Unruly Republic: Is AI the Next Climate Change?

"People should stop training radiologists now. It's just completely obvious that within five years, deep learning is going to do better than radiologists." So pronounced the cognitive scientist Geoffrey Hinton, colloquially known as the Godfather of AI, a decade ago. He was awarded the Nobel Prize in Physics in 2024 for his work in artificial intelligence. Mr. Hinton thought AI would make radiologists useless. They have since grown in number, demand and income.

 

Mr. Hinton's claim was among the earliest that AI would make a whole class of human practitioners redundant. Others have come at regular intervals since. In 2023 a Goldman Sachs study concluded that "roughly two-thirds" of U.S. jobs are "exposed to some degree of automation by AI" and that most of those "have a significant -- but partial -- share of their workload (25%-50%) that can be replaced." Hedged language aside, that sounds like a lot of people on the unemployment rolls.

 

Studies like Goldman's have generally shown more nuance than media reports and political pronouncements on supposed AI job loss. "Amazon axes 16,000 jobs as it pushes AI and efficiency," Reuters announced last week. Politicians of a progressive bent do their best to dramatize the threat. AI and automation "could replace nearly 100 million jobs over the next ten years, including 89% of fast food and counter workers, 64% of accountants and 47% of truck drivers," says a report by Democratic members of the Senate Health, Education, Labor and Pensions Committee.

 

Last year Anthropic CEO Dario Amodei predicted that AI could displace about half of all entry-level white-collar jobs over the next five years, even as it drives productivity and growth to new heights. He reiterated that prognosis last week in a 20,000-word essay.

 

Much of what he writes in this piece -- on the dangers of crazy people and rogue regimes accessing powerful AI tools, and on the ill-advisability of selling semiconductors to China -- is perceptive and interesting. His predictions about AI and the labor market, less so.

 

On the subject of AI replacing heretofore normal human activities, Mr. Amodei, born in 1983, writes with the wide-eyed wonder of a high-school-aged techno-utopian. "I'm sure there are some tasks for which a human touch really is important," he reflects, "but I'm not sure how many." When he states that "in the end AI will be able to do everything," he means it literally.

 

Mr. Amodei seems to base his prediction of massive and imminent job loss on his technological expertise, which is vast, rather than on his understanding of economics, which isn't. Nobody knows what AI's got in store for the economy, but less histrionic forecasts recommend themselves. Stephen Lewarne argued last week in these pages, cogently in my view, that new AI tools won't replace workers but allow them to perform higher-value tasks. Anyway if the threat to labor markets were as dire as Mr. Amodei supposes, job losses would be happening already. He acknowledges, in a parenthetical aside, that they likely aren't.

 

A generous donor to Democratic candidates, Mr. Amodei speaks vaguely of the need for "progressive taxation" as a "natural policy response" to the dire outcome he foresees, and of drastic "macroeconomic interventions": "I think the extreme levels of inequality predicted in this essay justify a more robust tax policy on basic moral grounds."

 

America's political class will happily supply the interventions he suggests. Sens. Josh Hawley (R., Mo.) and Mark Warner (D., Va.) want to require companies to report all AI-related layoffs, presumably with the aim of punitive taxation. Sen. Bernie Sanders (I., Vt.) favors a "robot tax" on companies that replace workers with AI -- an idea supported by Bill Gates, among other politically engaged billionaires. Proposals abound to boost funding for job-retraining programs -- their dismal record of failure notwithstanding. Expect more interventions as predictions grow more dire.

 

The world-weary consumer of news might reasonably wonder if we're headed into another version of climate alarmism. Consider:

 

A confederation of specialists -- climate scientists in one version, Silicon Valley geniuses in another -- joins with liberal politicians and nonprofit heads to warn of an impending catastrophe. The only moral response to this new situation, these Olympians tell us, is to transfer authority over large parts of the economy to people like themselves. That they would favor such a transfer under any circumstances, with or without a coming disaster, doesn't bother the mainstream press, which reports their predictions with credulity and fervor. Meanwhile ordinary people, lacking the specialized knowledge to draw their own conclusions, feel cowed into going along with it all.

 

The economic cataclysm caused by artificial intelligence may never come. But like Godot, its arrival will remain always imminent, never actual. Politicians, celebrities, journalists and scientists who have long predicted impending climate catastrophe have paid no price for their words. Nor will Mr. Amodei in the event that his prophecies come to nothing. Only the "macroeconomic interventions" his words inspire will survive.

 

As for Geoffrey Hinton, he barely admits he was wrong about radiologists. "In retrospect," says a 2025 New York Times report, "he believes he spoke too broadly in 2016. . . . He didn't make clear that he was speaking purely about image analysis, and was wrong on timing but not the direction."” [1]

 

1. Unruly Republic: Is AI the Next Climate Change? Swaim, Barton.  Wall Street Journal, Eastern edition; New York, N.Y.. 05 Feb 2026: A13.