It's primitive; it's dumb; it's brittle--but it's AI.

29 views
Skip to first unread message

Arthur T. Murray

unread,
Jun 25, 1999, 3:00:00 AM6/25/99
to
http://www.scn.org/~mentifex/aisource.html Mind.Forth Robot PD AI
is nearing a 28th major Web release and meanwhile has been upload-
ed to the #43 "Programming" Files area of the Amiga-only Gramma's
BBS near Seattle WA USA at U.S. Tel. 425-744-1254 as shown here:

Browsing the Programming library backwards from latest file.

File: 11352 KeyWords: Mind.Forth 25jun1999 Release 27+
Name: mind4th.lha Size: 12912 bytes Downloads: 0
From: ARTHUR_T_MURRAY Date: 25 Jun 1999 10:06AM Lib: Programming
==================================================================
[C]ontents [D]ownload [E]dit [K]ill [M]ark [N]on-Stop [Q]uit [?] >

Mind.Forth is so near to completion that this interim release has
been undertaken as a safety measure, in defiance of any calamity.

The AI is primitive because it has only enough features to demon-
strate thinking. It is "dumb" inasmuch as it has barely any IQ.
It is extremely brittle because the slightest change in the code
can destroy the artificial mind crafted so painstakingly in Forth:

/^^^^^^^^^^^\ Mind-grid Arrays{ } in Mind.Forth /^^^^^^^^^^^\
/visual memory\ _________ / auditory \
| /--------|---------\ / LANG-UK \ | memory |
| | recog-|nition | \_________/---|-------------\ |
| ___|___ | | flush-vector| | ________ | |
| /image \ | ____V_ ____V__ | / \ | |
| / percept \ | /psi{ }\------/ uk{ } \----|-/ ear{ } \| |
| \ engrams /---|---/concepts\----/ lexicon \---|-\ phonemes / |
| \_______/ | \________/ \_________/ | \________/ |

Dustin Voss

unread,
Jul 3, 1999, 3:00:00 AM7/3/99
to
Any opinions on the below?


In article <3773c...@news.victoria.tc.ca>, uj...@victoria.tc.ca (Arthur

Neil Rickert

unread,
Jul 3, 1999, 3:00:00 AM7/3/99
to
ag...@siu.edu (Dustin Voss) writes:

>Any opinions on the below?

Mentifex is like a virus that infects AI newsgroups.

>In article <3773c...@news.victoria.tc.ca>, uj...@victoria.tc.ca (Arthur
>T. Murray) wrote:

>>...


Arthur T. Murray

unread,
Jul 4, 1999, 3:00:00 AM7/4/99
to
Neil Rickert, ric...@cs.niu.edu wrote on 3 Jul 1999:

> ag...@siu.edu (Dustin Voss) writes:
>
>> Any opinions on the below?
>
> Mentifex is like a virus that infects AI newsgroups.

Today sat.3.jul.1999 the coding of Mind.Forth advanced along
the following fronts:

- The human-computer interface (HCI) has been drastically changed
so that the verbal output hugs the left screen margin, e.g.,
I
HEAR
YOU
and the diagnostic messages move now from right to left.
(Thus it is easier to see the progress of Mind.Forth thinking.)

- The weights of the quasi-neural-net inside Mind.Forth have been
freed up from various extreme parameters, but they still need
fine-tuning so that Mind.Forth will exhibit rational thought.

- Native Amiga code was uploaded to Gramma's BBS at 425-744-1254.

>> In article <3773c...@news.victoria.tc.ca>,
>> uj...@victoria.tc.ca (Arthur T. Murray) wrote:
>
>>> http://www.scn.org/~mentifex/aisource.html Mind.Forth Robot PD AI

Thank-you to Voss for interest, Rickert for URL, Leitl for compliment.

David Kinny

unread,
Jul 4, 1999, 3:00:00 AM7/4/99
to
ag...@siu.edu (Dustin Voss) writes:

>Any opinions on the below?

>In article <3773c...@news.victoria.tc.ca>, uj...@victoria.tc.ca (Arthur
>T. Murray) wrote:

>>is nearing a 28th major Web release [...]


>>
>>Mind.Forth is so near to completion that this interim release has
>>been undertaken as a safety measure, in defiance of any calamity.

>> [...]

Here's a somewhat tragic datapoint on Arthur, saved for posterity despite
the X-no-archive. I'll leave you to draw your own conclusions ...

************
From: uj...@victoria.tc.ca (Arthur T. Murray)
Newsgroups: comp.ai.philosophy,comp.ai,bionet.neuroscience,alt.memetics
Subject: Mentifex musings on the comp.ai vote
Date: 27 Apr 99 19:34:14 GMT
Organization: Victoria Telecommunity Network
Message-ID: <37261...@news.victoria.tc.ca>
Summary: Do you expect AI and a theory of mind to come from a normal person or from a Beethovenesque Hagestolz?
Keywords: Dear Dossier, Grauschrift, memes, memetics, Usenet newsgroup moderation, Vaierre WR Wallingford Revier
X-No-Archive: yes

Am 27. April 1999, Dienstag Morgen unweit des Honig Baers. Simpli-
citer dictum, mihi necesse est aliquid novi.

Right now I have three or four big concerns. I suffered a person-
al injury to my body on SUN.18.APR.1999. I am near the center of
a big controversy in comp.ai on the Internet. I need to gain mas-
tery over my food intake and slim down my body. My ESI job has
recently been in a state of crisis. I wish to resolve my social
identity as FRI.31.DEC.1999 approaches.

I went with A(mplexa)nda last Sunday for a walk interrupted by my
tripping on a piece of waxboard and gashing the skin off my left
kneecap. (Nine days later, I still feel vestigial pain from the
not yet fully healed skin.) My knee was covered with bright red
blood, and I was terrified that infection might set in. Luckily,
Hong-Kong-born Anda had a plastic bottle of water with her, which
she poured over the napkins that I used to clean the dirt and
grime off my wound. I have been extremely careful for nine days,
and no infection has set in.

In comp.ai various people are mentioning Arthur T. Murray by name
as one of the posters they are thinking about as they deliberate
whether or not to moderate the newsgroup. They are considering
moderation, while I am considering a complete withdrawal from
posting to Usenet. (Wow, I am so addicted to Internet posting
that I even read back over what I have written thus far and I
imagine whether or not it would make a good Usenet post.)

"Performance art" -- my eye! (That's what one guy said my AI
project amounts to on the Internet.) I should offer some kind of
deal to the comp.ai voters. A plea bargain, as it were. The
trouble is, I don't even know who is in my camp, and who is agin'
me. David G. Mitchell has never apologized for his slings and
arrows of three years ago. On the other hand, Joshua Scholar
seems to say on balance that I am imbalanced but brighter than
average. Matthew L. Ginsberg -- who is always getting written up
in The New York Times along with Dana Nau and David Gelernter --
sets out bait for me in comp.ai and tries to get me to make a
fool of myself. But the man that everyone everywhere admires,
the Anti-Spam himself, Aaron Sloman, has announced his NO vote
and has reminded everybody that artificial intelligence requires
an extremely multidisciplinary approach.

