AE Public BB

Post by The whole truth
Quote (midwifetoad @ May 23 2013,06:41)Despite his multi-thousand word posts,  KF only has one idea that has any utility,  and that is his isolated islands of function. It's not an argument invented by the ID movement; Darwin expressed it in its most cogent form in Origin.

He does seem to be fascinated by gay sex, and he spends an inordinate amount of time dwelling lovingly on the details.
"Darwin expressed it in its most cogent form in Origin."

Do you know which page numbers?

Post by midwifetoad
Despite his multi-thousand word posts,  KF only has one idea that has any utility,  and that is his isolated islands of function. It's not an argument invented by the ID movement; Darwin expressed it in its most cogent form in Origin.

He does seem to be fascinated by gay sex, and he spends an inordinate amount of time dwelling lovingly on the details.

Post by stevestory
Quote (GaryGaulin @ May 23 2013,00:33)It's what you are just like me obliged to explain in K-12 simple words and illustrations
That's the wrongest thing since Wrong came to Wrongtown.

Gary, if you were in a high school freshman comp class, and I was still a teacher, after about a week of reading your work I would go to home depot and rent a Troy-Bilt™ chipper-shredder, and everything you handed in, I would just toss in it.

Post by stevestory
Man, I almost feel bad about doing that to Joe. Some very smart people in GFLPC's time had some deep reservations about cardinality etc. Setting that trap for him was a bit like loading a coyote snare with a hydrogen bomb.

But who doesn't love fireworks?

Post by Arctodus23
Quote (The whole truth @ May 23 2013,04:46)Elizabeth, after seeing gordo's latest comments-off sermon at UD, do you still think that he couldn't be accurately labeled as psychotic and psychopathic by what he says on the internet? :)
He's correctly labeled, "An IGNORAMUS!!!!"

Post by Arctodus23
So, sets are sets, that are sets? Sets must be the new pokemon.

Post by Febble
sets are the new mudkips?

Post by k.e..
Yeah, all very honorable Wes but I'm guessing when the FCC fined his ass they're only plan was to fine him again so it hurt harder and harder each time he repeated his foolishness and then let the sheriff sort through his worldly goods and hold a bankrupt's auction.

GiGo is lucky not to be out on the street. I'll bet he doesn't have anything of value with his name on it. Pity he doesn't go back to spectrum spamming instead of this thread.

GiGo you are a total waste of bandwidth.

Post by k.e..
Quote (socle @ May 22 2013,19:19)Joe:
      Quote And not only that no oen can say what the practical application is for saying that the set of all non-negative integers and all positive integers have the same cardinality. It doesn't appear to have any use at all.
Dembski and Marks, in The Search for a Search: Measuring the Information Cost of Higher Level Search:
      Quote ... note also that such combinations, when restricted to a countable dense subset of Omega form a countable dense subset of M(Omega) in the
weak topology, showing that M(Omega) is itself separable in
the weak topology.

Post by Wesley R. Elsberry
Quote (GaryGaulin @ May 22 2013,23:33)     Quote (Wesley R. Elsberry @ May 22 2013,22:38)     Quote (GaryGaulin @ May 22 2013,22:12)             Quote (Nomad @ May 22 2013,21:43)             Quote (GaryGaulin @ May 22 2013,18:55)Wesley, Wesley, Wesley. If you want to impress the heck out of the Theory of Intelligent Design then you need to reduce down your self-organizing neural network into a "module" that in electronics and as in theory is most simply known as a RAM. Inputs and outputs can be analog or digital even angle or waveform whatever you want and I'll plug that in for you, right after I get it all back together with RAM computers already have and the update at PSC.
Gary, Gary, Gary.  If you want to impress the heck out of anybody other than yourself, you're going to at least have the courage of your convictions required to defend the concepts that you claim you've spend decades refining.  Remember how you claimed you should be paid for all the work you put into this?  Is this a reflection on your work ethic that you think you should be paid for concepts you've spent decades developing that won't withstand even casual questioning?

