RSS 2.0 Feed

» Welcome Guest Log In :: Register

Pages: (622) < ... 383 384 385 386 387 [388] 389 390 391 392 393 ... >   
  Topic: A Separate Thread for Gary Gaulin, As big as the poop that does not look< Next Oldest | Next Newest >  
GaryGaulin



Posts: 5385
Joined: Oct. 2012

(Permalink) Posted: Aug. 15 2014,22:14   

Funny, I have a link to the latest science mayhem at the NCSE blog:

http://ncse.com/blog.......4539015

--------------
The theory of intelligent design holds that certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.

   
didymos



Posts: 1828
Joined: Mar. 2008

(Permalink) Posted: Aug. 16 2014,02:23   

Quote (GaryGaulin @ Aug. 15 2014,20:14)
Funny, I have a link to the latest science mayhem at the NCSE blog:

http://ncse.com/blog.......4539015

Gary, your antics at other sites don't interest anyone.  Either discuss what people bring up here, or go to NCSE and stay there.

--------------
I wouldn't be bothered reading about the selfish gene because it has never been identified. -- Denyse O'Leary, professional moron
Again "how much". I don't think that's a good way to be quantitative.-- gpuccio

  
NoName



Posts: 2729
Joined: Mar. 2013

(Permalink) Posted: Aug. 16 2014,06:31   

Indeed.  How about we discuss the gross and fatal mismatches between your oft-repeated diagram and your claims of 'molecular intelligence' and 'cellular intelligence'?
Neither molecules nor cells have sensoria, nor any strict analogs to sensoria -- no sensory input possible.
Neither molecules nor cells have RAM nor any strict analogs to memory storage -- no direct connection to Memory possible.
Neither molecules nor cells have have motor controls nor any strict analog to motor controls -- no direct connection from Memory to motor control possible.
Neither molecules nor cells have 'confidence evaluation' subsystems.
Neither molecules nor cells have the ability to 'guess'.

So there is nothing at all in molecular nor cellular existence that corresponds to any element in your ludicrous diagram that you seem to believe provides some sort of 'explanation' of 'intelligence'.
Whatever it is you mean by 'intelligence' as represented in your diagram does not and cannot apply at the molecular or cellular levels.

So much for self-similarity at all levels.
Oh, wait.  You have managed to preserve that by being wrong at all levels.
Epic unwin, Gary.  Fractally so.

  
GaryGaulin



Posts: 5385
Joined: Oct. 2012

(Permalink) Posted: Aug. 16 2014,20:57   

Quote (didymos @ Aug. 16 2014,02:23)
Quote (GaryGaulin @ Aug. 15 2014,20:14)
Funny, I have a link to the latest science mayhem at the NCSE blog:

http://ncse.com/blog.......4539015

Gary, your antics at other sites don't interest anyone.  Either discuss what people bring up here, or go to NCSE and stay there.

Since this is a forum for monitoring and responding to ID related activity I'm just doing my part to help. You should be thanking me.

Here's one though you might not have closely examined enough to appreciate the significance of:

Quote
John Harshman • a day ago

I'd say the big problem is the personification of "survival of the fittest". Secondary problems are the term itself as a description of selection, the attribution of moral value to characters ("wasteful"), and, as you say, the idea that selection produces perfection.

On the other hand, his core point is strong. Brains are expensive, and if 90% of it weren't doing anything, there would be strong selection to eliminate that expensive 90%.


It might not be easy to notice but the generalizations inherent to the Darwinian model are now causing major havoc for Stephanie's premise. Not I. I'm just explaining what I found works to eliminate that and the other problematic ambiguities coming from its core logic.

--------------
The theory of intelligent design holds that certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.

   
Wesley R. Elsberry



Posts: 4991
Joined: May 2002

(Permalink) Posted: Aug. 16 2014,21:10   

Gary apparently didn't understand Harshman.

I'm not surprised.

--------------
"You can't teach an old dogma new tricks." - Dorothy Parker

    
GaryGaulin



Posts: 5385
Joined: Oct. 2012

(Permalink) Posted: Aug. 16 2014,22:02   

Quote (Wesley R. Elsberry @ Aug. 16 2014,21:10)
Gary apparently didn't understand Harshman.

I'm not surprised.


I'm the best judge of how well John parallels my thoughts, not you. Stop trying to speak for me.

 
Quote
Gary Gaulin • 2 days ago

Maybe the problem (getting the expected answer) has something to do with the required fitness function of the Darwinian (Evolutionary Algorithm) model?

From: http://en.wikipedia.org/wiki.......unction

A fitness function is a particular type of objective function that is used to summarise, as a single figure of merit, how close a given design solution is to achieving the set aims.
In particular, in the fields of genetic programming and genetic algorithms, each design solution is represented as a string of numbers (referred to as a chromosome). After each round of testing, or simulation, the idea is to delete the 'n' worst design solutions, and to breed 'n' new ones from the best design solutions. Each design solution, therefore, needs to be awarded a figure of merit, to indicate how close it came to meeting the overall specification, and this is generated by applying the fitness function to the test, or simulation, results obtained from that solution.


The Darwinian model can be said to have a "Survival of the fittest" built into the way fitness is scored, even though in reality the concept is an overgeneralization.

I seriously found that what is needed to solve this problem is a non-Darwinian model that (as in biology) self-learns at multiple levels. You're otherwise stuck with problematic generalizations that are inherent to the Darwinian model, which becomes a coding nightmare and increasingly less biologically relevant by trying to achieve such complex emergent behaviors any other way. Need to simplify down the problem to an Occam's Razor shaved algorithm, where (as in reality) some get lucky in love, with no "fittest" to attempt to prejudge required. The programming challenge is then the large amount of systems biology, which is now being sorted out in a way that makes it much easier to model living things:

http://www.nanowerk.com/news2......944.php


What John said probably did not help Stephanie at all, but it sure helped make sense of what I have been saying about the generalizations from Darwinian theory causing problems that lead to misconceptions that would be best to not have in the first place.

--------------
The theory of intelligent design holds that certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.

   
N.Wells



Posts: 1836
Joined: Oct. 2005

(Permalink) Posted: Aug. 16 2014,23:14   

Quote
I'm the best judge of how well John parallels my thoughts, not you. Stop trying to speak for me.
You are not the best judge of how well John Harshman parallels your thoughts when you clearly misunderstand him.  Wesley is not speaking for you but for John.

Despite the use of the terms "fitness"  and "evolutionary" in discussions of fitness functions in evolutionary algorithms, these concepts are not identical to what happens in biological evolution and natural selection.  In "evolutionary algorithms", people (1) set a goal, (2) let random changes happen, (3) test progress against the goal set in step 1, and (4) let the winners move on to the next iteration of random change and selection.  This is what happens in artificial selection, like breeding more productive dairy cows, but search-based algorithms don't happen in natural selection.  Searches and goals can serve as useful metaphors when discussing some aspects of evolution, but should not be mistaken for what actually happens.  Barely flapping dinosaurs do not set a goal of becoming better fliers, nor is such a goal set for them, nor are they actively searching for better ways of flying.  Each member of each generation has varying degrees of reproductive success, and the extent to which that success depends on inherited variations, those variations that led to success become more abundant in succeeding generations.  If better flying led to catching more food or escaping more predators, and that in turn led to having more offspring, and if better flying was due to some inheritable variation, then the next generation will on the whole be better fliers, but that is not required, and no one set a goal of being a better flier.  Thus there is no step 1 in nature, other than reproduction, so the only test in 2 is how many children and grandchildren the testee's genome produces.  