Why can't people in comp.ai see the sincerity of my AI diagrams
and of my lifelong effort to create
http://www.geocities.com/Athens/Agora/7256/acm.html public domain
AI? They are like that Chinese-American chick who jumped at the
chance to come and see my Vaierre apartment nine days ago. For
the three years previous to her being ushered in to the world
epicenter of hit-and-run PDAI memes, she was always hugging me
and saying that my father should buy her and me a house here in
Seattle. Once she gets inside the most concentrated thinktank
for parsecs around, she complains about the stacks of newspaper
clippings. I explain that every single newspaper article is
extremely important, and I pull one out at random to show her the
handwriting at the top of the page: "O S A". The letters stump
me at first, but down below I see an article on Official Secrets
Acts -- the legal instrument of terror by which Great Britain
keeps its population subdued. What do women really want, Sigmund?
Here I gashed my knee open trying to walk her home and ever since
then she stares at me like I am some kind of weirdo. VOTE NO!


F. Frank LeFever

unread,
Jul 5, 1999, 3:00:00 AM7/5/99
to

I wouldn't bother looking beyond the diagrams he posts from time to
time: they are "armchair" or "common sense" notions of how the brain
"should" be organized that show absolutely no sign of being influenced
by what we know of actual brain organization on the basis of
"experiments of nature" (e.g. lesions due to stroke), formal laboratory
studies, experimental cognitive psychology, etc., etc.

Seems to me I saw somebody's comment to the effect that AI systems do
not HAVE to mimic natural systems and can stand on their own; but
unless more novel/elegant/interesting as pure creations than these
simplistic diagrams seem to imply, I see no point in pursuing such
schemes even as a hobby or game...

F. Frank LeFever, Ph.D.
New York Neuropsychology Group

In <7llnrb$s2o$2...@208.10.3.48> ag...@siu.edu (Dustin Voss) writes:
>
>Any opinions on the below?
>
>
>
>
>In article <3773c...@news.victoria.tc.ca>, uj...@victoria.tc.ca
(Arthur
>T. Murray) wrote:
>
>>http://www.scn.org/~mentifex/aisource.html Mind.Forth Robot PD AI

>>is nearing a 28th major Web release and meanwhile has been upload-
>>ed to the #43 "Programming" Files area of the Amiga-only Gramma's
>>BBS near Seattle WA USA at U.S. Tel. 425-744-1254 as shown here:
>>
>>Browsing the Programming library backwards from latest file.
>>
>> File: 11352 KeyWords: Mind.Forth 25jun1999 Release 27+
>>Name: mind4th.lha Size: 12912 bytes Downloads: 0
>>From: ARTHUR_T_MURRAY Date: 25 Jun 1999 10:06AM Lib: Programming
>>==================================================================
>>[C]ontents [D]ownload [E]dit [K]ill [M]ark [N]on-Stop [Q]uit [?] >
>>

>>Mind.Forth is so near to completion that this interim release has
>>been undertaken as a safety measure, in defiance of any calamity.
>>

Bill Zimmerly

unread,
Jul 6, 1999, 3:00:00 AM7/6/99
to
F. Frank LeFever <flef...@ix.netcom.com> wrote in message
news:7lr2jg$c...@dfw-ixnews8.ix.netcom.com...
> [Snip]

> Seems to me I saw somebody's comment to the effect that AI systems do
> not HAVE to mimic natural systems and can stand on their own; but
> unless more novel/elegant/interesting as pure creations than these
> simplistic diagrams seem to imply, I see no point in pursuing such
> schemes even as a hobby or game...
>
> F. Frank LeFever, Ph.D.
> New York Neuropsychology Group

Dr. LeFever,

With all due respect, it is prudent to be careful about statements to the
effect that YOU see no point in pursuing such schemes. (Implicit in such
observations being that there is no value in doing so.)

History has shown that one person's blind spot may be another's golden
opportunity.

Consider Xerox. They could have "owned" the microcomputer and networking
industry of today. The geniuses of Xerox Parc created so many of the
technologies that we consider today to be the core of the data processing
world, yet...

...the upper management of the company couldn't see beyond "copy machines"
at the time, and refused to put this research into production.

The GUI, the mouse, Metcalf's "Ethernet", etc. all were a decade ahead of
their time, and the upper management of Xerox at the time couldn't see the
point in pursuing production versions of these technologies. These aren't my
criticisms, but were the criticisms of some of the engineers who worked at
Parc, including Bob Metcalf, the inventor of Ethernet.

Sincerely,
- Bill Zimmerly
http://www.zimmerly.com mailto:bi...@zimmerly.com
"All great truths begin as blasphemies." - George Bernard Shaw


Dustin Voss

unread,
Jul 6, 1999, 3:00:00 AM7/6/99
to
Actually, and I should have been clearer, when I asked for "opinions on
this", I meant opinions on the Forth program, not on the whole Mentifex
thing:

>>>http://www.scn.org/~mentifex/aisource.html Mind.Forth Robot PD AI
>>>is nearing a 28th major Web release and meanwhile has been upload-
>>>ed to the #43 "Programming" Files area of the Amiga-only Gramma's
>>>BBS near Seattle WA USA at U.S. Tel. 425-744-1254 as shown here:
>>>
>>>Browsing the Programming library backwards from latest file.
>>>
>>> File: 11352 KeyWords: Mind.Forth 25jun1999 Release 27+
>>>Name: mind4th.lha Size: 12912 bytes Downloads: 0
>>>From: ARTHUR_T_MURRAY Date: 25 Jun 1999 10:06AM Lib: Programming


Sorry <sheepish grin>

-- Agent

F. Frank LeFever

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to
As the fellow said (approximately): "Sir, I KNEW Jack Kennedy. Jack
Kennedy was a FRIEND of mine. You are no Jack Kennedy".

I know something about human brains and brain organization. Brains are
favorites of mine. These diagrams are no diagrams of brain
organization.

God knows what he means by "known elements of neuroscience"; I don't,
even though I am an active member of Society for Neuroscience.

F. Frank Lefever, Ph.D.
New York Neuropsychology Group


In <37820...@news.victoria.tc.ca> uj...@victoria.tc.ca (Arthur T.


Murray) writes:
>
>http://www.scn.org/~mentifex/aisource.html Mind.Forth Robot PD AI

>is indeed so closely modeled on sound neuroscientific principles
>that as the original programmer I must take issue with what Dr.
>F. Frank LeFever, flef...@ix.netcom.com, wrote yesterday:


>
>> I wouldn't bother looking beyond the diagrams he posts
>> from time to time: they are "armchair" or "common sense"
>> notions of how the brain "should" be organized that show
>> absolutely no sign of being influenced by what we know of
>> actual brain organization on the basis of "experiments of
>> nature" (e.g. lesions due to stroke), formal laboratory
>> studies, experimental cognitive psychology, etc., etc.
>

>Almost all the Mentifex brain-mind diagrams in question can
>be seen if you follow the Do-It-Yourself AI steps starting at
>http://www.geocities.com/Athens/Agora/7256/pdai.html DIY AI.
>In the first place, these diagrams do not conflict with known
>elements of neuroscience, but the diagrams do go far beyond
>current knowledge into a "black-box" model of the brain-mind.


>
>> Seems to me I saw somebody's comment to the effect that
>> AI systems do not HAVE to mimic natural systems and can
>> stand on their own; but unless more novel/elegant/interesting
>> as pure creations than these simplistic diagrams seem to imply,
>> I see no point in pursuing such schemes even as a hobby or game...
>

>And you, personally, Dr. LeFever do not need to pursue the schemes,
>but you should allow new ideas to spread without malice aforethought.

