I agree about your latter example, but about your first example: isn't it the case that these faulty AI systems for medical diagnosis have been rejected? Docto

I agree about your latter example, but about your first example: isn't it the case that these faulty AI systems for medical diagnosis have been rejected? Doctors don't like them because they don't want to be replaced or one-upped, and because they just don't trust them (rightly so, as it turns out). So the systems, which were put out for use on a trial basis, don't get used.