Why wouldn't God exist?

[Sisyphus]

I have looked very hard for an example of a known (documented as opposed to presumed) contiguous evolutionary pathway of greater than three stepwise mutations where each step has selectable advantage.

 

And why is this particular hoop the one to be jumped through?

 

Yes, I specified the kind of information (coherent, useful) earlier in the thread.

 

The point is that all information is useful information if it's embedded in the genome, inasmuch as any sequence has the potential to be expressed. Entropy increases the total amount of this information. If there were an original "message" being transmitted, then entropy would degrade that message, true, but if you define information in that restricted way, then you're not talking about how evolution functions. Evolution works with what it gets.

 

So yes, the "random" process of mutation does generate increased information. Whether that information is "coherent" is determined by how natural selection acts on the expression of that information, at that point or at some future point after it has spent time being transmitted as so-called "junk DNA."

[Cap'n Refsmmat]

You would do well to read the TalkOrigins articles on information theory rather than dismissing Sisyphus directly.

 

Incidentally, you are aware of the Lenski experiment, yes? You earlier dismissed it as off-topic, but now that you've asked directly for something matching its characteristics. (Particularly the directly observed two-step evolution of citrate metabolization.)

 

TalkOrigins does a good job of citing other studies in which evolution increases information within a population, through novel biological activities, increased biological variety, and so on. You are free to browse through the citations as well.

 

The prime point of the article, that Shannon-Weaver information theory shows that random mutation maximizes information, is supported by an open-access PNAS paper cited in Further Reading.

 

In particular, we show that, in fixed environments, for organisms whose fitness depends only on their own sequence information, physical complexity must always increase. That a genome's physical complexity must be reflected in the structural complexity of the organism that harbors it seems to us inevitable, as the purpose of a physically complex genome is complex information processing, which can only be achieved by the computer which it (the genome) creates.

 

Have a gander. I'm sure it'll be interesting. I'll read it when I have the time as well. (I think I'll just be sticking it on my Kindle later...)

[cypress]

You would do well to read the TalkOrigins articles on information theory rather than dismissing Sisyphus directly.

 

Incidentally, you are aware of the Lenski experiment, yes? You earlier dismissed it as off-topic, but now that you've asked directly for something matching its characteristics. (Particularly the directly observed two-step evolution of citrate metabolization.)

 

I have read the talk origins articles and I find they fall short of addressing square on the actual issue. They are fine for what they do explain, They just don't explain the substance of the issue. It seems to be a common approach for that site.

 

I also noted the topic change. Previously I attempted to stay with pure random processes but it is clear that most here accept the limits of random chance so if you wish to raise it again that's fine.

 

Edit: If you take a close look at Lenski's ecoli you will find that it involves no more than three mutations one of which is not advantageous. It does fit in the class of events that random chance can accomplish given the resources available. It does not violate the rules of information enthalpy. Since it does not fit the request I am asking for, I properly dismiss it for that reason. It is interesting in that it actually confirms my claims regarding what natural systems are an are not capable of accomplishing.

 

TalkOrigins does a good job of citing other studies in which evolution increases information within a population, through novel biological activities, increased biological variety, and so on. You are free to browse through the citations as well.

 

The examples are incomplete and contain large gaps in the steps. It is like coming to a wide river and noting three exposed rocks near the center but then claiming that since there are three rocks one can cross using only the rocks without getting wet. The examples aare nothing more than the usual observation that similarities exist in biological components. They do not explain how the differences occurred and it is the how that we are discussing. I do not dispute that changes happened, but i note that the how remains unaccounted for. Random error and selection seems to lack the ability to generate the changes detailed in the articles.

 

The prime point of the article, that Shannon-Weaver information theory shows that random mutation maximizes information, is supported by an open-access PNAS paper cited in Further Reading.

 

Have a gander. I'm sure it'll be interesting. I'll read it when I have the time as well. (I think I'll just be sticking it on my Kindle later...)

