RSS 2.0 Feed

» Welcome Guest Log In :: Register

Pages: (12) < [1] 2 3 4 5 6 ... >   
  Topic: Thread for Cryptoguru, Evolution, Evolutionary Computing, etc< Next Oldest | Next Newest >  
OgreMkV



Posts: 3668
Joined: Oct. 2009

(Permalink) Posted: Jan. 16 2015,08:04   

This is a thread for a creationist that has been visiting my blog. You can review his comments here: http://www.skepticink.com/smilodo....omputer

He seems to have some questions that he can't get answers to. Feel free to chime in.

He has stated that he will come by. We'll see.

--------------
Ignored by those who can't provide evidence for their claims.

http://skepticink.com/smilodo....retreat

   
cryptoguru



Posts: 53
Joined: Jan. 2015

(Permalink) Posted: Jan. 16 2015,09:09   

Hi Kevin,

thanks for the introduction.

I'm going to pose the following definition/algorithm of biological evolution, please feel free to disagree and post what you think it actually is.

Here are some assumptions:
1) DNA is a self-replicating computer code, a list of instructions that is used by a cell to build a specific organism by a very complex and highly non-linear process that involves transcription and translation.
2) DNA is randomly mutated at the level of nucleotides (i.e. the letters of the programming code, not the functional/gene level)
3) Natural selection is not an intelligent entity, therefore it cannot evaluate the performance of an organism. It is the name given to the process whereby more suitable organisms survive and others die ... achieved by a selective pressure being applied to the environment (e.g. availability of food/water, predators etc.)

Evolution is:
1) take an organism and let it replicate
2) random mutations should be introduced to the offspring DNA.
3) a selective pressure is applied to the environment the organisms live in
4) less fit organisms will die before replicating due to the pressure
5) most fit organisms will continue to replicate and mutate
6) eventually new functions and traits will emerge

Biological Evolution is presented as a scientific law, it is simple to define, so should be simple to model

My challenge is this.
Show me a computer model that models this process and shows new functions and information being added to the genome.

caveats:
- Mutation can't simply combine functions in different arrangements. Mutation must occur on the code that defines the functions themselves, and affect the ability of the organism to build itself, to function and to replicate.
- the selective process can't include measuring properties of the organism to remove them from the population, the selective process must rely on organisms dying off due to the environmental factors. (e.g. energy)

It may sound subtle ... but breaking these laws means that you're not modelling biological evolution, you're performing functional optimisation. This is VERY different, as the degrees of freedom of the problem have been limited so that a solution is either stochastically highly likely or definite.

So far I have seen no model that satisfies the conditions to be a true model of biological evolution. I am not doubting that there are numerous models that show the seeming emergence of complexity. However, when you limit the problem to randomly selecting functions to solve a problem, if the problem space is well defined and the functional space is limited then you are simply performing a stochastic search .. this is NOT evolution. You are not inventing functions out of nothing, you are finding the best way to perform a pre-determined function.

I assert that it is impossible for a truly random mutation process to create new functionality if that functionality doesn't exist already in component form.
EXAMPLE
Given all the parts for a clock in sensible quantities it is reasonable to assume that randomly choosing components until you get a functioning clock through an optimisation algorithm is completely achievable.
Blind Watchmaker Evolution
However, if you didn't define the component parts and allowed random shaped parts to be randomly mutated as well as selected and combined through a similar optimisation algorithm, you would never expect to get a working clock. (this is what biological evolution is claimed to do)

I know of AVIDA, but it uses 26 predefined functions to combine to solve optimisation problems ... it doesn't mutate the code that defines these functions or how they're executed. It also doesn't model real-world selection.

I am a computer modeller and Mathematical Biology researcher, so I'm writing my own code which I believe will demonstrate the proposed mechanism more accurately ... I will share (with source) when it's done.

I am interested in an intellectual discussion here, not a creationist-bashing flame-war ... if you don't have any scientific answers that hold up to scrutiny please don't just post insults as though that's going to help. And please don't just make unsubstantiated claims, back them up with logic or research that can be shown to be applicable to this issue.

Cheers

Crypto

  
JonF



Posts: 634
Joined: Feb. 2005

(Permalink) Posted: Jan. 16 2015,09:22   

Define "adding information".

  
midwifetoad



Posts: 4003
Joined: Mar. 2008

(Permalink) Posted: Jan. 16 2015,09:33   

Quote
- Mutation can't simply combine functions in different arrangements. Mutation must occur on the code that defines the functions themselves, and affect the ability of the organism to build itself, to function and to replicate.