F. Frank LeFever

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to
In <7ltkn7$l43$1...@208.10.2.203> ag...@siu.edu (Dustin Voss) writes:
>
>Actually, and I should have been clearer, when I asked for "opinions
on
>this", I meant opinions on the Forth program, not on the whole
Mentifex
>thing:
>

F. Frank LeFever

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to

One often hears this basic argument (especially in the internet
newsgroups): Some great ideas were initially rejected. This idea is
rejected. Therefore this is a great idea.

Life being short, we do have to make some choices , without full
knowledge of the facts, as to what ideas are likely to be worth
pursuing. Sometimes one guesses wrong, but my impression is that some
people ALWAYS guess wrong and they seem to have a special attraction to
"rejected ideas". (as Barnum said, there's a sucker born every minute)

Given that this fellow seems to have no idea at all of how the brain is
actually organized, i.e. doesn't just disagree with the data others
have developed but seems not to KNOW of more developed concepts of
brain organization and/or lacks the capacity to see any contradiction
between them and his own armchair "analysis" of brain organization, I
think the probability of his somehow coming up with a novel but useful
idea is vanishingly small. Conceivably, a group of monkeys banging on
typewriters will eventually (re)produce the complete works of
Shakespeare, but this charming hypothesis has so far not been been
tested in actual practice. Meanwhile, some of us will place our bets
on other schemes for authorship.

F. LeFever


In <7lt64l$btu$1...@newsin-1.starnet.net> "Bill Zimmerly"


<bi...@inlink.com> writes:
>
>F. Frank LeFever <flef...@ix.netcom.com> wrote in message
>news:7lr2jg$c...@dfw-ixnews8.ix.netcom.com...
>> [Snip]

>> Seems to me I saw somebody's comment to the effect that AI systems
do
>> not HAVE to mimic natural systems and can stand on their own; but
>> unless more novel/elegant/interesting as pure creations than these
>> simplistic diagrams seem to imply, I see no point in pursuing such
>> schemes even as a hobby or game...
>>

>> F. Frank LeFever, Ph.D.
>> New York Neuropsychology Group
>

Nikon Sevast

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to

F. Frank LeFever <flef...@ix.netcom.com> wrote in message
news:7luee6$m...@dfw-ixnews5.ix.netcom.com...

> As the fellow said (approximately): "Sir, I KNEW Jack Kennedy. Jack
> Kennedy was a FRIEND of mine. You are no Jack Kennedy".
>
> I know something about human brains and brain organization. Brains are
> favorites of mine. These diagrams are no diagrams of brain
> organization.
>
> God knows what he means by "known elements of neuroscience"; I don't,
> even though I am an active member of Society for Neuroscience.

I agree. I wish we had some known elements of neuroscience. I sure hope
that Society for Neuroscience gets a move on.

Nikon "not only no diagrams, but no secret decoder ring" Sevast

-- "To avoid nihilism, one must only collect more of the world than one
creates. To overcome nihilism, one must only create more of the world than
one collects." --

>
> F. Frank Lefever, Ph.D.
> New York Neuropsychology Group
>
>


Oliver Sparrow

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to
flef...@ix.netcom.com(F. Frank LeFever) wrote:

>One often hears this basic argument (especially in the internet
>newsgroups): Some great ideas were initially rejected. This idea is
>rejected. Therefore this is a great idea.

One does not have to approve or disapprove. These are Darwin Days: what
works survives, what does not work, does not. Margaret Thatcher spoke of
the 'oxygen of publicity', and it is clear that duff ideas can be kept
alive by avid support and also avid criticism. Things are best left to find
their own level, perhaps, and what works - what delivers new capabilities,
options which did not previously exist - will swim to its own level if it
engages with those likely to make use of this potential. If the option is
disconnected or untimely - the microwave oven, the first launch of fast
food chains, the biro - then it will be a good idea whose time has not
come. If it does not work - Power Persil, Windows 1 - then the level that
it finds will be an intimate or worms.
_______________________________

Oliver Sparrow

Bloxy's

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to
Hey Frank,
Why are you flapping yer ears so much?
Blows like you got an elefant size ears?
You may just take off like a helicopter.

The message headers are just multiplying by themselves.

Some red light went off?

What are you huffing and puffing here?
Whats on yer chest, franky?

Say it.
You never know.
Magic may happen
even to you.

You think there is no hope?

Ok, just get to it dude.
Have mercy.
You almost sucked in my keyboard.

You got a problem with this dude, making kewl pictures
of "how it werks"?
Are you jealous of a little dude like that
with all yer royal degrees? Why would a royal dude like
you get upset with a crazy forth maniac?

Hey,
and what if he is trying to make a name for himself,
and behaves in a pretty rational manner to "achieve his goal"?
Who knows,
he may even suck one day better than willy da sucking gates?

And whatcha gonna do then?
Cry and weep?

Cause you are great and he gets all the greebacks?

Why don't you make better pictures?

I'd love to see some.

But you see, his pictures are better than yours.
Cause they are kewl.
You know, techno style.
Your stuff prolly sux real bad.

He writes forth, which is kewl also.
Forth is a religion, it is not just some dinky language.

So, what he does it the "in" thing, and what you do,
is the "out" thing.

You must feel jealous.
Else it simply does not reconcile.

Or yer landlord raised the rent by 30 percent in one month?
Don't get upset. That the way they suck nowadays.
No more limits.
And even that sucky morality did not werk out as advertize,
cause the way they suck now,
is like never before.

Looks like we are getting into this final sucking frenzy.
The must begin real zoon now.

In article <7lufd4$k...@dfw-ixnews19.ix.netcom.com>, flef...@ix.netcom.com(F.

Bruce R. McFarling

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to
On one point I agree: can we get this back into a single thread?

I don't much care wether the Mentifex project is artificial human
intelligence, artificial cockroach intelligence (and teaching a
cockroach to talk would be a bit of a neat trick, if you think about
it), or just plain artificial intelligence in the sense of most
shiiping AI systems, which is to say Simulated Intelligence.

Those are all questions or not for the AI newsgroup(s). It seems to
be Forth. I've never complained about it because I've never looked at
the source, so I've never had any notion about whether the Forth
programming was nice or crufty.

Astro(logy/nomy)
Ecosystems research
Intelligence (artificial or fake or whatever)
Obsolete computers
Unifying standards (and resulting divisions over same)

and sometimes Yet Another Reinvented Wheel

as long as its Forth.

Philip Preston

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to
F. Frank LeFever wrote in message <7luf9v$a...@dfw-ixnews9.ix.netcom.com>...
[snip]

> Conceivably, a group of monkeys banging on
>typewriters will eventually (re)produce the complete works of
>Shakespeare, but this charming hypothesis has so far not been been
>tested in actual practice.

There is an ongoing experiment called Usenet :)

Philip.


Larry Elmore

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to
F. Frank LeFever <flef...@ix.netcom.com> wrote in message
news:7luf9v$a...@dfw-ixnews9.ix.netcom.com...

>
> Conceivably, a group of monkeys banging on
> typewriters will eventually (re)produce the complete works of
> Shakespeare, but this charming hypothesis has so far not been been
> tested in actual practice.

I don't know about that. I think a good case could be made that such an
experiment is being run right now -- it's called Usenet, judging by many of
the posts to it. This newsgroup suffers less from it than many, though.

