A story has been doing the rounds, and it taps something real. White-collar professionals, the argument goes, are not waiting to be replaced by AI. They are walking out first, retraining, downshifting, leaving careers they spent a decade building because they are convinced the work is about to vanish. The emblem of it is Mrinank Sharma, who left a research job at Anthropic to study poetry. His resignation letter, posted to X, has been viewed more than 15 million times.
The feeling behind this is not irrational, and I do not want to wave it away. But the claim underneath, that quitting now is the sensible response to an imminent collapse in white-collar work, deserves to be held up against the numbers. When you do, the picture gets less tidy.
The fear is well documented
On the mood, the surveys are consistent. Pew has found that 71 percent of Americans expect AI to mean fewer jobs over the next twenty years. In other research, 41 percent of Gen Z workers and 37 percent of millennials said they were considering a career change because of AI anxiety. The Guardian's reporting is full of people making real decisions on that basis, including one jobseeker who deliberately skipped 30 to 40 postings that mentioned AI. The anxiety is genuine and widespread. That part is not in question.
The losses, so far, are not
Here is where the story and the evidence part company. The mass displacement that the fear anticipates has not shown up in the aggregate figures. The Economist has estimated that AI has been associated with roughly one million new jobs in the United States against something like 200,000 cuts attributed to it. You can argue with any such estimate, but the direction is the opposite of a rout.
The most telling number is about quitting itself. If workers genuinely believed their industries were collapsing, you would expect them to bolt. Instead, ADP data shows quit rates in office-based roles such as finance and business services at the lowest levels ever recorded. People are not fleeing. They are gripping their desks. The same anxiety that pushes a handful of visible professionals to make a clean break is pushing a far larger number to stay exactly where they are, because the alternative feels worse.
So the headline figure, a wave of workers quitting over AI, sits on top of a reality where most workers are doing the opposite. A dramatic few resignations, amplified to millions of views, are easy to mistake for a trend. The quieter mass of people staying put is the actual trend, and it rarely makes the feed.
What the evidence actually supports
None of this means AI will leave work untouched. It is reshaping tasks now, and the people worried about entry-level roles closing to new graduates or about software engineering jobs thinning out are pointing at pressures that are real and worth watching. The honest reading is narrower than the panic and narrower than the dismissal. Specific tasks are under pressure. Whole careers vanishing on a timeline that justifies quitting today is a forecast, not a fact.
Preemptively abandoning a career is a bet that the worst case arrives soon and arrives for you. For a few people, in a few roles, that bet may pay off. As a general strategy, it asks you to act on a prediction that the current data does not yet support. The fear is reasonable. Treating it as settled fact is the part to resist.
Commentarii · 0