Well that's not exactly what happens in biological evolution. It is rare for the infrastructure of translation to change. It would seem that the likelihood of a mutation being tolerated is inversely proportional to its effect on the cell infrastructure.

That's a rather sloppy way of saying that vital functions are highly conserved.

--------------
Any version of ID consistent with all the evidence is indistinguishable from evolution.

  
NoName



Posts: 2729
Joined: Mar. 2013

(Permalink) Posted: Jan. 16 2015,09:33   

Many things are simple to define but appallingly complex, if not impossible, to model.  Fluid flow is one vast field where there are simple descriptions and definitions of phenomena that resist modeling.
The same goes for economics.
The same goes for such iterative functions as produce the Mandelbrot and Julia sets, as well as a host of other similar phenomena.  I add these because while we can 'model' them by virtue of producing the simple equation and iteration process involved, we cannot, in principle, predict what will be the result at any given point.  Models are not perfect predictors, nor is that their purpose.
Finally, let us note that the current absence of a model does not say anything about the possibility of ever modeling any particular phenomenon.  What we can say is that, in principle, it is impossible to assert that any given model can be predictive of real-world results, regardless of how many models are, in fact, predictive of such results.

  
midwifetoad



Posts: 4003
Joined: Mar. 2008

(Permalink) Posted: Jan. 16 2015,09:46   

Quote
However, if you didn't define the component parts and allowed random shaped parts to be randomly mutated as well as selected and combined through a similar optimisation algorithm, you would never expect to get a working clock. (this is what biological evolution is claimed to do)


If you intend to model evolution (possibly with the intention of demonstrating that life cannot arise "spontaneously," Your component parts will be expressed as chemistry.

Before wasting your time, please solve the problem of emergence. Demonstrate that you can predict, for example, the properties of water by deriving them from the properties of hydrogen and oxygen.

--------------
Any version of ID consistent with all the evidence is indistinguishable from evolution.

  
damitall



Posts: 331
Joined: Jan. 2009

(Permalink) Posted: Jan. 16 2015,09:51   

I believe that Dr Elizabeth Liddle ("Febble", and "Pingu") has ideed modelled the addition of new information in an computer simulation of evolutionary process - at least that type of information known to IDists as "CSI" or any one of a number of related acronyms.

I can't be arsed to look it out, but an intelligent search of The Skeptical Zone should reveal it.

It certainly caused an outbreak of the malice and vituperation to be expected from IDists with no answers

  
NoName



Posts: 2729
Joined: Mar. 2013

(Permalink) Posted: Jan. 16 2015,09:55   

Quote (cryptoguru @ Jan. 16 2015,10:09)
...
Here are some assumptions:
1) DNA is a self-replicating computer code, a list of instructions that is used by a cell to build a specific organism by a very complex and highly non-linear process that involves transcription and translation.

Questionable definition -- it is almost certainly question-begging, if not entirely incorrect, to assert that DNA is a 'computer code'.  To do so is to take metaphorical language and reify it to the point where the specific phrase becomes meaningless.
 
Quote
2) DNA is randomly mutated at the level of nucleotides (i.e. the letters of the programming code, not the functional/gene level)

OK, but this obscures the fact that function is tied to specific DNA arrangements.  Note, too, that those arrangements include physical folding and proximity to other DNA fragments.
 
Quote
3) Natural selection is not an intelligent entity, therefore it cannot evaluate the performance of an organism.

Half true, half radically false.  Yes, 'Natural Selection' is not an intelligent entity, but this is because it is not an entity at all.  It is a process.  The distinction matters.  Whether this process can, or need be, called 'intelligent' really depends on what your operational definition of 'intelligence' is.  I know researchers who take a stance towards intelligence that includes processes such as natural selection within the scope of intelligent behavior.  I know others who reject that judgement.  You appear to be taking an anthropomorphic, physicalist (i.e., embodied) approach to intelligence such that 'natural selection', not having a body (or a soul?) cannot be 'intelligent'.  What's your operational definition of 'intelligence' such that you can support your assertions using the term?
 
Quote
It is the name given to the process whereby more suitable organisms survive and others die ... achieved by a selective pressure being applied to the environment (e.g. availability of food/water, predators etc.)

Correct but misleading.  What natural selection does is always negative.  It weeds out failures.  The failures are always with respect to the organism's current environment and is only usefully evaluated at the group level.
 