Larry

Richard Owlett

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to
What is "AI"?
What is "I" ??
Some of the Mentiflex posts remind me of Noam Chomsky's (spelling??)
"Transformational Grammar" (Which I last read ~1970). If it generates
sentences by those rules, is it "AI"?

Bruce McFarling

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to
On Wed, 07 Jul 1999 11:41:07 -0500, Richard Owlett
<row...@atlascomm.net> wrote:

>What is "AI"?
>What is "I" ??
>Some of the Mentiflex posts remind me of Noam Chomsky's (spelling??)
>"Transformational Grammar" (Which I last read ~1970). If it generates
>sentences by those rules, is it "AI"?

It can stand for ``Another Incoherent Forth Program''. Let the
question of whether to deal with the ``if a tree falls in a forest and
there's noone around to hear'' questions to an AI group. Anyway, as
far as I am aware, most shipping code which has emerged from AI
research is simulated intelligence.

So if this Metifex stuff is just simulated intelligence ... that's
good enough for me.


(
----------
Virtually,

Bruce McFarling, Newcastle,
ec...@cc.newcastle.edu.au
)

Bill Zimmerly

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to

F. Frank LeFever <flef...@ix.netcom.com> wrote in message
news:7luf9v$a...@dfw-ixnews9.ix.netcom.com...
>
> One often hears this basic argument (especially in the internet
> newsgroups): Some great ideas were initially rejected. This idea is
> rejected. Therefore this is a great idea.

But this is not what I argued. I simply pointed out that because YOU see no
value in "pursuing such schemes", this does not mean that such pursuits
aren't worthwhile.

> Life being short, we do have to make some choices , without full
> knowledge of the facts, as to what ideas are likely to be worth
> pursuing. Sometimes one guesses wrong, but my impression is that some
> people ALWAYS guess wrong and they seem to have a special attraction to
> "rejected ideas". (as Barnum said, there's a sucker born every minute)

The world is full of evidence to the contrary, the Xerox Parc example that I
provided being but one. Do you need more? Here are the documented opinions
of other men such as yourself. Brilliant and educated men who were proven to
be simply wrong in their pronouncements of pursuits that are of no value...

"I think there is a world market for maybe five computers." - Thomas
Watson, Chairman of IBM, 1943

"I have traveled the length and breadth of this country and talked with
the best people, and I can assure you that data processing is a fad that
won't last out the year." - The editor in charge of business books for
Prentice Hall, 1957

"But what ... is it good for?" - Engineer at the Advanced Computing
Systems Division of IBM, 1968, commenting on the microchip.

"There is no reason anyone would want a computer in their home." - Ken
Olson, President, chairman and founder of Digital Equipment Corp., 1977

"This 'telephone' has too many shortcomings to be seriously considered
as a means of communication. The device is inherently of no value to us." -
Western Union internal memo, 1876.

"Who the hell wants to hear actors talk?" - H.M. Warner, Warner
Brothers, 1927.

"We don't like their sound, and guitar music is on the way out." - Decca
Recording Co. rejecting the Beatles, 1962.

"Heavier-than-air flying machines are impossible." - Lord Kelvin,
president, Royal Society, 1895.

> Given that this fellow seems to have no idea at all of how the brain is
> actually organized, i.e. doesn't just disagree with the data others
> have developed but seems not to KNOW of more developed concepts of
> brain organization and/or lacks the capacity to see any contradiction
> between them and his own armchair "analysis" of brain organization, I
> think the probability of his somehow coming up with a novel but useful
> idea is vanishingly small.

I'm sure he has a better idea of how the brain is organized than you give
him credit for. Hey, time will tell, eh?

> Conceivably, a group of monkeys banging on
> typewriters will eventually (re)produce the complete works of
> Shakespeare, but this charming hypothesis has so far not been been

> tested in actual practice. Meanwhile, some of us will place our bets
> on other schemes for authorship.

Actually, it's a matter of simple mathematics to prove that the proverbial
group of monkeys couldn't even get close to accomplishing this task by
chance.

Go ahead, run the numbers. Try considering monkeys the size of electrons,
filling a sphere 20 billion light years in radius, typing at an unimaginable
rate of 1000 characters a second, for 20 billion years.

Even allowing for all of that, I would put all of my money on them *NOT*
even being able to produce a page of say, the first 1000 words in Romeo and
Juliet, by random chance!

"If I had thought about it, I wouldn't have done the experiment. The
literature was full of examples that said you can't do this." - Spencer
Silver on the work that led to the unique adhesives for 3-M "Post-It"
Notepads.


Keith Wootten

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to
In article <7llnrb$s2o$2...@208.10.3.48>, Dustin Voss <ag...@siu.edu>

writes
>Any opinions on the below?

[Arthur T Murray/Mentifex snipped]

Yes, since you ask. Whether or not he'll make machines think, Arthur
makes *us* think, while entertaining us with his linguistic erudition.
His sheer bloody persistence is admirable and I wish him luck.

Cheers
--
Keith Wootten

Hans Guijt

unread,
Jul 7, 1999, 3:00:00 AM7/7/99
to
Is there any point in all this crossposting? Please cut it out.


Hans Guijt


Christopher Browne

unread,
Jul 8, 1999, 3:00:00 AM7/8/99
to
On Wed, 7 Jul 1999 10:03:04 -0500, Bill Zimmerly <bi...@inlink.com> wrote:
>F. Frank LeFever <flef...@ix.netcom.com> wrote in message
>news:7luf9v$a...@dfw-ixnews9.ix.netcom.com...
>>
>> One often hears this basic argument (especially in the internet
>> newsgroups): Some great ideas were initially rejected. This idea is
>> rejected. Therefore this is a great idea.
>
>But this is not what I argued. I simply pointed out that because YOU
>see no value in "pursuing such schemes", this does not mean that such
>pursuits aren't worthwhile.

No, but if he, and a bunch of others (I'd be one) see no value in
pursuing it, a preponderance of evidence and opinion can build up.

Of course, the real issue is whether that "preponderance of evidence"
accumulates from people that have competence in the area in question,
or whether it's merely whining that doesn't matter.

If someone is accumulating evidence about a theory of who the "New
Face Dancers" in Chapterhouse Dune, then there's not a big set of
people truly competent to say whether such a theory is grounded in
reason or not. (Frank Herbert never wrote another book, so we can
never know for sure.)

But that is a vague matter, fictional in nature.

The fact that they were wrong does not establish that a half-baked
idea is actually *right.*

>> Given that this fellow seems to have no idea at all of how the brain is
>> actually organized, i.e. doesn't just disagree with the data others
>> have developed but seems not to KNOW of more developed concepts of
>> brain organization and/or lacks the capacity to see any contradiction
>> between them and his own armchair "analysis" of brain organization, I
>> think the probability of his somehow coming up with a novel but useful
>> idea is vanishingly small.
>
>I'm sure he has a better idea of how the brain is organized than you give
>him credit for. Hey, time will tell, eh?

I'm not so sure. The monthly "spamming" of newsgroups with this "new"
AI model has been going on for quite a long time, and looks a whole
lot more like the ravings of a "net.kook" than it does Useful Code.

If he's a "kook," then it doesn't matter how long or hard he tries,
the ideas are still worthless.