 

I have read it a while ago. Shannon information theory deals with preservation/degradation of data sequences through communication networks. It is interesting, but it is not relevant to our discussion. The discussion here is the source of meaningful, functional coherent information used to generate and control functional systems. A very different situation. In the article the authors address this difference and make clear that they avoid this topic intentionally. Here is what they say.

 

In this paper, we skirt the issue of structural and functional complexity by examining genomic complexity. It is tempting to believe that genomic complexity is mirrored in functional complexity and vice versa. Such an hypothesis, however, hinges upon both the aforementioned ambiguous definition of complexity and the obvious difficulty of matching genes with function.

 

The paper once again is interesting for what it does address, it just does not address the topic here.

 

And why is this particular hoop the one to be jumped through?

 

Actually there are several good reasons to doubt the neo-Darwinian narrative. The fact that mind is capable of generating coherent, encoded, digital information and that all biological systems require coherent digitally encoded information to generate and mange the processes, and that we do not observe any natural processes currently in operation generating this kind of information is a strong indicator that there was a mind behind the cause of life. The same is true about systems that are fine tuned. An earlier post claimed that life is fine tuned and thus we are discussing whether or not it has been established that biological systems were derived from natural processes. I contend the answer is that we don't know, but we do know that thus far only mind is known to be capable of generating the fundamental aspects of 1) fine tuned systems and 2) Biological systems.

 

Now there are several other indicators of mind behind both this universe and life and we may yet come to those, but currently we are working these issues.

 

The point is that all information is useful information if it's embedded in the genome, inasmuch as any sequence has the potential to be expressed.

 

Entropy increases the total amount of this information. If there were an original "message" being transmitted, then entropy would degrade that message, true, but if you define information in that restricted way, then you're not talking about how evolution functions. Evolution works with what it gets.

 

It is only useful if it is functional. Biological systems must be functional or it is a waste of critical resources expressing junk. Ten years ago we did not know if every or even many mutation generates functional components. The theory predicted that a reasonable percentage of mutations should be functional and at least some should be advantageous. However research has continued in this front and has answered this question¹. It turns out that in a protein of 150 amino acids (small for most proteins) the ratio of functional to non-functional sequences is 1 in 10⁷⁴. While any sequence may be expressed, it is far from true that any sequence has the the potential to be biologically active. It is an understatement to say the most aren't and it makes your argument vacuous. The ratio continues to drop as sequence size increases as Axe demonstrates in his work.

 

1) Douglas Axe, "Estimating the Prevalence of Protein Sequencing", Journal of Molecular Biology, 2004, plus several other from Axe in earlier releases of this journal and also Proceeding of the Nation Academy of Sciences between 1996 and 2004.

 

So yes, the "random" process of mutation does generate increased information. Whether that information is "coherent" is determined by how natural selection acts on the expression of that information, at that point or at some future point after it has spent time being transmitted as so-called "junk DNA."

 

I find this argument completely ridiculous. Never mind the reality that biological systems quickly trim away junk to conserve energy, to equate junk with functional systems stretches credulity to the breaking point. I understand why talk Origins would do this. After all they are advocates of the neo-Darwinian evolutionary theory and present not only points that build the case but also do their best to obfuscate any argument against their presuppositions. They are hardly the fair minded unbiased site they purport to be. It's ok to be an advocate but when they try to shift the substance of the argument as in this example, it comes across as slimy. Let's stay focused on encoded information that drives functional systems please, since coherent information is a necessary requirement while noise (junk) is not only not required, it is detrimental.

Edited July 25, 2010 by cypress

[pioneer]

The random arguments of life work under the assumption that the organic nature of life occurs out of the context of water. For example, if we had a beaker of pure lipids (no water) random would be a useful assumption to help describe how the molecules move and orientate at room temperature. If you add water, so water becomes the main phase, the random of the lipids becomes reduced into a lipid bi-layer with a propensity to become a spherical shell, with water on both sides of shell. Once we add water, we can roll the lipid dice again and again, and double sixes appear again and again, because the dice are loaded because of the potentials within water.