Quote

Evolution is:
1) take an organism and let it replicate
2) random mutations should be introduced to the offspring DNA.
3) a selective pressure is applied to the environment the organisms live in
4) less fit organisms will die before replicating due to the pressure
5) most fit organisms will continue to replicate and mutate
6) eventually new functions and traits will emerge

Biological Evolution is presented as a scientific law, it is simple to define, so should be simple to model

See my previous reply.
 
Quote
My challenge is this.
Show me a computer model that models this process and shows new functions and information being added to the genome.

Why the insistence on a computer model?  Lenski et al have performed this over tens of thousands of generations in the lab and verified the process.  What more is required?
How is gene duplication not an increase of information in an organism?  Trivially, the information of the number of genes, and therefore DNA base pairs, has changed, in fact, increased.
 
Quote
caveats:
- Mutation can't simply combine functions in different arrangements. Mutation must occur on the code that defines the functions themselves, and affect the ability of the organism to build itself, to function and to replicate.

Why?
 
Quote
- the selective process can't include measuring properties of the organism to remove them from the population, the selective process must rely on organisms dying off due to the environmental factors. (e.g. energy)

Weasel-worded and a distinction without a difference.
If it's done in a computer model, measurement will happen, albeit properly the measurement does indeed need to be of survival(reproduction)/failure(to survive and to reproduce).  How else do you propose to determine the whether selection occurred at all in a computer model?  Randomly?  That is invalid for ecosystems due to the inherently constrained nature of the 'change space'.  See below.
 
Quote
It may sound subtle ... but breaking these laws means that you're not modelling biological evolution, you're performing functional optimisation. This is VERY different, as the degrees of freedom of the problem have been limited so that a solution is either stochastically highly likely or definite.

As I said before, weasel-worded.  A distinction without a difference.  Natural selection is a functional optimization.  And the work has been done in the lab that verifies it.

Quote
So far I have seen no model that satisfies the conditions to be a true model of biological evolution. I am not doubting that there are numerous models that show the seeming emergence of complexity. However, when you limit the problem to randomly selecting functions to solve a problem, if the problem space is well defined and the functional space is limited then you are simply performing a stochastic search .. this is NOT evolution. ...

No, that is simply confused.
The problem, in part, is that the random changes are inherently constrained by the laws of physics, chemistry, and thermodynamics.  The 'change space' is not infinite although it may be unimaginably large.  The ever changing 'problem space' against which the specific changes which occur are evaluated is also inherently constrained by the laws of physics, chemistry, and thermodynamics.  It is not infinite even if it might be unimaginably large.  
Neither the 'change space' nor the 'problem space' is free to vary perfectly randomly.  
There is also the factor of time.  Either space may hold relatively constant while the other space changes at a greater or lesser rate.  Both spaces may change relatively quickly or slowly.  Neither space may vary much.  Over varying spans of time, these temporal factors matter tremendously.

  
cryptoguru



Posts: 53
Joined: Jan. 2015

(Permalink) Posted: Jan. 16 2015,10:06   

My field of research originally was fluid mechanics, lately I've applied that to biological chaotic flows. I've written a LOT of programs that model fluids. We can accurately model the components of fluid flow (as well as chemical reactions with them), sure it's complicated but we can do it ... and we can compare them to real-world observations. (Watch any Disney 3D animated movie). We can model complex environmental interactions too ... it's done everyday and applied to thousands of useful real-world problems.

The definition of evolution is simple and it's implementation is simple too ... a self-building, self-replicating entity is mutated so that the code that runs it is mutated at an atomic (nucleotide) level. That's easy to model, I can write code to do that in a couple of hours ... my point is the other models that are used as "proofs" of the theory evolution do not prove anything, they simply solve optimisation problems, which is not the same thing.

JonF: "adding information" - new and novel functions that didn't exist before that solve a problem that wasn't solved previously.

midwifetoad: I'm not just asking that vital functions be conserved I'm asking that the overall net effect is improvement after mutation occurs on the definition of the code that is being run. DNA is an information system, it defines the process by which something is built ... we are mutating THAT. Not simply selecting between functions.

midwifetoad: the properties of water are observed, we understand the physics and chemistry behind them. We don't need to model things from atoms ... but we do need a model that reflects the theory to test the theory is sound, we don't have that with evolution.

  
NoName



Posts: 2729
Joined: Mar. 2013

(Permalink) Posted: Jan. 16 2015,10:19   

Quote (cryptoguru @ Jan. 16 2015,11:06)
... That's easy to model, I can write code to do that in a couple of hours ... my point is the other models that are used as "proofs" of the theory evolution do not prove anything, they simply solve optimisation problems, which is not the same thing.