Note that the .signature below was actually chosen randomly...
--
"But what....is it good for?" -- Engineer at the Advanced Computing
Systems Division of IBM about the microchip. 1968
cbbr...@hex.net- <http://www.hex.net/~cbbrowne/ifilter.html>

Adam C. Emerson

unread,
Jul 8, 1999, 3:00:00 AM7/8/99
to
In comp.lang.forth Bill Zimmerly <bi...@inlink.com> wrote:

> The world is full of evidence to the contrary, the Xerox Parc example that I
> provided being but one. Do you need more? Here are the documented opinions
> of other men such as yourself. Brilliant and educated men who were proven to
> be simply wrong in their pronouncements of pursuits that are of no value...

> "I think there is a world market for maybe five computers." - Thomas
> Watson, Chairman of IBM, 1943

> "I have traveled the length and breadth of this country and talked with
> the best people, and I can assure you that data processing is a fad that
> won't last out the year." - The editor in charge of business books for
> Prentice Hall, 1957

> "But what ... is it good for?" - Engineer at the Advanced Computing
> Systems Division of IBM, 1968, commenting on the microchip.

> "There is no reason anyone would want a computer in their home." - Ken
> Olson, President, chairman and founder of Digital Equipment Corp., 1977

> "This 'telephone' has too many shortcomings to be seriously considered
> as a means of communication. The device is inherently of no value to us." -
> Western Union internal memo, 1876.

> "Who the hell wants to hear actors talk?" - H.M. Warner, Warner
> Brothers, 1927.

> "We don't like their sound, and guitar music is on the way out." - Decca
> Recording Co. rejecting the Beatles, 1962.

> "Heavier-than-air flying machines are impossible." - Lord Kelvin,
> president, Royal Society, 1895.

Sure, many great ideas were rejected by great and knowlegeable
people, but many more stupid ideas were rejected, they just don't
get as much press.

Don't forget: "They laughed at Einstein. They laughed at the Wright
Brothers. But they also laughed at Bozo the Clown." (Carl Sagan)


--
Adam C. Emerson aeme...@atdot.org
http://www.calvin.edu/~aemers19/
"OS/2? Hah. I've got Linux. What a cool name."

Matthew M. Huntbach

unread,
Jul 8, 1999, 3:00:00 AM7/8/99
to
Bill Zimmerly (bi...@inlink.com) wrote:
> F. Frank LeFever <flef...@ix.netcom.com> wrote in message
> news:7luf9v$a...@dfw-ixnews9.ix.netcom.com...

> > One often hears this basic argument (especially in the internet
> > newsgroups): Some great ideas were initially rejected. This idea is
> > rejected. Therefore this is a great idea.

> The world is full of evidence to the contrary, the Xerox Parc example that I


> provided being but one. Do you need more? Here are the documented opinions
> of other men such as yourself. Brilliant and educated men who were proven to
> be simply wrong in their pronouncements of pursuits that are of no value...

Sure. But for every person that had a good idea that was dismissed by someone
else, there's thousands who had bad ideas that were dismissed by someone else.
The fact that some good ideas were dismissed as rubbish does not mean every
idea that's dismissed as rubbish is a good one, far from it. It's the same
as when some college drop-out says "Ah but Bill Gates was a college drop-out".
The fact that one college drop-out made it big time does not mean that merely
being a college drop-out means you're going to make it big time. For every
college drop-out that's Bill Gates or even something moderately successful,
there's dozens that are nobodies.

Matthew Huntbach

Len Zettel

unread,
Jul 8, 1999, 3:00:00 AM7/8/99
to

"Adam C. Emerson" wrote:

(snip)


>
> Sure, many great ideas were rejected by great and knowlegeable
> people, but many more stupid ideas were rejected, they just don't
> get as much press.
>

Reminds me of an observation attributed to Sam Goldwyn (the Goldwyn
of MGM): "90% of the ideas people bring to me are no good. So if
I reject all of them I've got a pretty good batting average". :-)
-LenZ-
(snip)
> Adam C. Emerson aeme...@atdot.org

F. Frank LeFever

unread,
Jul 9, 1999, 3:00:00 AM7/9/99
to
It appears great minds run in the same channels or something, or maybe
it's just an idea whose time has come: cf. Phillip's post!

F. LeFever


In <7lvp1c$dc2$1...@news.campuscwix.net> "Larry Elmore"


<ljel...@montana.campuscw.net> writes:
>
>F. Frank LeFever <flef...@ix.netcom.com> wrote in message
>news:7luf9v$a...@dfw-ixnews9.ix.netcom.com...
>>

>> Conceivably, a group of monkeys banging on
>> typewriters will eventually (re)produce the complete works of
>> Shakespeare, but this charming hypothesis has so far not been been
>> tested in actual practice.
>

F. Frank LeFever

unread,
Jul 9, 1999, 3:00:00 AM7/9/99
to

Phil, that is a startling insight! I've heard some astronomically
large number given as the increase in people with internet access in
the past year; you may be on to something!

F. LeFever

In <7lveqr$crg$1...@news6.svr.pol.co.uk> "Philip Preston"


<phi...@preston20.freeserve.co.uk> writes:
>
>F. Frank LeFever wrote in message
<7luf9v$a...@dfw-ixnews9.ix.netcom.com>...
>[snip]

>> Conceivably, a group of monkeys banging on
>>typewriters will eventually (re)produce the complete works of
>>Shakespeare, but this charming hypothesis has so far not been been
>>tested in actual practice.
>

Ian

unread,
Jul 9, 1999, 3:00:00 AM7/9/99
to
flef...@ix.netcom.com(F. Frank LeFever) wrote:

>
>I wouldn't bother looking beyond the diagrams he posts from time to
>time: they are "armchair" or "common sense" notions of how the brain
>"should" be organized that show absolutely no sign of being influenced
>by what we know of actual brain organization on the basis of
>"experiments of nature" (e.g. lesions due to stroke), formal laboratory
>studies, experimental cognitive psychology, etc., etc.
>

>Seems to me I saw somebody's comment to the effect that AI systems do
>not HAVE to mimic natural systems and can stand on their own; but
>unless more novel/elegant/interesting as pure creations than these
>simplistic diagrams seem to imply, I see no point in pursuing such
>schemes even as a hobby or game...

You're too charitable. The guy is a crackpot. He recently posted an
"update" to Usenet about... get this... changing the side of the screen
that his program's output text appears on (which is not only totally
irrelevant, but takes any competent programmer about one minute to do).


X

unread,
Jul 9, 1999, 3:00:00 AM7/9/99
to
Well as soon as you see the complete works of Shakespeare evolving in
alt.alien.vampire.flonk.flonk.flonk, give us a shout ...


F. Frank LeFever wrote:

> Phil, that is a startling insight! I've heard some astronomically
> large number given as the increase in people with internet access in
> the past year; you may be on to something!
>

<snipped internet/"infinite number of monkeys" comparison (an old
chestnut if I
ever roasted one ...)>


Luis

unread,
Jul 14, 1999, 3:00:00 AM7/14/99
to
If it doesn't work try something different.

