Good point. I've got to rephrase things a bit, because I didn't come across as saying what I meant. They claim that it COULD replace it. That it has the potential to. To me, that's akin to'not yet currently'. It's like Tesla and their self driving anthics: they throw out a beta and imply that it's a good replacement while it's never tested under scrutiny.
AI has no medical, psychological or other degree. And when I use it in a field I know something about, it's often dead wrong on empirical things (as I find out later). There's no reason to blindly assume it knows better in fields I don't know anything about (okay... I assume it's better than what I know on the subject, but that's not to say that's it's therefore correct).
