Eighty per cent of healthcare support work cannot be sped up, and the largest occupation in it employs 1.32 million people at 79 per cent protected.
Exposure means published research judged an AI could halve the time a task takes, at the same quality. It is not a prediction about any job. How to read this
Eighteen per cent sits in the band needing a tool, mostly record keeping and scheduling. The remaining 79 per cent is direct physical care.
Personal Care Aides are the least protected occupation in the group at 64 per cent, and Endoscopy Technicians the most at 100 per cent.
Solid is reachable with a chat window today, pale needs software building first, grey cannot be sped up. Task counts are shown because several of these rest on fewer than twenty.
These occupations are among the hardest to fill in most economies and among the least touched by what this report measures.
Whatever solves that shortage, this evidence says it is not going to be a language model doing the work instead.
There is no data here on what anybody actually does with AI. A dataset that appears to measure it was removed from this report because its meaning could not be stated. The reasoning is on the limits page.
The exposure ratings cover all 350 tasks in these 19 occupations with no gaps, and were made in 2023 against the model capabilities of that year. Read every figure as a floor.
Exposure ratings from Eloundou, Manning, Mishkin and Rock, GPTs are GPTs: Labor market impact potential of LLMs, Science 384, 1306–1308, 2024, used under the MIT licence. This page includes information from the O*NET Database by the U.S. Department of Labor, Employment and Training Administration (USDOL/ETA), used under the CC BY 4.0 license. O*NET® is a trademark of USDOL/ETA. People Team AI has modified all or some of this information. USDOL/ETA has not approved, endorsed, or tested these modifications. Employment figures from the US Bureau of Labor Statistics.