes short-term job searches between positions, or frictional unemployment, from structural unemployment caused by mismatches between workers’ skills or locations and available jobs. Cyclical unemployment rises when a recession reduces demand for workers. Technological job displacement is generally treated as structural, the article says, because it can leave workers’ skills out of step with employers’ needs.
Why Long Job Searches Matter
The distinction matters because the headline unemployment rate does not capture everyone who lacks work or wants a job. People who have stopped searching are excluded from the labor force, while the share of unemployed people out of work for a long period can reveal strain that the overall rate alone does not show. Meyer’s cited 27.6% long-term unemployment share is therefore a separate indicator, not proof of an AI-driven jobs crisis.
The categories also point to different causes. A recession can produce cyclical unemployment across industries; a person between jobs may be experiencing frictional unemployment; and a persistent skills mismatch may be structural. If AI replaces tasks faster than workers can move into other roles, that could add to structural pressures. The article explains that classification, but does not establish the scale of AI-related displacement or show that it has caused the reported labor-market figures.
For readers, this is a useful distinction when evaluating claims that AI is “taking jobs.” A change in tasks, a job loss and a sustained rise in unemployment are not the same measure. The source’s figures provide a snapshot of US unemployment in October 2026, while leaving the causes of long job searches open.
Popular right now

VIZIO 65-Inch V-Series 4K HDR Smart TV

KitchenAid 5.5-Quart Bowl-Lift Stand Mixer, Porcelain White

Ninja Foodi DZ401 DualZone XL Air Fryer, 10 Qt, 6-in-1, Grey
As an affiliate, we earn on qualifying purchases.
How Unemployment Is Counted
The article traces the modern US unemployment count to the Great Depression. The Bureau of Labor Statistics says there was no monthly survey of unemployed people before the 1930s; mass unemployment increased demand for regular statistics. Meyer cites historical estimates that unemployment rose from about 3.2% in 1929 to 24.9% in 1933. The monthly survey behind the familiar figures began in 1940, according to the article, and the BLS says the definition has changed little since.
Meyer also recalls John Maynard Keynes’s 1930 essay, “Economic Possibilities for our Grandchildren,” which described “technological unemployment” as labor-saving advances outpacing the creation of new uses for workers. Keynes called it a temporary phase of maladjustment. That historical observation frames the article’s question about AI; it does not settle whether today’s technology will produce the same adjustment or how long it might take.
As evidence that work changes over time, the article cites research by David Autor and colleagues estimating that about 60% of US jobs in 2018 were in occupational titles that did not exist in 1940. That comparison concerns the evolution of job titles across decades, not a guarantee that workers displaced by current AI systems will find equivalent work quickly or in the same places.
“Before the 1930s, there was no monthly survey to count the unemployed.”
— Thorsten Meyer, citing the US Bureau of Labor Statistics
The article does not quantify how many workers have lost jobs because of AI, identify affected occupations, or establish a causal link between AI adoption and the October 2026 unemployment figures. It also does not provide a forecast of how quickly displaced workers might find other jobs. The long-term unemployment figures are indicators of duration, not evidence on their own of the reason people remain unemployed.
International comparisons require care. Meyer reports that 36.0% of registered unemployed people in Germany—1,063,000 people—had been out of work for more than a year, citing the Federal Employment Agency. The US and German figures use different definitions and time thresholds, so they should not be treated as directly comparable rates. The article’s central question—whether AI will lead to lasting joblessness at scale—remains unresolved.
What Future Data Can Show
The next evidence needed is a clearer record of how AI adoption affects hiring, layoffs, hours and workers’ ability to move into new roles. Labor-market statistics can show whether unemployment and job-search duration change, but attributing a change to AI requires more than observing that the technology is spreading at the same time.
Meyer’s article offers a framework for reading those developments: distinguish short periods between jobs from persistent skills or location mismatches, and separate both from recession-driven losses. Future unemployment releases and research tracking specific industries will help establish whether AI is changing the number and type of available jobs. The source does not identify a scheduled next report or make a forecast.
Key Questions
What are the three main forms of unemployment described?
Frictional unemployment occurs while people search between jobs; structural unemployment reflects mismatches in skills or location; and cyclical unemployment is associated with job losses during economic downturns.
How does the article classify technological unemployment?
Meyer says economists usually group technological unemployment under structural unemployment, because technology can change the skills employers need and the work available. The article does not estimate how many current job losses are caused by AI.
Who is counted as unemployed in the US measure?
Under the definition described in the article, someone must have had no job during the reference week, been available to work and actively searched in the previous four weeks. People who want work but have stopped searching are not counted as unemployed in the headline rate.
Does the 4.1% October 2026 rate show that AI is causing mass unemployment?
No. The article reports a 4.1% US unemployment rate in October 2026 but does not attribute it to AI. It says there is no mass unemployment in its account and notes that the causes of the increase in long-term job searches are not established.
Source: ThorstenMeyerAI.com
