Doing it through AGI verses AI is a challenge.



Sure, if your algorithm does not know of English then it would be more difficult to extract the word “Google” but given multiple images it could retrieve the graphically probabilistic pattern “Google” and perhaps hieroglyphically and individually extract the letters… maybe, if it knew it was searching for symbols…



Still though – narrow AI – I think, someone correct me if I am wrong. All this can be done now with narrow AI, though better and more applicable for us with AGI.



John



From: Mike Tintner [mailto:tintner@blueyonder.co.uk]





What word "GOOGLE"?



You're v. confused, I suggest, here.



You seem to be dealing in Platonic mystical "ideas" here - some magic essence of "GOOGLE."



The reality is your base schema above consists of



G - O - O - G - L - E.



Those precise letters/forms. (And you have to have some concrete formal schema for object recognition to proceed).



Now how do you relate those specific forms/letters to the forms of the logos below? Your "G" say to a Dr Seuss figure? And then an infinity of figuratively related human and animal forms? Because there's no math. operation to do it, and there's no algorithm to do it.



Except perhaps in your imagination. I repeat, this is the challenge of AGI. Instead of trying to brush it away, why not actually engage with it?





From: John G. Rose <mailto:johnrose@polyplexic.com&gt;

Sent: Tuesday, September 07, 2010 7:01 PM

To: AGI <mailto:agi@listbox.com&gt;

Subject: RE: [agi] And now the quintessential AGI font



The pattern is the word “GOOGLE”. It doesn’t matter if you throw mud on the wall to spell it or construct it out of pig intestines. The algorithm is narrow AI based.



Those narrow AI programs are getting better and better to the point where you have to choose multiple CAPTCHAs until you actually can recognize one since the bots can recognize more than a human at times. So new CAPTCHAs are being developed often that are not OCR related.



John



From: Mike Tintner [mailto:tintner@blueyonder.co.uk]



Nope



Your examples are useful {incl clock} - they show the **limits** of patterns - wh. can indeed be fluidly morphed and stretched and still remain patterns..



But you are always morphing the *same parts* arranged in the *same structures*. - basically homogeneous, homeomorphic forms - wh. are essential for any set of patterns



The Google logos show the difference -



you are recognizing



*different diverse parts* arranged in **different diverse structures**.- wh. can in fact be endlessly diverse. The letters of the Google logo can in fact be almost any form in the universe from letters to snakes to naked Adam's. If your clock's letters suddenly changed into snakes or naked Adams, you would cease to be dealing in patterns.



If you cannot see that, you have a truly major visual/conceptual blindness.



And the challenge is to understand how the mind actually processes and compares these diverse forms - because there clearly are underlying principles to how the mind achieves this, but they ain't the principles of math. patterns.



From: John G. Rose <mailto:johnrose@polyplexic.com&gt;

Sent: Tuesday, September 07, 2010 6:09 PM

To: AGI <mailto:agi@listbox.com&gt;

Subject: RE: [agi] And now the quintessential AGI font



Isn’t this just pattern recognition? Narrow AI?



Like CAPTCHA is use to deter bots:



http://en.wikipedia.org/wiki/CAPTCHA



John



From: Mike Tintner [mailto:tintner@blueyonder.co.uk]
Sent: Tuesday, September 07, 2010 10:57 AM
To: AGI
Subject: Re: [agi] And now the quintessential AGI font



John: <http://www.physorg.com/news203011174.html&gt; nothing to do with AGI... really.



Cor blimey, you don't know to look. You're looking at that logo and seeing ONE logo, right?



Wrong.



You have to see an extensive, and in principle endless series of alternatives, and ask: how does the mind recognize them ALL? That is the essence of AGI and of course visual perception, But nothing in maths can handle it. Sorry, John. Art & AGI are not maths. Hope you can see the difference.


Description: Description: Description: cid:6F0F3D448963485D9B5F1A06EB6D2462@MikePC






Description: Description: Description: cid:27FD4421261D4D688A07F0A57419B59E@MikePC






Description: Description: Description: cid:DC70104C2F6446C9A7640DCCB395407A@MikePC






Description: Description: Description: cid:5BEF1645863A4155A1CC9DFFF9BA7C58@MikePC






Description: Description: Description: cid:EEF64A755878467BA6616042666A5B23@MikePC






Description: Description: Description: cid:0BB33CDA98BE4E46A6A8C241A6205A8F@MikePC






Description: Description: Description: cid:3FCA4AF59E71452799F61D0917BBF4A0@MikePC






Description: Description: Description: cid:5F1D1457005D44C0A99EC698208E3C7B@MikePC






Description: Description: Description: cid:ED0FD1B7CF54450680C8AC2459679F9D@MikePC






Description: Description: Description: cid:58C4D40E61944F9FA4916BD96A860182@MikePC






Description: Description: Description: cid:D6F4E75F8F1C436990212F3DC0AFBB47@MikePC






Description: Description: Description: cid:EE89465626FE4FE6AA1F58B175B4F345@MikePC






Description: Description: Description: cid:CCC2E2F7CB844A2F97BA73184572F67F@MikePC






Description: Description: Description: cid:BA1107351AEB41558E416A0CE5841BBF@MikePC






AGI | <https://www.listbox.com/member/archive/303/=now&gt; Archives <https://www.listbox.com/member/archive/rss/303/&gt; Description: Description: Description: Image removed by sender.| <https://www.listbox.com/member/?&&gt; Modify Your Subscription

<http://www.listbox.com/&gt; Description: Description: Description: Image removed by sender.




AGI | <https://www.listbox.com/member/archive/303/=now&gt; Archives <https://www.listbox.com/member/archive/rss/303/&gt; Description: Description: Image removed by sender.| <https://www.listbox.com/member/?&&gt; Modify Your Subscription

<http://www.listbox.com/&gt; Description: Description: Image removed by sender.


AGI | <https://www.listbox.com/member/archive/303/=now&gt; Archives <https://www.listbox.com/member/archive/rss/303/&gt; Description: Description: Image removed by sender.| <https://www.listbox.com/member/?&&gt; Modify Your Subscription

<http://www.listbox.com/&gt; Description: Description: Image removed by sender.




AGI | <https://www.listbox.com/member/archive/303/=now&gt; Archives <https://www.listbox.com/member/archive/rss/303/&gt; Description: Image removed by sender.| <https://www.listbox.com/member/?&&gt; Modify Your Subscription

<http://www.listbox.com/&gt; Description: Image removed by sender.


AGI | <https://www.listbox.com/member/archive/303/=now&gt; Archives <https://www.listbox.com/member/archive/rss/303/248029-d0d678e2&gt; Description: Image removed by sender.| <https://www.listbox.com/member/?&&gt; Modify Your Subscription

<http://www.listbox.com/&gt; Description: Image removed by sender.

Go back to my original fonts problem - wh. constitutes excellent, hard evidence.

"Visually object recognizing" those v. different letter A's, is essentially the same as conceptualising "the letter A" is the same as "creating"/designing a new font A". The only differences are that in recognition, the objects are presented to you, whereas in conceptualisation, examples of A fonts may have to be sought from memory, whereas in creation, a new font has to be constructed and/or sought from the world of shapes at large. The basic operation of matching diverse forms is the same in all,

And this is no fundamentally different from understanding or creating a metaphor like "mushroom cloud", for visually mapping the distinct forms of a mushroom and an A-bomb cloud together is much the same as visually mapping the distinct forms of diverse A fonts, or 'O' fonts. They are more or less the same graphical distances apart.

