Teach AI Your Job, Then Ask If Your Job Is Safe

Smart Ass Take:

The 60 Minutes segment opens with Jon Wertheim speaking to the camera, except it isn’t Jon Wertheim, it’s AI Jon Wertheim, which is a hell of a way for 60 Minutes to admit it’s nervous too. Then we meet Mercor, run by a 23-year-old billionaire, which pays more than 100,000 experts to teach AI everything they know, at night, after their real jobs, for side money. Every one of them insists their career is safe. My favorite bit is the Columbus trial lawyer whose AI clone read her draft opening statement and told her it was too stiff. She says clients will always want to “look their lawyer in the eye.” Maybe. I’d have said the same about bank tellers, and I haven’t looked one in the eye since the Clinton administration. This is a LABOR problem disguised as a productivity miracle: if your work is your identity and someone offers to digitize the identity, read the fine print. Who needs humans? The machines do, apparently. For now. By the hour.

Article Excerpt:

“So far this year, U.S. companies have cited AI as the No. 1 reason for layoffs. AI, after all, doesn’t sleep, or take sick days.“

Article Summary:

60 Minutes correspondent Jon Wertheim asks the question everyone with a paycheck is quietly googling: will AI take my job? He opens with an AI-generated version of himself and notes that AI is the No. 1 reason U.S. companies have cited for layoffs this year.

First stop: Mercor, a San Francisco startup whose 23-year-old CEO, Brendan Foody, is now a billionaire. It pays more than 100,000 experts, including a winemaker, a music producer and a physician, to train models like ChatGPT and Claude. They insist their careers are safe. Foody calls the shift in human work a natural “evolution.” Next, Columbus law firm Vorys has built AI clones of its senior lawyers. Partner Kim Herlihy’s double critiques her opening statement, yet she says clients will always want to look their lawyer in the eye.

Then the segment turns. Hiring has slowed and wages have dropped in the most AI-exposed jobs, and 20-somethings are hit hardest. Former Salesforce and Meta AI executive Clara Shih quit to help young people navigate it, warning that the entry-level work AI now does well is what young hires used to get, and that job loss means lost self-worth and fractured communities. She doesn’t believe the AI labs’ new “people at the center” message. Nearly three-fourths of Americans fear for their jobs. Mercor trainers earn $20 to $200 an hour, but many say the work is erratic, part-time and benefit-free.

Back to Foody, who claims top economists expect more jobs in 10 years. Cut to Nobel laureate Daron Acemoglu of MIT says: “No. Not on our current trajectory.” AI is moving in one or two years, across many sectors, what took the Industrial Revolution 80. Training AI is a small job sector next to the number of people who will be replaced, which he calls B.S. Unemployment could triple in a decade if we do nothing.

Acemoglu isn’t anti-technology. He wants AI that makes humans better rather than expendable, and a tax code that rewards hiring real people. But the window is closing, and he fears for society’s stability. Should that scare the hell out of us? “It should,” he says, and it should push us toward action, because the future is ours to choose.

Watch the Full Episode: