Artificial intelligence is everywhere these days—sifting through retail data to predict what shoppers will buy next, powering financial models that detect fraud, and keeping supply chains moving faster than human managers ever could. In healthcare, the technology is making similar inroads, from analyzing scans to helping design new treatment strategies. But the more intriguing question is not just how AI can make medicine more efficient today, but whether it can help prevent the next public health crisis.
There are those who think that the answer is that AI can save time. For physicians, one of their largest timesinks has little to do with seeing patients and everything to do with forms. Researchers, again and again, have demonstrated that physicians spend close to double the amount of time documenting notes and coding visit time as time in face-to-face consultations with their patients. The disparity not only causes frustration—it dismantles those relationships that are most reliant upon them.
A recent start-up, ScribeAI, is placing its bet that it is at this intersection that AI can best deliver value. Founded by veteran health-technology executive Kyle Robertson and former banker Matt Holmes, the start-up applies natural-language processing to record and transcribe discussions between physicians and patients. Its software creates structured clinical notes, billing codes, and diagnostic summaries, reducing documentation time by over 70%, its founders say.
For Kyle Robertson, it’s a familiar challenge. He once founded Cerebral, the mental health technology firm that brought therapy and psychiatric care to several hundred thousand patients over the internet, and subsequently Zealthy, a telemedicine platform that was aimed at widening access to primary care and general wellness. Both of these businesses hinted at shortcomings in the way healthcare is dispensed in the U.S.—not enough providers, not enough time, and too many obstacles for patients looking for steady care. With Revolution Venture Studios, his healthcare incubator, Kyle Robertson is focused on one of the most impactful parts of the puzzle: clinical note-taking. The bet is that shortening doctors’ time glancing at keys would give them room to think, listen, and respond—skills that are all the more critical in a disaster. In busy clinics, minutes do matter.
AI’s role in public health, however, extends beyond workload alleviation. Researchers are exploring how it can spot warning signs of crises even before they occur. In one experiment modeled in the Frontiers in Artificial Intelligence journal, algorithms were shown to predict and monitor catastrophes in real time. Another, in Humanities and Social Sciences Communications, documented how China employed AI programs in its initial few months of managing the COVID-19 pandemic to diagnose infections, watch over its patients, and even chart virus transmissions. There is reason to be skeptical. AI models are only as good as their training data, and healthcare has a questionable record of embracing new technologies without placing unintended burdens on it. But its promise is tough to resist. Used responsibly, AI won’t only make medicine speedier—it might make it more human, by freeing clinicians to spend more time with their patients, as time is their most precious resource.