In 2016, when a Texas Tech researcher first asked a national sample how afraid they were of various things, from heights to spiders to artificial intelligence, about one in six people said they feared AI would take their job. Back then, as professor Austin Lee points out, the AI most people met was Siri or Alexa. By 2025 the same survey found the figure had climbed to 49.2 percent. In under a decade, dread of being automated out of work has gone from a fringe worry to something close to a coin flip, and this week that anxiety collided with a fresh round of corporate layoffs that seemed to confirm it.
The layoffs came from Oracle, which reportedly began cutting staff again in mid-September, only months after its headcount fell by roughly 21,000, about 13 percent, over its last fiscal year. What makes Oracle notable is its candour: in regulatory filings the company states plainly that "the adoption and deployment of AI technologies across our operations have resulted, and may continue to result, in reductions to our workforce." It is pouring billions into AI data centres while trimming the humans, and it has raised the projected cost of its restructuring plan to around $2.8 billion. For anyone worried that AI is coming for jobs, Oracle reads like exhibit A.
And yet the researchers who track the fear are careful to say it is not the same thing as the fact. Lee and his colleague Milad Jalalian Ebrahimi compared their fear data against economic indicators such as unemployment and found no significant correlation between how scared people are and what the labour market actually does. Fear, it turns out, tracks the headlines. It dipped during the pandemic, when disease crowded out other worries, then spiked after ChatGPT arrived. The top-line numbers remain stubbornly calm: US unemployment sits at 4.1 percent, and more than 160,000 jobs were added in August.
The gap between the two matters, because fear does not fall evenly. The Texas Tech team found the anxiety gap between men and women has nearly doubled over ten years, a sign, Lee warns, that AI could deepen inequalities that already exist. The old divide by education, in which less-educated workers feared automation most, has narrowed to the point where blue-collar and white-collar workers now report roughly equal dread. Into that mood arrive listicles like one circulating this week naming 65 jobs supposedly at "zero percent" risk, heavy on nurses, therapists and surgeons, the roles that lean on empathy, physical dexterity and judgement.
The honest reading is that both things are true at once. Individual companies really are using AI to justify cuts, and the aggregate data really does not yet show a machine-driven jobs collapse. That is an uncomfortable place to sit, because it means the fear is neither vindicated nor imaginary. The researchers offer a historical shrug: this is how people greeted computers too, resisting first and adjusting later. Whether AI follows that pattern or breaks it is, for now, still a question the numbers cannot answer, which may be exactly why the fear keeps climbing.