What am I to think when you put these ideas out there, invite people to use them, and then suddenly want nothing to do with them anymore and desperately try to talk about anything else?

I'm particularly amused to see you accusing Wes of moving the goalposts.  You yourself did that recently as well.  If it's a bad thing then why did you do it?  Are you special and allowed to do things like that that no one else is?

If you hate goalpost shifting then you'll be eager to get back to discussing your four requirements, the goalpost that you first erected days ago.

But no, after slinging accusations you suddenly have to run away:

                Quote I will now [at least try to] go get some more work done on the new ID Lab.

This impresses no one.  We all have other things we must do, but most of us do not need to call attention to it.  We all know what this really means is "I've got too many questions that I dare not answer building up, so I'm going to run away for a while and then come back and pretend that none of that ever happened".

So far as I know, none of the participants in this discussion have a severe enough problem with their short term memory that they'll forget what happened several days ago.  So this strategy is somewhat baffling.
If ADALINE/LMS is a comparable model then Wesley should have no problem explaining how it easily forms a hypothesis, and where where hooked into a word/vocal motor system would be forming sentences to test while searching for an answer on Jeopardy or how something being sensed works (science).
Gary not only can't quote me saying that ADALINE/LMS is a "comparable model" in the sense of doing the things Gary's PSC VB code does, he had to either fail to read my posts related to ADALINE or completely discomprehend them. There was this just above that Gary doesn't take note of:

        Quote
I never even implied that ADALINE was suitable as "RAM" to fit into Gary's PSC VB code. I presented ADALINE as one of the very simplest of artificial neural systems. It was an example of what would appear in the code base where an artificial neural system was implemented, and as such a pattern so that others could examine Gary's PSC VB code to see what I saw, which was a complete absence of artificial neural system implementations.


As such, I'm not on the hook to explain "how it easily forms a hypothesis", since I've never claimed any such thing. And I'll note that I have seen nothing to indicate that Gary's PSC VB code has any capability to "form a hypothesis". Gary is invited to give file name and line numbers of where such a capability is supposed to be implemented if he is claiming his code does such.

Nor have I compared ADALINE to Watson.

Gary's final "or" clause is just gibberish.

Now, Gary did offer a comparison of his code to ADALINE/LMS.

Gary:

        Quote
In both cases there is an "approximation" of real neurons and I doubt yours could beat the performance of the simple method I use.


Since I've pointed out how ludicrous Gary's claim is on its face, Gary has done his usual routine of digression, character attacks on his critics, and bogus claims of what other people supposedly are obliged to do. What is conspicuous by its absence is any defense whatsoever by Gary of his actual claim.
I already went into detail about the GUESS block shown in the diagram where random or good guess hypotheses of any motor system kind are produced.



Whatever model you want to be taken seriously must as simply as possible account for what IBM Watson has that makes it able to win a complex game show against world champion humans. Not being able to do so only indicates that you did not present a comparable model, in the first place.

This is not a situation where it's what you said your model can do. It's what you are just like me obliged to explain in K-12 simple words and illustrations or the model that the theory now uses remains standing.
Bizarre. When I point out faults in Gary's response, his next response is to reiterate the faults rather than address his claim.

Gary:

        Quote
In both cases there is an "approximation" of real neurons and I doubt yours could beat the performance of the simple method I use.


This claim isn't about a particular application, it is about general performance compared between Gary's "simple method" and ADALINE/LMS. No matter how often this is pointed out, Gary seems unable to comprehend it.

My aim is much more modest than Gary's, apparently. All I'm out to accomplish here is noting that one particular claim of Gary's is astoundingly ludicrous. Since that doesn't require any further action on Gary's part, he could consider the notion that he might stop digging.

