Ai Forums Home Welcome Guest    Saturday, October 25, 2014
Ai Site > Ai Forums > The Artificial Intelligence Forum > Mentifex Theory of Mind Last PostsLoginRegisterWhy Register
Topic: Mentifex Theory of Mind

Pennywise
posted 7/20/2003  19:27Reply with quote
I'm starting this thread in hopes that Arthur will walk us through his AI documentation and so that we may come to understand his point of view of why and how he has "solved AI".

I believe that the homepage for his work in this respect is http://mentifex.virtualentity.com which contains voluminous work.

My question to Arthur is this: of all the work on this page, which is the paper to start with that describes the initial assumptions and the inital development of your theory of mind? If your response is "there is no single paper" could you perhaps write a summary for us here to get us started in understanding your work?


Rob Hoogers
posted 7/20/2003  22:35Send e-mail to userReply with quote
Mentifex has a distinct profile on more than one mad scientist afoot/debunking site:

Anyone giving a grant would do a google first, wouldn't they, these days? Works like a very effective blacklist, even. Which makes it quite a responsibility. You could ruin a career easily.

Although some people are quite capable of doing that totally unaided.

 Radio Brain Control - loony pages
Last edited by Rob Hoogers @ 7/20/2003 10:37:00 PM

Raphael
posted 7/20/2003  23:04Send e-mail to userReply with quote
If the purpose of this thread is to investigate and discuss Arthur Murray's AI work, let's see some responsible and intelligent discussion raised based on the review of his work that you've (assumedly) already done. Don't wait for Arthur himself to come answer your questions; he might, but he is certainly not obligated to do so whatsoever.

If the purpose of this thread is to quote "loony" lists to hold up ANY persons and their work for ridicule and scorn - you are SO in the wrong forum.


Rob Hoogers
posted 7/20/2003  23:32Send e-mail to userReply with quote
quote: "I go rather personally into such tangential narrations because this Nolarbeit is really a personal, unshared project (I can find no collaborators) and all this writing serves the personal purpose of capturing ideas in order somewhat laboriously to build a structure of ideas. I really rush to get certain things down. Ideally I might write while a tape recorder runs so that I can blurt things onto the tape before I forget the wording that came to me. This is a personal, hobby project, but I would like to achieve some results that would get someone else interested in what I'm doing. Then they might like to read my notebooks, but even so I am continuing to write what to myself is clearest and feels most comfortable, such as my slangy vocabulary and my concealing use of "we." Then let no one complain, because right now this material isn't written for a second party. I don't even have to separate the theoretical from the personal content, because later on I can index the content by date plus or minus paragraph. To indulge further, the beauty of it is that I am generating a mass of ideas in an utterly free process, with no constraints of deadline, money, or practical purpose, and it is not a sand-castle situation, but rather a real-life endeavor because everything is tied in with the idea that the proper study of man is man. Furthermore, it amazes me that I generate all this glib verbiage with slang, termini technici, and neologisms year after year."

Dated 9 AUG 1976.

Anyone must admire the tenacity of the man, if nothing else.

But 'bouleumatic accumulator', Arthur? If anything invites attacks, this type of neologistic black-boxing sure does the trick.

 The full text
Last edited by Rob Hoogers @ 7/20/2003 11:34:00 PM

Pennywise
posted 7/21/2003  05:56Reply with quote
 
Raphael wrote @ 7/20/2003 11:04:00 PM:
If the purpose of this thread is to quote "loony" lists to hold up ANY persons and their work for ridicule and scorn - you are SO in the wrong forum.

 
Good point. Given the current context of conversations and the replies this thread has already received, I guess it would be better to ask my questions by email.