 

Experiments several decades ago, proved that lightning (simulated) can form the building blocks of life from basic gas molecules such as methane and ammonia. If you look at lightning, this is also an output feature from water, since lightning is generated by the water in certain types of clouds. Right from the beginning water is forging the basic building blocks of life, due its many common and anomalous properties.

 

In Genesis 20; And God said, Let the waters bring forth abundantly the moving creature that hath life. This ancient thinking is actually the future state of the art, if we can ever get past the erroneous assumption of random, that appears to occurs, when we leave out the potentials of water. This old thinking could create the illusion the basic cell membranes had to go through a long process of trial error and selective advantage, when water defined the way.

Edited July 25, 2010 by pioneer

[Cap'n Refsmmat]

Let me get a handle on your position. If I'm not mistaken, you're saying:

 

1. Evolution does happen.
2. We have indeed observed organisms to evolve new behaviors and abilities.
3. Random mutation and natural selection can't be the sole cause.

Is this accurate?

[Mr Skeptic]

You can doubt all you like the abilities of a process beyond your comprehension, but there is undeniable evidence that evolution, or something very much like it, did indeed happen. Evidence that our "creator" is profoundly stupid and did their creating via copying and mutating. The recurrent laryngeal nerve, for example, a relic we have from our fish ancestors, is a nerve that goes down into the thorax, loops around the aorta, and then goes back up the neck to they larynx. It made sense in fish, but was kept in mammals where it does not make sense. In fact, for the giraffe, that loop is several extra feet long.

 

Another and more obvious relic is the presence of broken retrovirus fragments in our DNA and the DNA of our ancestors. We know that the DNA is of a retrovirus because said retrovirus was reconstructed and resurrected. In phone books and certain similar databases, the creators intentionally introduce certain mistakes. The presence of the same mistakes in other databases is nearly undeniable proof that they were copied from the original, as the chance of making the same mistakes in the same places is astronomically small. The same principle applies to DNA, only more so since there is so much more DNA.

 

And regardless of whatever you may believe about the capabilities of evolution, the same algorithm was used to design several devices. Perhaps you're even using one of them. When used as a design algorithm it is called either an "evolutionary algorithm" or a "genetic algorithm".

 

So not only do we know evolution works but there is evidence that can only be explained by evolution. It is about as certain as anything can be for those who do not have absolute faith that certain books are the words of omnipotent beings.

[Cap'n Refsmmat]

The first minute or two don't entirely apply here; it's the simulation I wanted to point at. The simulation of numerous random mutations occurring to bring about functional improvements to a device, in only a few thousand generations.

(Actually, I'd love to get the code of this simulator running for fun, and perhaps work on improving it...)

[cypress]

Let me get a handle on your position. If I'm not mistaken, you're saying:

 

1. Evolution does happen.
2. We have indeed observed organisms to evolve new behaviors and abilities.
3. Random mutation and natural selection can't be the sole cause.

Is this accurate?

 

Organisms indeed have and do change over time and have diversified over time.

I am not certain that we have observed new behaviors and abilities.

Experimental results indicate that random mutation and natural selection does not explain how diversification of life occurred.

[Klaplunk]

Organisms indeed have and do change over time and have diversified over time.

I am not certain that we have observed new behaviors and abilities.

Experimental results indicate that random mutation and natural selection does not explain how diversification of life occurred.

 

Chaos Evolution

[Cap'n Refsmmat]

Organisms indeed have and do change over time and have diversified over time.

I am not certain that we have observed new behaviors and abilities.

Experimental results indicate that random mutation and natural selection does not explain how diversification of life occurred.

 

I see. Could you clarify what you mean by "observed"? I am not sure how to make sense of your comment in light of, say, the Lenski experiment. (It may not match your previous criteria, but metabolization of citrate was a new behavior or ability.)

 

I am also curious what experimental results you cite -- so far I have only seen arguments from information theory.

[cypress]

You can doubt all you like the abilities of a process beyond your comprehension, but there is undeniable evidence that evolution, or something very much like it, did indeed happen.

