My motivation, first as an academic physician-scientist, and then as a founder of a health-tech startup, has been about wanting to solve the big problems in medicine. At first, I was very science-focused. I remember distinctly that, after finishing my PhD and returning to the clinical wards, I was horrified by most routine clinical decision-making, specifically how variable and unscientific it appeared.
Later, over the next 20 years, I became bothered by other larger trends: lack of access, high costs, the slow pace of learning. And I’ve watched the supposed panaceas, genetics, genomics, and now artificial intelligence, fall short, though usually after much fanfare with many grants awarded and papers published.
Today, artificial intelligence, a field I’ve been actively involved in since my postdoc in machine learning back in 2006, has dominated the news cycle. And in many fields (self-driving cars, for example) the advances are incredible. But if you look at the advances in healthcare, in contrast, they are quite pedestrian, particularly in light of the major crises we face regarding costs and quality.
We now have lookup tools that are better at going through medical reference material. And we have AI scribes that shave off one minute per encounter on a good day. It’s not that I think these aren’t useful (and the company valuations are enviable!). We’re surely improving physicians’ quality of life.
I just don’t anticipate much impact on the major healthcare problems we’re facing as a society.