Comment by FpUser

I am curious how would they "explain" AI algorithms where it is impossible to explain how / why the decision has been made.