Taha Kass-Hout, the former U.S. Food and Drug Administration chief health informatics officer who joined Amazon in March, didn't reveal any trade secrets in the interview with the health website Stat, but indicated that Amazon sees big potential in developing AI tools for health.
"AI is medicine is not a new concept," Kass-Hout said, explaining that the earliest uses of the technology in health dates to the 1960s when it was first used to navigate patients to the right place. He noted that the past year has seen a massive uptick in research papers that use "prediction and natural language understanding" in a variety of new ways.
Despite that, he also cautioned that AI tools have been overhyped.
"I'm really happy to see that in 2018, a lot of that dust settled and we started to see real, concrete examples of its use," he said.
One of these, which Kass-Hout also touched on in the interview, is Comprehend Medical, a machine learning service for health information that Amazon launched in late 2018. One early partner, Change Healthcare, which processes health claims for pharmacies, uses Comprehend Medical to predict whether an insurance claim will likely be denied.