Arthur T. Murray
[Guest]
posted 7/21/2003  06:14Reply with quote
No, wait, James, here is my reply to the earlier posts :-)

James Pennywise graciously posted 7/20/2003 19:27:

> I'm starting this thread in hopes that Arthur will walk us through
> his AI documentation and so that we may come to understand his
> point of view of why and how he has "solved AI".
>
> I believe that the homepage for his work in this respect is
> http://mentifex.virtualentity.com which contains voluminous work.
>
> My question to Arthur is this: of all the work on this page,
> which is the paper to start with that describes the initial
> assumptions and the inital development of your theory of mind?

http://mentifex.virtualentity.com/theory5.html -- also online as
http://mind.sourceforge.net/theory5.html -- "describes the initial
assumptions" and serves as a "popularizing" intoduction to the
theory of mind (TOM).

> If your response is "there is no single paper" could you perhaps
> write a summary for us here to get us started in understanding
> your work?

The theory5.html document, "Brain-Mind: Know Thyself!" (BMKT),
is the summary written and published in 1981 -- two years after
the development of the theory3.html breakthrough achieved in 1979.

Recently now in mid-2003 I have decided to call the whole package
the Concept-Fiber Theory of Mind so as to emphasize the idea of
neuronal fibers (or redundant gangs of fibers) holding a concept
by virtue of associative tags from the concept-fiber to sensory
engrams, to related concepts, and to a linguistic superstructure.

The theory3.html document is more important as the actual creative
instrument of developing the TOM through written gedankenexperiments,
but part three of the Nolarbeit Theory Journal will probably not
make much sense to any reader who has not yet read the BMKT text.

On 19 April 1972 I burned the first seven years of naive theorizing
and started over with the five documents theory1-5 now on-line.

Except for publishing the BMKT theory5 document in 1981, the
nineteen eighties were a wasted decade for my AI project because
I did not have Internet access and Tim Berners-Lee had not yet
created the World Wide Web. However, in 1986 I got into Amiga
computing and in July of 1993 I began coding Amiga Mind.REXX AI.

Rob Hoogers sagaciously posted 7/20/2003 22:35:

> Mentifex has a distinct profile on more than one mad scientist
> afoot/debunking site: [...]

The Radio Brain Control loony pages showed up quite soon after
the five Mentifex theory documents were first Web-published
around November of 1995. Several other sites over the years
tried to stigmatize Mentifex as a crank. It can't be helped.


Raphael so nicely posted 7/20/2003 23:04 that it needs no comment :-)


Rob Hoogers laboriously posted 7/20/2003 23:32:

> But 'bouleumatic accumulator', Arthur? If anything invites attacks,
> this type of neologistic black-boxing sure does the trick.

The Greek word "bouleuma" means "will" or volition, and at the time
(1973-1977) I thought that I was writing for myself alone, so I rushed
to give names to gedankenexperiment constructs that rapidly emerged.

TO SUM UP: It is awfully decent of all of you in this thread to
hear me out and to let me present my ideas and software as links.

IMHO the Mentifex body of work will get its just desserts in time,
but time for me personally is running out and so I try to accelerate
the process of having the Mentifex theory, software and book evaluated.

'Twas not I (was it Martin?) who came up with "AI has been solved"
as a thread title, but I love it -- it tacitly poses AI as a problem
that has been solved. It is a perfect formulation for claiming what
I have clumsily tried to claim for years: that here is a useful
theory of mind, software to implement the TOM, and the AI4U book.

As we all continue to post and discuss things, please simply
tolerate my claims of "AI has been solved" and let those who
investigate the Mentifex body of work come to their own conclusions.
I am not trying to swindle anyone, but rather to spread ideas.
The AI project grew out of my Panama Canal Zone boyhood at
Ft. Gulick CZ, where my eight-yeat-old self, usually out
exploring the jungle, found an olive-green Army battery
on the ground near the Bachelor Officers Quarters (BOQ)
and entered a lifetime quest to understand electronic devices.
You know where that sort of thing leads, don't you? To AI.


 AI Mind Project weblog at SourceForge open-source software site

Rob Hoogers
posted 7/21/2003  10:58Send e-mail to userReply with quote
 
Arthur T. Murray wrote @ 7/21/2003 6:14:00 AM:

Rob Hoogers laboriously posted 7/20/2003 23:32:

> But 'bouleumatic accumulator', Arthur? If anything invites attacks,
> this type of neologistic black-boxing sure does the trick.

