Will AI Make Doctors Better—or Make Them Less Skilled?

Résumé détaillé

As AI takes on complex medical tasks, health leaders debate whether it will erode doctors’ skills or enhance them.

As artificial intelligence becomes capable of taking on more complex medical tasks, health leaders are grappling with whether the technology could erode doctors’ skills – or make clinicians who learn to use it more valuable.

Dr Eric Poon, chief health information officer at Duke Health, raised the prospect of “deskilling” as AI advances from performing simple tasks to helping diagnose difficult cases.

“When the AI tools can start prompting clinicians not to just do simple things like order a mammogram for women over age 40, it can perhaps even diagnose difficult cases much more accurately at lower costs,” Poon said at Newsweek’s AI Health Summit in New York.

“We have to think about what would happen to our clinicians at the bedside,” he said. “Would this lead to some kind of deskilling?”

The concern comes as a Harvard-led study earlier this year found that an advanced AI agent outperformed human doctors in a series of demanding tests that assess the ability to correctly diagnose patient illnesses.

OpenAI’s “o1-preview” demonstrated its ability to handle emergency-room triage scenarios, recommend diagnostic tests and manage clinical cases at a level that matched or surpassed well-trained physicians.

Poon said the question has implications for how future doctors are trained.

“What is our obligation as we prepare our next generation of clinicians?” Poon asked. “I want great clinicians to take care of me when I am old. So, how should we train the next generation?”

Poon said the issue becomes more important as clinicians grow increasingly dependent on technology and AI.

Début de l'article, extrait de Newsweek. Pour lire la suite, ouvrez l'article chez le média.