Not testing at all. You know what that leads to? Bad outcomes. I'm all for selection pressure toward more efficacious, less invasive, less dangerous tests. Writing an insurer a blank check to blanket deny millions of claims without checking or reading up though? Nah. That's profiteering.
You want to run that algo in parallel and dift through the claims that would change, individually weeding out the actual false positives? Kosher. Just YOLO'ing it on prod and trottingbit out in front of investors? Hell no.
You know what that incentivizes?
Not testing at all. You know what that leads to? Bad outcomes. I'm all for selection pressure toward more efficacious, less invasive, less dangerous tests. Writing an insurer a blank check to blanket deny millions of claims without checking or reading up though? Nah. That's profiteering.
You want to run that algo in parallel and dift through the claims that would change, individually weeding out the actual false positives? Kosher. Just YOLO'ing it on prod and trottingbit out in front of investors? Hell no.