Problem of the Day: AI Job Loss

April 2026 Goldman Sachs research finds that AI isn’t just helping workers, it’s replacing them — 16,000 net lost jobs every month in the U.S. alone. 40 percent of companies, according to Beautiful.ai, are using AI to shrink staff and lower wages. With Boston Consulting Group (BCG) forecasting that up to 25 million positions could vanish by 2031, the door is shutting on entry-level workers and stripping trillions from household incomes. Without a clear plan, the effects could go beyond individual poverty, depression, substance abuse, and violence, to even societal dissolution, chaos.

We’re not just talking about manual laborers. College graduates, even in traditionally safer fields like STEM, business, or accounting are vulnerable in a world which AI agents can do so much and getting ever better, and if not, plenty bright people are happy to work remotely for a relative pittance, witness, for example, all the offshoring to India.

Often-proposed solutions are too problematic:

Universal basic income. Unless standard of living provided by UBI is dropped below what activist groups and many politicians would accept, UBI is too expensive even if taxes were increased significantly, e.g., AI “excess profits” tax and increased income tax. For the revenue to be sufficient, it would eat corporations’ seed corn: money that would have been used for R&D, hiring people, and investor return. And the middle class is already hollowing out, with ever fewer perks of working hard and delaying gratification compared with welfare recipients. Tax them yet more and it unfairly hits the people who make this world run, and voters would opt for other candidates, which would make it a non-starter for many incumbent politicians.

30-hour workweek. That would redistribute hours from the most desirable: those that employers or customers are willing to pay to work 50 hours a week: e.g., top cardiologists, AI software engineers, haircutters, etc., to people unable to get even 30 hours a week. That reduces all of our lives to a lower common denominator.

Partial fixes

Ivy League Plan. Students at Ivy League Columbia and Penn pay $400,000 sticker price for four years to live two or three to a dorm room, eat cafeteria food, and get just basic health care in a high-crime location. The argument would go, “If it’s good enough for Ivy Leaguers, it’s good enough for people living off the taxpayer.” Yes, the formerly employed middle-class people would object, but its fairness would, ahem, trump.

SmallService Corps. Many people don’t adequately consider hiring help such as tech tutor, homework helper, block beautifier, even pothole repairer. The government could promote the idea, including a match.com-like website that would pair such people with job seekers. Those workers putting together a few such gigs, might well make a sustainable living doing varied tasks more pleasant than traditional jobs.

Entrepreneurship education. New and expanding businesses create jobs. But most students go through even college with little training in the art and science of starting a business. A K-20 entrepreneurship curriculum would help. In a world of very high potential unemployment, teaching quadratic equations, stochastic processes, and the intricacies of Shakespeare is education malpractice.

WPA II. To help the U.S. out of the Depression, FDR created a massive taxpayer-paid jobs program called the Work Projects Administration (WPA.) Those jobs gave meaning, stability, income, and national beautification for millions of people. A WPA II would help. Make it voluntary and it attracts better workers than if mandated. Also, that avoids worker resentment. Of course, given the likely high percentage of unemployed people, the cost could be prohibitive, even with major tax hikes that would eat our seed corn: private-sector created jobs and R&D.

Even if all the above were adopted, which is unlikely, it wouldn’t completely solve. But the perfect is the enemy of the good, and this is the best plan I can come up with. Have any better ideas?

I read this aloud on YouTube.

Is there a problem — policy or personal — that you’d like to see me tackle as a Problem of the Day?

Marty Nemko holds a Berkeley Ph.D specializing in the evaluation of innovation. You can reach him at mnemko@gmail.com