And all this - although the following is a much greater leap - is much the same problem as finding a freeform alternative to a program/algorithm - for if you are to map two or more very diverse forms like mushroom and cloud, or A fonts, onto each other, it must be a freeform operation.You can, and must be able to begin, continue and end anywhere, at any points or sectors of these forms, in your mapping. Algorithms are setform [not freeform] in structure, and tell you where to begin, continue and end - something that is impossible if you are to have a single basic operation that can map, say, a G onto a Dr Seuss, or a naked coiled woman, or a twisting snake, or a hedgerow or any other of an infinity of such diverse forms..

So what I've done here is to show that the operations of

1.visually recognizing
2.conceptualising
3.creating a new form of... any given class of objects

and...

4.metaphorically comparing.,.. one class of objects with a diverse class of objects

all depend on the same basic operation of matching/mapping diverse forms together.

And since that operation is necessarily freeform (as explained above) it provides a loose model for a freeform computational alternative to algorithms, just as patterns provide a loose model for algorithms. [This last proposition is obviously much more controversial and v. loosely outlined - but it will hold].

Try and find the umbilical connection of visual object recognition, conceptualisation, creativity and metaphor elsewhere.




From: David Jones
Sent: Tuesday, September 07, 2010 8:15 PM
To: AGI
Subject: Re: [agi] And now the quintessential AGI font


This assumption is neither supported by you through any evidence, nor is it true in my opinion.

There is a TON of evidence that this assumption is false. In fact, I believe that many AGI researchers disagree with me on this point, but my opinion is that general intelligence is not all based on a general learning method. The generality requires a combination of very specialized functions, such as for sensory perception as a foundation. Even if there is a part that is general, I'm not sure which part you could j

See http://en.wikipedia.org/wiki/Evolutionary_psychology The field of Evolutionary Psychology is a theory that I subscribe to that suggests that the brain is not just one big pattern analysis machine. I also have a very explanatory theory about how such specialization is accomplished and self organized within the brain. It is achieved through cell specialization, which occurs in every cell in the body. I theorize that the same sort of specialization occurs within the brain, but in a unique way. It may also specializes based on input patterns from various other specialized cells.

So, no. You can't just say that one problem represents the whole AGI problem. The evidence says otherwise.

Dave


On Tue, Sep 7, 2010 at 3:01 PM, Mike Tintner <tintner@blueyonder.co.uk> wrote:

Dave:It is worth analyzing as a significant principle/part of AGI, but it is not THEE only core of AGI or the bottleneck to understanding AGI

You're missing a v. important point I made earlier - wh. perhaps I didn't make clearly enough. The point is: it doesn't really matter where or with what particular kind of problem you start - because, in the final analysis, all (or at any rate most of) the problems of AGI are interchangeable.

Visual object recognition is conceptualisation is creativity [deliberate creation] is analogy is metaphor is language understanding is navigating or manipulating a real world environment. They all centre on patching diverse objects together that don't naturally, formally, mathematically fit together. G's and Dr Seusses. Chalk and cheese.

Ben and most of you have a point when you say everything can be reduced to patterns. A pattern is a structure is a journey is a program/algorithm is a way of comparing things is a way of combining things. The trouble is a pattern is everything *narrow AI* and the opposite of everything AGI. Patterns are all about fitting objects together that belong to the same class and do naturally, formally, mathematically fit together. Bricks and bricks. Dogs and bones.

Once you have an overview of how all the problems connect, the kind of visual examples I'm giving are probably the easiest places to start. No doubt most AGI-ers don't make the overall connection but just see them as isolated problems of "design". They're not. They're more or less universal.


From: David Jones
Sent: Tuesday, September 07, 2010 6:16 PM
To: AGI
Subject: Re: [agi] And now the quintessential AGI font


What I was thinking is that the software used to generate the font has nothing to do with AGI... clearly someone is designing the font and that someone is of general intelligence. which is why I agreed that it had nothing to do with AGI.

If you mean recognition of the font... ok, that would be AGI, but not without knowledge, experience and contextual analysis. We have a lot of experience that goes into this such recognition. For example, if you go to the google home page where the regular font used to be, you clearly expect the image there to be of a google font, even if it's drawn using other forms. The letter's method of recognition is based on shape, size, curves and other such elements. In the brain, these are likely processed in parallel. So, although you recognize the lady bug below, you also expect the letters that makeup the letters in the name "google". And when you see something that doesn't make sense in an image... a coincidence that is just too coincidental, then it is natural for us to look for a match and explanation. Thus we are able to see the font, even though it is drawn using an odd symbol/form.

It is worth analyzing as a significant principle/part of AGI, but it is not THEE only core of AGI or the bottleneck to understanding AGI. There is a hell of a lot of other things going on. Nonetheless, it should be studied.

Dave


On Tue, Sep 7, 2010 at 12:56 PM, Mike Tintner <tintner@blueyonder.co.uk> wrote:

John: nothing to do with AGI... really.

Cor blimey, you don't know to look. You're looking at that logo and seeing ONE logo, right?

Wrong.

You have to see an extensive, and in principle endless series of alternatives, and ask: how does the mind recognize them ALL? That is the essence of AGI and of course visual perception, But nothing in maths can handle it. Sorry, John. Art & AGI are not maths. Hope you can see the difference.





























AGI | Archives | Modify Your Subscription



AGI | Archives | Modify Your Subscription

AGI | Archives | Modify Your Subscription



AGI | Archives | Modify Your Subscription

This assumption is neither supported by you through any evidence, nor is it
true in my opinion.

There is a TON of evidence that this assumption is false. In fact, I believe
that many AGI researchers disagree with me on this point, but my opinion is
that general intelligence is not all based on a general learning method. The
generality requires a combination of very specialized functions, such as for
sensory perception as a foundation. Even if there is a part that is general,
I'm not sure which part you could j

See http://en.wikipedia.org/wiki/Evolutionary_psychology The field of
Evolutionary Psychology is a theory that I subscribe to that suggests that
the brain is not just one big pattern analysis machine. I also have a very
explanatory theory about how such specialization is accomplished and self
organized within the brain. It is achieved through cell specialization,
which occurs in every cell in the body. I theorize that the same sort of
specialization occurs within the brain, but in a unique way. It may also
specializes based on input patterns from various other specialized cells.

So, no. You can't just say that one problem represents the whole AGI
problem. The evidence says otherwise.