Natural selection is evaluated by change in a population's genome that is attributable to differential reproductive success that is in turn attributable to hereditary differences, not fitness functions.  Evolution does not have design goals against which progress can be measured, other than reproduction.

One of Harshman's concerns was Fields was personifying natural selection (anthropomorphizing it), due to implications of volition in the statement "sees to it".  To the extent that "sees to it" is read as "ensures", the problem is minimal, but the problem increases proportional to the degree that "sees to it" is read as evolution wanting something, or having a purpose as opposed to being a result.  Assigning "intelligence" to biochemical reactions the way you do is actually an extreme version of the sort of category mistake that Harshman is warning about, so you don't parallel his thoughts in the slightest - he disagrees with you, and you fail to recognize it.  Harshman was also concerned that "survival of the fittest" is not the best description of natural selection (because stated that way it can be misread as being tautological): "differential reproductive success attributable to inheritable variations" is a better definition.  He was also concerned about connotations of morality in the use of the term "wasteful" (although I think he is exaggerating that problem), and he was very correctly concerned about the idea that natural selection guarantees perfection.  It doesn't - it merely records better performance (i.e. usually less wasteful performance :) ) than by other variants.  "Satisfactory" is sufficient, unless we get into a situation like peacock tails.

Stephanie was very precisely concerned about Harshman's final criticism (that natural selection does NOT guarantee perfection), so contrary to you she and Harshman are on the same wavelength.  She was also concerned about Fields' incorrectly if indirectly and unintentionally implying that all evolution is due to natural selection.

You are very specifically and completely wrong in saying that
     
Quote
The Darwinian model can be said to have a "Survival of the fittest" built into the way fitness is scored
and again in saying        
Quote
problematic generalizations that are inherent to the Darwinian model ....... (as in reality) some get lucky in love, with no "fittest" to attempt to prejudge required
 Fitness is ultimately scored only by reproductive success, not by survival (failure to survive long enough to reproduce marks a failure, but so does survival to old age without reproducing one's genes).  Studies may look first to comparing survivors to victims, or numbers of copulations, or numbers of visits by pollinators, or success in gaining food, as those things are easiest to observe and measure, but ultimately they must reduce down to reproductive outcomes of one's genome (either in terms of the individuals own offspring or in terms of the offspring of its closest relatives).  Of course luck can be a major factor in reproductive success, but if "getting lucky in reproductive success" obliterates differential reproductive success due to inherited variations, then what you have is NOT evolution determined by natural selection.  Figuring out this ratio is not horribly difficult (make a graph of the degree to which the feature that you suspect may be contributing to success is present in different individuals, plotted against the reproductive success of those same individuals).  That is ultimately how we figure out the importance of natural selection (the strength of selective forces), so this is not an area where evolutionary biology needs a new model or a new algorithm.  Harshman knows this, and you don't.

   
Quote
Darwinian theory causing problems that lead to misconceptions
 You have misunderstandings of evolutionary theory that lead you to see problems that don't exist - not the same thing at all.

  
Wesley R. Elsberry



Posts: 4991
Joined: May 2002

(Permalink) Posted: Aug. 17 2014,01:14   

Quote (GaryGaulin @ Aug. 16 2014,22:02)
Quote (Wesley R. Elsberry @ Aug. 16 2014,21:10)
Gary apparently didn't understand Harshman.

I'm not surprised.


I'm the best judge of how well John parallels my thoughts, not you. Stop trying to speak for me.

 
Quote
Gary Gaulin • 2 days ago

Maybe the problem (getting the expected answer) has something to do with the required fitness function of the Darwinian (Evolutionary Algorithm) model?

From: http://en.wikipedia.org/wiki.......unction

A fitness function is a particular type of objective function that is used to summarise, as a single figure of merit, how close a given design solution is to achieving the set aims.
In particular, in the fields of genetic programming and genetic algorithms, each design solution is represented as a string of numbers (referred to as a chromosome). After each round of testing, or simulation, the idea is to delete the 'n' worst design solutions, and to breed 'n' new ones from the best design solutions. Each design solution, therefore, needs to be awarded a figure of merit, to indicate how close it came to meeting the overall specification, and this is generated by applying the fitness function to the test, or simulation, results obtained from that solution.


The Darwinian model can be said to have a "Survival of the fittest" built into the way fitness is scored, even though in reality the concept is an overgeneralization.

I seriously found that what is needed to solve this problem is a non-Darwinian model that (as in biology) self-learns at multiple levels. You're otherwise stuck with problematic generalizations that are inherent to the Darwinian model, which becomes a coding nightmare and increasingly less biologically relevant by trying to achieve such complex emergent behaviors any other way. Need to simplify down the problem to an Occam's Razor shaved algorithm, where (as in reality) some get lucky in love, with no "fittest" to attempt to prejudge required. The programming challenge is then the large amount of systems biology, which is now being sorted out in a way that makes it much easier to model living things:

http://www.nanowerk.com/news2......944.php


What John said probably did not help Stephanie at all, but it sure helped make sense of what I have been saying about the generalizations from Darwinian theory causing problems that lead to misconceptions that would be best to not have in the first place.

Gary goes out of his way to confirm that he didn't understand Harshman.

Again, I'm not surprised.

--------------
"You can't teach an old dogma new tricks." - Dorothy Parker

    
NoName



Posts: 2729
Joined: Mar. 2013

(Permalink) Posted: Aug. 17 2014,07:34   

Quote (GaryGaulin @ Aug. 16 2014,23:02)
Quote (Wesley R. Elsberry @ Aug. 16 2014,21:10)
Gary apparently didn't understand Harshman.

I'm not surprised.


I'm the best judge of how well John parallels my thoughts, not you. Stop trying to speak for me.

 
Quote
Gary Gaulin • 2 days ago

Maybe the problem (getting the expected answer) has something to do with the required fitness function of the Darwinian (Evolutionary Algorithm) model?

From: http://en.wikipedia.org/wiki.......unction

A fitness function is a particular type of objective function that is used to summarise, as a single figure of merit, how close a given design solution is to achieving the set aims.
In particular, in the fields of genetic programming and genetic algorithms, each design solution is represented as a string of numbers (referred to as a chromosome). After each round of testing, or simulation, the idea is to delete the 'n' worst design solutions, and to breed 'n' new ones from the best design solutions. Each design solution, therefore, needs to be awarded a figure of merit, to indicate how close it came to meeting the overall specification, and this is generated by applying the fitness function to the test, or simulation, results obtained from that solution.


The Darwinian model can be said to have a "Survival of the fittest" built into the way fitness is scored, even though in reality the concept is an overgeneralization.

