How AI Is Bringing the Human Touch Back to Healthcare
Artificial intelligence has been quietly transforming industries for decades. Retailers deploy it to forecast what customers will grab next from a shelf. Banks rely on it to identify suspicious activity or give meaning to vast financial datasets. In healthcare, however, where stakes are high and demands on providers unremitting, the technology is beginning to transition from novel to necessary.
For physicians, one of the biggest drains on time and energy isn’t the work most people imagine. It’s not the high-stakes decisions in an exam room or the delicate balance of prescribing the right treatment. It’s paperwork. Studies consistently show doctors spend nearly twice as many hours on administrative tasks as they do with patients. That imbalance fuels burnout, shortens visits, and strains relationships at a time when healthcare workers are already stretched thin.
That’s the problem Kyle Robertson and Matt Holmes are aiming at with ScribeAI, a clinical documentation tool that relies on machine learning to take the notes so doctors don’t have to. The software listens in on patient visits, transcribes the conversation, and organizes it into structured medical documentation—SOAP notes, billing codes, after-visit summaries. The pitch is simple: free up hours that providers would otherwise spend hunched over a keyboard.
Kyle Robertson is no novice in navigating the obstacles related to personal health. He started the telehealth company Zealthy to increase access to care and has run his previous company, Cerebral, to bring virtual mental health services nationwide. With his incubator, Revolution Venture Studios, he’s aiming his attention at the less flashy but no less impactful part of healthcare: the back office. There are early indications that the bet is paying off. ScribeAI has already generated an eight-figure pipeline, insiders close to the business say. But of more consequence, clinicians working with the software say hours being reduced are benefitting patients. It’s not efficiency alone. It’s giving time to doctors to spend time with people.
ScribeAI belongs to a larger influx of businesses that are experimenting with deploying AI not to replace doctors, but to augment their abilities. Flatiron Health applies algorithms to assist oncologists in sorting through cancer data. Cleerly utilizes machine learning for cardiology. Enlitic targets radiology. In all instances, expertise is not replaced, but complemented by technology—in some instances, AI can even identify patterns just as well as human experts do. Flatiron even printed studies demonstrating that large language models are capable of documenting cancer progress with rates similar to those of oncologists.
That matters because healthcare has historically been resistant to hyped-up technology solutions. Electronic medical records promised to revolutionize medicine; they often made it more agonizing to take notes. Instead of devising new visions of care from scratch, businesses like ScribeAI begin with something simpler: eliminate the friction that makes medicine more difficult than it needs to be.
The bigger question is how quickly these technologies get deployed—and whether they’ll end up helping patient outcomes in a dramatic way. For the moment, however, doctors who use Kyle Robertson’s ScribeAI call it something that’s all too rare in medical technology: a product that simplifies their lives. And in a system for which time all too often is the most sparse resource, that’s radical enough.