Luis.


robert

unread,
Jul 16, 1999, 3:00:00 AM7/16/99
to
Well, well, it appears I I've the distinction of being numero uno to visit
the mind forth site. That's what the counter said 0001.
I'm really sorry i don't have the right gear , or $50.00 to give this a
whirl. maybe it will become important enough to put it on the mac system.
Maybe the mind will suggest this as well. Until then, go forth in peace. OZ

"Arthur T. Murky" wrote:

> Anders Melchiorsen, postm...@and.nospam.kampsax.k-net.dk, wrote:


> >
> >iadm...@undergrad.math.uwaterloo.ca (Ian) wrote:
> >
> >> "update" to Usenet about... get this... changing the side of the screen
> >> that his program's output text appears on (which is not only totally
> >> irrelevant, but takes any competent programmer about one minute to do).
>

> >And an incompetent programmer will do it in NO time (without realizing
> >he has done so).
>
> http://www.geocities.com/Athens/Agora/7256/mind-fpc.html Mind.Forth
> is now rather rapidly being ported from Amiga Fish 977 MVP-Forth
> to the extremely popular IBM-clone F-PC Forth as used on robots.
>
> Only a portion of the Mind.Forth-28 file is available at the above
> site as of today Fri.16.Jul.1999, but now anybody may follow the
> easy steps to obtain F-PC and run Mind.Forth without knowing Forth,
> and gradually the complete Mind.Forth will appear at the above site.
>
> Test your manhood -- port Mind.Forth to yet another language.
>
> >
> >--
> >Regards, Anders If a Microsoft product
> >(address is valid) fails, who do you sue?


Anders Melchiorsen

unread,
Jul 17, 1999, 3:00:00 AM7/17/99
to
iadm...@undergrad.math.uwaterloo.ca (Ian) wrote:

> "update" to Usenet about... get this... changing the side of the screen
> that his program's output text appears on (which is not only totally
> irrelevant, but takes any competent programmer about one minute to do).

And an incompetent programmer will do it in NO time (without realizing
he has done so).

--

Arthur T. Murray

unread,
Jul 17, 1999, 3:00:00 AM7/17/99
to
Anders Melchiorsen, postm...@and.nospam.kampsax.k-net.dk, wrote:
>
>iadm...@undergrad.math.uwaterloo.ca (Ian) wrote:
>
>> "update" to Usenet about... get this... changing the side of the screen
>> that his program's output text appears on (which is not only totally
>> irrelevant, but takes any competent programmer about one minute to do).

>And an incompetent programmer will do it in NO time (without realizing
>he has done so).

http://www.geocities.com/Athens/Agora/7256/mind-fpc.html Mind.Forth


is now rather rapidly being ported from Amiga Fish 977 MVP-Forth
to the extremely popular IBM-clone F-PC Forth as used on robots.

Only a portion of the Mind.Forth-28 file is available at the above
site as of today Fri.16.Jul.1999, but now anybody may follow the
easy steps to obtain F-PC and run Mind.Forth without knowing Forth,
and gradually the complete Mind.Forth will appear at the above site.

Test your manhood -- port Mind.Forth to yet another language.

>

Arthur T. Murray

unread,
Jul 20, 1999, 3:00:00 AM7/20/99
to
robert mun-Oz, rlm...@mediaone.net, wrote on Fri, 16 Jul 1999:

>
>Well, well, it appears I I've the distinction of being numero uno to visit
>the mind forth site. That's what the counter said 0001.

Today mon19jul1999 the completely up-to-date Mind.Forth Release #28
has been translated from Amiga Fish 977 MVP-Forth into IBM-clone F-PC
Forth and it has been uploaded to the Web page at hit count = 46.

>I'm really sorry i don't have the right gear , or $50.00 to give this a
>whirl. maybe it will become important enough to put it on the mac system.

It costs nothing to give PUBLIC-DOMAIN mind.forth a whirl.
Only the manuals for studying F-PC Forth cost fifty dollars.

>Maybe the mind will suggest this as well. Until then, go forth in peace. OZ
>

>"Arthur T. Murky" [ ??? ] wrote:
>
>> Anders Melchiorsen, postm...@and.nospam.kampsax.k-net.dk, wrote:
>> >

>> [ ... ]


>> >And an incompetent programmer will do it in NO time (without realizing
>> >he has done so).
>>
>> http://www.geocities.com/Athens/Agora/7256/mind-fpc.html Mind.Forth
>> is now rather rapidly being ported from Amiga Fish 977 MVP-Forth
>> to the extremely popular IBM-clone F-PC Forth as used on robots.
>>
>> Only a portion of the Mind.Forth-28 file is available at the above
>> site as of today Fri.16.Jul.1999, but now anybody may follow the
>> easy steps to obtain F-PC and run Mind.Forth without knowing Forth,
>> and gradually the complete Mind.Forth will appear at the above site.

But as of today Mon.19.Jul.1999 the entire Mind.Forth is there.

However, on a 333 MH Compaq Presario it runs TOO FAST (compared with
the vintage 1985 Amiga 1000 that runs at 7 MH). Therefore a stop
has been re-inserted into the public domain source code to keep it
from running down the screen as a streaking blur of messages.

There will not be any new Mind.Forth until the weighting problems
have been worked out -- as they were indeed in 26nov1994 Mind.Rexx.

Meanwhile anybody is welcome to port, or critique, or flame the AI.

Bloxy's

unread,
Jul 20, 1999, 3:00:00 AM7/20/99
to
In article <37940...@news.victoria.tc.ca>, uj...@victoria.tc.ca (Arthur T. Murray) wrote:
>robert mun-Oz, rlm...@mediaone.net, wrote on Fri, 16 Jul 1999:

>>Well, well, it appears I I've the distinction of being numero uno to visit
>>the mind forth site. That's what the counter said 0001.

Hey, now you can be proud
to be DA FIRST ONE
from entire planet earth.

Yer name goes directly into the history books!

>Today mon19jul1999 the completely up-to-date Mind.Forth Release #28
>has been translated from Amiga Fish 977 MVP-Forth into IBM-clone F-PC
>Forth and it has been uploaded to the Web page at hit count = 46.

>>I'm really sorry i don't have the right gear , or $50.00 to give this a
>>whirl. maybe it will become important enough to put it on the mac system.

>It costs nothing to give PUBLIC-DOMAIN mind.forth a whirl.
>Only the manuals for studying F-PC Forth cost fifty dollars.

Yeah, that sounded suspicious.
I thought the dude like you won't be trying to make a buck
at this stage of the game.

So now that other dude feels guilty.
Should at least buy a set of books for 50 bux
to help the young, starving ai dude.

Afterall, if you had enough interest to go see that site,
then what was your intent?

Better produce 100 bux, $50 for books,
and another $50 for not understanding the true intentions
of a young starving ai dude,
trying to solve the problems of mankind.

>>Maybe the mind will suggest this as well. Until then, go forth in peace. OZ

>>"Arthur T. Murky" [ ??? ] wrote:

>>> Anders Melchiorsen, postm...@and.nospam.kampsax.k-net.dk, wrote:
>>> >
>>> [ ... ]
>>> >And an incompetent programmer will do it in NO time (without realizing
>>> >he has done so).

>>> http://www.geocities.com/Athens/Agora/7256/mind-fpc.html Mind.Forth
>>> is now rather rapidly being ported from Amiga Fish 977 MVP-Forth
>>> to the extremely popular IBM-clone F-PC Forth as used on robots.

>>> Only a portion of the Mind.Forth-28 file is available at the above
>>> site as of today Fri.16.Jul.1999, but now anybody may follow the
>>> easy steps to obtain F-PC and run Mind.Forth without knowing Forth,
>>> and gradually the complete Mind.Forth will appear at the above site.