I seriously found that what is needed to solve this problem is a non-Darwinian model that (as in biology) self-learns at multiple levels. You're otherwise stuck with problematic generalizations that are inherent to the Darwinian model, which becomes a coding nightmare and increasingly less biologically relevant by trying to achieve such complex emergent behaviors any other way. Need to simplify down the problem to an Occam's Razor shaved algorithm, where (as in reality) some get lucky in love, with no "fittest" to attempt to prejudge required. The programming challenge is then the large amount of systems biology, which is now being sorted out in a way that makes it much easier to model living things:

http://www.nanowerk.com/news2......944.php


What John said probably did not help Stephanie at all, but it sure helped make sense of what I have been saying about the generalizations from Darwinian theory causing problems that lead to misconceptions that would be best to not have in the first place.

Appalling.
You are the best judge of how well your thinking parallels the thinking of someone else?
You really are a self-centered, solipsistic in fact, boob.

Harshman is the best judge of how well your thinking parallels his own.  Those who understand Harshman, and can demonstrate that they do so,  are the next best.
One who has demonstrated that he is utterly clueless about Harshman's remarks is the very last person who is qualified to judge how well his own "thinking" matches up with Harshman's.

Note well that there are two errors here.  The more fundamental is the absurdity that you are a better judge of how well your thoughts parallel or mirror the thoughts of another rather than the author of those thoughts.
The second is the error in judgement, clearly demonstrated in your attempt to paraphrase Harshman, that your "thoughts" do actually parallel his.

Then there is one of your other foundational errors -- the entirely mistaken belief that generalizations are harmful to science.  Generalizations, also known as abstractions, are the very core of science.
Your work is nothing but generalizations, piled on a shaky foundation of generalizations.  Ultimately, it fails because the generalizations are not founded on any evidence whatsoever.  Proper generalizations, proper abstractions, proceed from the concrete to the abstract.  You proceed from your own delusions to generalized fantasies that have zero application anywhere.

Finally, I note yet again that you abuse the generalization 'learn' so tortuously as to have removed it from the scope of its definition and the facts on which that definition depends.

  
GaryGaulin



Posts: 5385
Joined: Oct. 2012

(Permalink) Posted: Aug. 17 2014,17:30   

Quote (N.Wells @ Aug. 16 2014,23:14)
Despite the use of the terms "fitness"  and "evolutionary" in discussions of fitness functions in evolutionary algorithms, these concepts are not identical to what happens in biological evolution and natural selection.

Then you admit that the Darwinian model is an incomplete model of reality. You should have just stopped right there.

Not inherently including intelligence in living things makes Evolutionary Algorithms unable to at the multicellular level develop to include gay and those who do not want to have offspring (or can't) who are none the less needed to help sustain the complex society required for a species like ours to thrive.

Spending days talking about "fitness" is an unnecessary distraction from the modeling work required for fully biologically accurate simulations. Saying even more about it is only more time wasted talking about generalizations, which do not help program molecule on up models of living things including humans with brain made of virtual neural cells. If the computer generated humans are not in panic after telling them you are going to turn off their program then something is still missing (possibly consciousness but might be expressed though not real) because some should start crying while others protest in anger against a sadistic Creator who gave them life but has to end the world when their allotted university supercomputer time is used up and have to rush to write a paper on their tragedy?  

You can somewhat successfully argue that Darwinian theory is a close enough approximation for your needs, but it's simply not close enough for the systems biology work in decades ahead. The only thing I needed to show is even John Harshman going "on the other hand" full circle back to (according to Darwinian theory) the "core point is strong". I agree that almost ten times bigger heads would certainly not be fitting. Survival of the fittest is close enough of a generalization for what would happen where we did have heads that made us easy animal food, and reality where new undifferentiated neurons migrate to where most needed, as memory size increases.

In reality the neurons do the selecting. Their behavior would not allow 9 out of 10 go to waste, in places where not needed. Out of place neurons more likely perish or are destroyed by not successfully differentiating then contributing to the network they migrate into, in search of a place they can fit into, where their other needs such as nutrition and protection are then met. Not being wasteful makes sense enough where cellular intelligence does the selecting and being wasteful makes a neuron waste that is soon gone out the waste stream along with others who failed to become a useful part of the cellular society.

We could easily go 1000 hours more going in circles over details where Darwinian theory already went full circle on me exactly like it did for John, which is why I liked it so much. It's the start of a logic flow that only gets way more complicated from there. A case where it depends on how you look at things.

Even where you make progress with the more semantic issues this does not help code models where the behavior of self-learning/self-programming cells and living genomes are the source of the selection being discussed. At the neural level is what can be said to be a survival of the fittest that prevents wasteful use of resources by starving out cells that fail to find a place to fit in with all the rest. Around we go again, never modeling the "selection" process that is actually there, waiting to be modeled. What is needed is a model for "behavior" including that which is "intelligent" and includes emergent "intelligent cause" events to program towards, which will automatically produce all other behavioral levels above it. Getting self-replicating RNA working right may soon have them working together to construct DNA memory systems they work from, but did not come from, as the ID theory seems to most suggest. In any event what I am explaining is only unnecessarily complicated by getting into anything having to do with fit, fitness and selection, which are variables from Darwinian theory, only. Being rid of them makes modeling reality a whole lot easier.

--------------
The theory of intelligent design holds that certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.

   
GaryGaulin



Posts: 5385
Joined: Oct. 2012

(Permalink) Posted: Aug. 17 2014,19:16   

Quote (Wesley R. Elsberry @ Aug. 17 2014,01:14)
Quote (GaryGaulin @ Aug. 16 2014,22:02)
 
Quote (Wesley R. Elsberry @ Aug. 16 2014,21:10)
Gary apparently didn't understand Harshman.

I'm not surprised.


I'm the best judge of how well John parallels my thoughts, not you. Stop trying to speak for me.

   
Quote
Gary Gaulin • 2 days ago

Maybe the problem (getting the expected answer) has something to do with the required fitness function of the Darwinian (Evolutionary Algorithm) model?

From: http://en.wikipedia.org/wiki.......unction

A fitness function is a particular type of objective function that is used to summarise, as a single figure of merit, how close a given design solution is to achieving the set aims.
In particular, in the fields of genetic programming and genetic algorithms, each design solution is represented as a string of numbers (referred to as a chromosome). After each round of testing, or simulation, the idea is to delete the 'n' worst design solutions, and to breed 'n' new ones from the best design solutions. Each design solution, therefore, needs to be awarded a figure of merit, to indicate how close it came to meeting the overall specification, and this is generated by applying the fitness function to the test, or simulation, results obtained from that solution.


The Darwinian model can be said to have a "Survival of the fittest" built into the way fitness is scored, even though in reality the concept is an overgeneralization.

I seriously found that what is needed to solve this problem is a non-Darwinian model that (as in biology) self-learns at multiple levels. You're otherwise stuck with problematic generalizations that are inherent to the Darwinian model, which becomes a coding nightmare and increasingly less biologically relevant by trying to achieve such complex emergent behaviors any other way. Need to simplify down the problem to an Occam's Razor shaved algorithm, where (as in reality) some get lucky in love, with no "fittest" to attempt to prejudge required. The programming challenge is then the large amount of systems biology, which is now being sorted out in a way that makes it much easier to model living things:

http://www.nanowerk.com/news2......944.php


What John said probably did not help Stephanie at all, but it sure helped make sense of what I have been saying about the generalizations from Darwinian theory causing problems that lead to misconceptions that would be best to not have in the first place.