The Greek word "bouleuma" means "will" or volition, and at the time
(1973-1977) I thought that I was writing for myself alone, so I rushed
to give names to gedankenexperiment constructs that rapidly emerged.



 
I was with you as far as the Greek, Arthur ;)

I've also read you've more or less ditched it, since it would be far from practical to even try to build one, for obvious reasons. It wouldn't work. Technonightmare.

It's what you did with the Will-processes afterwards, is what I object to: you've made it disappear into the woodwork, by saying it will happen all by itself for certain reasons/processes.

This I find a gross oversimplification, probably based on your wishful thinking, and an aversity/awe of parallel processes. What you describe cannot and will not work as advertised.

I'm afraid it's exactly there where your basic error lies. You've begun such a Herculean task though, that admission to such errors would threaten the whole shebang, which is nonsense, there's oodles of good stuff too. But I can imagine criticism becomes harder to handle when there is so much work involved. We're all human.

Last edited by Rob Hoogers @ 7/21/2003 10:59:00 AM

Arthur T. Murray
[Guest]
posted 7/21/2003  14:13Reply with quote
Rob Hoogers engagingly wrote:

> [...] I've also read you've more or less ditched it,

Pray tell, what is "it"?

> since it would be far from practical to even try to build one,

And what does "one" refer to :-)

> for obvious reasons. It wouldn't work. Technonightmare.

I need to know, what "It" do you refer to? Volition? The AI?

Please, everybody, try to use nouns at first, then please switch
to pronouns only after the intended references are quite clear.

> It's what you did with the Will-processes afterwards,
> is what I object to: you've made it disappear into the woodwork,
> by saying it will happen all by itself for certain reasons/processes.

It is perhaps a curio that the volition work, such as at

http://mind.sourceforge.net/theory2.html#1978mar17 -- NTJ2

(Nolarbeit Theory Journal, part two) was done before the NTJ3
breakthroughs establishing the design for a linguistic mindgrid.

Going back post-breakthrough over the work on volition, I felt
that the ideas on the will were already sound, for the time being.

I don't mind correcting errors, though; come, let us reason together.

> [...] I'm afraid it's exactly there where your basic error lies.
> You've begun such a Herculean task though, that admission to
> such errors would threaten the whole shebang, which is nonsense,
> there's oodles of good stuff too. But I can imagine criticism
> becomes harder to handle when there is so much work involved.
> We're all human.

Nope. "Bring 'em on." Criticisms are welcome. Let's build AI!

The active hyperlink below provides a lot of background data :-)

 Nanomagazine interview with Arthur T. Murray (24 October 2001)

Rob Hoogers
posted 7/21/2003  17:00Send e-mail to userReply with quote
The only word with Greek origin 'it' could refer to was your so-called 'bouleumatic accumulator'.

I'm still very interested in your answer. Your idea's on volition *have* become very unclear/unconvincing to me after you dropped *it*. ;)


Ribald
[Guest]
posted 7/21/2003  22:23Reply with quote
What evidence is their to support the claim that the software build qualifies as a "solution" to artificial intelligence?

I see that a system is being created that has interacting components, but what can it actually do in the real world sense?

If the current example is to simple or basic to be used in a real robot or application, what would it take to "scale up" to such degree that it would be useful?


Pennywise
posted 7/22/2003  05:00Reply with quote
Arthur, here are my initial impressions.

To start with, it took me a few minutes to understand the opening ASCII picture... mostly because it was ASCII. I strongly recommend that you download Gimp (gimp.org) and create a jpeg or something. Gopher disappeared 10 years ago for a reason, long live the world wide web. :)

I could go on with a list of nit-picking of individual phrases and concepts, but that would be distracting from the overall point the paper is trying to make.

I'll admit that I skimmed the paper due to time constraints, but it would seem that these are it's major points (please correct me if I'm wrong):
1) The mind works through information flows in the brain.
2) Information flows are based on different three dimentional axes, eg - the time dimentional axis.
3) These information flows lead to fibers which store memory.