Dave

On Tue, Sep 7, 2010 at 3:01 PM, Mike Tintner <tintner@blueyonder.co.uk>wrote:

> Dave:It is worth analyzing as a significant principle/part of AGI, but it
> is not THEE only core of AGI or the bottleneck to understanding AGI
>
> You're missing a v. important point I made earlier - wh. perhaps I didn't
> make clearly enough. The point is: it doesn't really matter where or with
> what particular kind of problem you start - because, in the final analysis,
> all (or at any rate most of) the problems of AGI are interchangeable.
>
> Visual object recognition is conceptualisation is creativity [deliberate
> creation] is analogy is metaphor is language understanding is navigating or
> manipulating a real world environment. They all centre on patching diverse
> objects together that don't naturally, formally, mathematically fit
> together. G's and Dr Seusses. Chalk and cheese.
>
> Ben and most of you have a point when you say everything can be reduced to
> patterns. A pattern is a structure is a journey is a program/algorithm is a
> way of comparing things is a way of combining things. The trouble is a
> pattern is everything *narrow AI* and the opposite of everything AGI.
> Patterns are all about fitting objects together that belong to the same
> class and do naturally, formally, mathematically fit together. Bricks and
> bricks. Dogs and bones.
>
> Once you have an overview of how all the problems connect, the kind of
> visual examples I'm giving are probably the easiest places to start. No
> doubt most AGI-ers don't make the overall connection but just see them as
> isolated problems of "design". They're not. They're more or less universal.
>
>
> *From:* David Jones <davidhere40@gmail.com>
> *Sent:* Tuesday, September 07, 2010 6:16 PM
> *To:* AGI <agi@listbox.com>
> *Subject:* Re: [agi] And now the quintessential AGI font
>
> What I was thinking is that the software used to generate the font has
> nothing to do with AGI... clearly someone is designing the font and that
> someone is of general intelligence. which is why I agreed that it had
> nothing to do with AGI.
>
> If you mean recognition of the font... ok, that would be AGI, but not
> without knowledge, experience and contextual analysis. We have a lot of
> experience that goes into this such recognition. For example, if you go to
> the google home page where the regular font used to be, you clearly expect
> the image there to be of a google font, even if it's drawn using other
> forms. The letter's method of recognition is based on shape, size, curves
> and other such elements. In the brain, these are likely processed in
> parallel. So, although you recognize the lady bug below, you also expect the
> letters that makeup the letters in the name "google". And when you see
> something that doesn't make sense in an image... a coincidence that is just
> too coincidental, then it is natural for us to look for a match and
> explanation. Thus we are able to see the font, even though it is drawn using
> an odd symbol/form.
>
> It is worth analyzing as a significant principle/part of AGI, but it is not
> THEE only core of AGI or the bottleneck to understanding AGI. There is a
> hell of a lot of other things going on. Nonetheless, it should be studied.
>
> Dave
>
> On Tue, Sep 7, 2010 at 12:56 PM, Mike Tintner <tintner@blueyonder.co.uk>wrote:
>
>> John: nothing <http://www.physorg.com/news203011174.html&gt; to do with
>> AGI... really.
>>
>> Cor blimey, you don't know to look. You're looking at that logo and seeing
>> ONE logo, right?
>>
>> Wrong.
>>
>> You have to see an extensive, and in principle endless series of
>> alternatives, and ask: how does the mind recognize them ALL? That is the
>> essence of AGI and of course visual perception, But nothing in maths can
>> handle it. Sorry, John. Art & AGI are not maths. Hope you can see the
>> difference.
>>
>> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
>> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
>> <http://www.listbox.com&gt;
>>
>
> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
> <http://www.listbox.com&gt;
> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
> <https://www.listbox.com/member/archive/rss/303/16338999-6d5f576b&gt; |
> Modify<https://www.listbox.com/member/?&>Your Subscription
> <http://www.listbox.com&gt;
>

Dave:It is worth analyzing as a significant principle/part of AGI, but it is not THEE only core of AGI or the bottleneck to understanding AGI

You're missing a v. important point I made earlier - wh. perhaps I didn't make clearly enough. The point is: it doesn't really matter where or with what particular kind of problem you start - because, in the final analysis, all (or at any rate most of) the problems of AGI are interchangeable.

Visual object recognition is conceptualisation is creativity [deliberate creation] is analogy is metaphor is language understanding is navigating or manipulating a real world environment. They all centre on patching diverse objects together that don't naturally, formally, mathematically fit together. G's and Dr Seusses. Chalk and cheese.

Ben and most of you have a point when you say everything can be reduced to patterns. A pattern is a structure is a journey is a program/algorithm is a way of comparing things is a way of combining things. The trouble is a pattern is everything *narrow AI* and the opposite of everything AGI. Patterns are all about fitting objects together that belong to the same class and do naturally, formally, mathematically fit together. Bricks and bricks. Dogs and bones.

Once you have an overview of how all the problems connect, the kind of visual examples I'm giving are probably the easiest places to start. No doubt most AGI-ers don't make the overall connection but just see them as isolated problems of "design". They're not. They're more or less universal.


From: David Jones
Sent: Tuesday, September 07, 2010 6:16 PM
To: AGI
Subject: Re: [agi] And now the quintessential AGI font


What I was thinking is that the software used to generate the font has nothing to do with AGI... clearly someone is designing the font and that someone is of general intelligence. which is why I agreed that it had nothing to do with AGI.

If you mean recognition of the font... ok, that would be AGI, but not without knowledge, experience and contextual analysis. We have a lot of experience that goes into this such recognition. For example, if you go to the google home page where the regular font used to be, you clearly expect the image there to be of a google font, even if it's drawn using other forms. The letter's method of recognition is based on shape, size, curves and other such elements. In the brain, these are likely processed in parallel. So, although you recognize the lady bug below, you also expect the letters that makeup the letters in the name "google". And when you see something that doesn't make sense in an image... a coincidence that is just too coincidental, then it is natural for us to look for a match and explanation. Thus we are able to see the font, even though it is drawn using an odd symbol/form.

It is worth analyzing as a significant principle/part of AGI, but it is not THEE only core of AGI or the bottleneck to understanding AGI. There is a hell of a lot of other things going on. Nonetheless, it should be studied.

Dave


On Tue, Sep 7, 2010 at 12:56 PM, Mike Tintner <tintner@blueyonder.co.uk> wrote:

John: nothing to do with AGI... really.

Cor blimey, you don't know to look. You're looking at that logo and seeing ONE logo, right?

Wrong.

You have to see an extensive, and in principle endless series of alternatives, and ask: how does the mind recognize them ALL? That is the essence of AGI and of course visual perception, But nothing in maths can handle it. Sorry, John. Art & AGI are not maths. Hope you can see the difference.





























AGI | Archives | Modify Your Subscription



AGI | Archives | Modify Your Subscription

What word "GOOGLE"?

You're v. confused, I suggest, here.

You seem to be dealing in Platonic mystical "ideas" here - some magic essence of "GOOGLE."

The reality is your base schema above consists of

G - O - O - G - L - E.

Those precise letters/forms. (And you have to have some concrete formal schema for object recognition to proceed).

Now how do you relate those specific forms/letters to the forms of the logos below? Your "G" say to a Dr Seuss figure? And then an infinity of figuratively related human and animal forms? Because there's no math. operation to do it, and there's no algorithm to do it.

Except perhaps in your imagination. I repeat, this is the challenge of AGI. Instead of trying to brush it away, why not actually engage with it?


From: John G. Rose
Sent: Tuesday, September 07, 2010 7:01 PM
To: AGI
Subject: RE: [agi] And now the quintessential AGI font


The pattern is the word “GOOGLE”. It doesn’t matter if you throw mud on the wall to spell it or construct it out of pig intestines. The algorithm is narrow AI based.



Those narrow AI programs are getting better and better to the point where you have to choose multiple CAPTCHAs until you actually can recognize one since the bots can recognize more than a human at times. So new CAPTCHAs are being developed often that are not OCR related.



John



From: Mike Tintner [mailto:tintner@blueyonder.co.uk]





Nope



Your examples are useful {incl clock} - they show the **limits** of patterns - wh. can indeed be fluidly morphed and stretched and still remain patterns..