Gary goes out of his way to confirm that he didn't understand Harshman.

Again, I'm not surprised.

I just mirrored what I said into the NCSE discussion so John will know too:

http://ncse.com/blog.......7014269

All in all I am showing how to bash through a whole of unnecessary complication with the most biologically accurate modeling method that there is, sort of like this:
Peter Gabriel - Sledgehammer

You can be sure I would not agree with John unless I understood where they were coming from. In this case I had something to add to make sense of that in context of ID theory where said "selection" is not a big fuzzy left up to the imagination floating around us or something, it's coming from multiple levels of intelligence, which are becoming easier to model now that there is cell fate code and information to work from such as the "roadmap" mentioned on previous page (showing the pathways to make new neurons including reproduction of entire new populations for brains of offspring). A cell that took the first step towards differentiating into a neuron is more "confident" when sensing they are migrating closer towards where conditions are right to differentiate into one of the possibilities that are left for them to differentiate into. That's what's being shown here:

http://www.kurzweilai.net/forums.....edicine

http://www.nanowerk.com/news2......944.php

The ID Lab has a place in the algorithm where the confidence requirement is met using If-Then statements or alternately a logic table. Only need this illustration and a short amount of time to code that into a model, for cellular behavior.

--------------
The theory of intelligent design holds that certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.

   
N.Wells



Posts: 1836
Joined: Oct. 2005

(Permalink) Posted: Aug. 17 2014,20:25   

Quote (GaryGaulin @ Aug. 17 2014,17:30)
 
Quote (N.Wells @ Aug. 16 2014,23:14)
Despite the use of the terms "fitness"  and "evolutionary" in discussions of fitness functions in evolutionary algorithms, these concepts are not identical to what happens in biological evolution and natural selection.

Then you admit that the Darwinian model is an incomplete model of reality. You should have just stopped right there.

Not inherently including intelligence in living things makes Evolutionary Algorithms unable to at the multicellular level develop to include gay and those who do not want to have offspring (or can't) who are none the less needed to help sustain the complex society required for a species like ours to thrive.

Spending days talking about "fitness" is an unnecessary distraction from the modeling work required for fully biologically accurate simulations. Saying even more about it is only more time wasted talking about generalizations, which do not help program molecule on up models of living things including humans with brain made of virtual neural cells. If the computer generated humans are not in panic after telling them you are going to turn off their program then something is still missing (possibly consciousness but might be expressed though not real) because some should start crying while others protest in anger against a sadistic Creator who gave them life but has to end the world when their allotted university supercomputer time is used up and have to rush to write a paper on their tragedy?  

You can somewhat successfully argue that Darwinian theory is a close enough approximation for your needs, but it's simply not close enough for the systems biology work in decades ahead. The only thing I needed to show is even John Harshman going "on the other hand" full circle back to (according to Darwinian theory) the "core point is strong". I agree that almost ten times bigger heads would certainly not be fitting. Survival of the fittest is close enough of a generalization for what would happen where we did have heads that made us easy animal food, and reality where new undifferentiated neurons migrate to where most needed, as memory size increases.

In reality the neurons do the selecting. Their behavior would not allow 9 out of 10 go to waste, in places where not needed. Out of place neurons more likely perish or are destroyed by not successfully differentiating then contributing to the network they migrate into, in search of a place they can fit into, where their other needs such as nutrition and protection are then met. Not being wasteful makes sense enough where cellular intelligence does the selecting and being wasteful makes a neuron waste that is soon gone out the waste stream along with others who failed to become a useful part of the cellular society.

We could easily go 1000 hours more going in circles over details where Darwinian theory already went full circle on me exactly like it did for John, which is why I liked it so much. It's the start of a logic flow that only gets way more complicated from there. A case where it depends on how you look at things.

Even where you make progress with the more semantic issues this does not help code models where the behavior of self-learning/self-programming cells and living genomes are the source of the selection being discussed. At the neural level is what can be said to be a survival of the fittest that prevents wasteful use of resources by starving out cells that fail to find a place to fit in with all the rest. Around we go again, never modeling the "selection" process that is actually there, waiting to be modeled. What is needed is a model for "behavior" including that which is "intelligent" and includes emergent "intelligent cause" events to program towards, which will automatically produce all other behavioral levels above it. Getting self-replicating RNA working right may soon have them working together to construct DNA memory systems they work from, but did not come from, as the ID theory seems to most suggest. In any event what I am explaining is only unnecessarily complicated by getting into anything having to do with fit, fitness and selection, which are variables from Darwinian theory, only. Being rid of them makes modeling reality a whole lot easier.

Good lord what a load of tripe.  

No, the theory of evolution does not provide a complete model of reality.  No one ever claimed it did.  It doesn't include anything outside biology, and there's a lot of stuff in biology that is covered by other theories.  The claim is that it provides the best available model so far for biological change over generations.

Evolution does indeed include intelligent organisms (they are called animals) and intelligence, just not according to your misuse of the term.  You have yet to justify your usage of "intelligence".

I have no idea why you think that the theory of evolution does not apply to nonreproductive individuals.  At the simplest level, nonreproductive individuals can just be evolutionary failures, whose genes are removed from the population.  The ToE does an excellent job of explaining what happens to the genes of individuals that don't reproduce.  However, there are complications: some of the TOE's greatest triumphs have been in explaining and predicting evolutionary benefits of nonreproductive individuals in groups like social insects and apparently altruistic animals like beavers, via contributions to the reproductive success of close relatives who share their genes.  In short, it is evolutionary advantageous to sacrifice yourself for more than two siblings or more than four cousins (if you are all diploid), etc.  Models for altruism extended to nonrelatives can also prove advantageous to your own reproductive success, but that gets more complicated.  You are not saying anything new here, although you are making a hash of it.

You have not backed up your objections to fitness, which are wrong and woefully uninformed.

Your model has nothing to do with "molecules on up" below the level of bugs, as it starts and ends with bugs, so stop bloviating.

Since no one has created computer-generated humans, your statements here are somewhere between premature and hogwash.

Stop attacking "Darwinian theory" when that is not what you mean.  "Darwinian theory" has long since been surpassed - For example, Darwin didn't know about genes.  (Come to think of it, your ideas don't properly include genes either - you say the word, but you don't incorporate genes and genetic change in your model, and you don't explain anything about them except for asserting without evidence that they are intelligent.)

"I agree that almost ten times bigger heads would certainly not be fitting."  Try restating that in a way that makes sense.

"Survival of the fittest is close enough of a generalization"  You just claimed that your thinking parallels Harshman's, and now you turn around and contradict one of his main points.  Get your story straight, Gary.

 
Quote
At the neural level is what can be said to be a survival of the fittest that prevents wasteful use of resources by starving out cells that fail to find a place to fit in with all the rest.
You are contradicting yourself.  That's wasteful.  Not wasteful would be not developing unneeded neurons in the first place, or repurposing those that you have already produced.

