Patient safety
Medication adherence: the invisible lever with outsized impact
One WHO conclusion reframed the field: adherence may matter more than any single new treatment.
5 min read
Every few weeks a new claim insists AI will transform medicine overnight. The credible research is more useful, and more specific, than the hype.
The grounded estimates
McKinsey Global Institute's analysis of the economic potential of generative AI places a large share of its value in a handful of functions, and the Stanford HAI AI Index tracks where adoption and capability are genuinely advancing versus stalling. Read together, they separate signal from marketing.
Where it lands in a clinic
The durable value shows up in the unglamorous places: drafting documentation, summarising a long history, suggesting codes, and surfacing a relevant result. Not diagnosis-by-robot, augmentation of the tasks that eat a clinician's day, with the doctor still deciding.
The evidence
McKinsey Global Institute and the Stanford HAI AI Index provide the sourced framing of where generative AI creates real value, and where the hype outpaces reality. Sources: McKinsey Global Institute, “The economic potential of generative AI: The next productivity frontier”, 2023; Stanford HAI, “AI Index Report 2025”.
The trillion-dollar number is real in aggregate. In your clinic, the value is measured in minutes returned to the visit.