But you are always morphing the *same parts* arranged in the *same structures*. - basically homogeneous, homeomorphic forms - wh. are essential for any set of patterns



The Google logos show the difference -



you are recognizing



*different diverse parts* arranged in **different diverse structures**.- wh. can in fact be endlessly diverse. The letters of the Google logo can in fact be almost any form in the universe from letters to snakes to naked Adam's. If your clock's letters suddenly changed into snakes or naked Adams, you would cease to be dealing in patterns.



If you cannot see that, you have a truly major visual/conceptual blindness.



And the challenge is to understand how the mind actually processes and compares these diverse forms - because there clearly are underlying principles to how the mind achieves this, but they ain't the principles of math. patterns.



From: John G. Rose

Sent: Tuesday, September 07, 2010 6:09 PM

To: AGI

Subject: RE: [agi] And now the quintessential AGI font



Isn’t this just pattern recognition? Narrow AI?



Like CAPTCHA is use to deter bots:



http://en.wikipedia.org/wiki/CAPTCHA



John



From: Mike Tintner [mailto:tintner@blueyonder.co.uk]
Sent: Tuesday, September 07, 2010 10:57 AM
To: AGI
Subject: Re: [agi] And now the quintessential AGI font



John: nothing to do with AGI... really.



Cor blimey, you don't know to look. You're looking at that logo and seeing ONE logo, right?



Wrong.



You have to see an extensive, and in principle endless series of alternatives, and ask: how does the mind recognize them ALL? That is the essence of AGI and of course visual perception, But nothing in maths can handle it. Sorry, John. Art & AGI are not maths. Hope you can see the difference.





















































































AGI | Archives | Modify Your Subscription





AGI | Archives | Modify Your Subscription



AGI | Archives | Modify Your Subscription





AGI | Archives | Modify Your Subscription

Mike, btw, the brain handles this stuff using top down processing in my
opinion. That is... it suggests, from experience or knowledge what to look
for and then matches it to what it sees, even if the matches are very loose
and sketchy. How a computer would do this is an interesting thing to study.
But, like I said, you can't just say that this is *thee* most important
problem to analyze when designing AGI. that is unjustified... you need a bit
more support to claim that we should spend so much time analyzing this.


On Tue, Sep 7, 2010 at 1:16 PM, David Jones <davidhere40@gmail.com> wrote:

> What I was thinking is that the software used to generate the font has
> nothing to do with AGI... clearly someone is designing the font and that
> someone is of general intelligence. which is why I agreed that it had
> nothing to do with AGI.
>
> If you mean recognition of the font... ok, that would be AGI, but not
> without knowledge, experience and contextual analysis. We have a lot of
> experience that goes into this such recognition. For example, if you go to
> the google home page where the regular font used to be, you clearly expect
> the image there to be of a google font, even if it's drawn using other
> forms. The letter's method of recognition is based on shape, size, curves
> and other such elements. In the brain, these are likely processed in
> parallel. So, although you recognize the lady bug below, you also expect the
> letters that makeup the letters in the name "google". And when you see
> something that doesn't make sense in an image... a coincidence that is just
> too coincidental, then it is natural for us to look for a match and
> explanation. Thus we are able to see the font, even though it is drawn using
> an odd symbol/form.
>
> It is worth analyzing as a significant principle/part of AGI, but it is not
> THEE only core of AGI or the bottleneck to understanding AGI. There is a
> hell of a lot of other things going on. Nonetheless, it should be studied.
>
> Dave
>
>
> On Tue, Sep 7, 2010 at 12:56 PM, Mike Tintner <tintner@blueyonder.co.uk>wrote:
>
>> John: nothing <http://www.physorg.com/news203011174.html&gt; to do with
>> AGI... really.
>>
>> Cor blimey, you don't know to look. You're looking at that logo and seeing
>> ONE logo, right?
>>
>> Wrong.
>>
>> You have to see an extensive, and in principle endless series of
>> alternatives, and ask: how does the mind recognize them ALL? That is the
>> essence of AGI and of course visual perception, But nothing in maths can
>> handle it. Sorry, John. Art & AGI are not maths. Hope you can see the
>> difference.
>>
>> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
>> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
>> <http://www.listbox.com&gt;
>>
>
>

John,

The various logos incl. the Think/animated particles illustrate the most basic principle of an AGI mind, which is "cogitation". The human mind literally "tosses ideas (/forms) around" - and can recognize ideas/forms that have been tossed around. It cogitates, wh. literally means co-agitare - to shake ideas up/together - and also to mix in new things.

You can take a Mandelbrot fractal and squeeze it and variously morph it - and you're still in the realms of geometry and maths. You can't toss and shake the parts of the fractal around, let alone cut and paste new parts - it ceases to be a fractal/pattern. But the new form (provided the shaking isn't too severe) will remain an AGI form/patchwork that the mind can recognize as belonging with the original Mandelbrot fractal.

Now the challenge of AGI is to recognize how that's done - how the brain can visually and conceptually toss ideas/forms around - and still recognize their commonality.

All this is fundamental to engaging with the real world, which itself is continually tossing things around - we continually have to navigate territories, like streets and rooms, where the parts are tossed around and changed from one day to the next.

The pattern is the word “GOOGLE”. It doesn’t matter if you throw mud on the wall to spell it or construct it out of pig intestines. The algorithm is narrow AI based.



Those narrow AI programs are getting better and better to the point where you have to choose multiple CAPTCHAs until you actually can recognize one since the bots can recognize more than a human at times. So new CAPTCHAs are being developed often that are not OCR related.



John



From: Mike Tintner [mailto:tintner@blueyonder.co.uk]





Nope



Your examples are useful {incl clock} - they show the **limits** of patterns - wh. can indeed be fluidly morphed and stretched and still remain patterns..



But you are always morphing the *same parts* arranged in the *same structures*. - basically homogeneous, homeomorphic forms - wh. are essential for any set of patterns



The Google logos show the difference -



you are recognizing



*different diverse parts* arranged in **different diverse structures**.- wh. can in fact be endlessly diverse. The letters of the Google logo can in fact be almost any form in the universe from letters to snakes to naked Adam's. If your clock's letters suddenly changed into snakes or naked Adams, you would cease to be dealing in patterns.



If you cannot see that, you have a truly major visual/conceptual blindness.



And the challenge is to understand how the mind actually processes and compares these diverse forms - because there clearly are underlying principles to how the mind achieves this, but they ain't the principles of math. patterns.



From: John G. Rose <mailto:johnrose@polyplexic.com&gt;

Sent: Tuesday, September 07, 2010 6:09 PM

To: AGI <mailto:agi@listbox.com&gt;

Subject: RE: [agi] And now the quintessential AGI font



Isn’t this just pattern recognition? Narrow AI?



Like CAPTCHA is use to deter bots:



http://en.wikipedia.org/wiki/CAPTCHA



John



From: Mike Tintner [mailto:tintner@blueyonder.co.uk]
Sent: Tuesday, September 07, 2010 10:57 AM
To: AGI
Subject: Re: [agi] And now the quintessential AGI font



John: <http://www.physorg.com/news203011174.html&gt; nothing to do with AGI... really.



Cor blimey, you don't know to look. You're looking at that logo and seeing ONE logo, right?



Wrong.