My comments on the paper itself is that it is extremely well written and clear. It gets it's points across in a manner that anyone interested -- newbie to expert -- could understand. The material is a bit slow going for the experts among us, as the paper reviews basic scientific concepts, such as the senses of the human body. I find this to be a reasonable tradeoff so that newbies can understand the material, but some of the philosophy could cut out to make the paper more tolerable to the experts.
The definition of dimentionality is critical to the rest of the paper, and unfortunately redundant. Although, since the word isn't being used in any uncommon fashion and introduced in a intuitive way, it's not hard to overcome.
I'm left wondering if the paper is supposed to be a theoretical model of the brain, or a supposed explanation of the brain. If it's an explanation of the brain, it seems like it could use some findings from MRIs and such to back it up.

Just my two cents.


Arthur T. Murray
[Guest]
posted 7/22/2003  11:03Reply with quote
Ribald posted 7/21/2003 22:23

> What evidence is there to support the claim that the software build
> qualifies as a "solution" to artificial intelligence?

http://mentifex.virtualentity.com/jsaimind.html is a solution because
it is a thinking, functioning Mind that illustrates a Theory of Mind.
The Mind-1.1 AI is very primitive, but it does indeed think. It is
not a standard "chatterbot" that "fakes" the thinking in order to
carry on a pleasant and amusing conversation with the interlocutor.
The True AI carries on a maddeningly stupid conversation because
it knows so little: a few innate concepts, and what you tell it.

> I see that a system is being created that has interacting components,
> but what can it actually do in the real world sense?

Even before installation in a robot, an AI going just a little beyond
http://mind.sourceforge.net/jsaimind.html -- AI4U Mind-1.1 -- can be
built to embody a special knowledge base (KB) in any chosen domain.

A mentifex-class AI such as Mind.Forth or a Mind.XYZ derivative
can be installed in a robot as a mind aloof from its own motorium
at first, then awakening like a human baby to its motor options
as the quasi-parent human-in-charge "scales up" the motorium.

> If the current example is too simple or basic to be used in
> a real robot or application, what would it take to "scale up"
> to such a degree that it would be useful?

On the contrary, I (Arthur) am actually waiting for the various
commercially available "hobby" robots to "scale up" to the basic
http://mind.sourceforge.net/mind4th.html memory requirements :-)

The popular LEGO (R) Mindstorms robot, which can already run
http://www.hempeldesigngroup.com/lego/pbForth/index.html Forth,
will be an ideal platform for Mind.Forth as soon as a scaled-up
Mindstorms robot comes out with substantially more user-programmable
memory space than the current (ca. 2003) limit of about 28K RAM.

Give me a LEGO (R) Mindstorms that affords, say, 128 K of RAM
for the robot hobbyist to load in both an AI Mind and its data,
and I will hasten to release an initial AI native to the robot,
i.e. one which thousands of robot-loving tinkerers will soup up,
scale up and use for oneupsmanship at all the robotics meetings.

The AI4U Mind-1.1 currently on the Web is merely a curiosity,
a novelty, something worth of a desultory click-to-enliven.
As soon as the commercially available hobby robots catch up
just a little with the RAM memory and CPU width of a personal
computer (PC) or even a personal digital assistant (PDA),
then a basic Seed AI released into the robotics gene pool
will become an uncontrollable branch of AI Mind evolution.

Do NOT underestimate the unstoppable power of the amateur
roboticist to take a primitive but free artificial Mind and
to "Frankenstein" it into: Behold! The Singularity.

 The Art of Computer Mindmaking

Ribald
[Guest]
posted 7/22/2003  18:13Reply with quote
Why don't you take the time to build an example that is capable of showing proof of concept for a limited realm of knowledge? Just pick a subject, and demonstrate that it does more than act like a chatterbot with a more abstract sentence selection method...

Also, there are robot kits out there that provide a platform capable supporting a notebook, so I don't think the hardware capability is lacking as you describe.