Why and how not?

Evolution is the solution of an optimization problem.  Optimize the reproductive success of ever changing, in constrained ways, collections of organisms in an every changing, in constrained ways, environment.

As to 'easy to model, I can write code to do that in a couple of hours', well, no you can't.  Not and model all the factors involved in physically functional (in all senses) genes or chromosomes.  If you could, we'd have a trivial solution to the folding problem, and we don't.  Physical structure is inherently bound to a wide variety of factors impacting the existence and functionality of specific DNA segments in situ.  Slight changes to chemical structure can have anywhere from minor to overwhelming impact on the physical structure, and vice versa.  Your 'couple of hours effort' model is so flawed as to be meaningless for an actual simulation of the reality you are attempting to assess or reject.  It provides essentially zero guidance to what can and does happen in the living environment in the chemical, physical, and thermodynamic realities constantly at play at the level of DNA.  You might as well be solving the 3 body problem by recourse to the behavior of balls swung on the end of strings.

Also, I note you ignored a couple of my key points in your rejoinder:  the success of any number of computer molding strategies says nothing whatsoever about the in principle solvability of all problems with computer modeling strategies.  This has been proven as well as demonstrated, countless times in countless fields.
Nor does the present absence of any particular model or computer simulation speak to the in principle possibility or lack thereof of a future model.

  
midwifetoad



Posts: 4003
Joined: Mar. 2008

(Permalink) Posted: Jan. 16 2015,10:19   

It has already been noted that you are reifying a metaphor. DNA is not a computer program, It is a molecule.

The map is not the territory. Dawkins' Weasel does not prove evolution (although it would be interesting if it didn't work).

Now if you, in your brilliance, can model the HOX gene and can model how variations in form and body plan can be made by varying the timing of developmental events, you might have something useful.

But what will be demonstrated if you fail to model biological evolution? I can write a program that doesn't work.

I'm trying to get a grip on your hypothesis. What are you trying to demonstrate?

--------------
Any version of ID consistent with all the evidence is indistinguishable from evolution.

  
cryptoguru



Posts: 53
Joined: Jan. 2015

(Permalink) Posted: Jan. 16 2015,10:33   

Quote
if not entirely incorrect, to assert that DNA is a 'computer code'


I've been over this ground on the other site.
DNA is a computer code ... a living cell is a computer, not analogous to a computer, but IS a computer.
This is a true statement because it fulfils the definition of a computer as a Turing machine.
This is also proved by the fact that we use the living cell as a computer ourselves now
DNA computing - Wikipedia
There are LOTS of evolutionists who attest to this fact.

The reason you're avoiding this definition is because you want to keep it mysterious and difficult to pin down.

Quote
OK, but this obscures the fact that function is tied to specific DNA arrangements.  Note, too, that those arrangements include physical folding and proximity to other DNA fragments.

I don't see how what I said contradicts that? I'm just speaking in generic terms.



Quote
You appear to be taking an anthropomorphic, physicalist (i.e., embodied) approach to intelligence such that 'natural selection', not having a body (or a soul?) cannot be 'intelligent'.

I think you're jumping the gun here, I'm not claiming Natural Selection is powerless to do anything, I'm just asserting that it is not like a god ... it can't measure and evaluate, it's not magic. It is simply the process of filtering organisms that don't work in their environment. It's not the Grim Reaper going around choosing who to kill. Therefore modelling processes that attempt to model natural selection as though it is intelligent are incorrect.

Quote
Correct but misleading

Again, I complete agree with your definition there ... I am not trying to be sneaky. Nothing I said disagrees with your understanding of natural selection, unless you think it IS magic.


Quote
Lenski et al have performed this over tens of thousands of generations in the lab and verified the process

Lenski has demonstrated E. Coli doing exactly what E Coli is programmed to do, it adapts (this is true of all viruses and bacteria) ... this does not demonstrate new features. They've performed an equivalent of millions of years of evolution ... where's the more complex life-forms? They're still bacteria!

Again ... a computer model should be able to at least prove the theory is possible.


Quote
The problem, in part, is that the random changes are inherently constrained by the laws of physics, chemistry, and thermodynamics

You're grasping at straws ... random mutations are random. Natural selection is purely the ability for the organism to keep on reproducing. It can fail to do that for many reasons e.g. competition for food, disease etc
But the process is simple


Quote
If it's done in a computer model, measurement will happen, albeit properly the measurement does indeed need to be of survival(reproduction)/failure(to survive and to reproduce).  How else do you propose to determine the whether selection occurred at all in a computer model?  Randomly?