 

This discussion is about what processes are capable of generating new coherent information since it is a necessary component of the observed diversity. I do not dispute that diversification happened, but I do note that random mutation and selection does not generate the requisite precursor components in a timeframe sufficient to account for the diversity recorded in the fossil record.

 

And regardless of whatever you may believe about the capabilities of evolution, the same algorithm was used to design several devices. Perhaps you're even using one of them. When used as a design algorithm it is called either an "evolutionary algorithm" or a "genetic algorithm".

 

Yes I am aware of genetic algorithms. I have studied them in detail and I note that every one I have looked at the designers have imported active information sufficient to allow the system to function. In other words they work because the designer designed them to work. If you doubt this let's deconstruct one and I will show you where the designer has imported information. Should we conclude from this that evolution was designed?

 

So not only do we know evolution works but there is evidence that can only be explained by evolution.

 

Unfortunately your evidence points to design by virtue of smuggled in design as I explained above.

 

It is about as certain as anything can be for those who do not have absolute faith that certain books are the words of omnipotent beings.

 

It can only certain if your prior commitment has eliminated all other possibilities because unfortunately for you your strong evidence was designed. Your other evidence of retroviruses and supposed suboptimal design is off topic and unconvincing because both could be exactly as you imply and yet neither offer any method for natural processes to generate coherent information. Edit: It is also somewhat useless to presuppose what might or might not be a "good" design as in the case of the larygeanal nerve. If the nerve is suboptimal as you suppose and if natural selection selects advantageous systems then you would not expect to see it if selection had the capability the narrative supposes it does. One can play this game endlessly, so I suggest we focus on the arguments at hand rather than raising red herrings.

 

I see. Could you clarify what you mean by "observed"? I am not sure how to make sense of your comment in light of, say, the Lenski experiment. (It may not match your previous criteria, but metabolization of citrate was a new behavior or ability.)

 

I suppose I bolded the wrong word. I took exception to characterizing it as "observing new" since metabolization of citrate is not new. It has been observed in the past and is even metabolized by the bacteria under other environmental circumstances so it is more properly characterized as modification of existing function. It is an adaptation. Furthermore that particular example does not fall outside the boundaries of what should be expected of a blind search given the very modest increase in coherent information and thus small decrease in entropy more than covered by the resources brought to bear on the problem.

 

I am also curious what experimental results you cite -- so far I have only seen arguments from information theory.

 

This thread is primarily about evidence for a creator of the universe and life and the evidence is generation of coherent information. Lenski's work with metabolism of Citric acid in an oxygenated environment is one such example of the limits of mutation and selection. Axe's work outlining the rarity of functional protein systems is another that provides evidence of severe limitations on stepwise searches through protein sequence space. There are many others but I think they are more appropriate for another thread. Start one on experimental evidence suggesting limits on random mutation and select ion and I will be happy to contribute.

Edited July 26, 2010 by cypress

[Mr Skeptic]

Yes I am aware of genetic algorithms. I have studied them in detail and I note that every one I have looked at the designers have imported active information sufficient to allow the system to function. In other words they work because the designer designed them to work. If you doubt this let's deconstruct one and I will show you where the designer has imported information. Should we conclude from this that evolution was designed?

 

Genetic algorithms only need the same attributes as evolution has, to function. The most complicated would be the fitness function; but in evolution this comes naturally via death and reproduction. Likewise, you could implement a genetic algorithm with very little information, if you were to build every prototype yourself and input the outcome to the program; however, this would be absurdly time-consuming and expensive to do.

 

So because we design things to look like something natural, you conclude that the natural thing must be designed, is that it? I can make a rock, therefore rocks are designed?

 

Unfortunately your evidence points to design by virtue of smuggled in design as I explained above.

 

What design? Who designed evolution, or what makes evolution not be a genetic algorithm?

 

It can only certain if your prior commitment has eliminated all other possibilities because unfortunately for you your strong evidence was designed. Your other evidence of retroviruses and supposed suboptimal design is off topic and unconvincing because both could be exactly as you imply and yet neither offer any method for natural processes to generate coherent information so I did not respond to them.

