Re: [unrev-II] Re: Towards an atomic data structure.

From: Eric Armstrong (
Date: Tue May 02 2000 - 13:02:47 PDT

  • Next message: Eric Armstrong: "Re: [unrev-II] Re: Towards an atomic data structure."

    > Gil Regev wrote:
    > Eric wrote
    > > One thing I like about Tractions ability to *add*
    > categories
    > > to an information node is that it makes it possible to
    > > reframe points in a discussion from a different point.
    > Yes, this is nice but not very different from what programs
    > such as Lotus Notes or Outlook do which is to put a node in
    > multiple categories.
    To which I say: Great argument for investigating those two apps
    further. I think this is critical functionality. That makes it
    necessary to "get it right".

    > ...we are unable to see things with other people's category and
    > value systems. It is extremely difficult to externalize
    > one's category system in ways that make it understandable by
    > others.
    Putting things in the repository, and categorizing them, and
    categorizing other people's additions makes some of that explicit,
    doesn't it?

    > It is one thing to flesh out some of it in a
    > conversation, it is another to make it explicit in a way
    > that makes it understandable or even simply viewable by a
    > machine.
    You know what, I really don't care what a machine understands, if
    ever. I want to build systems that *humans* can use to reach
    deeper understanding. Deep Blue plays a marvelous game of chess.
    But it does absolutely zero for me. It has not improved my
    understanding of the game, or improved my ability to play, or
    given me the most remote opportunity to do so.

    Kasparov "took on the world" in a chess game. (Any chess player
    in the world could vote on what to do next.) That game went much
    worse than Deep Blue. The "world" was unable to match the
    capability of a single program, or even a computer. The system
    I'm interested in will allow the "world" to collaborate so
    effectively that it will manifestly better than either of them.

    > Machines also don't have much of a sense of humor :) Would a
    > machine understand anything from the Forrest Gump Story?
    > Absolutely not, but you all did. Naratives and jokes are
    > very powerful. How do we integrate them into the OHS or DKR?
    > I understand that such a question goes beyond the tidepool
    > boat but if we don't tackle it soon enough, aren't we just
    > going to produce another Lotus Notes?
    It's worth examining what Lotus Notes does *wrong*. Because if all we
    ever did was to produce a Lotus Notes that really "took off" and
    fulfilled its promise by becoming close-to-universally used, then
    I'd say we had succeeded in augmenting *human* intelligence.

    With eVoice Now you can keep in touch with clients, vendors, co-workers,
    friends and family ANYTIME, ANYWHERE. Sign Up Today for FREE!

    Community email addresses:
      Post message:
      List owner:

    Shortcut URL to this page:

    This archive was generated by hypermail 2b29 : Tue May 02 2000 - 13:10:24 PDT