Hardly a week goes by without a new warning that AI will cost millions of jobs – heck, maybe even 300 million jobs by 2030. One early report (2018) predicted a loss of up to 800 million global jobs by 2030.
A more modest 2023 report forecast AI could impact 300 million jobs, and more recently, Anthropic CEO Dario Amodei predicted AI could wipe out half of all entry-level jobs from 2025 to 2030. Also, the latest World Economic Forum estimated 92 million jobs would be displaced by 2030, although they were sane enough to counter this with 170 million new jobs being created, but firings, not hiring, make headlines.
Last week’s Wall Street Journal covered a new Goldman Sachs report saying AI will eventually displace 11 million jobs, above 6% of U.S. workers. The report also talked about a corresponding 30% gain in productivity, boosting and corporate earnings, but the loss of 6% of workers earned the headline, since bad news sells. The report also said companies that discussed AI in the fourth quarter “cut their job openings by 12%,” but any list of job openings implies new hires, more jobs ahead, replacing posts lost. (“Goldman Sachs Predicts AI Will Eventually Displace 6% of U.S. Workers,” Heather Gillers, WSJ, March 3, 2026).
Long-term, as the following chart shows, America has created an average of 1.5 million net new jobs per year since 1940, as our robust economy easily absorbed massive new waves of workers: Baby boomers (born 1946 to 1964), former housewives entering the labor force, and millions of new immigrants:
This trend began long before 1940. John D. Rockefeller’s low-cost kerosene put whale hunters in dry dock. Autos and trains put horseshoe makers and carriage makers out of work. Telephones put telegram clickers on hold. Companies are always at risk if they don’t keep up with newer and better technologies.
Back in the 1930s, John Maynard Keynes, the dominant economist of that decade, warned of a coming era of “technological unemployment.” Others called industrial automation a “Frankenstein monster” destined to destroy jobs, and a committee of dominant Keynesian economists of the 1960s warned President Lyndon Johnson about a “cyber-nation revolution” (computers), wiping out manual labor.
Last Friday’s Jobs Report Deepened the Job Market Gloom
Last Friday, the market continued its gloomy late-winter mood after the came out, saying we’ve seen a job loss of 32,000 since June, while the rose from 4.1% in July to 4.4%:
One thing to notice in this list is the alternating rising months (July, September, November and January) with answering negative months. There’s no consistency in this data, although it says we’re losing jobs.
Turning to the kinds of jobs lost, the information sector only lost 11,000 jobs in February, far exceeded by health services (-28k), due mostly to labor union actions, and leisure and hospitality (-27k), so the loss of jobs by computer programmers to “AI” bots needs some rethinking (as jobs are needed to monitor AI).
Almost daily, we read of companies laying off 4,000 or 10,000 or more, but mention is seldom made of the millions of new jobs these companies created during the previous decade. For instance, the big headlines declared 4,000 layoffs at FedEx, but do you recall any headlines covering their 100,000 net new hires from 2020 to 2022? In a more extreme case, we’ve been assaulted by news that Amazon has eliminated approximately 30,000 “corporate roles” in 2025 and early 2026, but do you recall any headline covering the 1.4 million new jobs Amazon created, 2015 to 2022? (Hiring seldom makes the headlines).
Based on history, AI may reduce demand for certain jobs, but it will inevitably create more positions than it eliminates, as savvy workers and companies will adapt to these emerging technologies.
Jobs are never guaranteed – except for tenured college professors, I suppose – but this sort of dynamic job market is also a recipe for national growth. Any guaranteed lifetime employment leads to what Soviet workers mumbled: “They pretend to pay us, so we pretend to work.” Nothing creates an economic crisis more than “guaranteed employment” (translation: few are hired). Europe’s were in double-digits from 2001 to 2015, but the EU has recently learned how to rethink their “guaranteed jobs.”
Smart workers will always look for greater job security by being irreplaceable in the talents they offer, showing willingness to learn, and serving the customer well.