@ulintl i suspect that if AI is going to replace these jobs we will see regular cases of those AI doing the wrong thing and the companies being damaged in the process. AI is bullshit. nobody in their sane minds would give a heuristic algorithm power, but if you call it AI everyone seems to ignore that it's just a statistical model which will have edge cases where it fails.
combine that with adversary input, which is guaranteed to be possible, and it's going to be really interesting :)
@ulintl i think it will just end with "it was a software error!" if something goes wrong. nobody is to blame then ;)
@bonifartius I am not so sure about it.
Of course automatization has limits, but IMHO routine-jobs they do better than humans still.
So for most part humans could only become supervisors and operative case-handler only in special cases.