Skip to main content

View Diary: Bookflurries-Bookchat: Featuring Daily Kos Writers and their New Books (106 comments)

Comment Preferences

  •  Define "conscious." (3+ / 0-)
    Recommended by:
    cfk, MT Spaces, Dumbo

    We may never have a "conscious" AI for no other reason than the definition may lack rigor.  On the other hand, we're already computationally modeling several areas of human cognition and personality that can be defined materially.  Do you mean we may never see a combination of such aspects such that the "mind" as a whole is indistinguishable from a human's to some observer?  Maybe, but I don't think we have any reason to rule it out.

    •  oh, dear (2+ / 0-)
      Recommended by:
      MT Spaces, Dumbo

      I was just typing a long answer with quotes from Michio Kaku's book Physics of the Future from the chapter Future of AI...when my computer crashed.

      It is worth reading.

      I won't retype...but it has to do with pattern recognition, common sense and emotions.

      I agree that the definition of conscious could vary a lot.

      The programming just to get a robot to bring juice to you is tremendous.

      I am always interested in this question.

      Join us at Bookflurries-Bookchat on Wednesday nights 8:00 PM EST

      by cfk on Wed Dec 05, 2012 at 07:37:22 PM PST

      [ Parent ]

    •  The definition is DOOMED to lack rigor because (3+ / 0-)
      Recommended by:
      cfk, MT Spaces, Pete Cortez

      the real standard for consciousness seems to be "kinda like you and me," which isn't very rigorous at all.  Any attempt at a rigorous definition would fail that in some way.

      For instance, what if we came up with a rigorous definition of consciousness, but when it was implemented, the conscious computer didn't care to communicate with us?  Maybe it just sat there and tuned us out, trying to solve some math conjecture that interests it through brute force that might not have any solution.  Without the need for sleep or food or drink, it could just do it forever.  Or what if it was just plain sullen and unhelpful.  Or conscious but too alien to interact with.  See what I mean?  There seems to be a dumb assumption that if we made a computer conscious, it would say hello and want to talk about life with us and crack jokes.  But why would it, if it doesn't have the urge of even the lowest primates to socialize and form communities?  

      So I think any consciousness app in computers will have to be a huge kluge of disparate parts that give it the semblance of emotions and needs, while the actual consciousness itself may be something simpler and purer but not very human.

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site