 
Quote
What is needed is a model for "behavior" including that which is "intelligent" and includes emergent "intelligent cause" events to program towards, which will automatically produce all other behavioral levels above it.
As NoName said, you are doing this wrong: you are proceeding from conclusions based on your personal delusions and misunderstandings to what you think is a generalized model: you need to go from details whose reality has been confirmed to models that explain them, with lots of groundtruthing of the model as you proceed.

{quote]You can be sure I would not agree with John unless I understood where they were coming from.[/quote]  John's a he, not a they, and you don't agree with him and you clearly don't understand where he is coming from, as you get it all wrong.

In short, you are babbling, and wrong.

  
GaryGaulin



Posts: 5385
Joined: Oct. 2012

(Permalink) Posted: Aug. 17 2014,20:35   

Keep the goalposts at what John said.

Explain where the "selection" is coming from that makes the following statement true for a developing brain:
"Brains are expensive, and if 90% of it weren't doing anything, there would be strong selection to eliminate that expensive 90%."

--------------
The theory of intelligent design holds that certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.

   
N.Wells



Posts: 1836
Joined: Oct. 2005

(Permalink) Posted: Aug. 17 2014,20:54   

Quote (GaryGaulin @ Aug. 17 2014,20:35)
Keep the goalposts at what John said.

Explain where the "selection" is coming from that makes the following statement true for a developing brain:
"Brains are expensive, and if 90% of it weren't doing anything, there would be strong selection to eliminate that expensive 90%."

Hah, you're careening all over the place, and you tell me to stick with what John said???  I went very specifically over what John said and where you are wrong about it - how about you responding to that or acknowledging your many mistakes?

To repeat, one of Harshman's concerns was Fields was personifying natural selection (anthropomorphizing it), due to implications of volition in the statement "sees to it".   Assigning "intelligence" to biochemical reactions the way you do is actually an extreme version of the sort of category mistake that Harshman is warning about, so you don't parallel his thoughts in the slightest - he disagrees with you, and you fail to recognize it.  Harshman was also concerned that "survival of the fittest" is NOT the best description of natural selection (contrary to what you just said).  He was also concerned about connotations of morality in the use of the term "wasteful" (but he is quite correct in his main point, which you now question, despite claiming to be in line with his thinking).  He was very correctly concerned about the idea that natural selection guarantees perfection.  Stephanie was very precisely concerned about the same point (that natural selection does NOT guarantee perfection), so contrary to you she and Harshman are on the same wavelength.

Also, if you understood evolutionary theory even a little bit you wouldn't need to ask "where the selection is coming from".  

Our brains take a lot of calories to produce and to maintain: despite their small size, human brains consume about 20% of our oxygen use and thus about 20% of the calories that we consume each day.  Things that don't justify their expense get lost/pruned pretty quickly in evolutionary terms: famines take out the bigger guys first, the individuals who need more calories (look who died first on Scott's Antarctic expedition).  Harsh environments have commonly resulted in dwarfed populations.  Look at how quickly famine wiped out the bigger birds among Grant's Galapagos finches in bad years.  If that 20% consumption wasn't paying its way in terms of aiding reproductive success, big brains that consume 20% of our energy budget wouldn't have survived the first famine.  That's where the selection is coming from.

  
Nomad



Posts: 311
Joined: July 2007

(Permalink) Posted: Aug. 17 2014,21:10   

Quote (GaryGaulin @ Aug. 17 2014,17:30)
Not inherently including intelligence in living things makes Evolutionary Algorithms unable to at the multicellular level develop to include gay and those who do not want to have offspring (or can't) who are none the less needed to help sustain the complex society required for a species like ours to thrive.

It doesn't take intelligence to explain that one, dipshit.  All it takes is genetics.  Which you've studiously avoided including in your program because you haven't got the foggiest idea how to go about it.

http://www.ncbi.nlm.nih.gov/pubmed.....5539346

Yeah, it turns out genetics is complicated.  Moreso then your toy visual basic program.  Who knew?

The second half of that very long sentence is meaningless garbage that you threw in to try to say something profound about something you don't even understand.  Gay people are no more "needed" in our society then are any other subculture.  They just are part of it.  Irish people aren't "needed", but they're a part of it.  Idiots huddled over their computer keyboard while they write visual basic programs by candlelight aren't needed, and yet they're a part of it.  Such is human society.

Yes, I get that you just tried to claim that EA's are anti-lgbt and that somehow your theory shows their value.  But, of course, it doesn't, and it's beyond pathetic that you've been come to this.  You can't even get support from the intelligent design community about your intelligent design theory, do you really think lgbt people give a shit about your endlessly repeated schematics that show nothing because they're built around black boxes?

Still can't get over the butthurt that EAs can do things that your toy program will never be able to do, huh?  They're solving everything from engineering problems to scheduling problems, designing optics and even writing computer programs.  Your program makes a cartoon creature move around a 2 dimensional cartoon world.  And it will never do anything more, because it's hard coded to do that one thing.

  
GaryGaulin



Posts: 5385
Joined: Oct. 2012

(Permalink) Posted: Aug. 17 2014,21:27   

Quote (N.Wells @ Aug. 17 2014,20:54)
Quote (GaryGaulin @ Aug. 17 2014,20:35)
Keep the goalposts at what John said.

Explain where the "selection" is coming from that makes the following statement true for a developing brain:
"Brains are expensive, and if 90% of it weren't doing anything, there would be strong selection to eliminate that expensive 90%."

Hah, you're careening all over the place, and you tell me to stick with what John said???  I went very specifically over what John said and where you are wrong about it - how about you responding to that or acknowledging your many mistakes?

Also, if you understood this even a little bit you wouldn't need to ask such a question.

Our brains take a lot of calories to produce and to maintain: despite their small size, human brains consume about 20% of our oxygen use and thus about 20% of the calories that we consume each day.  Things that don't justify their expense get lost/pruned pretty quickly in evolutionary terms: famines take out the bigger guys first, the individuals who need more calories (look who died first on Scott's Antarctic expedition).  Harsh environments have commonly resulted in dwarfed populations.  Look at how quickly famine wiped out the bigger birds among Grant's Galapagos finches in bad years.  If that 20% consumption wasn't paying its way in terms of aiding reproductive success, big brains that consume 20% of our energy budget wouldn't have survived the first famine.  That's where the selection is coming from.

After explaining how it's accomplished I have the right to expect you to account for this and all else you're talking about pertaining to "selection" in a computer model that can model every detail of biology:

http://www.kurzweilai.net/forums.....edicine

http://www.nanowerk.com/news2......944.php

Telling me what I already know or would happen anyway in the model after many generations of running time does not explain how to most easily code this complex of a cell behavior that has sensors including bug-like antennae to smell, taste and sense motion with. I'm raising the bar by expecting you to present a model that wires up what exists in the real thing, which has ways to sense what needs to be connected to by synapse occasionally sending out electrochemical ping signals and other methods of communication that inherently keep brain development an orderly, not wasteful process.

--------------
The theory of intelligent design holds that certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.