I completely agree ... but we should be doing as you suggest measuring survival and not measuring for example how fast something goes. If we want to measure speed competition, we need to let the organisms have speed and see which survive, not calculate speeds and set an arbitrary threshold ... does that make sense to what I'm saying now?

  
Cubist



Posts: 558
Joined: Oct. 2007

(Permalink) Posted: Jan. 16 2015,10:37   

Quote (cryptoguru @ Jan. 16 2015,09:09)
My challenge is this.
Show me a computer model that models this process and shows new functions and information being added to the genome. [emphasis added]

Whether or not you, yourself, are a Creationist, cryptoguru, this challenge of yours is (whether you realize it or not) the hoary old mutations can't create new information argument beloved of Creationists, just dressed up in computer terminology. So please, tell us if you will: What does "new information" look like? Given an arbitrary string of nucleotides, and a mutation which alters that string of nucleotides, how can you tell whether or not the post-mutation version of that string contains any "new" information?

  
cryptoguru



Posts: 53
Joined: Jan. 2015

(Permalink) Posted: Jan. 16 2015,10:43   

midwifetoad:

I explained to you why it is a computer program. You call DNA a double helix, it is a definition of it's shape. A computer program is a definition of its function.
You need to understand the definition of a computer.
Nature article on cell being Turing machine

I'm not trying to model the WHOLE of proposed biological evolution ... what nonsense are you suggesting?

I'm modelling the key components ... a key component is the code that is executed by the cell's operating system is mutated, which gives the possibility of it crashing (death) ... this is what we observe through genetic mutations.

I'm suggesting we model the core tennets of evolution ... existing models do not do this.

  
midwifetoad



Posts: 4003
Joined: Mar. 2008

(Permalink) Posted: Jan. 16 2015,10:51   

When you build such a model that works, I'll be impressed.

Explain why I should be impressed by a model that fails.

--------------
Any version of ID consistent with all the evidence is indistinguishable from evolution.

  
Wesley R. Elsberry



Posts: 4991
Joined: May 2002

(Permalink) Posted: Jan. 16 2015,10:52   

Avida.

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

    
midwifetoad



Posts: 4003
Joined: Mar. 2008

(Permalink) Posted: Jan. 16 2015,10:55   

Quote
I explained to you why it is a computer program. You call DNA a double helix, it is a definition of it's shape. A computer program is a definition of its function.


The word program is a metaphor. You cannot determine the limitations of DNA from the limitations of a computer program.

Just as you cannot determine the existence or non-existence of a landscape feature from a map.

--------------
Any version of ID consistent with all the evidence is indistinguishable from evolution.

  
Doc Bill



Posts: 1039
Joined: April 2007

(Permalink) Posted: Jan. 16 2015,10:55   

Greetings, NotGuru, we meet again!

Perhaps a full disclosure of your real purpose is in order.  I present to you NotGuru's Agenda:

Quote
I paint evolution (non-science) as faith ... because it is.
It's an absurd unsupported set of stories, based on the premise that there is no possibility of non-material causation. It satisfies a huge amount of self-centred god-haters, who've wanted nothing more than a decent excuse to pretend that there are no eternal consequences for their actions.

You still don't address any of the gaping black holes that I've pointed out.


I'm thinking this one is a Gary Gaulin clone, but super tard charged.  If you don't mind I'll duck out and get some popcorn.

  
rossum



Posts: 289
Joined: Dec. 2008

(Permalink) Posted: Jan. 16 2015,11:03   

Quote (cryptoguru @ Jan. 16 2015,09:09)
My challenge is this.
Show me a computer model that models this process and shows new functions and information being added to the genome.

Certainly.  Dr. Lenski used the Avida program to show the evolution of new functions.

The cover page is The Evolutionary Origin of Complex Features

The detail is line of descent.

For "new informaiton" you will need to tell us how you are calculating the quantity of information present.  Evolution can increase both Shannon and Kolmogorov information.  Are you using a different measure?

rossum

--------------
The ultimate truth is that there is no ultimate truth.

  
OgreMkV



Posts: 3668
Joined: Oct. 2009

(Permalink) Posted: Jan. 16 2015,11:18   

Perhaps, he will listen to you guys. I've said all this already.

crypto, perhaps you should explain, in detail why AVIDA does not do what the writers of the software say it does... and show that it does in the paper I provided you.

