From: tim@iss.nus.sg (Tim Poston)
Subject: Re: MISC: The End of the World (Wide Web) / Part II
Date: 10 Oct 1996 09:39:36 GMT
Organization: Institute Of Systems Science, NUS



Gerry McKiernan (JL.GJM@ISUMVS.IASTATE.EDU) wrote:

:         Instead,
:             can we not consider extracting or summarizing the full
:             text of these documents, and offer to the user the
:             'concepts' that the words of the document represent.

Lovely -- once we get close to whatever a `concept' is.  The
AIntelligentsia has not gotten significantly closer to understanding
that since it got rolling in the 50s; it has regularly promised the
Moon, but (unlike Kennedy) failed to deliver.  Back then they thought
they would have automatic English-French translation as soon as they
had machine-readable dictionaries.  Outside of narrow technical
domains like weather forecasts, that's _still_ a pipe dream.  Language
AI is still mostly manipulating strings, and that i s n o t
 w h a t   h u m a n   l a n g u a g e   d o e s.  Word and letter
strings are a dead skeleton, with the muscles (tone, volume, rhythm, non-verbal
association) all rotted off.

Take the concept of `shuttle'.
Anybody who has ever watched a weaving shuttle bang to and fro
would understand the expression `shuttle bus' without extra
words -- it's bus that behaves similarly.  That similarity
is not perceived as "the same expression, `to and fro', applies
to both", it is at the level of imagery and the muscular sense
of movement.  No software yet written can make the same
non-verbal connection between words.

Human concepts are full of non-string understanding.
Software is empty of it.  Until it fills, search tools
will miss connections that humans want to follow.
One way to fill in for this digital blindness is
hand-inserted pointers -- and we're back to
hypertext links.  They'll be around for quite
a while.

They should be a lot more powerful.  (I'd like to be able
to click on any symbol in a mathematical expression and
see its definition, for instance.)  But my dissatisfaction
with them means I want them strengthened and streamlined.
They won't be obsolete until a lot of very old Real Soon
Now promises are finally kept: and there will be a lot
more white in my beard before then.

Tim
tim@iss.nus.sg (Tim Poston)
____________________________________________________________________________
Tim Poston    Institute of Systems Science, National University of Singapore
            Ask not what your time-zone can do for you:
              ask what you can do for your time-zone.