I think you would get a lot farther by showing evidence of your solution, rather than just making grandiose claims about seeding the singularity, etc.


Arthur T. Murray
[Guest]
posted 7/22/2003  18:51Reply with quote
Ribald posted 7/22/2003 18:13

> Why don't you take the time to build an example that is capable
> of showing proof of concept for a limited realm of knowledge?

http://mentifex.virtualentity.com/jsaimind.html does exactly that,
with a bootstrap sequence that contains a limited set of ~35 concepts.

> Just pick a subject, and demonstrate that it does more than act
> like a chatterbot with a more abstract sentence selection method...

It is not an "abstract sentence selection method" but it
is rather an "abstract concept selection method." By using
concepts rather than parts of sentences, the AI Mind-1.1 thinks.

> Also, there are robot kits out there that provide a platform
> capable supporting a notebook, so I don't think the hardware
> capability is lacking as you describe.

It's not the physical strength that is lacking, or the ability
to hold a notebook computer. For rapid replication of the AI Mind
among amateur roboticists, we need mass-produced robots such as the
iRobot or the Aibo of the LEGO (R) Mindstorms, with plenty of
user-programmable memory space. Then we load in the AI Mind.

> I think you would get a lot farther by showing evidence of your
> solution, rather than just making grandiose claims about seeding
> the singularity, etc.

"The squeaking wheel gets the grease."

 AI has been solved (the original thread)

Ribald
[Guest]
posted 7/22/2003  20:44Reply with quote
 
Arthur T. Murray wrote @ 7/22/2003 6:51:00 PM:

> Just pick a subject, and demonstrate that it does more than act
> like a chatterbot with a more abstract sentence selection method...

[1] It is not an "abstract sentence selection method" but it
is rather an "abstract concept selection method." By using
concepts rather than parts of sentences, the AI Mind-1.1 thinks.

> Also, there are robot kits out there that provide a platform
> capable supporting a notebook, so I don't think the hardware
> capability is lacking as you describe.

[2] It's not the physical strength that is lacking, or the ability
to hold a notebook computer. For rapid replication of the AI Mind
among amateur roboticists, we need mass-produced robots such as the
iRobot or the Aibo of the LEGO (R) Mindstorms, with plenty of
user-programmable memory space. Then we load in the AI Mind.

> I think you would get a lot farther by showing evidence of your
> solution, rather than just making grandiose claims about seeding
> the singularity, etc.

[3] "The squeaking wheel gets the grease."

 
I thought you said: "On the contrary, I (Arthur) am actually waiting for the various
commercially available "hobby" robots to "scale up" to the basic
http://mind.sourceforge.net/mind4th.html memory requirements :-)"

Here are a few links that would certainly meet or exceed the requirements of your software, and are being mass produced commercially:

http://www.evolution.com/product/consumer/?source=google&term=robot+kit

http://www.acroname.com/robotics/info/PPRK/PPRK.html

http://www.acroname.com/robotics/parts/A11-WIRELESS-LAN.html

http://www.acroname.com/garcia/garcia.html

http://www1.tpgi.com.au/users/p8king/releas12.htm

[1] There are other symbolic logic processing systems out there. What I am trying to suggest is that if your idea is capable of doing more than your self-admittedly limited demonstration, then why don't you make a better demo?

[2] I already gave a bunch of links that address this, including PDA driven robot kits that are currently available. There are plenty of wireless solutios that would allow the "brain" to be in your desktop PC while the robot roamed around. Wouldn't this present an opportunity for you to prove that your software is a meaningful contribution?

[3] What is more likely to happen is that people will "throw the baby out with the bath water" if they feel you are erroneously promoting yourself as the person that "solved" AI. I think that your valid contributions will then be lost in people's reaction to your over-stated claims.

You have produced a large volume of work, and I think it contains value, so I think it would be a shame if your desire for recognition lead to just the opposite.


Pennywise
posted 7/23/2003  05:20Reply with quote
 
Ribald wrote @ 7/22/2003 8:44:00 PM:
I think that your valid contributions will then be lost in people's reaction to your over-stated claims.