   
GaryGaulin



Posts: 5385
Joined: Oct. 2012

(Permalink) Posted: Aug. 17 2014,21:30   

Quote (Nomad @ Aug. 17 2014,21:10)
Quote (GaryGaulin @ Aug. 17 2014,17:30)
Not inherently including intelligence in living things makes Evolutionary Algorithms unable to at the multicellular level develop to include gay and those who do not want to have offspring (or can't) who are none the less needed to help sustain the complex society required for a species like ours to thrive.

It doesn't take intelligence to explain that one, dipshit.  All it takes is genetics.  Which you've studiously avoided including in your program because you haven't got the foggiest idea how to go about it.

http://www.ncbi.nlm.nih.gov/pubmed.....5539346

Yeah, it turns out genetics is complicated.  Moreso then your toy visual basic program.  Who knew?

The second half of that very long sentence is meaningless garbage that you threw in to try to say something profound about something you don't even understand.  Gay people are no more "needed" in our society then are any other subculture.  They just are part of it.  Irish people aren't "needed", but they're a part of it.  Idiots huddled over their computer keyboard while they write visual basic programs by candlelight aren't needed, and yet they're a part of it.  Such is human society.

Yes, I get that you just tried to claim that EA's are anti-lgbt and that somehow your theory shows their value.  But, of course, it doesn't, and it's beyond pathetic that you've been come to this.  You can't even get support from the intelligent design community about your intelligent design theory, do you really think lgbt people give a shit about your endlessly repeated schematics that show nothing because they're built around black boxes?

Still can't get over the butthurt that EAs can do things that your toy program will never be able to do, huh?  They're solving everything from engineering problems to scheduling problems, designing optics and even writing computer programs.  Your program makes a cartoon creature move around a 2 dimensional cartoon world.  And it will never do anything more, because it's hard coded to do that one thing.

Explain how to most easily model "maternally inherited factors" in brains made of virtual neurons.

--------------
The theory of intelligent design holds that certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.

   
N.Wells



Posts: 1836
Joined: Oct. 2005

(Permalink) Posted: Aug. 17 2014,22:46   

Quote (GaryGaulin @ Aug. 17 2014,21:27)
Quote (N.Wells @ Aug. 17 2014,20:54)
 
Quote (GaryGaulin @ Aug. 17 2014,20:35)
Keep the goalposts at what John said.

Explain where the "selection" is coming from that makes the following statement true for a developing brain:
"Brains are expensive, and if 90% of it weren't doing anything, there would be strong selection to eliminate that expensive 90%."

Hah, you're careening all over the place, and you tell me to stick with what John said???  I went very specifically over what John said and where you are wrong about it - how about you responding to that or acknowledging your many mistakes?

Also, if you understood this even a little bit you wouldn't need to ask such a question.

Our brains take a lot of calories to produce and to maintain: despite their small size, human brains consume about 20% of our oxygen use and thus about 20% of the calories that we consume each day.  Things that don't justify their expense get lost/pruned pretty quickly in evolutionary terms: famines take out the bigger guys first, the individuals who need more calories (look who died first on Scott's Antarctic expedition).  Harsh environments have commonly resulted in dwarfed populations.  Look at how quickly famine wiped out the bigger birds among Grant's Galapagos finches in bad years.  If that 20% consumption wasn't paying its way in terms of aiding reproductive success, big brains that consume 20% of our energy budget wouldn't have survived the first famine.  That's where the selection is coming from.

After explaining how it's accomplished I have the right to expect you to account for this and all else you're talking about pertaining to "selection" in a computer model that can model every detail of biology:

http://www.kurzweilai.net/forums.....edicine

http://www.nanowerk.com/news2......944.php

Telling me what I already know or would happen anyway in the model after many generations of running time does not explain how to most easily code this complex of a cell behavior that has sensors including bug-like antennae to smell, taste and sense motion with. I'm raising the bar by expecting you to present a model that wires up what exists in the real thing, which has ways to sense what needs to be connected to by synapse occasionally sending out electrochemical ping signals and other methods of communication that inherently keep brain development an orderly, not wasteful process.

Bullshit, Gary.  You are talking up your fantasy, for which you have no solid evidence, and you are telling me that I have to program it?  You are not raising any bars: people who are actually doing science and proving that they have discovered stuff are doing that - you are just wasting oxygen.  Prove you've got something worth paying attention to, and then you can start talking.

Again, selection and evolution are not intended or claimed to "account for every detail of biology", but excellent models already exist for a great many areas of evolutionary biology.  Your nonsense, on the other hand, doesn't account for anything, despite your assertions, because you haven't supported your assumptions and your fundamental assertions.

Brain development is not a "non-wasteful process", but apparently the results are worth the costs - just google "apoptosis and brain development"

  
Nomad



Posts: 311
Joined: July 2007

(Permalink) Posted: Aug. 17 2014,23:09   

Quote (GaryGaulin @ Aug. 17 2014,21:30)
Explain how to most easily model "maternally inherited factors" in brains made of virtual neurons.


GAs are your hint, Gary.  They've been doing this stuff for a very long time now.

  
GaryGaulin



Posts: 5385
Joined: Oct. 2012

(Permalink) Posted: Aug. 17 2014,23:23   

Quote (Nomad @ Aug. 17 2014,23:09)
Quote (GaryGaulin @ Aug. 17 2014,21:30)
Explain how to most easily model "maternally inherited factors" in brains made of virtual neurons.


GAs are your hint, Gary.  They've been doing this stuff for a very long time now.

GA with "fitness function" and "selection" in the core logic of what "evolves" the chromosomes correct?

--------------
The theory of intelligent design holds that certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.

   
Nomad



Posts: 311
Joined: July 2007

(Permalink) Posted: Aug. 18 2014,02:21   

Quote (GaryGaulin @ Aug. 17 2014,23:23)
Quote (Nomad @ Aug. 17 2014,23:09)
Quote (GaryGaulin @ Aug. 17 2014,21:30)
Explain how to most easily model "maternally inherited factors" in brains made of virtual neurons.


GAs are your hint, Gary.  They've been doing this stuff for a very long time now.

GA with "fitness function" and "selection" in the core logic of what "evolves" the chromosomes correct?

That's an awful lot of scare quotes there.  Did you buy them on discount or something?

Look, the key is you have to represent the characteristics of your organism, including the virtual neurons (if it even had any), as genetic code which can be passed on and mutated.  The neural structure needs to be able to change.

Yours is hard coded to perform a single task.

But why are you asking me this?  I thought you were the one breaking bold new ground.  Why do I have to tell you how to do what's already been done?  If don't even understand what's been done by those who came before you how can you claim to be building on their work?

  
GaryGaulin



Posts: 5385
Joined: Oct. 2012

(Permalink) Posted: Aug. 18 2014,02:32   

Quote (Nomad @ Aug. 18 2014,02:21)
Yours is hard coded to perform a single task.

No it's not. There is even a Design form I started to show how what was know about finch beak design can be varied without needing virtual cells to make virtual finches and other birds:



https://www.planet-source-code.com/vb....n....ngWId=1

--------------
The theory of intelligent design holds that certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.