This ought to be entertaining.

We desperately need a popcorn smiley.

--------------
Ignored by those who can't provide evidence for their claims.

http://skepticink.com/smilodo....retreat

   
NoName



Posts: 2729
Joined: Mar. 2013

(Permalink) Posted: Jan. 16 2015,11:47   

Quote (cryptoguru @ Jan. 16 2015,11:33)
   
Quote
if not entirely incorrect, to assert that DNA is a 'computer code'


I've been over this ground on the other site.
DNA is a computer code ... a living cell is a computer, not analogous to a computer, but IS a computer.

Prove it.  Don't assert it, prove it.
You can't because there simply are no significant respects in which it is true.
The entire realms of thermodynamic effects and of self-assembly, with constrained errors, and self-repair, with constrained errors, simply have no equivalent in computers.  At all.
This metaphor is wildly less accurate and wildly less appropriate to the cases at hand than 'natural selection'.
So, once again, prove it.  Don't assert it, don't point me to garbage sites where you've also asserted it.
It's a simple proposition, it should be simple to prove, without abstracting away all the differences that make a difference.
   
Quote
This is a true statement because it fulfils the definition of a computer as a Turing machine.

A computer is a Turing machine.  Are all Turing machines computers?  Proof please.
   
Quote
This is also proved by the fact that we use the living cell as a computer ourselves now
DNA computing - Wikipedia
There are LOTS of evolutionists who attest to this fact.

Irrelevant and absurd.  We can use strings and beads as calculating machines, this does not mean that strings and beads are calculators.

   
Quote
The reason you're avoiding this definition is because you want to keep it mysterious and difficult to pin down.

No.  And offensive.
We're 'avoiding' this definition because it is not a definition and because it is incorrect.  It is analogy run amuck.
   
Quote

     
Quote
OK, but this obscures the fact that function is tied to specific DNA arrangements.  Note, too, that those arrangements include physical folding and proximity to other DNA fragments.

I don't see how what I said contradicts that? I'm just speaking in generic terms.

And the genericity is where you go wrong.  The same set of DNA patterns operate, or fail to operate, depending on the physical arrangement of other DNA patterns, which are dynamic over time.
That you don't understand this fundamental fact demonstrates that you are unprepared to make the claims you are in fact making.
Do you even know what the folding problem is?  Or why it matters, and matters particularly in this problem space?

   
Quote
     
Quote
You appear to be taking an anthropomorphic, physicalist (i.e., embodied) approach to intelligence such that 'natural selection', not having a body (or a soul?) cannot be 'intelligent'.

I think you're jumping the gun here, I'm not claiming Natural Selection is powerless to do anything, I'm just asserting that it is not like a god ... it can't measure and evaluate, it's not magic.

Disagreement without basis.  No one is claiming it is, or 'is like' a god (for one thing, it is not fictional or mythical, it does not operate by miracles, and it can be seen in action in controlled and well-defined tests, none of which can be done with any god ever proposed).  It is, in the sense it which you are using it here, a description of a process of evaluation, of measurement.  It measures, it evaluates, reproductive failure.  That's all it has to do.
We have operational definitions.  It appears you do not.
   
Quote
It is simply the process of filtering organisms that don't work in their environment. It's not the Grim Reaper going around choosing who to kill. Therefore modelling processes that attempt to model natural selection as though it is intelligent are incorrect.

     
Quote
Correct but misleading

Again, I complete agree with your definition there ... I am not trying to be sneaky. Nothing I said disagrees with your understanding of natural selection, unless you think it IS magic.


     
Quote
Lenski et al have performed this over tens of thousands of generations in the lab and verified the process

Lenski has demonstrated E. Coli doing exactly what E Coli is programmed to do, it adapts (this is true of all viruses and bacteria)

Can you say 'begging the question'?  You appear to be arguing that cells/organisms are programmed by asserting as one of your premises that cells/organisms are programmed.
You are also making recourse to a particularly problematic undefined metric -- what and how do you determine which organism is 'more advanced' than some other?  What is the metric of advancement?  It is, in fact, a holdover from superstition that saw all life as a glorious unfolding of lower creatures to higher creatures, culminating in mankind.  Which is indefensibly stupid not least because it is entirely unsupported by facts of any sort.
Either all creatures are equally advanced, or bacteria are far more advanced than humans because they've been evolving for more generations.
Either drop this 'more advanced' crap or provide facts, evidence, logic, and operational definitions to make it clear and unequivocal for all to see.
   
