Re: Changing ones mind

From: Anders Sandberg (asa@nada.kth.se)
Date: Mon Apr 07 2003 - 15:47:22 MDT

  • Next message: Matthew Gingell: "Re: Help with a Minimum Wage Model"

    On Mon, Apr 07, 2003 at 12:08:51PM -0700, Adrian Tymes wrote:
    > --- Anders Sandberg <asa@nada.kth.se> wrote:
    > > I would love to build a graphical computer-aided
    > > system for building
    > > rational arguments, even if they were lacking in the
    > > depth or complexity
    > > we see in ordinary discussions. It would be great to
    > > use as a skeleton
    > > for building scientific theories or thinking about
    > > what to research
    > > right now.
    >
    > Umm...out of curiosity, what parts of an argument could be reduced to
    > graphics? An argument is a collection of words - maybe verbal, maybe
    > written, but still words. I can see, in a high level, breaking out
    > lists of facts and drawing the dependencies between them (so you can
    > see what parts of an argument shatter when a basic fact is disproven),
    > along the lines of how an AI might formalize its reasoning...

    You are probably more of a verbal thinker than me, I'm quite visual. As
    I envisioned the system, different concepts (small groups of words)
    would be linked by lines in different colors and with annotations
    showing their relationship. Of course, these links in turn can be
    connected in different ways to other concepts. A bit like a 3D mindmap,
    but with stronger functional relations. Maybe a bit like the semantic
    webs studied in AI and cognitive psychology.

    But this is mostly a daydream about the system, I have no blueprints.
    Yet.

    > Hmm. Now that I think about it a bit more, wouldn't someone probably
    > already have done something like this, in the pursuit of top-down AIs?
    > I wonder which keywords would get a good google for this - and, at the
    > end of that, if the software someone wrote for that is publically
    > available (as in, public domain and runs on today's computers)?

    Look for semantic nets. There is lots of them within the AI research
    community. The problem is that it is mostly internal representations,
    not anything you actually write or use.

    -- 
    -----------------------------------------------------------------------
    Anders Sandberg                                      Towards Ascension!
    asa@nada.kth.se                            http://www.nada.kth.se/~asa/
    GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y
    


    This archive was generated by hypermail 2.1.5 : Mon Apr 07 2003 - 15:50:51 MDT