Post by NoName
Quote (GaryGaulin @ May 22 2013,23:33)...
Whatever model you want to be taken seriously must as simply as possible account for what IBM Watson has that makes it able to win a complex game show against world champion humans. Not being able to do so only indicates that you did not present a comparable model, in the first place.
Gary, Gary, Gary, you pretentious whining loon.

Whatever model you want to be taken seriously must be defended on the merits of the model.  NOT on the mendacious crap you've been dishing out.
Defended by the modeler, directly, rather than by reference to those whose work the modeler barely begins to comprehend.  (Do you move your lips when you look at pictures?)

Try addressing Nomad's questions.  Try demonstrating that you have an actual bona fide theory rather than codswallop.
Try being coherent.
Try anything other than what you've been up to, because quite clearly what you've been up to isn't working.  In terms of your model, all you've got are bad guesses.  Why is that?

Post by k.e..
Quote (GaryGaulin @ May 18 2013,12:03) Quote (Nomad @ May 18 2013,03:01)Gary.. I'm confused.. you say:

      Quote Anthropomorphic generalizations are not needed

And then in the description for your first requirement, all you do is talk about how humans act.  You even feel the need to start talking about religion.  There is nothing in that "requirement" that isn't anthropomorphic.

The first requirement is baloney anyway.  "Something to control"?  Really?  I guess neural nets that do nothing but analyze data and come to conclusions can never be declared intelligent then.  Woops, there goes Watson.  You referred approvingly to it before, but apparently you didn't attempt to find your four requirements in it.

This has been touched on by others.  Concepts like neural nets and machine intelligence really aren't my area, I'm not a programmer at all, beyond messing with basic years ago, but it seems you've skipped out on a great deal of stuff.  You don't seem to be defining intelligence at all.  Let's apply your four requirements to my camera, or more specifically only the autofocus system on my camera.

#1 something to control.

Yep, the autofocus motor on the lens.

#2 sensory addressed memory

Yep.  The autofocus sensor feeds data into memory inside the camera.

#3 confidence to gauge failure or success

Yep.  The AF sensor tells the camera how well focused the scene is, providing a sense of better or worse that it uses to converge on the best possible setting.

#4 ability to take a guess

Yep.  It doesn't necessarily know how far it has to adjust the focus, so it takes a guess and evaluates the results.

Well alrighty then.  My camera is intelligent.

I dunno, Gary.  It kind of sounds like you've so loosely defined intelligence that many things that wouldn't normally be considered intelligent qualify.  Preprogrammed behaviors intended to achieve simple tasks qualify as intelligent now.  I suspect this is why you're having such a hard time understand when people tell you that your code is not a neural net, because your code fulfills your notion of requirements of intelligence you've concluded that it must be intelligent and is therefore a neural net.  But it doesn't work like that.  Or else my camera has a neural net AF system.
It's relatively common knowledge that a camera focus system uses a circuit called a "servo"

The rest of your hand-waving mess of scientifically useless generalizations are equally ridiculous.
GiGo a servo is NOT a circuit you don't have clue.

Common knowledge?

On this forum it is common knowledge that all your claims are just blatherings from a deranged mind.

Hand waving?

You would know.

Post by k.e..
Quote (Henry J @ May 18 2013,08:37) Quote (k.e.. @ May 17 2013,23:08)I wonder if Joe would do any better than AFDave on imaginary numbers?
That's a complex subject.
yeah he couldn't conjugate them

Post by timothya
KF included this in one of his comments at UD:
Quote FSCO/I is a good example, such as is manifest in this post.
I couldn't agree more.

Post by GaryGaulin
Quote (Nomad @ May 18 2013,03:01)Gary.. I'm confused.. you say:

      Quote Anthropomorphic generalizations are not needed

And then in the description for your first requirement, all you do is talk about how humans act.  You even feel the need to start talking about religion.  There is nothing in that "requirement" that isn't anthropomorphic.