Quote
... this does not demonstrate new features. They've performed an equivalent of millions of years of evolution ... where's the more complex life-forms? They're still bacteria!

They are 'more advanced' bacteria in that they demonstrate features strictly absent from the parent population.
How does this not meet your challenge?  You're refusing to accept it because it leaves you without a foundation for your creationism.
 
Quote
Again ... a computer model should be able to at least prove the theory is possible.

See responses from others.

     
Quote
 
Quote
The problem, in part, is that the random changes are inherently constrained by the laws of physics, chemistry, and thermodynamics

You're grasping at straws ... random mutations are random.
No, they are only random with respect to certain factors, such as future needs of the organism.  They are highly constrained.  You are handwaving away the hard problem, essentially by ignoring it.  The results of that act of ignorance are errors which permeate your argument from the point of deciding these critical factors don't matter.  It's as if you were arguing that the SR-71 could not be supersonic because your plastic model would melt at the friction imposed at mach speeds.
 
Quote
Natural selection is purely the ability for the organism to keep on reproducing. It can fail to do that for many reasons e.g. competition for food, disease etc
But the process is simple


     
Quote
If it's done in a computer model, measurement will happen, albeit properly the measurement does indeed need to be of survival(reproduction)/failure(to survive and to reproduce).  How else do you propose to determine the whether selection occurred at all in a computer model?  Randomly?

I completely agree ... but we should be doing as you suggest measuring survival and not measuring for example how fast something goes. If we want to measure speed competition, we need to let the organisms have speed and see which survive, not calculate speeds and set an arbitrary threshold ... does that make sense to what I'm saying now?

And we're back to the critical issue of exactly what do you mean by 'intelligent'.  As you use it here, you appear to  mean that intelligence is inherently purposive, that it operates with 'purpose aforethought'.  That is a peculiarly anthropomorphic, and ultimately too tightly restrictive, a definition of 'intelligence'.
It's clear you throw the word around as a magic wand because you do not, and likely cannot, provide an operational definition that will suit in all cases of your use of the term.
NO ONE is asserting that 'natural selection' is any sort of Grim Reaper going around and thinking about which individuals out of which groups are going to live or die at any given time.  Why would anyone assert that when it is so clearly incorrect?
What 'natural selection' is is simply a description of the failure of certain contained-random changes in the genome to persist in a population across time in a given constrained-random environment.  It requires no purpose, no planning, no intention or intentionality.  It's how the world works.  No one has ever shown otherwise.
Nor has anyone shown that changes to the genome do not occur, randomly with respect to their environment and constrained by the laws of physics, chemistry, and thermodynamics.
Nor has anyone shown that information cannot increase due to simply chemical and physical processes, constrained in various natural ways including thermodynamically and by containment.
You've ignored the examples that demonstrate this.
Why is that?
Is 'information' another magic word that you throw around without being able to provide a consistent coherent operational definition for?

  
qetzal



Posts: 311
Joined: Feb. 2006

(Permalink) Posted: Jan. 16 2015,11:53   

Quote (cryptoguru @ Jan. 16 2015,9:09)
My challenge is this.
Show me a computer model that models this process and shows new functions and information being added to the genome.


Quote (cryptoguru @ Jan. 16 2015,10:33)
[A] living cell is a computer, not analogous to a computer, but IS a computer.


Perfect. Lenski already demonstrated new functions being added to the genome using E. coli, which are living cells, and therefore ARE computers.

You're welcome.

  
NoName



Posts: 2729
Joined: Mar. 2013

(Permalink) Posted: Jan. 16 2015,12:03   

Quote (qetzal @ Jan. 16 2015,12:53)
Quote (cryptoguru @ Jan. 16 2015,9:09)
My challenge is this.
Show me a computer model that models this process and shows new functions and information being added to the genome.


 
Quote (cryptoguru @ Jan. 16 2015,10:33)
[A] living cell is a computer, not analogous to a computer, but IS a computer.


Perfect. Lenski already demonstrated new functions being added to the genome using E. coli, which are living cells, and therefore ARE computers.

You're welcome.

Heh.  ha.  haHaHa!  Bwahahahahahahaha!
ROFL

I think we're done here.

Except, of course, he's not accepting Lenski's work as having done what it has so clearly done.
What's wrong with Lenski's work, crypto?  How are your challenges not already met?

  
OgreMkV



Posts: 3668
Joined: Oct. 2009

(Permalink) Posted: Jan. 16 2015,12:04   

I'd like to add that natural selection is not always a death.