You have to see an extensive, and in principle endless series of alternatives, and ask: how does the mind recognize them ALL? That is the essence of AGI and of course visual perception, But nothing in maths can handle it. Sorry, John. Art & AGI are not maths. Hope you can see the difference.


Description: Description: cid:6F0F3D448963485D9B5F1A06EB6D2462@MikePC






Description: Description: cid:27FD4421261D4D688A07F0A57419B59E@MikePC






Description: Description: cid:DC70104C2F6446C9A7640DCCB395407A@MikePC






Description: Description: cid:5BEF1645863A4155A1CC9DFFF9BA7C58@MikePC






Description: Description: cid:EEF64A755878467BA6616042666A5B23@MikePC






Description: Description: cid:0BB33CDA98BE4E46A6A8C241A6205A8F@MikePC






Description: Description: cid:3FCA4AF59E71452799F61D0917BBF4A0@MikePC






Description: Description: cid:5F1D1457005D44C0A99EC698208E3C7B@MikePC






Description: Description: cid:ED0FD1B7CF54450680C8AC2459679F9D@MikePC






Description: Description: cid:58C4D40E61944F9FA4916BD96A860182@MikePC






Description: Description: cid:D6F4E75F8F1C436990212F3DC0AFBB47@MikePC






Description: Description: cid:EE89465626FE4FE6AA1F58B175B4F345@MikePC






Description: Description: cid:CCC2E2F7CB844A2F97BA73184572F67F@MikePC






Description: Description: cid:BA1107351AEB41558E416A0CE5841BBF@MikePC






AGI | <https://www.listbox.com/member/archive/303/=now&gt; Archives <https://www.listbox.com/member/archive/rss/303/&gt; Description: Description: Image removed by sender.| <https://www.listbox.com/member/?&&gt; Modify Your Subscription

<http://www.listbox.com/&gt; Description: Description: Image removed by sender.




AGI | <https://www.listbox.com/member/archive/303/=now&gt; Archives <https://www.listbox.com/member/archive/rss/303/&gt; Description: Image removed by sender.| <https://www.listbox.com/member/?&&gt; Modify Your Subscription

<http://www.listbox.com/&gt; Description: Image removed by sender.


AGI | <https://www.listbox.com/member/archive/303/=now&gt; Archives <https://www.listbox.com/member/archive/rss/303/&gt; Description: Image removed by sender.| <https://www.listbox.com/member/?&&gt; Modify Your Subscription

<http://www.listbox.com/&gt; Description: Image removed by sender.

Nope

Your examples are useful {incl clock} - they show the **limits** of patterns - wh. can indeed be fluidly morphed and stretched and still remain patterns..

But you are always morphing the *same parts* arranged in the *same structures*. - basically homogeneous, homeomorphic forms - wh. are essential for any set of patterns

The Google logos show the difference -

you are recognizing

*different diverse parts* arranged in **different diverse structures**.- wh. can in fact be endlessly diverse. The letters of the Google logo can in fact be almost any form in the universe from letters to snakes to naked Adam's. If your clock's letters suddenly changed into snakes or naked Adams, you would cease to be dealing in patterns.

If you cannot see that, you have a truly major visual/conceptual blindness.

And the challenge is to understand how the mind actually processes and compares these diverse forms - because there clearly are underlying principles to how the mind achieves this, but they ain't the principles of math. patterns.


From: John G. Rose
Sent: Tuesday, September 07, 2010 6:09 PM
To: AGI
Subject: RE: [agi] And now the quintessential AGI font


Isn’t this just pattern recognition? Narrow AI?



Like CAPTCHA is use to deter bots:



http://en.wikipedia.org/wiki/CAPTCHA



John



From: Mike Tintner [mailto:tintner@blueyonder.co.uk]
Sent: Tuesday, September 07, 2010 10:57 AM
To: AGI
Subject: Re: [agi] And now the quintessential AGI font



John: nothing to do with AGI... really.



Cor blimey, you don't know to look. You're looking at that logo and seeing ONE logo, right?



Wrong.



You have to see an extensive, and in principle endless series of alternatives, and ask: how does the mind recognize them ALL? That is the essence of AGI and of course visual perception, But nothing in maths can handle it. Sorry, John. Art & AGI are not maths. Hope you can see the difference.





















































































AGI | Archives | Modify Your Subscription





AGI | Archives | Modify Your Subscription

What I was thinking is that the software used to generate the font has
nothing to do with AGI... clearly someone is designing the font and that
someone is of general intelligence. which is why I agreed that it had
nothing to do with AGI.

If you mean recognition of the font... ok, that would be AGI, but not
without knowledge, experience and contextual analysis. We have a lot of
experience that goes into this such recognition. For example, if you go to
the google home page where the regular font used to be, you clearly expect
the image there to be of a google font, even if it's drawn using other
forms. The letter's method of recognition is based on shape, size, curves
and other such elements. In the brain, these are likely processed in
parallel. So, although you recognize the lady bug below, you also expect the
letters that makeup the letters in the name "google". And when you see
something that doesn't make sense in an image... a coincidence that is just
too coincidental, then it is natural for us to look for a match and
explanation. Thus we are able to see the font, even though it is drawn using
an odd symbol/form.

It is worth analyzing as a significant principle/part of AGI, but it is not
THEE only core of AGI or the bottleneck to understanding AGI. There is a
hell of a lot of other things going on. Nonetheless, it should be studied.

Dave

On Tue, Sep 7, 2010 at 12:56 PM, Mike Tintner <tintner@blueyonder.co.uk>wrote:

> John: nothing <http://www.physorg.com/news203011174.html&gt; to do with
> AGI... really.
>
> Cor blimey, you don't know to look. You're looking at that logo and seeing
> ONE logo, right?
>
> Wrong.
>
> You have to see an extensive, and in principle endless series of
> alternatives, and ask: how does the mind recognize them ALL? That is the
> essence of AGI and of course visual perception, But nothing in maths can
> handle it. Sorry, John. Art & AGI are not maths. Hope you can see the
> difference.
>
> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
> <http://www.listbox.com&gt;
>

"I know from experience"

Now that was funny.

L.

On Tue, Sep 7, 2010 at 2:16 PM, Leandro <l.detetive@gmail.com> wrote:

