Artificial intelligence is already helping to analyse scans and support diagnostic decisions. But when does it actually improve patient care? Eric Topol, one of the world’s most prominent physician-scientists, argues that medical AI should be judged by clinical outcomes — and by whether it gives doctors more time for their patients.
Stories about medical technology often centre on a single inventor credited with a breakthrough and millions of lives saved. The reality is more complicated. Discoveries must pass through research, trials, implementation and evaluation in ordinary clinical settings. At every stage, the promise may prove greater than the demonstrated benefit.
