Wouldn’t it be better if AI bots got real with us, instead of being only sophisticated parrots? A conscious AI would seem to be present, to have an inner life, and would act as if we did, too. You could actually converse with one.
It would be fun! But how effective would it be to judge the capability of an interviewee is open to discussion.
You are right, it’s arguable whether or not conversation would be useful in measuring other abilities or qualities of an AI. I chose this somewhat science fiction-y opener just to be provocative. My main point is that creation of consciousness needs to be informed by our stronger theories about how consciousness arises. On the conversation issue I have a non-scientific bias: that realistic conversation is hard to fake and readily recognized (guess I’m with Turing on that one). One can tell a quality novel or movie from junk by entirely how like real life the dialog is. And I think that an artificial consciousness (IF one could be made) would converse pretty well. Especially if it spent a lot of time reading novels. I tried to extend my ideas about a conscious AI in another piece on Hacker Noon. And elsewhere I offered a lot of speculation about bots that could fake conversation.