> "I'm constantly amazed at how you construe the evidence to meet your pet
> theory. "
>
> And I'd add more: the article clearly talks about GEOMETRIC patterns, in
> contrast to NON-GEOMETRIC patterns (your (Mike's) "natural" thing). Both are
> patterns, even in the mathematical sense. But it's easier to just skip some
> words, conveniently.
>
> L.
>
> On Tue, Sep 7, 2010 at 1:41 PM, David Jones <davidhere40@gmail.com> wrote:
>
>> I'm constantly amazed at how you construe the evidence to meet your pet
>> theory.
>>
>> Just put the autistic artist "Richard Wawro" into google images. The first
>> thing you see is a painting of himself! LOL. I thought he wasn't able to
>> understand and recognize such complex shapes? I'm sure there are many, many
>> autistic savants that have the ability to draw amazingly complex things,
>> including animals and people, which are too complex according to you.
>>
>> Dave
>>
>> On Tue, Sep 7, 2010 at 12:35 PM, Mike Tintner <tintner@blueyonder.co.uk>wrote:
>>
>>> It's pretty clear that autistics prefer patterns because they are
>>> **simpler** than natural shapes. Even when changing and moving, those
>>> patterns are simpler and more structured than the irregular, twisted and
>>> often jerky paths of movement of natural forms, like human and animal
>>> figures as in the vids.
>>>
>>> Autistics have some kind of perceptual disorder which means that they
>>> have problems processing complicated and unpredictable irregular forms as
>>> fast as the ordinary person does - and this often makes dealing with people
>>> esp. much too threatening for them.
>>>
>>> The challenge for AGI, as I keep repeating, is to deal with the
>>> complicated irregular, unpredictable natural forms (& also complicated
>>> *collections* of artificial forms) as successfully and fast as real AGI's
>>> do.
>>>
>>> The font I just instanced, is extremely important because it gives, I
>>> believe, vital clues to how the brain achieves that - it breaks figures (as
>>> the IBM logo does) into their component particles and shows how the mind is
>>> capable of fluidly, hetero- rather than homostructurally rearranging
>>> ("patching") those particles into new forms, and still recognizing their
>>> similarities..
>>>
>>> In general, I know from experience, that AGI-ers will always choose
>>> narrow AI, artificial patterns over complicated real AGI, natural patchworks
>>> to study
>>>
>>> *From:* David Jones <davidhere40@gmail.com>
>>> *Sent:* Tuesday, September 07, 2010 4:12 PM
>>> *To:* AGI <agi@listbox.com>
>>> *Subject:* Re: [agi] Patterns vs Natural Shapes
>>>
>>> You know what is interesting. Smart people tend to want to analyze
>>> problems, even as children. They are more analytical. So, it has also been
>>> noticed that smart parents are more likely to have autistic children. It
>>> seems to me that the parents both have analytical tendencies that are sort
>>> of rare in the population, but they meet each other because of similar
>>> vocations and interests. By having children, they concentrate this genetic
>>> tendency to be analytical and it may result in a combination of multiple
>>> recessive genes for overly analytical kids. The problem occurs when the kids
>>> ignore social cues which are very, very important for learning. They are
>>> also overly focused in what they choose to learn about... overly analytical
>>> and hyperfocusing. This combination of problems is a perfect match for the
>>> cause of autism as others have also observed. That also explains why they
>>> like the patterns and don't care as much about social images... that part of
>>> the brain or those genes are inhibited.
>>>
>>> Dave
>>>
>>> On Tue, Sep 7, 2010 at 11:04 AM, Mike Tintner <tintner@blueyonder.co.uk>wrote:
>>>
>>>> Visual pattern preference may be indicator of autism in toddlers
>>>> http://www.physorg.com/news203011174.html
>>>> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
>>>> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
>>>> <http://www.listbox.com&gt;
>>>>
>>>
>>> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
>>> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
>>> <http://www.listbox.com&gt;
>>> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
>>> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
>>> <http://www.listbox.com&gt;
>>>
>>
>> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
>> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
>> <http://www.listbox.com&gt;
>>
>
>

"I'm constantly amazed at how you construe the evidence to meet your pet
theory. "

And I'd add more: the article clearly talks about GEOMETRIC patterns, in
contrast to NON-GEOMETRIC patterns (your (Mike's) "natural" thing). Both are
patterns, even in the mathematical sense. But it's easier to just skip some
words, conveniently.

L.

On Tue, Sep 7, 2010 at 1:41 PM, David Jones <davidhere40@gmail.com> wrote:

> I'm constantly amazed at how you construe the evidence to meet your pet
> theory.
>
> Just put the autistic artist "Richard Wawro" into google images. The first
> thing you see is a painting of himself! LOL. I thought he wasn't able to
> understand and recognize such complex shapes? I'm sure there are many, many
> autistic savants that have the ability to draw amazingly complex things,
> including animals and people, which are too complex according to you.
>
> Dave
>
> On Tue, Sep 7, 2010 at 12:35 PM, Mike Tintner <tintner@blueyonder.co.uk>wrote:
>
>> It's pretty clear that autistics prefer patterns because they are
>> **simpler** than natural shapes. Even when changing and moving, those
>> patterns are simpler and more structured than the irregular, twisted and
>> often jerky paths of movement of natural forms, like human and animal
>> figures as in the vids.
>>
>> Autistics have some kind of perceptual disorder which means that they have
>> problems processing complicated and unpredictable irregular forms as fast as
>> the ordinary person does - and this often makes dealing with people esp.
>> much too threatening for them.
>>
>> The challenge for AGI, as I keep repeating, is to deal with the
>> complicated irregular, unpredictable natural forms (& also complicated
>> *collections* of artificial forms) as successfully and fast as real AGI's
>> do.
>>
>> The font I just instanced, is extremely important because it gives, I
>> believe, vital clues to how the brain achieves that - it breaks figures (as
>> the IBM logo does) into their component particles and shows how the mind is
>> capable of fluidly, hetero- rather than homostructurally rearranging
>> ("patching") those particles into new forms, and still recognizing their
>> similarities..
>>
>> In general, I know from experience, that AGI-ers will always choose narrow
>> AI, artificial patterns over complicated real AGI, natural patchworks to
>> study
>>
>> *From:* David Jones <davidhere40@gmail.com>
>> *Sent:* Tuesday, September 07, 2010 4:12 PM
>> *To:* AGI <agi@listbox.com>
>> *Subject:* Re: [agi] Patterns vs Natural Shapes
>>
>> You know what is interesting. Smart people tend to want to analyze
>> problems, even as children. They are more analytical. So, it has also been
>> noticed that smart parents are more likely to have autistic children. It
>> seems to me that the parents both have analytical tendencies that are sort
>> of rare in the population, but they meet each other because of similar
>> vocations and interests. By having children, they concentrate this genetic
>> tendency to be analytical and it may result in a combination of multiple
>> recessive genes for overly analytical kids. The problem occurs when the kids
>> ignore social cues which are very, very important for learning. They are
>> also overly focused in what they choose to learn about... overly analytical
>> and hyperfocusing. This combination of problems is a perfect match for the
>> cause of autism as others have also observed. That also explains why they
>> like the patterns and don't care as much about social images... that part of
>> the brain or those genes are inhibited.
>>
>> Dave
>>
>> On Tue, Sep 7, 2010 at 11:04 AM, Mike Tintner <tintner@blueyonder.co.uk>wrote:
>>
>>> Visual pattern preference may be indicator of autism in toddlers
>>> http://www.physorg.com/news203011174.html
>>> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
>>> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
>>> <http://www.listbox.com&gt;
>>>
>>
>> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
>> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
>> <http://www.listbox.com&gt;
>> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
>> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
>> <http://www.listbox.com&gt;
>>
>
> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
> <http://www.listbox.com&gt;
>

Isn’t this just pattern recognition? Narrow AI?



Like CAPTCHA is use to deter bots:



http://en.wikipedia.org/wiki/CAPTCHA



John



From: Mike Tintner [mailto:tintner@blueyonder.co.uk]
Sent: Tuesday, September 07, 2010 10:57 AM
To: AGI
Subject: Re: [agi] And now the quintessential AGI font



John: <http://www.physorg.com/news203011174.html&gt; nothing to do with AGI... really.



Cor blimey, you don't know to look. You're looking at that logo and seeing ONE logo, right?



Wrong.