The first requirement is baloney anyway.  "Something to control"?  Really?  I guess neural nets that do nothing but analyze data and come to conclusions can never be declared intelligent then.  Woops, there goes Watson.  You referred approvingly to it before, but apparently you didn't attempt to find your four requirements in it.

This has been touched on by others.  Concepts like neural nets and machine intelligence really aren't my area, I'm not a programmer at all, beyond messing with basic years ago, but it seems you've skipped out on a great deal of stuff.  You don't seem to be defining intelligence at all.  Let's apply your four requirements to my camera, or more specifically only the autofocus system on my camera.

#1 something to control.

Yep, the autofocus motor on the lens.

#2 sensory addressed memory

Yep.  The autofocus sensor feeds data into memory inside the camera.

#3 confidence to gauge failure or success

Yep.  The AF sensor tells the camera how well focused the scene is, providing a sense of better or worse that it uses to converge on the best possible setting.

#4 ability to take a guess

Yep.  It doesn't necessarily know how far it has to adjust the focus, so it takes a guess and evaluates the results.

Well alrighty then.  My camera is intelligent.

I dunno, Gary.  It kind of sounds like you've so loosely defined intelligence that many things that wouldn't normally be considered intelligent qualify.  Preprogrammed behaviors intended to achieve simple tasks qualify as intelligent now.  I suspect this is why you're having such a hard time understand when people tell you that your code is not a neural net, because your code fulfills your notion of requirements of intelligence you've concluded that it must be intelligent and is therefore a neural net.  But it doesn't work like that.  Or else my camera has a neural net AF system.
It's relatively common knowledge that a camera focus system uses a circuit called a "servo"

The rest of your hand-waving mess of scientifically useless generalizations are equally ridiculous.

Post by Nomad
Gary.. I'm confused.. you say:

    Quote Anthropomorphic generalizations are not needed

And then in the description for your first requirement, all you do is talk about how humans act.  You even feel the need to start talking about religion.  There is nothing in that "requirement" that isn't anthropomorphic.

The first requirement is baloney anyway.  "Something to control"?  Really?  I guess neural nets that do nothing but analyze data and come to conclusions can never be declared intelligent then.  Woops, there goes Watson.  You referred approvingly to it before, but apparently you didn't attempt to find your four requirements in it.

This has been touched on by others.  Concepts like neural nets and machine intelligence really aren't my area, I'm not a programmer at all, beyond messing with basic years ago, but it seems you've skipped out on a great deal of stuff.  You don't seem to be defining intelligence at all.  Let's apply your four requirements to my camera, or more specifically only the autofocus system on my camera.

#1 something to control.

Yep, the autofocus motor on the lens.

#2 sensory addressed memory

Yep.  The autofocus sensor feeds data into memory inside the camera.

#3 confidence to gauge failure or success

Yep.  The AF sensor tells the camera how well focused the scene is, providing a sense of better or worse that it uses to converge on the best possible setting.

#4 ability to take a guess

Yep.  It doesn't necessarily know how far it has to adjust the focus, so it takes a guess and evaluates the results.

Well alrighty then.  My camera is intelligent.

I dunno, Gary.  It kind of sounds like you've so loosely defined intelligence that many things that wouldn't normally be considered intelligent qualify.  Preprogrammed behaviors intended to achieve simple tasks qualify as intelligent now.  I suspect this is why you're having such a hard time understand when people tell you that your code is not a neural net, because your code fulfills your notion of requirements of intelligence you've concluded that it must be intelligent and is therefore a neural net.  But it doesn't work like that.  Or else my camera has a neural net AF system.

Post by Quack
Indeed, thanks for reminding me, I'll look into it.

Post by Quack
I'd rather see you tone down your anti-science a lot!

Post by Richardthughes
LOL@mullings bellyaching on UD because he can't do the math like Lizzie!  

Post by Henry J
Quote (k.e.. @ May 17 2013,23:08)I wonder if Joe would do any better than AFDave on imaginary numbers?
That's a complex subject.