-Caveat Lector-

You are on the right track and such people are extremely rare. That means
you are a "target" (for discrediting, ignore-tactics etc.) by the
Technolords and Technopriests. Read "The Psychology of Everyday Things" by
computer scientist/psychologist Don Norman on the psychology and
psychopathology of everyday things. Then ask yourself why Don Norman, now
a professor at Northwestern, won't try out the modern teaching machine
which is capable of rendering the college-level teaching of computer
programming, obsolete. He has replied by email (see IMP list archives)
that he doesn't want to put himself out of a job.

On Tue, 30 Mar 2004, Damon Richter wrote:

> In news:internal.ml.new-age.ctrl, Adam Ness <[EMAIL PROTECTED]> posted
> on Mon, 29 Mar 2004 08:04:54 -0800:
>
> > Squad helps dog bite victim
> > Red-hot star to wed astronomer
> > Helicopter powered by human flies
> > Once-sagging cloth diaper industry saved by full dumps
> > America pushes bottle up Iraqis
> >
> > Obviously, each of those sentences can have at least two, if not more
> > meanings.  Humans have enough knowledge about the world to decided that the
> > squad wasn't helping the dog bite people, or that the red hot star was a
> > celebrity rather than an actual celestial body.  A computer doesn't have that
> > knowledge though, and can't interpret those sentences correctly.
>
> Well, those aren't really grammatically correct unless you're Tarzan or
> an Indian Chief.  I think the languaged used would have to be literal.
> And debugging might be even easier.  The software could simply report to
> the programmer, "I don't understand this sentence" and highlight the
> appropriate statement.  The programmer would examine his sentence and
> find a better way to word it.

You got it. Again, have a look at the IMP archives. Why should we spend
our time writing NLPs to built the confusion and stupidities of the
English language into English-speaking computers? Why not just use the
sub-set of English which is clear and logical. It exists.

  And it wouldn't be that the computer
> requires very strict adherence to a set of rules, which is the problem
> with current programming languages, but the sentences used would have to
> be literal and grammatically correct, and the spelling would have to be
> correct, too, but we already have spell checkers.  We already have
> grammar checkers, too.

Google on the work of Professor David G. Lowe at UBC on object-recognition
from local scale-invariant features. If a robot can recognize objects in
its surroundings, then it is easy to teach it basic point-and-tell
language. Next build some basic logic around that point-and-tell language.
The words of basic logic mean the same to a machine as they do to us.
Words like and, or, not, if-then are nice natural language words just
begging to be put into NLP. That gives you the start on "Robo-speak".
Spend a little quality time (lots of TLC) with your pc or robot and you
can build up its language capabilities from there.

> > Additionally, passive voice is often difficult for computer natural language
> > systems.
> >
> > "Could you open my file for me?" be answered with a verbal "Yes" or by
> > opening the file is the most simplistic example, but I'm sure this
> > intelligent crowd can come up with some better examples.
>
> The computer could be told to operate in "polite mode" which would accept
> an asking as a commmand.  Statements beginning with "Would you" or "Could
> you" or "Can you" would be executed based optionally on an assumption
> that if the human unit is asking if a file can be opened, then he
> obviously wants it to be opened.
>
> Those who like to feel powerful can operate in standard command mode.
> That way, the computer would interpret "Could you open my file for me" as
> a question about its ability to open the file, and the computer would
> respond with "Yes".  If the human really wants the file opened, he would
> say, "Computer.  Open my file."  He could name the computer anything if
> he doesn't want to sound like Captain Picard all the time.  "Computer.
> Your name is now Jason.  Jason.  Open my file!"
>
> ...your file is opened.
>
> "Jason!  Open the letter I wrote last night to Mother."
>
> Well, if there's no indication of who your mom is in the filename, then
> the computer would search all files created by your word processor last
> night.  If it can't find a reference to "Mom" then it would reply, "I
> can't seem to find that file.  Can you be more specific?"  Hey, a
> computer that can interpret English should certainly be able to speak it
> in a conversational manner.
>
> The most depressing part about this conversation is the fact that I
> listened to an archived Coast to Coast AM show with Alex Jones of
> Infowars as the guest.  Good grief!  If everything he's talking about is
> really happening, I'd be surprised to see 6 GHz Athlon CPU's on the
> market.

You have to be careful to proceed very slowly with the development of
Robo-speak. Otherwise you tangle yourself in new verbal knots which
replace the existing ones of our language.

ZG

www.ctrl.org
DECLARATION & DISCLAIMER
==========
CTRL is a discussion & informational exchange list. Proselytizing propagandic
screeds are unwelcomed. Substance—not soap-boxing—please!   These are
sordid matters and 'conspiracy theory'—with its many half-truths, mis-
directions and outright frauds—is used politically by different groups with
major and minor effects spread throughout the spectrum of time and thought.
That being said, CTRLgives no endorsement to the validity of posts, and
always suggests to readers; be wary of what you read. CTRL gives no
credence to Holocaust denial and nazi's need not apply.

Let us please be civil and as always, Caveat Lector.
========================================================================
Archives Available at:

http://www.mail-archive.com/[EMAIL PROTECTED]/
<A HREF="http://www.mail-archive.com/[EMAIL PROTECTED]/">ctrl</A>
========================================================================
To subscribe to Conspiracy Theory Research List[CTRL] send email:
SUBSCRIBE CTRL [to:] [EMAIL PROTECTED]

To UNsubscribe to Conspiracy Theory Research List[CTRL] send email:
SIGNOFF CTRL [to:] [EMAIL PROTECTED]

Om

Reply via email to