You have to see an extensive, and in principle endless series of alternatives, and ask: how does the mind recognize them ALL? That is the essence of AGI and of course visual perception, But nothing in maths can handle it. Sorry, John. Art & AGI are not maths. Hope you can see the difference.


Description: cid:6F0F3D448963485D9B5F1A06EB6D2462@MikePC






Description: cid:27FD4421261D4D688A07F0A57419B59E@MikePC






Description: cid:DC70104C2F6446C9A7640DCCB395407A@MikePC






Description: cid:5BEF1645863A4155A1CC9DFFF9BA7C58@MikePC






Description: cid:EEF64A755878467BA6616042666A5B23@MikePC






Description: cid:0BB33CDA98BE4E46A6A8C241A6205A8F@MikePC






Description: cid:3FCA4AF59E71452799F61D0917BBF4A0@MikePC






Description: cid:5F1D1457005D44C0A99EC698208E3C7B@MikePC






Description: cid:ED0FD1B7CF54450680C8AC2459679F9D@MikePC






Description: cid:58C4D40E61944F9FA4916BD96A860182@MikePC






Description: cid:D6F4E75F8F1C436990212F3DC0AFBB47@MikePC






Description: cid:EE89465626FE4FE6AA1F58B175B4F345@MikePC






Description: cid:CCC2E2F7CB844A2F97BA73184572F67F@MikePC






Description: cid:BA1107351AEB41558E416A0CE5841BBF@MikePC






AGI | <https://www.listbox.com/member/archive/303/=now&gt; Archives <https://www.listbox.com/member/archive/rss/303/&gt; Description: Image removed by sender.| <https://www.listbox.com/member/?&&gt; Modify Your Subscription

<http://www.listbox.com/&gt; Description: Image removed by sender.

Well at some point when it is self-organizing itself into the word "Google"
the eyes/mind pattern recognizer does converge to a mental focus and you see
that the word is Google.



The code is similar to this:



http://rainbow.arch.scriptmania.com/scripts/mouse_clock3.html



John



From: deepakjnath [mailto:deepakjnath@gmail.com]



But it is very very smooth!

On Tue, Sep 7, 2010 at 9:14 PM, John G. Rose <johnrose@polyplexic.com>
wrote:

This just some JavaScript that moves some dots around in reaction to mouse
pointer movements and rearranges them to the shape of the word "Google".
This technique has been used on a number of websites, nothing ground
breaking here just some fancy script code.



Nothing to do with AGI.really.



John



From: Mike Tintner [mailto:tintner@blueyonder.co.uk]
Sent: Tuesday, September 07, 2010 3:49 AM
To: AGI
Subject: [agi] And now the quintessential AGI font



And another mobile Google font today -check it out because it's
quintessential AGI. A completely mobile dot Google dot logo. Beautiful
because it illustrates the essential requirement of AGI. An AGI must be able
to navigate and construct structures that are *freeform *freestructure
*freecontent. Just about everything can change in the territory you're
navigating, the materials or fonts you are working with - and yet you can
still succeed in navigating, recognizing and constructing with them . A real
AGI can navigate the same street if it was all buildings yesterday, and is
blown up rubble today - or recognize that a text about Obama is basically
the same even though the wording has completely changed.



The beauty of this font is that it works on exactly the same principle as
IBM's "Think" logo - where the mobile dots make up not letters but the
continents of the world.



It's a nice illustration too of the complete absurdity of all complexity
approaches to AGI. In order to recognize a new font, you don't need to
search/consult a vast space of previous fonts - you just need a few at most.
"Think".


AGI | <https://www.listbox.com/member/archive/303/=now&gt; Archives
<https://www.listbox.com/member/archive/rss/303/&gt; Description: Description:
Image removed by sender.| <https://www.listbox.com/member/?&&gt; Modify Your
Subscription

<http://www.listbox.com/&gt; Description: Description: Image removed by
sender.




AGI | <https://www.listbox.com/member/archive/303/=now&gt; Archives
<https://www.listbox.com/member/archive/rss/303/&gt; Description: Image removed
by sender.| <https://www.listbox.com/member/?&&gt; Modify Your Subscription

<http://www.listbox.com/&gt; Description: Image removed by sender.




cheers,
Deepak


AGI | <https://www.listbox.com/member/archive/303/=now&gt; Archives
<https://www.listbox.com/member/archive/rss/303/&gt; Description: Image removed
by sender.|
<https://www.listbox.com/member/?&&gt;
Modify Your Subscription

<http://www.listbox.com/&gt; Description: Image removed by sender.

John: nothing to do with AGI... really.

Cor blimey, you don't know to look. You're looking at that logo and seeing ONE logo, right?

Wrong.

You have to see an extensive, and in principle endless series of alternatives, and ask: how does the mind recognize them ALL? That is the essence of AGI and of course visual perception, But nothing in maths can handle it. Sorry, John. Art & AGI are not maths. Hope you can see the difference.

I'm constantly amazed at how you construe the evidence to meet your pet
theory.

Just put the autistic artist "Richard Wawro" into google images. The first
thing you see is a painting of himself! LOL. I thought he wasn't able to
understand and recognize such complex shapes? I'm sure there are many, many
autistic savants that have the ability to draw amazingly complex things,
including animals and people, which are too complex according to you.

Dave

On Tue, Sep 7, 2010 at 12:35 PM, Mike Tintner <tintner@blueyonder.co.uk>wrote:

> It's pretty clear that autistics prefer patterns because they are
> **simpler** than natural shapes. Even when changing and moving, those
> patterns are simpler and more structured than the irregular, twisted and
> often jerky paths of movement of natural forms, like human and animal
> figures as in the vids.
>
> Autistics have some kind of perceptual disorder which means that they have
> problems processing complicated and unpredictable irregular forms as fast as
> the ordinary person does - and this often makes dealing with people esp.
> much too threatening for them.
>
> The challenge for AGI, as I keep repeating, is to deal with the
> complicated irregular, unpredictable natural forms (& also complicated
> *collections* of artificial forms) as successfully and fast as real AGI's
> do.
>
> The font I just instanced, is extremely important because it gives, I
> believe, vital clues to how the brain achieves that - it breaks figures (as
> the IBM logo does) into their component particles and shows how the mind is
> capable of fluidly, hetero- rather than homostructurally rearranging
> ("patching") those particles into new forms, and still recognizing their
> similarities..
>
> In general, I know from experience, that AGI-ers will always choose narrow
> AI, artificial patterns over complicated real AGI, natural patchworks to
> study
>
> *From:* David Jones <davidhere40@gmail.com>
> *Sent:* Tuesday, September 07, 2010 4:12 PM
> *To:* AGI <agi@listbox.com>
> *Subject:* Re: [agi] Patterns vs Natural Shapes
>
> You know what is interesting. Smart people tend to want to analyze
> problems, even as children. They are more analytical. So, it has also been
> noticed that smart parents are more likely to have autistic children. It
> seems to me that the parents both have analytical tendencies that are sort
> of rare in the population, but they meet each other because of similar
> vocations and interests. By having children, they concentrate this genetic
> tendency to be analytical and it may result in a combination of multiple
> recessive genes for overly analytical kids. The problem occurs when the kids
> ignore social cues which are very, very important for learning. They are
> also overly focused in what they choose to learn about... overly analytical
> and hyperfocusing. This combination of problems is a perfect match for the
> cause of autism as others have also observed. That also explains why they
> like the patterns and don't care as much about social images... that part of
> the brain or those genes are inhibited.
>
> Dave
>
> On Tue, Sep 7, 2010 at 11:04 AM, Mike Tintner <tintner@blueyonder.co.uk>wrote:
>
>> Visual pattern preference may be indicator of autism in toddlers
>> http://www.physorg.com/news203011174.html
>> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
>> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
>> <http://www.listbox.com&gt;
>>
>
> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
> <http://www.listbox.com&gt;
> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
> <http://www.listbox.com&gt;
>

see my last post - think of your capacity to recognize shopping bags of all shapes and forms filled with all manner of objects - as all "shopping bags" - despite their vastly different forms; ditto to recognize rocks of all shapes and forms as "rocks"; ditto to recognize amoebas and islands of all shapes and forms as "amoebas" and "islands". No narrow AI method can handle this. These are the unsolved problems of visual object recognition, conceptualisation, and creativity (in this case formal design/drawing).

