Several occupations named as near-certain casualties of earlier automation are still here in numbers. Paralegals and legal assistants are projected at 376,200 in 2024 and 376,800 in 2034. Radiologic technologists grow 4.3 per cent. Travel agents grow 2.2 per cent. Bank tellers do decline, at 12.9 per cent, but from 347,400 people — a slow contraction rather than the disappearance that was forecast.
Paralegals were the textbook case
Document review was the example everyone reached for: repetitive, rule-governed, high-volume and expensive. Software genuinely did absorb most of it.
The occupation is projected at 376,200 people in 2024 and 376,800 in 2034. Flat, not gone. The automatable part left and the rest stayed — client contact, case management, knowing which document matters and why, and being the person an attorney can ask a question.
Three more that were written off
Radiologic technologists. Imaging was supposed to be an early casualty of pattern recognition. Projected to grow 4.3 per cent to about 237,800.
Travel agents. Written off completely when booking moved online, and the occupation did shrink hard in the 2000s. Now projected to grow 2.2 per cent, having settled into complex, multi-leg and high-value trips that a search box handles badly.
Bank tellers. The classic case, and the one where the prediction was directionally right and wrong about pace by a generation. Down 12.9 per cent over the coming decade from 347,400 people — decades after the cash machine arrived.
Where the forecasts landed hard
Word processors and typists are projected down 36.1 per cent, to about 25,600 people. Data entry keyers down 25.9 per cent. Those predictions were accurate and then some, and it would be dishonest to write this article as though the forecasters were simply wrong.
The pattern separating hits from misses is clean: the forecasts were right where an occupation’s entire output was a single automatable task, and wrong wherever the job was a bundle of several. Typing is a task. Being a paralegal is not.
Why the misses are systematic, not unlucky
Scoring a whole occupation assumes it is one thing. Most occupations are five or six things, and the automatable ones leave first while the rest consolidate into a changed job with the same title.
That is the same methodological criticism that took headline automation risk estimates from 47 per cent down to single digits when the analysis moved from occupations to tasks. The old forecasts and the modern risk indices made the identical error for the identical reason, roughly a decade apart.
What survives is consistent across every case
The occupation keeps the part requiring judgment, accountability or a relationship, and sheds the part that was mechanical. That is true of paralegals, radiologic technologists, travel agents and tellers alike.
No exposure score can represent it, because a score attaches to a job title and the title is precisely what stays constant while the content underneath changes completely.
The uncomfortable half
An occupation surviving in changed form still requires the person in it to change. Somebody excellent at document review and indifferent at client management is worse off, even though “paralegal” is a stable occupation and the statistics look reassuring.
Aggregate survival is not individual security, and the aggregate is what gets quoted when somebody wants to be encouraging.
Running the check yourself
Take a projection published a decade ago for a year now past, find the same occupation in current employment data, and compare. Every release is archived, so this is genuinely available rather than a rhetorical suggestion.
Watch for occupational reclassification, which explains a share of apparent misses that are really definitional — an occupation split into two is not comparable with its earlier self. And expect the demographic projections to have held up considerably better than anything resting on how fast a technology was adopted.
What this justifies believing
That occupations rarely vanish and frequently change beyond recognition. That confident write-offs are usually about a task and get reported as being about a job. And that the useful planning question concerns your week rather than your job title.
It also guards against the opposite error. These forecasts were right about typists and data entry, and being right about the clear cases is exactly what makes them worth reading carefully on the unclear ones.
Common questions
Did paralegals get automated away?
No. Projected at 376,200 in 2024 and 376,800 in 2034 — essentially flat, after document review genuinely was absorbed by software.
What about bank tellers?
Declining at 12.9 per cent over the decade from 347,400 people. Directionally right, and far slower than predicted, decades after the cash machine.
Were any predictions accurate?
Yes, where an occupation's entire output was one automatable task. Word processors and typists are projected down 36.1 per cent.
Why were the misses systematic?
Because scoring a whole occupation assumes it is one thing. Most are bundles, and the automatable parts leave first.
What should I take from this?
Occupations rarely vanish and often change beyond recognition. Ask which of your tasks stay, not whether the title survives.
Did paralegals get automated away?
No. Projected at 376,200 in 2024 and 376,800 in 2034 — flat, after software genuinely did absorb most of document review.
Which write-offs turned out to be accurate?
Word processors and typists, projected down 36.1 per cent to about 25,600, and data entry keyers down 25.9 per cent.
What separates the accurate forecasts from the wrong ones?
They were right where an occupation's whole output was one automatable task, and wrong wherever the job was a bundle of several.