Workers are concerned that they’ll lose their jobs to AI—and Bill Gates thinks they are right to worry. He has identified a loophole which may be incentivizing businesses to shift away from human capital to robots.

Last week, Pew Research released a study finding that 71% of adults think AI will lead to fewer jobs in the United States over the next two decades, up from 64% in 2024—only 5% think it will lead to more jobs. And young people, those whose job prospects are most likely to be impacted over the long run, are equally as concerned as their older counterparts: 73% believe they’ll get fewer career opportunities because of the transformative technology over the next 20 years.

Despite the concerns of the public, and the watchful eye of policymakers like former Fed chairman Jerome Powell, Microsoft co-founder Bill Gates has suggested that employers’ bottom lines, under current tax frameworks, may actually benefit from using AI-empowered machines rather than human workers.

In a new essay posted to his blog, the entrepreneur and philanthropist wrote: “Right now, if you’re an employer and you hire someone, you pay payroll taxes on their earnings. But if you buy a robot, you can usually write it off right away as a business expense. The tax system nudges you toward replacing people with machines.”

As a result, Gates suggests taxing AI tokens and robots.

This revenue generation is also necessary for governments, he suggests, which will be subject to lower revenues from income tax if fewer people are working, combined with greater demand for retraining and social security benefits.

“The funds will have to come from somewhere at a time when budgets are stretched,” Gates writes. His nearly 6,000-word essay came hot off the news that U.S. government borrowing has hit $40 trillion.

“A tax would slow the rush away from human labor a little and raise money for retraining and a stronger safety net,” Gates explained, “It would need to be targeted so it does not slow down the purely beneficial uses of AI, like making medicine and education cheaper.”

Mixed response

Gates made the same suggestion nearly a decade ago and was subject to robust criticism at the time. Writing before the current AI boom, former Treasury Secretary Larry Summers said in 2017 that Gates was “seriously astray” with the suggestion, adding: “Gates’s robot tax risks essentially being protectionism against progress.”

Robert Seamans, a professor at NYU Stern, similarly chimed in 2017 that “there’s no question that the potential increase in robots and automation requires policymakers to rethink fiscal policy for the 21st century (and other policy as well, such as education and retraining policy).”

But he added that “based on the data we currently have, a tax on robots would be bad policy. Robot taxes would dissuade firms from investing in robots, which would lower economic growth, and, to the extent that robots complement labor in some cases, would lead to less hiring and lower wage growth.”

Gates, writing in the piece shared with Fortune ahead of publication this week, is aware of such criticism. But he insists that critics are “not considering the broader value of work for individuals and society. And with all the accelerated innovation we will have, we’ll be able to afford a little inefficiency as the price for keeping people employed.”

He added that while tax is not the whole solution to the threat of AI (the three major risks of which he outlines in the rest of the essay), it is a suggestion that forms part of a “wise response.”

Gates adds: “However we raise money for more assistance, it needs to reach the people who need it most, including workers who lose their jobs to AI and robots, people whose hours or wages decline, and communities where the losses are concentrated.”

This story was originally featured on Fortune.com