A dialog between you and the search engine gets to
understand
better what you are searching for, and will search deeper,
going into a second and third
stage search, searching INSIDE the results for you, giving you
a
comprehensive description of what it finds instead of a list of
sites.
Of course at first this won't work very well but as time
passes it will get better at understanding you, and better at
telling you what it found out.
So each search has a list of fields around it. When you ask a
question it understands that your search is in the field of
philosophy or cognitive science, that you are NOT searching
for
a certain phrase in a book, but rather trying to find the
"name"
or title of one of the theories in the physical explanation for
consciousness. It is important to note that when you initiated
the search you didn't say a word about all those terms in
your question.
But RUS4 already had you searching and watching Jeff
Kaplan's
intruductory videos on this topic, so it knows you aren't a
complete novice.
In order to give you: You Are Searching For: A theory of
physicalism in the philosophy of mind?
(logical behaviorism? functionalism? The mind-brain
identity?)
or are you looking for the names of people who gave these
ideas? (David Chalmers' Golem? Mary's room by Frank
Jackson?
Hilary Putnam's super Spartans?)
Are You Searching For a sequence of actions to walk you
through
fixing a broken applience?
Something like that?
Something else completely?
This is similar to the NOT THAT idea that I posted here a few
years ago, but its not that.
Our program will need to artificially (but with human
assistance) know how to categorize, get closer through
keyword
clouds, do secondary automated self searches, and finally
automatically and manually give comprehensive titles and
subtitles to pieces of information.
It will also need a way to learn when it is not being
comprehensive, and "how to talk" with the different human
searchers, each with their own and their group vocabulary,
knowledge and fields of interest.