ChatGPT is coming for your job.

That’s the fear about the rapid advances in artificial intelligence.

In a headline the other day, Axios warned of a “white-collar bloodbath.” The CEO of the artificial intelligence firm Anthropic told the publication that AI could destroy half of all entry-level white-collar jobs in the next one to five years and drive the unemployment rate up to 10-20%, or roughly Great Depression levels.

This sounds dire, but we’ve been here before. In the 1930s, John Maynard Keynes thought that labor-saving devices were “outrunning the pace at which we can find new uses for labor.” Analysts thought the same thing in the 1960s, when John F. Kennedy warned “the automation problem is as important as any we face,” and in our era, too.

If a prediction has been consistently wrong, it doesn’t necessarily mean that it will forever be wrong. Still, we shouldn’t have much confidence in the same alarmism, repeated for the same reasons.

If technological advance was really a net killer of jobs, the labor market should have been in decline since the invention of the wheel.

Instead, we live in a time of technological marvels, and the unemployment rate is 4.2%. Rob Atkinson of the Information Technology and Innovation Foundation points out that the average unemployment rate in the United States hasn’t changed much over the last century, despite productivity — the ability to produce more with the same inputs — increasing almost 10 times.

Technology increases productivity, driving down costs and making it possible to invest and spend on other things, creating new jobs that replace the old. This is the process of a society becoming wealthier, and it’s why nations that innovate are better off than those that don’t.

The rise of personal computers collapsed the demand for typists and word processors. These positions were often held by women. Did this decimate the economic prospects of women in America? No, they got different, and frequently better, jobs.

Spreadsheets drastically reduced the demand for bookkeepers and accounting clerks. Did this end the profession of accounting? No, there was an increase in more sophisticated accounting roles.

The job market has never been stuck in amber. The MIT economist David Autor co-authored a study that found that the majority of current jobs are in occupational categories that arose since 1940.

It’s true that artificial intelligence is projected to affect white-collar jobs — computer programming, consulting, law and the like — more than prior waves of technological change. But these kinds of jobs shouldn’t be immune from the effects of automation any more than factory work has been.

AI will end up augmenting many jobs — helping workers become more efficient — and there will be a limit to how much it can encroach on human work.

It’s hard to imagine, say, Meta ever giving over its legal representation in an antitrust case to artificial intelligence. Lawyers handling such a case will, however, rely on AI for more and more support, diminishing the need for junior lawyers.

This will be a significant disruption for the legal profession, yet legal services will also become cheaper and more widely available, in a benefit to everyone else.

There’s no doubt that the changes wrought by technology can be painful, and it’s possible that artificial intelligence eventually gets so good at so many tasks that people have no ready recourse to new, better jobs, as has always happened in the past.

The potential upside, though, is vast. After strong productivity growth for about a decade beginning in the mid-1990s, we shifted into a lower gear in the mid-2000s. It will be a boon if artificial intelligence puts us on a better trajectory. An era of high productivity growth will, among other things, make it easier to deal with the budget deficit and the fiscal strain of retiring baby boomers.

Like anything else, AI will have its downsides, but it’s not an inherent threat any more than computers or the internet.

© 2025 by King Features Syndicate