but rather that will activate a device upon hearing the command, e.g. "Start," spoken by any speaker and will embark upon a subsequent activity, associating that activity with a unique individual on the hearing of a dictated set of numbers,
So you not only want to recognise anybody saying something, but when you do you want to recognise the individuals that has said it?
If so carry on reading science fiction while the world catches up with you.