The font here gives you a mode of analysing what is going on both externally in the objects-to-be-dealt-with and internally in the mind processing them - break them down into component particles that can be fluidly rearranged, morphed, and edited within certain limits and according to certain principles.

It presents a creative challenge for anyone actually interested in solving the problems of AGI. ...


From: John G. Rose
Sent: Tuesday, September 07, 2010 4:44 PM
To: AGI
Subject: RE: [agi] And now the quintessential AGI font


This just some JavaScript that moves some dots around in reaction to mouse pointer movements and rearranges them to the shape of the word “Google”. This technique has been used on a number of websites, nothing ground breaking here just some fancy script code.



Nothing to do with AGI…really.



John



From: Mike Tintner [mailto:tintner@blueyonder.co.uk]
Sent: Tuesday, September 07, 2010 3:49 AM
To: AGI
Subject: [agi] And now the quintessential AGI font



And another mobile Google font today -check it out because it's quintessential AGI. A completely mobile dot Google dot logo. Beautiful because it illustrates the essential requirement of AGI. An AGI must be able to navigate and construct structures that are *freeform *freestructure *freecontent. Just about everything can change in the territory you're navigating, the materials or fonts you are working with - and yet you can still succeed in navigating, recognizing and constructing with them . A real AGI can navigate the same street if it was all buildings yesterday, and is blown up rubble today - or recognize that a text about Obama is basically the same even though the wording has completely changed.



The beauty of this font is that it works on exactly the same principle as IBM's "Think" logo - where the mobile dots make up not letters but the continents of the world.



It's a nice illustration too of the complete absurdity of all complexity approaches to AGI. In order to recognize a new font, you don't need to search/consult a vast space of previous fonts - you just need a few at most. "Think".

AGI | Archives | Modify Your Subscription





AGI | Archives | Modify Your Subscription

It's pretty clear that autistics prefer patterns because they are **simpler** than natural shapes. Even when changing and moving, those patterns are simpler and more structured than the irregular, twisted and often jerky paths of movement of natural forms, like human and animal figures as in the vids.

Autistics have some kind of perceptual disorder which means that they have problems processing complicated and unpredictable irregular forms as fast as the ordinary person does - and this often makes dealing with people esp. much too threatening for them.

The challenge for AGI, as I keep repeating, is to deal with the complicated irregular, unpredictable natural forms (& also complicated *collections* of artificial forms) as successfully and fast as real AGI's do.

The font I just instanced, is extremely important because it gives, I believe, vital clues to how the brain achieves that - it breaks figures (as the IBM logo does) into their component particles and shows how the mind is capable of fluidly, hetero- rather than homostructurally rearranging ("patching") those particles into new forms, and still recognizing their similarities..

In general, I know from experience, that AGI-ers will always choose narrow AI, artificial patterns over complicated real AGI, natural patchworks to study


From: David Jones
Sent: Tuesday, September 07, 2010 4:12 PM
To: AGI
Subject: Re: [agi] Patterns vs Natural Shapes


You know what is interesting. Smart people tend to want to analyze problems, even as children. They are more analytical. So, it has also been noticed that smart parents are more likely to have autistic children. It seems to me that the parents both have analytical tendencies that are sort of rare in the population, but they meet each other because of similar vocations and interests. By having children, they concentrate this genetic tendency to be analytical and it may result in a combination of multiple recessive genes for overly analytical kids. The problem occurs when the kids ignore social cues which are very, very important for learning. They are also overly focused in what they choose to learn about... overly analytical and hyperfocusing. This combination of problems is a perfect match for the cause of autism as others have also observed. That also explains why they like the patterns and don't care as much about social images... that part of the brain or those genes are inhibited.

Dave


On Tue, Sep 7, 2010 at 11:04 AM, Mike Tintner <tintner@blueyonder.co.uk> wrote:

Visual pattern preference may be indicator of autism in toddlers
http://www.physorg.com/news203011174.html
AGI | Archives | Modify Your Subscription



AGI | Archives | Modify Your Subscription

On Tue, Sep 7, 2010 at 12:35 AM, A. T. Murray <mentifex@scn.org> wrote:
> This "Singularity Alert" from Mentifex
> may generate a collective "Huh?" from
> the list readership...


I just wanted to highlight the only part of your post that made sense.

--
J. Andrew Rogers

But it is very very smooth!

On Tue, Sep 7, 2010 at 9:14 PM, John G. Rose <johnrose@polyplexic.com>wrote:

> This just some JavaScript that moves some dots around in reaction to mouse
> pointer movements and rearranges them to the shape of the word “Google”.
> This technique has been used on a number of websites, nothing ground
> breaking here just some fancy script code.
>
>
>
> Nothing to do with AGI…really.
>
>
>
> John
>
>
>
> *From:* Mike Tintner [mailto:tintner@blueyonder.co.uk]
> *Sent:* Tuesday, September 07, 2010 3:49 AM
> *To:* AGI
> *Subject:* [agi] And now the quintessential AGI font
>
>
>
> And another mobile Google font today -check it out because it's
> quintessential AGI. A completely mobile dot Google dot logo. Beautiful
> because it illustrates the essential requirement of AGI. An AGI must be able
> to navigate and construct structures that are *freeform *freestructure
> *freecontent. Just about everything can change in the territory you're
> navigating, the materials or fonts you are working with - and yet you can
> still succeed in navigating, recognizing and constructing with them . A real
> AGI can navigate the same street if it was all buildings yesterday, and is
> blown up rubble today - or recognize that a text about Obama is basically
> the same even though the wording has completely changed.
>
>
>
> The beauty of this font is that it works on exactly the same principle as
> IBM's "Think" logo - where the mobile dots make up not letters but the
> continents of the world.
>
>
>
> It's a nice illustration too of the complete absurdity of all complexity
> approaches to AGI. In order to recognize a new font, you don't need to
> search/consult a vast space of previous fonts - you just need a few at most.
> "Think".
>
> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt; [image:
> Description: Image removed by sender.]<https://www.listbox.com/member/archive/rss/303/&gt;|
> Modify <https://www.listbox.com/member/?&&gt; Your Subscription
>
> [image: Description: Image removed by sender.] <http://www.listbox.com/&gt;
>
>
> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now&gt;
> <https://www.listbox.com/member/archive/rss/303/&gt; | Modify<https://www.listbox.com/member/?&>Your Subscription
> <http://www.listbox.com&gt;
>



--
cheers,
Deepak