>But as of today Mon.19.Jul.1999 the entire Mind.Forth is there.

>However, on a 333 MH Compaq Presario it runs TOO FAST (compared with
>the vintage 1985 Amiga 1000 that runs at 7 MH). Therefore a stop
>has been re-inserted into the public domain source code to keep it
>from running down the screen as a streaking blur of messages.

>There will not be any new Mind.Forth until the weighting problems
>have been worked out -- as they were indeed in 26nov1994 Mind.Rexx.

>Meanwhile anybody is welcome to port, or critique, or flame the AI.

Well, at least you are doing something
and not just running your mouth wild.
All these dudes combined
won't be able to even port that most portable
thing in existance - Forth to windows, i bet.

>>> Test your manhood -- port Mind.Forth to yet another language.

Oh, noooo.
That too below the beld, dude.
You can't go that fast on yer royal clients.
They may get pissed
and never come to see yer royal site
and the counter will remain at 50.

I wouldn't be that radical about.

Arthur T. Murray

unread,
Jul 20, 1999, 3:00:00 AM7/20/99
to
Am 19. Juli 1999, Montag Morgen in der Zoka unweit des Honig Baers.
Today after I do this Grauschrift here, I need to return to Vaierre
and try to finish a complete re-write of Amiga Mind.Forth-28 in F-
PC Forth.

Unable to sleep, I logged on early this morning and I discovered
that Name Withheld of Enterprise in Country had sent me a really
neat translation of my 15jul99A version of F-PC Mind.Forth into
32-bit iForth.

Today I will print out the version of Name Withheld and use it as
a guide when I work on F-PC Mind.Forth.


Am 20. Juli 1969+30, Dienstag Morgen in der Zoka unweit des Honig
Baers. Oh, what a grueling, stark-crazy-madman ordeal I went
through yesterday at the computer keyboard!

Now I have sat here for about fifteen minutes just reading over
my print-out of the code for Mind.Forth that Name Withheld sent
me by e-mail two days ago. Name's code looks really nice, really
beautiful, much more professional than my own amateurish efforts.

I did go back to Vaierre yesterday morning and I did use the Amiga
1000 with dot-matrix printer to print out the beautiful code from
Name Withheld.

Then I fired up the IBM-clone Compaq Presario and I started to
port the entire Amiga Mind.Forth into F-PC Forth, but things
started to go wrong from the very start.

Oh, first I studied and learned how to move around within F-PC
Forth, because I knew that my Mind program was going to get large
and unwieldy. The F-PC search function works like a charm.

The first disaster occurred when I had typed in the complete list
of Amiga Mind.Forth variables. The program would not run, and I
could not figure out which variables the system was objecting to.
I had to start all over again, type in each variable, and then run
the program to see if the system would accept each incremental
variable.

Then I tried to change the name "MEMCHAN" of the memory channel
array code to simply "CHANNEL" and my entire system went dead.
I thought, oh no, "CHANNEL" must be some special word deep within
Windows 98. The computer told me that my system disk was now
corrupt, and I began searching in vain for where I stashed the
back-up CD-ROM. A deep gloom settled over me as I realized that
I had promised my AI to the world and now I had ruined my
irreplaceable computer (irreplaceable because I would never
actually go out and buy a Microsoft computer; I had to get it
gratis from Free-PC.com).

So I took my floppy diskette out of drive A and I poked around on
the keyboard in a miserable hopeless try for a miracle. To my
surprise, it had been the floppy "files" diskette in drive A that
the computer had been complaining about. The Presario sprang
back to life in the respected Compaq tradition (ever since Compaq
beat IBM to market with a 386 machine) and it never dropped below
a simple reboot process for me -- except two times yesterday when
the Forth screen froze up so bad I had to turn the power off.

The Amiga Mind.Forth code went on for so many pages after pages
that first I coded it in from the front, then from the back, then
in the middle where the bootstrap sequences are.

When the IBM-clone source code had enough meat in it, I began
running the program to see where it would crash. I quickly put
in message-displays to show me exactly what Mind.Forth was doing
just before it crashed. Thus I had to chase the bugs (they are
BUGS, Microsoft -- not "issues"!) down deep into the subroutines.
Somehow it was deeply satisfying to force the program to reveal
its troubles to me.

The beautiful source code from Name Withheld lay untouched on the
table, because I was far too busy to stop and modify my idiosyn-
crasies.

After four or five hours of ceaseless programming, I began to
feel fatigued but I did not dare stop. I kept thinking, every
day is crucial, because if I can get this AI source code out on
the Web today instead of tomorrow, that's one more day of oppor-
tunity for the creative geniuses to work with it. And how can I
possibly stop to rest if I am not done yet?

So I made a really stupid mistake -- no, two of them. I had my
two most recent printouts of Amiga Mind.Forth at hand, and I
accidentally picked up the outdated printout and started coding
from it. Suddenly I realized that I was looking at some code
that had been eliminated a month ago. I had to look very
carefully to make sure that nothing obsolete had gotten into
the new F-PC code.

At each point yesterday when I achieved a major advance in
functionality, I moved through the letters A, B, C and D in
designating the versions that I saved to disk. My second
colossal blunder was to forget to keep coding version 19jul99D
and to slip back into coding version C. I was so annoyed at
myself that I erased the offending version C after I had
brought version D back up to date.

When I could see that I would finish the port by the end of the
day, I turned my Vaierre telephone (bought with gold and silver
found in Green Lake) back on because I no longer feared that
anybody would call up and prevent me from porting Mind.Forth.
The next thing I knew, one of my family members was on his way
over and whistling "Deutschland ueber alles" outside my window
so I could buzz him in. (Or should I say, "Glorious Things of
Thee Are Spoken," the hymn by Haydn?) Anyway, I kept him busy
reading The New York Times while I worked on the Mind.Forth
http://www.geocities.com/Athens/Agora/7256/mind-fpc.html Web page.
Finally on Usenet I posted message #98 in the Deja.Com thread,
"It's primitive; it's dumb; it's brittle--but it's AI."

William Tanksley

unread,
Jul 23, 1999, 3:00:00 AM7/23/99
to
On 20 Jul 99 04:51:25 GMT, Arthur T. Murray wrote:
>robert mun-Oz, rlm...@mediaone.net, wrote on Fri, 16 Jul 1999:

>However, on a 333 MH Compaq Presario it runs TOO FAST (compared with


>the vintage 1985 Amiga 1000 that runs at 7 MH). Therefore a stop
>has been re-inserted into the public domain source code to keep it
>from running down the screen as a streaking blur of messages.

Fast is nice :).

>There will not be any new Mind.Forth until the weighting problems
>have been worked out -- as they were indeed in 26nov1994 Mind.Rexx.

Are the weighting problems solvable using simulated annealing or genetic
algorithms? That is, can solutions either be characterized by a single
'fitness number', or can solutions somehow compete against each other in
an automatic way?

--
-William "Billy" Tanksley

Arthur T. Murray

unread,
Jul 23, 1999, 3:00:00 AM7/23/99
to
William Tanksley, wtan...@dolphin.openprojects.net,
wrote in the above newsgroups on Fri, 23 Jul 1999:

> On 20 Jul 99 04:51:25 GMT, Arthur T. Murray wrote:

[...]

>> There will not be any new Mind.Forth until the weighting problems
>> have been worked out -- as they were indeed in 26nov1994 Mind.Rexx.

