The more automated the process, the less intelligence is
  required to operate the system.*

  This is a good thing BUT then it places the burden of accuracy
  NOT upon the system itself but upon the initiators and
  interpreters.   Plus, types of questions asked WITHIN the
  process itself should require greater and greater scrutiny. The
  system itself is very functional... and the "How" is very
  workable. But the Why's then matter more.