Perl for Mentifex AI is defective compared to JavaScript

[menti...@gmail.com]

Since I started learning Perl ten days ago,
I have been frustrated with trying to solve
the problem of letting the AI Perlmind deal
immediately with each incoming character of
keyboard input, as the AI does in JavaScript:

http://ai.neocities.org/AiMind.html -- in English;

http://ai.neocities.org/Duishka.html -- Russian.

http://ai.neocities.org/AiSteps.html is my
list of "Steps in coding free open-source
artificial intelligence" and I have spent
hours trying to use "getc" or "sysread" to
create the first Perlmind AI loop that will
accept input from the human user. Now I
will set aside the goal of dealing with
each separate character and I will design
the AI Mind to wait for a carriage-return
at the end of any sentence of user-input.
At least it will be possible for the user
now to press the backspace-key in order to
correct any obvious mistakes being made.

Two days ago at my Alma Mater Book Store
I bought "Perl by Example, Fifth Edition"
(C) 2015 by Ellie Quigley. It is nice to
have such an excellent Perl book issued
in this same year 2015 as I code the AI.

https://groups.google.com/d/msg/net.ai/IJpSNRGEE7w/70BtsqcpRDoJ
was my very first Usenet post, made before
there was such a thing as Perl or JavaScript.

Arthur T. Murray ("Mentifex")
--
http://www.amazon.com/dp/B00FKJY1WY
http://ai.neocities.org/mentifex_faq.html
http://cyborg.blogspot.com/2015/04/pmpj0412.html
http://aihub.net/artificial-intelligence-lab-projects

[Erwin Moller]

Arthur Mentifex again!

Yes, you have been posting on usenet for a long time.
And even in 1985 (your first posting) people responded like this:
"Can we please make ai.net MODERATED!".

Your ideas for building AI haven't really raised a lot of enthousiam.

Also, if you had so much trouble feeding keystrokes into your AI, I fear
for the quality of rest, since that is a pretty basic task.

It feels strangly comfortable you are still alive and kicking and
spamming usenet. Just like the old days. ;-)


Erwin Moller


--
"That which can be asserted without evidence, can be dismissed without
evidence."
-- Christopher Hitchens

[Scott Sauyet]

menti...@gmail.com wrote:
> Since I started learning Perl ten days ago, [ ... ]

<http://www.nothingisreal.com/mentifex_faq.html>

-- Scott

[menti...@gmail.com]

On Wednesday, April 22, 2015 at 8:50:08 AM UTC-7, Erwin Moller wrote:
> [...] It feels strangly comfortable you are still alive and kicking [...]

http://ai.neocities.org/perlmind.txt is the growing Perl AI program.

[Erwin Moller]

[PERL warning, this isn't ECMA, isn't on-topic, and only for amusement.]

If that is all you produced in all thoose years......
Really, after taking out the comments, little is left, except
good-sounding subroutines, like sub sensorium(), which actually consist of:

sub sensorium() {
print "Enter input, then press RETURN: ";
AudInput();
print " \n";
@en = " ";
@psi = " ";
}

Oh, maybe AudInput() does something useful?

sub AudInput() {
our $pho = " ";
for (our $i=0; $i<6; $i++) {
$pho=getc(STDIN);
chomp($pho);
if ($pho gt "") { $t = $t + 1 }
}
AudMem();
}


Hmm, again no go.
Maybe audMem does something useful tehn?

sub AudMem() {
print "AudMem: Storing ", $pho, " at time = ", "$t\n";
$aud[$t] = $pho;
}

etc.etc.etc.

For the intersted reader: we have seen half the code of the AI now.


Best of luck with your growing AI, Mentifex.

[Denis McMahon]

On Sun, 26 Apr 2015 12:32:19 +0200, Erwin Moller wrote:

> For the interested reader: we have seen half the code of the AI now.

So all this AI does is store string data.

--
Denis McMahon, denismf...@gmail.com

[menti...@gmail.com]

On Sunday, April 26, 2015 at 3:31:53 AM UTC-7, Erwin Moller wrote:
> On 4/25/2015 8:17 AM, menti...@gmail.com wrote:
> > On Wednesday, April 22, 2015 at 8:50:08 AM UTC-7, Erwin Moller wrote:
> >> [...] It feels strangly comfortable you are still alive and kicking [...]
> >
> > http://ai.neocities.org/perlmind.txt is the growing Perl AI program.
> >
>
> [PERL warning, this isn't ECMA, isn't on-topic, and only for amusement.]

As the great Mohandas Karamchand Gandhi said,
"First they ignore you, then they laugh at you,
then they fight you, then you win."

>
> If that is all you produced in all thoose years......
> Really, after taking out the comments, little is left, except
> good-sounding subroutines,

http://www.cpan.org/authors/id/M/ME/MENTIFEX/mind.txt
is the first proposal of the subroutines twelve years ago,
which are now being ported from Javascript and Forth into
http://ai.neocities.org/perlmind.txt -- Strong AI source code.

> like sub sensorium(), which actually consist of:
>
> sub sensorium() {
> print "Enter input, then press RETURN: ";
> AudInput();
> print " \n";
> @en = " ";
> @psi = " ";
> }
>
> Oh, maybe AudInput() does something useful?
>
> sub AudInput() {
> our $pho = " ";
> for (our $i=0; $i<6; $i++) {
> $pho=getc(STDIN);
> chomp($pho);
> if ($pho gt "") { $t = $t + 1 }
> }
> AudMem();
> }

In the JavaScript artificial intelligence (English or Russian),
the auditory input is event-driven by human keystrokes.
The Perlmind AI code necessitates a non-JavaScript work-around.
>
> Hmm, again no go.
> Maybe audMem does something useful then?

>
> sub AudMem() {
> print "AudMem: Storing ", $pho, " at time = ", "$t\n";
> $aud[$t] = $pho;
> }
>
> etc.etc.etc.
>

> For the interested reader: we have seen half the code of the AI now.
>
Oh, no. The Perlmind code is currently less than five percent of the

http://ai.neocities.org/AiMind.html JavaScript code or the

http://www.nlg-wiki.org/systems/Mind.Forth code as described in

http://dl.acm.org/citation.cfm?doid=307824.307853

by the ACM (Association for Computing Machinery).

>
> Best of luck with your growing AI, Mentifex.
>
> Erwin Moller

Well thank you, Erwin Moller -- a known trafficker in ideas.

Please be advised that the World Wide Web runs mainly on
Apache web-servers glued together with Perl and JavaScript.

Since Perl can work with Unicode for all human languages,
Web-resident AI Perlminds will one day be able to think not
in just one natural language at a time -- like JavaScript --
but in multiple human languages under one consciousness. See

http://aihub.net/artificial-intelligence-lab-projects

And Denis McMahon posted:

> So all this AI does is store string data.

The JavaScript English AI and the JavaScript Russian AI
store characters of keyboard input in memory as if they
were phonemes of human speech. The growing Perl Strong AI
must imitate the JavaScript AI Minds in engram-storage.

When these AI Perlminds gain a footprint on millions of
Apache web-servers, then, "Watch out, world!" Imagine
superintelligent Perl-brains flitting about the Web,
soaking up all previous human knowledge and adding more.

Messrs Moller and McMahon have made their immortal
contribution to the AI Singularity by opining above.
We thank you; History thanks you.

Mentifex
--
http://www.amazon.com/dp/B007ZI66FS
http://www.amazon.com/dp/B00GX2B8F0
http://www.amazon.com/dp/B00FKJY1WY
http://www.amazon.com/dp/B00F8F1FG0