 
I'll drink to that.


Thomas Gail Haws
posted 7/28/2003  20:43Send e-mail to userReply with quote
The Mind demo, or any other demonstration of the Mentifex Theory of Mind, doesn't need to come pre-loaded with a knowledge base. But it does need the ability to save its accumulated knowledge for future tutoring sessions. One way to accomplish this would be to present the current JavaScript Mind demonstration along with a download link for an executable Mind demonstration to be run on a user's local computer.

The Hal project (at this site) can be inspected by a newbie like me because its advanced tutorial states are open to public inspection. Hal's strengths and deficiencies are readily apparent without inspecting the code or the theory. Hal's full maturity is open to the public. Mind's is not. This is a handicap, since the proof is in the pudding.

If Arthur is to have the impact on AI I would hope he wants to have, he needs to find a way to demonstrate the Mentifex TOM not only in its infancy, but also in varying levels of maturity. I find myself hesitating to give Mind a serious test because I fear my computer may lock up at any moment.

Arthur, what is the chance of offering a downloadable executable or a demo like Hal that could demonstrate long-term maturation of Mind?


Arthur
[Guest]
posted 7/29/2003  00:44Reply with quote
Both the code and the theory of the Mind-1.1 software are totally open for inspection.

True, the AI Mind is not yet mature, but it is already extremely sophisticated and intricate -- in order to hew closely to the "Fiber-Concept Theory of Mind."

It is hard to bring the AI Mind to further maturity, because it is not a mattera language of enhancing just a module or two. Instead, quite a few modules have to be enhanced more or less simultaneously.

The Sun.27.OCT.2002 JavaScript version is probably at the limits of its viabilty. If I ever get back on my feet financially, I will either code the AI further in Forth, or in a language new to me, such as Perl or Java.

Today on Usenet (q.v.) tI have begun a major initiative of asking people to code just the Main Program Loop in their favorite programming language and to put thaet ostensibly simple code up on a Web site for others to copy and improve upon.

The idea is that any initial implementation will encourage people to participate. It is good if some coders maintain a steady corpus of Mind.XYZ code, while others "fork" the code in order to try new ideas.

I am sorry and I apologize for working so slowly.

 DIY AI Steps -- please code and Web-host the first Main Loop

Thomas Gail Haws
posted 7/29/2003  23:40Send e-mail to userReply with quote
Well, unlike Artificial Minds, we are alive and have other missions to perform, such as loving our children and feeding them and ourselves. I am the one who should apologize for pushing a little too hard.

The body of your work is ample. It shows long dedication and ingenuity. The very fact that you are willing to entertain the thoughts of a layman like me is most generous.


Thomas Gail Haws
posted 7/31/2003  20:07Send e-mail to userReply with quote
I admit I am a layman (not even a newbie), but I would like to make a performance based evaluation of the Mind 1.1 and the Mentifex Theory of Mind. I spent a little more time with Mind 1.1 and I have the following feedback:

1) http://mind.sourceforge.net/jsaimind.html does not respond very well--at least not on my computer. http://www.aibokennelclub.org/mind.html works much better. You might want to investigate the difference.

2) I don't think it would require much effort on your part to post some exemplary transcripts of Mind 1.1 sessions. Such postings would serve two purposes: a) They could serve as tutorials for Mind evaluators and b) They could serve as case studies on the performance of Mind 1.1. Would you be able to do that?

3) I am not getting very far trying to tutor Mind 1.1. I would like to forward a transcript to you for troubleshooting. Would that be possible?

4) You may not be appreciating how difficult it is to evaluate Mind as a working solution. Yes, the code is open, and yes, the theory is documented. But if nobody knows yet what the solution to AI is, we need to be able to evaluate your working model.

Please let me know if I can send you a transcript for troubleshooting.

Tom

  1  2  3  
'Send Send email to user    Reply with quote Reply with quote    Edit message Edit message

Forums Home    Hal and other child machines    Alan and other chatbots    Language Mind and Consciousness  
Contact Us Terms of Use