JAMA Paper Says AI Could Replace Doctors by 2030 as Studies Show 60% Diagnostic Hit Rate
Updated
Updated · Forbes · Aug 17
JAMA Paper Says AI Could Replace Doctors by 2030 as Studies Show 60% Diagnostic Hit Rate
3 articles · Updated · Forbes · Aug 17
Summary
A JAMA-published paper by Neal Khosla, Ezekiel Emanuel and Abe Baker-Butler argues top-tier care should shift to fully autonomous AI, not doctor-AI teams, as soon as 2030.
The authors cite studies showing Google’s AMIE beat physicians at taking complaints and histories, while ChatGPT o3 named the final diagnosis first in 60% of 377 complex cases versus about 16% for 20 internists on 302 cases.
A 2024 study they reference found adding human doctors to AI-led case reviews significantly worsened results, underpinning their claim that physicians may create more errors than they prevent.
The paper says legal, liability, reimbursement and regulatory barriers still prevent non-human clinicians from practicing independently, even as the authors warn doctors could lose skills through AI-driven deskilling.
That stance cuts against the American Medical Association, the American College of Physicians and AI scholar Robert Wachter, who argue physicians should keep AI in a supportive role rather than be replaced.