Every warning about AI and jobs assumes the technology takes work away from people. Healthcare breaks that pattern. The World Health Organization projects the world will be short 11 million health workers by 2030, with the worst gaps in Africa and the Middle East, and more than half of that shortfall made up of missing nurses.
This is not a demand problem. It is a training problem. Back in 2016, AI pioneer Geoffrey Hinton told medical schools to stop training radiologists, predicting machines would fully replace them within five years. Almost a decade later, more radiologists are working today than before AI tools arrived, not fewer.
The real bottleneck sits earlier in the pipeline. In the last school year alone, US nursing programs turned away more than 80,000 qualified applicants, not because those students failed to qualify, but because schools ran out of teachers and clinical training seats. Meanwhile healthcare keeps hiring: it added roughly 82,000 jobs in January 2026 alone, nearly two thirds of every new job created in the country that month. The openings exist. The training slots do not.
This is where AI is actually being pointed: at training, not at replacing. Companies are building simulated patient scenarios that let a nursing student practice a rare emergency dozens of times instead of waiting years for it to happen on an actual shift. One such company, Stepful, just raised 55 million dollars from investors including Oak HC/FT, pushing its total funding past 100 million dollars, to build AI-based training directly inside hospitals and clinics rather than on a college campus.
The person who wrote the original piece making this argument works for one of these training companies, so the case for AI-powered training doubles as a pitch for that business. That does not make the underlying shortage numbers wrong, but it means any single vendor's promise to turn training into "months instead of years" deserves the same scrutiny you would give a sales deck.
The money at stake is real regardless of who is selling what. US health spending already swallows 18 percent of the country's entire economic output and is on track to hit 20 percent within eight years. McKinsey estimates that closing the worldwide worker shortage could add 1.1 trillion dollars to the global economy. A new federal program called Workforce Pell, launching in 2026, will let students use financial aid for short job-training courses instead of full degrees, which lines up neatly with this shift toward faster, employer-based training.
For any business, healthcare or not, the lesson is that AI's effect on jobs depends entirely on where it gets pointed. Aimed at diagnosis, AI has barely dented headcount in ten years. Aimed at training, it can shrink the years it takes to build a qualified worker down to months. Any industry with a similar training bottleneck, from skilled trades to aviation maintenance, should be asking whether the same approach applies to them.