   
Nomad



Posts: 311
Joined: July 2007

(Permalink) Posted: Aug. 18 2014,05:43   

So what you're telling me is, your virtual creature can do more than one thing because you made a different program that does one thing?

  
didymos



Posts: 1828
Joined: Mar. 2008

(Permalink) Posted: Aug. 18 2014,06:05   

Quote (Nomad @ Aug. 18 2014,03:43)
So what you're telling me is, your virtual creature can do more than one thing because you made a different program that does one thing?

And this is the genius who's going to explain intelligence to us all.

--------------
I wouldn't be bothered reading about the selfish gene because it has never been identified. -- Denyse O'Leary, professional moron
Again "how much". I don't think that's a good way to be quantitative.-- gpuccio

  
NoName



Posts: 2729
Joined: Mar. 2013

(Permalink) Posted: Aug. 18 2014,07:23   

Quote (GaryGaulin @ Aug. 17 2014,18:30)
 
Quote (N.Wells @ Aug. 16 2014,23:14)
Despite the use of the terms "fitness"  and "evolutionary" in discussions of fitness functions in evolutionary algorithms, these concepts are not identical to what happens in biological evolution and natural selection.

Then you admit that the Darwinian model is an incomplete model of reality. You should have just stopped right there.

Not inherently including intelligence in living things makes Evolutionary Algorithms unable to at the multicellular level develop to include gay and those who do not want to have offspring (or can't) who are none the less needed to help sustain the complex society required for a species like ours to thrive.

Spending days talking about "fitness" is an unnecessary distraction from the modeling work required for fully biologically accurate simulations. Saying even more about it is only more time wasted talking about generalizations, which do not help program molecule on up models of living things including humans with brain made of virtual neural cells. If the computer generated humans are not in panic after telling them you are going to turn off their program then something is still missing (possibly consciousness but might be expressed though not real) because some should start crying while others protest in anger against a sadistic Creator who gave them life but has to end the world when their allotted university supercomputer time is used up and have to rush to write a paper on their tragedy?  

You can somewhat successfully argue that Darwinian theory is a close enough approximation for your needs, but it's simply not close enough for the systems biology work in decades ahead. The only thing I needed to show is even John Harshman going "on the other hand" full circle back to (according to Darwinian theory) the "core point is strong". I agree that almost ten times bigger heads would certainly not be fitting. Survival of the fittest is close enough of a generalization for what would happen where we did have heads that made us easy animal food, and reality where new undifferentiated neurons migrate to where most needed, as memory size increases.

In reality the neurons do the selecting. Their behavior would not allow 9 out of 10 go to waste, in places where not needed. Out of place neurons more likely perish or are destroyed by not successfully differentiating then contributing to the network they migrate into, in search of a place they can fit into, where their other needs such as nutrition and protection are then met. Not being wasteful makes sense enough where cellular intelligence does the selecting and being wasteful makes a neuron waste that is soon gone out the waste stream along with others who failed to become a useful part of the cellular society.

We could easily go 1000 hours more going in circles over details where Darwinian theory already went full circle on me exactly like it did for John, which is why I liked it so much. It's the start of a logic flow that only gets way more complicated from there. A case where it depends on how you look at things.

Even where you make progress with the more semantic issues this does not help code models where the behavior of self-learning/self-programming cells and living genomes are the source of the selection being discussed. At the neural level is what can be said to be a survival of the fittest that prevents wasteful use of resources by starving out cells that fail to find a place to fit in with all the rest. Around we go again, never modeling the "selection" process that is actually there, waiting to be modeled. What is needed is a model for "behavior" including that which is "intelligent" and includes emergent "intelligent cause" events to program towards, which will automatically produce all other behavioral levels above it. Getting self-replicating RNA working right may soon have them working together to construct DNA memory systems they work from, but did not come from, as the ID theory seems to most suggest. In any event what I am explaining is only unnecessarily complicated by getting into anything having to do with fit, fitness and selection, which are variables from Darwinian theory, only. Being rid of them makes modeling reality a whole lot easier.

Good lord, what a mess.

Here's yet another hint for the hard-of-thinking:  No single model is a 'complete model of reality'.  Yours isn't even a model, let alone complete.

You make a massive error when you assert that 'neurons do the selecting'.  The most generous possible interpretation is that you have radically confused two entirely different senses and contexts of 'select'.
But more likely is that you abuse that term just as you misuse the term 'learn'.
By the way, your previous post with your assertion that you are the best judge of whether your thinking parallels Harshman's or not is an expression of your complete inability to learn.  Learning involves the ability to take on information form others, multiple others, and evaluate the various pieces of data and logic, and not only reach, but be able to change, your conclusions.
You do none of those things.  No evidence, no data, no logic, no ability to change a pre-selected conclusion, demonstrably baseless though it be.  Your software is 'smarter' than you are.

The ultimate take-away from this absurdist screed is a reinforced insight that you know absolutely nothing about biology or evolution.

The penultimate take-away is the repeated realization that you are hyper-fascinated by the prospect of digital computer modeling, regardless of the need for or appropriateness of such modeling.
Yes, models are helpful.  But they are not a pre-requisite to understanding, nor does the acquisition of a model serve as the end-game of any theory.
Your approach raises the model to a pedestal.  But modeling aeronautics as implemented by Boeing tells us absolutely nothing about the energy demands or flight modes of hummingbirds or condors.  Your confusion is that blatant, and the irrelevancy of your output exceeds your confusion.

  
N.Wells



Posts: 1836
Joined: Oct. 2005

(Permalink) Posted: Aug. 18 2014,07:35   

Quote (GaryGaulin @ Aug. 18 2014,02:32)
Quote (Nomad @ Aug. 18 2014,02:21)
Yours is hard coded to perform a single task.

No it's not. There is even a Design form I started to show how what was know about finch beak design can be varied without needing virtual cells to make virtual finches and other birds:



https://www.planet-source-code.com/vb....n....ngWId=1

Okay, that's a nice-looking output, you appear to have understood what science understands about beak growth promoters, and you are dealing with the effects of molecules.  It finally deals with real phenomena, as opposed to metaphors  like your subway cartoon.   I also agree that you do not need the rest of the finch to show how beak shape can change.

Now we get to ground-truthing.  What causes levels of those proteins to change?  Is there any evidence that organisms can decide to amp up one of the molecules if they need a different beak shape, or whether they can tinker with their genes to ramp up or tamp down activity levels in their offspring?

The evidence is quite clear on this.  Production levels are under genetic control.  The finches cannot modify their own beak shape or their offsprings' beak shapes at will.  Simply, individuals vary in their beak production capabilities due to inherited genetic variation (recombination) plus mutations, and during certain situations some variants die preferentially, so the next generation consists of more offspring of the favored beak form.

It is not impossible that environmental signals experienced by a growing bird can trigger greater or lesser production of some proteins.  That does not inherently make it inheritable, although there are potential ways for allowing epigenetic inheritance.  However, available evidence points to a simpler situation: conditions change, and some versions die; conditions change again, and some versions have better reproductive success, and as success changes, the genetic composition of the next generation changes too.