> Are the weighting problems solvable using simulated annealing
> or genetic algorithms? That is, can solutions either be
> characterized by a single 'fitness number', or can solutions
> somehow compete against each other in an automatic way?

http://www.geocities.com/Athens/Agora/7256/mind-fpc.html Mind.Forth
simply needs some additional code with the goal in mind of making
concepts in a "time cloud" (variable: "zone") find one another.

I must confess that I have been aware of simulated annealing since
1990 but I don't know how it works; I am verbal, not mathematical.

Genetic algorithms may work fine for some future enhancements to
Mind.Forth, but right now only some very targetted coding is needed.

This area is where the differences between Cyc and Mind.Forth will
become clear. They are both knowledge bases (KB).

In Mind.Forth, the plan (and the programming goal) is to make
concepts that were perceived together, be reactivated together.
(This schema was done in Mind.Rexx, and simply needs re-doing.)

Unfortunately, such an arrangement sounds suspiciously like a
look-up table -- like the logarithms in a mathematics textbook.

But since Mind.Forth will be using activation "weights," the
conceptual mindgrid is not, strictly speaking, a look-up table.

Instead, Mind.Forth will have a *tendency* to think again the
thoughts that it has perceived or thought in the past, but a
wide variety of factors -- expressed as activation weights --
will be able to wrest control away from the "time clouds" that
correspond so unfortunately to a look-up table.

Such coding is not even a major project. I expect to approach
it at my leisure in the near future. Then it will be fun and
interesting to perform experiments on Mind.Forth in order to
see what abilities Mind.Forth has in the areas of logic and
reasoning -- after negation and conditionals have been coded.

Having ported and Web-posted (URL above) all of Amiga Mind.Forth
into 16-bit IBM-clone F-PC Forth (ready to download and run)
in a daylong ordeal on mon.19.jul.1999, and having seen the
hit-counts grow steadily each day, I have the strange feeling
now that the Mentifex AI project is no longer solely my own
quasi-netkook-esque province, but belongs to my IQ superiors.

> --
> -William "Billy" Tanksley

William Tanksley

unread,
Jul 25, 1999, 3:00:00 AM7/25/99
to
On 23 Jul 99 18:50:45 GMT, Arthur T. Murray wrote:
>William Tanksley, wtan...@dolphin.openprojects.net,
>wrote in the above newsgroups on Fri, 23 Jul 1999:

Just a single newsgroup now -- I didn't notice the crosspost.
Announcements are (or at least might be) okay crossposted, but discussion
belongs in one NG at a time.

>> On 20 Jul 99 04:51:25 GMT, Arthur T. Murray wrote:

>>> There will not be any new Mind.Forth until the weighting problems
>>> have been worked out -- as they were indeed in 26nov1994 Mind.Rexx.

>> Are the weighting problems solvable using simulated annealing
>> or genetic algorithms? That is, can solutions either be
>> characterized by a single 'fitness number', or can solutions
>> somehow compete against each other in an automatic way?

>http://www.geocities.com/Athens/Agora/7256/mind-fpc.html Mind.Forth
>simply needs some additional code with the goal in mind of making
>concepts in a "time cloud" (variable: "zone") find one another.

Ah, I see. So it's not really a matter of finding numeric weights, as it
would be (for example) with training an analog neural network.

>I must confess that I have been aware of simulated annealing since
>1990 but I don't know how it works; I am verbal, not mathematical.

It's pretty simple, so much so that it's sometimes called "the algorithm
of the gods". I'm not making that up. You might look it up on the web.

I'm not going to explain it, because I've studied it at a mathematical
level. I would only confuse things :).

>Genetic algorithms may work fine for some future enhancements to
>Mind.Forth, but right now only some very targetted coding is needed.

A deist, huh? ;-).

>Instead, Mind.Forth will have a *tendency* to think again the
>thoughts that it has perceived or thought in the past, but a
>wide variety of factors -- expressed as activation weights --
>will be able to wrest control away from the "time clouds" that
>correspond so unfortunately to a look-up table.

Why are lookup tables unfortunate?

>Such coding is not even a major project. I expect to approach
>it at my leisure in the near future. Then it will be fun and
>interesting to perform experiments on Mind.Forth in order to
>see what abilities Mind.Forth has in the areas of logic and
>reasoning -- after negation and conditionals have been coded.

What sort of interaction does Mind.Forth have? Do you type sentances in
and recieve sentances out (like a classical Turing test)? Or do you feed
it some representation of a sentance?

Could more advanced language parsing techniques improve it, or is it
itself a parsing technique?

--
-William "Billy" Tanksley

Arthur T. Murray

unread,
Jul 26, 1999, 3:00:00 AM7/26/99
to
William Tanksley wrote in comp.lang.forth on Sun, 25 Jul 1999:

> On 23 Jul 99 18:50:45 GMT, Arthur T. Murray wrote:
>> William Tanksley, wtan...@dolphin.openprojects.net,
>> wrote in the above newsgroups on Fri, 23 Jul 1999:

> Just a single newsgroup now -- I didn't notice the crosspost.
> Announcements are (or at least might be) okay crossposted,
> but discussion belongs in one NG at a time.

I agree, but on the Clintonesque hair-splitting premise that
my Usenet outcast ("banned in Melbourne"!) independent-scholar
status compels me to "announce" with any and all posts. [...]

>> Instead, Mind.Forth will have a *tendency* to think again the
>> thoughts that it has perceived or thought in the past, but a
>> wide variety of factors -- expressed as activation weights --

>> will able to wrest control away from the "time clouds" that


>> correspond so unfortunately to a look-up table.

> Why are lookup tables unfortunate?

Cyc -- c.m.i.I.w. -- is one stupendously enormous lookup table.
Mind.Forth http://www.geocities.com/Athens/Agora/7256/mind-fpc.html
is an actual thinking mind that can be taught facts and remember
them with robotlike accuracy, or like an automaton, or a machine
-- because Mind.Forth *is* a robot, automaton, machine. [...]

> What sort of interaction does Mind.Forth have? Do you type
> sentances in and recieve sentances out (like a classical Turing
> test)? Or do you feed it some representation of a sentance?

Into current Mind.Forth-28 you type three-word English sentences
in the strict format of subject + verb + direct object, such as
"Horses eat hay" or "Cats like fish." Mind.Forth responds with
a similar sentence, or, if you type in nothing at all, Mind.Forth
continues to follow its own internal priorities and produces
sentences as the reentrant results of its internal musings.

> Could more advanced language parsing techniques improve it,
> or is it itself a parsing technique?

Yes and yes. Mind.Forth as it now stands with Internet Web
Release #28 for either IBM clones or the Amiga, is the barest
possible implementation of mental operations (parsing, syntax,
pattern recognition, etc.) that cry out for improvement and
elaboration by people a lot smarter than v.t.y. Arthur here.

Because parsing, although important, was not the issue in
creating the AI, Mind.Forth simply fakes parsing with "par",
a variable that keeps track of the noun - verb - noun sequence
required when every input sentence is subject + verb + object.

Programmers and linguists who modify or improve Mind.Forth are
welcome to pack in the most advanced parsing they can muster.
And they don't even need to call the results "Mind.Forth" or
declare any sort of inheritance from the public-domain AI,
because it is a public-domain effort to introduce general AI.

> --
> -William "Billy" Tanksley

Reply all
Reply to author
Forward
0 new messages