For decades, automation has chipped away at manufacturing and other blue-collar roles, while office workers largely watched from the sidelines. That dynamic is now shifting, according to MIT economist David Autor, who says generative AI is beginning to target the kind of high-paying, degree-requiring jobs that were once considered safe from mechanization.
In an interview with The Atlantic's Annie Lowrey, Autor explained that earlier computer systems followed fixed procedures, but the latest AI models can learn and improvise. "Before, progress was linear and predictable," he said. "You figured out the steps and the computer followed them. It followed the procedure; it didn't learn and it didn't improvise." This new capability, he argues, makes previously non-automatable tasks vulnerable to replacement.
The warning comes amid a week of controversy at CNET, where the tech news site was found to have quietly published AI-generated financial explainers without disclosing the lack of human authorship. The move drew criticism from journalists and raised questions about the reliability of automated content. SEO spammers, however, have embraced the new text-generating tools to churn out low-quality articles at scale.
Journalism is just one example. Autor's analysis points to a broader impact across white-collar fields, including computer science and healthcare. The shift represents a departure from the past, where automation primarily affected routine manual labor, leaving cognitive work largely untouched.
Why This Time Is Different
The key difference, according to Autor, is that generative AI doesn't just follow a script. It can absorb vast amounts of data, recognize patterns, and produce original output—tasks that were once the exclusive domain of human expertise. This has led to predictions that entry-level positions, often the gateway to specialized careers, may be the first to disappear.
Autor does see a silver lining. He told The Atlantic that AI could help people use their expertise more effectively, potentially leading to greater specialization. "In many ways, AI will help people use expertise better," he said. "It means that we'll specialize more." But he acknowledged that securing a specialty role is already difficult, and the loss of entry-level jobs could make it even harder to break into those fields.
The long-term outlook remains uncertain. Historically, the job market has adapted to major technological shifts, but the pace and scope of AI-driven change are unprecedented. For now, workers in white-collar industries may have reason to feel uneasy as they watch the machines learn new tricks.
