Is clinical AI safe as a human doctor?

Maybe that's not the right question. Maybe the question is whether clinical AI is safer than no doctor at all. Because that's the case for so many patients throughout the United States and moreso throughout the world.

Is that the right framing? If clinical AI access is better than no access, what are we saying about ourselves when we decide that we're willing to lower the threshold of safety for the patients most likely to need this new resource? Patients who are also the most vulnerable, and least empowered to question this new technology or advocate for themselves when something goes awry?

There are no easy answers, but these are the kinds of questions we need to ask ourselves. And the type of hard work that is desperately needed in the field of clinical AI and health care