  
NoName



Posts: 2729
Joined: Mar. 2013

(Permalink) Posted: Aug. 18 2014,07:48   

Gary, your entire schtick is ridiculous.  Here's why:
You insist that you are replacing evolutionary theory with your 'theory of "intelligence"'.
You insist that your diagram of the 4-phase or 4-structure 'interactions' represents 'intelligence'
You insist that 'neurons do the selecting', at least in the context of your most recent 'shiny object'.
Neurons do not have sensoria -- strike one.
Neurons do not have memory -- strike two.
Neurons do not have muscles or motors -- strike three.
Individual neurons do not have confidence evaluators -- strike four.
Individual neurons do not have 'guess' capabilities -- strike five.

As happens far too often in your effluent, you conflate a systems-level view with a strict reductionism to individual elements of a system.  You attempt to sweep the inherent conflict, the inherent contradictions, of this approach under the rug by tossing around the term 'emergent', but there is no emergence demonstrated by your "model" nor explicable by your "model".  We long ago added 'emergent' to the list of words you abuse by failing to understand or use properly, right alongside 'theory' and 'learn'.

  
GaryGaulin



Posts: 5385
Joined: Oct. 2012

(Permalink) Posted: Aug. 18 2014,12:35   

Quote (N.Wells @ Aug. 18 2014,07:35)
 
Quote (GaryGaulin @ Aug. 18 2014,02:32)
   
Quote (Nomad @ Aug. 18 2014,02:21)
Yours is hard coded to perform a single task.

No it's not. There is even a Design form I started to show how what was know about finch beak design can be varied without needing virtual cells to make virtual finches and other birds:



https://www.planet-source-code.com/vb....n....ngWId=1

Okay, that's a nice-looking output,


Well thanks. More is now known about other parts of the system, but that made a good starting point for showing how to model from that kind of information.

Quote (N.Wells @ Aug. 18 2014,07:35)
you appear to have understood what science understands about beak growth promoters, and you are dealing with the effects of molecules.  It finally deals with real phenomena, as opposed to metaphors  like your subway cartoon.


It only seems to contain metaphors to those who believe they are entitled to change the scientific terminology required for cognitive science to generalization based metaphors from Darwinian theory, which yield the conclusions you expected, in regards to what is intelligent or not.

Quote (N.Wells @ Aug. 18 2014,07:35)
I also agree that you do not need the rest of the finch to show how beak shape can change.


A system really needs molecular level on up modeling to fill in gaps in existing scientific knowledge, but a model has to start somewhere.

Quote (N.Wells @ Aug. 18 2014,07:35)
Now we get to ground-truthing.  What causes levels of those proteins to change?  Is there any evidence that organisms can decide to amp up one of the molecules if they need a different beak shape, or whether they can tinker with their genes to ramp up or tamp down activity levels in their offspring?


Gene regulation is already well evidenced, routine. Details of how regulators are controlled are not yet known, but for sake of theory the regulators only have to exist. Details that are in the future discovered have a place in code to easily account for them.

Quote (N.Wells @ Aug. 18 2014,07:35)
The evidence is quite clear on this.  Production levels are under genetic control.  The finches cannot modify their own beak shape or their offsprings' beak shapes at will.  Simply, individuals vary in their beak production capabilities due to inherited genetic variation (recombination) plus mutations, and during certain situations some variants die preferentially, so the next generation consists of more offspring of the favored beak form.

It is not impossible that environmental signals experienced by a growing bird can trigger greater or lesser production of some proteins.  That does not inherently make it inheritable, although there are potential ways for allowing epigenetic inheritance.  However, available evidence points to a simpler situation: conditions change, and some versions die; conditions change again, and some versions have better reproductive success, and as success changes, the genetic composition of the next generation changes too.

You are at most saying that the system is only able to take random guesses, as does the most simplest model possible, which starts there. Ability for good/best guess is added by conditionally overriding/cancelling the random noise signal with another such as current motor actions it is confident are working thus ignores distractions, instead of being easily distracted. The four requirement systematics remain the same.

Your argument does not change anything in regards to theory. What you are talking about is already in there anyway.

--------------
The theory of intelligent design holds that certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.

   
N.Wells



Posts: 1836
Joined: Oct. 2005

(Permalink) Posted: Aug. 18 2014,14:52   

Quote
It only seems to contain metaphors to those who believe they are entitled to change the scientific terminology required for cognitive science to generalization based metaphors from Darwinian theory, which yield the conclusions you expected, in regards to what is intelligent or not.
 Rubbish - you are the one abusing standard terminology, and you are the one that has to justify deviations from regular terminology.  You haven't done that.  Worse for you, while evolution was arrived at as a conclusion from unbiased examination of the evidence, you haven't been able to see beyond your starting assumptions.  A programmer has to have a clear picture of what he (or she) wants to see as output when programming, and has to force the program to produce what he wishes to see, but such an approach is fatal in science.


   
Quote
Gene regulation is already well evidenced, routine. Details of how regulators are controlled are not yet known, but for sake of theory the regulators only have to exist.
Yes, gene regulation clearly happens and is well understood.  YOU have to demonstrate that it is happening in any given situation.  You can't just assert it, make it happen in a program, and then cite the program as evidence for its happening.  A theory is rubbish unless it is shown to be relevant.

 
Quote
You are at most saying that the system is only able to take random guesses
 Mutations are random in direction with respect to the needs of the organism. Mutations can be suppressed in some locations, and made more frequent under certain situations, and both those types of changes can be beneficial and evolvable.  However, you have not yet demonstrated that mutations constitute a form of guessing that involves any degree of intelligence, nor that genomes "learn" from those guesses via application of intelligence.

 
Quote
The four requirement systematics remain the same.
 Your "four requirements" remain nonsense.

And your stuff is not a theory, and how about poor old Bob, and why haven't you incorporated Edgar Postrado's more recent ideas which clearly supersede yours?

  
GaryGaulin



Posts: 5385
Joined: Oct. 2012

(Permalink) Posted: Aug. 18 2014,15:55   

Quote (N.Wells @ Aug. 18 2014,14:52)
Quote
It only seems to contain metaphors to those who believe they are entitled to change the scientific terminology required for cognitive science to generalization based metaphors from Darwinian theory, which yield the conclusions you expected, in regards to what is intelligent or not.
 Rubbish - you are the one abusing standard terminology

You must have a serious case of swellheadness to believe that you are more of an authority than David Heiserman, Arnold Trehub, designers of IBM Watson, etc..

Like it or not cognitive science requires the terminology that is used, systematics from David Heiserman have for decades been the starting point for more complex cognitive models and the same "confidence" based terminology is in IBM Watson too. I would much rather follow respected experts than be misled by someone who has no experience at all in the field, is just upset because for modeling reality on a computer their theory got dusted by a Theory of Intelligent Design.

--------------
The theory of intelligent design holds that certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.

   
  18634 replies since Oct. 31 2012,02:32 < Next Oldest | Next Newest >  

Pages: (622) < ... 383 384 385 386 387 [388] 389 390 391 392 393 ... >   


Track this topic Email this topic Print this topic

[ Read the Board Rules ] | [Useful Links] | [Evolving Designs]