It can be an effective removal of the gene pool. For example, the strongest, fittest, most efficient lion ever born might have a problem reproducing. Which means that he is removed from the gene pool. He can't pass his genes on to the next generation, so he's out of contention.

I'll note that the paper by Lenski (http://myxo.css.msu.edu/papers....lex.pdf specifically mentions that happening in some of their digital organisms.

Quote
However, two mutations reduced fitness by >50%.
One was a point mutation that disrupted replication efficiency. Its
harmful effect was eliminated by the next mutation in the line of
descent, which occurred at a distant site in the genome. The other
very deleterious step was a point mutation, at depth 110, that
knocked out NAND, one of the simplest logic functions. Only two
individuals had this maladapted genotype, yet their descendants
emerged as eventual winners. In fact, in the very next step, this
genotype produced the mutation that gave rise to EQU. Was that
deleterious mutation extremely lucky to hitchhike with such a
beneficial mutation? Or was the deleterious mutation a prerequisite
for producing the EQU function within that genome context? To
distinguish between these hypotheses, we reversed this one-stepprior
mutation in the genotype that first expressed EQU. This
reversal eliminated the EQU function. Therefore, a mutation that
was highly deleterious when it appeared was highly beneficial in
combination with a subsequent mutation. The evolution of a
complex feature, such as EQU, is not always an inexorably upward
climb toward a fitness peak, but instead may involve sideways and
even backward steps, some of which are important.


Note that the "deleterious" mutation, which dropped fitness by greater than 50% was one mutation that resulted in the most fit individuals.

If you program this kind of thing OUT of an evolutionary algorithm, then it is not matching biology (as crypto demands of it).

I will ask a third time, crypto, have you read this paper? It's sort of a massively important paper in the field.

--------------
Ignored by those who can't provide evidence for their claims.

http://skepticink.com/smilodo....retreat

   
midwifetoad



Posts: 4003
Joined: Mar. 2008

(Permalink) Posted: Jan. 16 2015,12:13   

First, I think Crypto needs to model and replicate in silico, Lenski's result. That would be the bare minimum for a competent model.

If that isn't sufficient, we could discuss the insufficiency later. But first things first. Let's see if we can model microevolution.

--------------
Any version of ID consistent with all the evidence is indistinguishable from evolution.

  
OgreMkV



Posts: 3668
Joined: Oct. 2009

(Permalink) Posted: Jan. 16 2015,12:17   

BTW: Here's an interesting interpretation of that duon article you linked to.

http://sandwalk.blogspot.ca/2014....ion.htm

--------------
Ignored by those who can't provide evidence for their claims.

http://skepticink.com/smilodo....retreat

   
stevestory



Posts: 13407
Joined: Oct. 2005

(Permalink) Posted: Jan. 16 2015,12:18   

i wonder if he has something not already covered in the Index to Creationist Claims

I doubt it.

   
midwifetoad



Posts: 4003
Joined: Mar. 2008

(Permalink) Posted: Jan. 16 2015,12:22   

The most important thing that needs to be modeled is the robustness of biochemistry, it's ability to tolerate variation.

Living populations contain multitudes of unique mutations. Individuals are seldom exact copies of other individuals.

So the model operating system and cpu need to have this kind of flexibility built in.

That's why I said early on, that if one is trying to model the origin of life or early evolution, one needs to model chemistry.

Current living things rarely (successfully) change their operating system.

--------------
Any version of ID consistent with all the evidence is indistinguishable from evolution.

  
midwifetoad



Posts: 4003
Joined: Mar. 2008

(Permalink) Posted: Jan. 16 2015,12:55   

Quote (OgreMkV @ Jan. 16 2015,12:17)
BTW: Here's an interesting interpretation of that duon article you linked to.

http://sandwalk.blogspot.ca/2014.......ion.htm

That doesn't work for me. I'll try again.

http://sandwalk.blogspot.com/2014....on.html

--------------
Any version of ID consistent with all the evidence is indistinguishable from evolution.

  
midwifetoad



Posts: 4003
Joined: Mar. 2008

(Permalink) Posted: Jan. 16 2015,13:16   

I guess he'll be back in a couple of hours with his model.

--------------
Any version of ID consistent with all the evidence is indistinguishable from evolution.

  
  336 replies since Jan. 16 2015,08:04 < Next Oldest | Next Newest >  

Pages: (12) < [1] 2 3 4 5 6 ... >   


Track this topic Email this topic Print this topic

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