Sound capture

Sound capture

Post by uj79 » Fri, 30 Jul 2004 12:00:21

Brad Eckert wrote on 28 Jul 2004:

The answer to this question will potentially be of use in coding ~mentifex/mind4th.html -- Mind in Win32Forth. ~mentifex/jsaimind.html is in JavaScript:

____________ ____________
AI4U textbook / \ / \
User's Manual ( Motorium ) ( Security )
Add your link \____________/ \____________/
____________ ____________
/ \ / Sensorium \
( Volition ) Listening... ( --> Audition )
\____________/ \____________/
____________ ____________
/ \ / \
( Think ) ( Emotion )
\____________/ \____________/

This program, the first true artificial intelligence on the Web,
cycles through Thinking, Listening, Rejuvenating and Terminated.

Sound capture

Post by aiiadic » Sat, 31 Jul 2004 14:39:13


can you supply a link with your AI software bundled with a FORTH
compiler/interpreter. I'd like the following:

download your software.
RUN it.

I'd like to see an EXE so I can turn the thing on and see what it
does, and I don't want to go through the process of getting compiler/
pasting code/compile/run. just EXE. executable.

I think many more people would check it out if you did this.

I checked out the webpage/javascript. breaks IE 6.0.2800 very

but, I typed in a few things and seemed to get a stupid (but ?intelligent?
answer).... It asked me what I was talking about and asked
me for more information. I've seen dumb-text-parsers which would
do a horrible job of "appearing" intelligent. ELIZA from way back
and HAL I downloaded a few years ago. both not making any attempt
to actually understand what you typed in, but outputting part of
your text in question form.

The text that your software spit back out at me was impressive.
I want an EXE so I can play with it some more.

eliza and HAL seemed programs. I got the impression
(from VERY limited use) that your code is doing something that the
others aren't.

aiiadict AT hotmail DOT com


Sound capture

Post by Chris S » Sat, 31 Jul 2004 15:59:33

Indeed, a stable program is a fair request. However, I don't see the
wisdom in any AI project that doesn't incorporate any kind of memory. I
managed to get the script working in IE 5 and although I'm not
especially impressed with the results, I couldn't help but imagine what
might be going to waste after I terminate the program. If you have
"solved" AI (a claim I heavily discount), it seems cruel to bring to
life some intelligence only to throw it away after you're done with it.

Sound capture

Post by Stephan M. » Sat, 31 Jul 2004 16:09:40

On 2004-07-30 08:59:33 +0200, "Chris S." < XXXX@XXXXX.COM > said:

Hmm, hearing you talk like this makes me wonder if you're a vegetarian....?


Stephan M. Bernsee

Sound capture

Post by Alex McDon » Sat, 31 Jul 2004 19:02:32

For Win32Forth you do. It's in the public domain, apart from certain pieces
which are GPL/LGPL.

Alex McDonald

Sound capture

Post by uj79 » Sat, 31 Jul 2004 19:34:14

XXXX@XXXXX.COM (Rich J.) wrote on 29 Jul 2004:

The problem is, I don't think that I ("ATM") have the legal right
to grab Win32Forth (or any other Forth compiler) and include it in
a ZIPped distribution of my Win32Forth AI Mind available at and elsewhere.

True, the code has to be pasted as one long block from that page
and the Win32Forth has to be downloaded separately by FTP, but
afterwards any Forth programmer has the code ready to work with.

Unfortunately, the MS IE 5 JavaScript code is now more advanced
than the Win32Forth Mind.Forth from August of 2002. So it might
be easier simply to scout out somebody's computer with IE 5.
The same has been reported recently by Richard Cornford, who seems
to be, on the one hand, a genuine JavaScript expert, but who, on
the other hand, rapidly seems to be losing faith in Mentifex Ai,
perhaps under the negative influence of the anti-Mentifex
pack of jackals that like to attack my every Usenet post :-( is the mind-module
designed to solicit further information about unknown concepts. is not a chatbot,
but rather a genuine attempt to mimic the thinking brain-mind.

MS IE 5 JavaScript obviates the need for an EXE, because the
compiler (interpreter) is contained inside the IE 5 browser.
After REXX and Forth, I coded the AI in JavaScript for exactly
that reason, as Richard Cornford has recently and correctly
speculated in a pertinent, recent Usenet thread elsewhere.
JavaScript is not just one more programming language for me
to code in, but a special case, because it obviates the EXE.

My code is *thinking* -- stupidly, primitively, but real-ly.
It thinks by tracking, on an uppper tier, the spreading activation
among concepts on a lower tier. The linguistic upper tier
translates conceptual thinking into linguistic expression.

Right now I am mulling over and trying to design a "Calibrate"
mind-module that shall automatically adjust the activation
variables so as to prevent them from spiraling out of whack.
The very idea of "calibrating" a mind seems anti-free-thinking,
but I think that it can be done; I just work rather slowly.

Thank you very much for examining and reporting on your visit to -- the AI Mind. -Arthur

Sound capture

Post by uj79 » Sat, 31 Jul 2004 20:29:46

"Chris S." writes on Fri, 30 Jul 2004:
ATM: (the Rejuvenate
module) finally works reliably, i.e., interminably in the IE 5 ~mentifex/jsaimind.html AI Mind software.

ATM: (Mind.REXX) from
26 November 1994 had a system of saving memories on the Amiga. ~mentifex/mind4th.html (1998+) does not save
states of mind to disk because I was too hard at work on the
basic AI algorithms to sidestep into learning Forth disk-saves. ~mentifex/jsaimind.html run forever in IE 5
(but of course nobody keeps their browser open forever) and
saves *internally* those memory-engrams that re-occur to the AI
in the brief time alloted before the Rejuvenate module kicks in. ~mentifex/theory5.html -- solved in *theory*.

I like the way you think. -Arthur

Sound capture

Post by Jon Harri » Sun, 01 Aug 2004 05:26:33

eating != throwing away ;-)

Sound capture

Post by Chris S » Sun, 01 Aug 2004 06:31:48

After reading over your javascript code, your implementation appears to
be a mix between a knowledge base and a natural language parser.
Unfortunately, it doesn't seem to have any general IO capabilities. This
is a common pitfall to many amateur AI researchers. Language alone
doesn't define knowledge. Without a physical perception of the world, a
computer will never "know" what 'soft', 'pain', or 'up' mean. I've often
thought vision algorithms, speech synthesizers, NLPs, and such without a
general form of processing and representing sensory data was like trying
to run before we could walk.

But if I were to recommend an improvement in your system, I would
suggest you try coding it in Python ( Compared to
Javascript, and especially Forth, Python is remarkably easy to learn.
Python would allow you to represent your idea in an object-oriented
framework, making it much easier for others to understand. It would also
make persisting dynamic data much easier, so you wouldn't have to hard
code so much of your lexicon. Plus you should never run into any
platform issues, since Python runs everywhere. Python's quickly becoming
a serious tool for AI research. For instance, "AI: A Modern Approach" is
currently offering both Lisp and Python source code for many of its

Sound capture

Post by Stephan M. » Sun, 01 Aug 2004 16:58:43

On 2004-07-30 22:26:33 +0200, "Jon Harris" < XXXX@XXXXX.COM > said:

Yes I know. My reasoning is that someone who has a m *** problem
shutting an AI off should have at least *some* second thoughts when
killing/eating a *real* animal... :-)
Stephan M. Bernsee