 

My prior commitment is a Young Earth Creationist, a position which I was forced after decades of kicking and struggling, to abandon. Look in the archives of this forum if you do not believe me, and you will see that for yourself. It is your prior commitments which drag you down, not mine.

 

I suppose I bolded the wrong word. I took exception to characterizing it as "observing new" since metabolization of citrate is not new. It has been observed in the past and is even metabolized by the bacteria under other environmental circumstances so it is more properly characterized as modification of existing function. It is an adaptation. Furthermore that particular example does not fall outside the boundaries of what should be expected of a blind search given the very modest increase in coherent information and thus small decrease in entropy more than covered by the resources brought to bear on the problem.

 

Oh, no no no no. This ability did not already exist; it was newly created during the experiment. The ability to use citrate may have existed elsewhere, but not in the sample in the experiment (not under oxidizing conditions). Just like if aliens 1 trillion light-years from here invented a lightbulb, doesn't mean that we didn't invent it too.

[cypress]

Genetic algorithms only need the same attributes as evolution has, to function. The most complicated would be the fitness function; but in evolution this comes naturally via death and reproduction. Likewise, you could implement a genetic algorithm with very little information, if you were to build every prototype yourself and input the outcome to the program; however, this would be absurdly time-consuming and expensive to do.

 

So because we design things to look like something natural, you conclude that the natural thing must be designed, is that it? I can make a rock, therefore rocks are designed?

 

Show me a genetic algorithm that does not import active information and I might agree with your claim. The reality is that they do not function without the designer artificially adding information to the search. In other words they function because the designer designed them to function. Likewise natural selection can only derive the diverse organisms we observe if there exists contiguous advantageous stepwise evolutionary pathways from one organism to another. I have asked for an example of a pathway just four steps long and I am certain that nobody will be able to provide even one. Yet if natural selection generated diversity then there would have to exist hundreds of millions of them each one of them hundreds of thousands long.

 

Actually I conclude that some things are natural because natural processes in operation today are observed to generate them. It is because life has the appearance of and contents that we observe only design processes construct and never natural processes that I conclude that design is a possible explanation for life. I personally believe there may be other processes yet to be discovered so I have not "concluded" that life was designed, however I note that currently design is the better explanation since there is no coherent natural explanation at this time that is causally adequate in that it cites known processes in operation today that have been shown to be capable of generating the effects ascribed to it.

 

 

What design? Who designed evolution, or what makes evolution not be a genetic algorithm?

 

Genetic algorithms are all designed and purposely import active information. Are you sure you want to place evolution into that category?

 

My prior commitment is a Young Earth Creationist, a position which I was forced after decades of kicking and struggling, to abandon. Look in the archives of this forum if you do not believe me, and you will see that for yourself. It is your prior commitments which drag you down, not mine.

 

I am speaking of your prior commitment (and certainty) to natural processes that have not been shown capable of generating the effects you have assigned to them. I know of no other prior commitments you hold and they don't seem relevant to this discussion. What prior commitment have you noticed I subscribe to because I don't think I have offered any.

 

Oh, no no no no. This ability did not already exist; it was newly created during the experiment. The ability to use citrate may have existed elsewhere, but not in the sample in the experiment (not under oxidizing conditions). Just like if aliens 1 trillion light-years from here invented a lightbulb, doesn't mean that we didn't invent it too.

 

Sure I don't see that I implied that the experiment previously contained the derived mutations, but the mutations were an alteration of a pre-existing capability to metabolize citrate in deoxygenated conditions. This is clearly an adaptation. Furthermore the mutations involved fewer than four steps and was well within the reach of a blind random search.

 

 

The first minute or two don't entirely apply here; it's the simulation I wanted to point at. The simulation of numerous random mutations occurring to bring about functional improvements to a device, in only a few thousand generations.

 

(Actually, I'd love to get the code of this simulator running for fun, and perhaps work on improving it...)

 

I don't find the slide show or the code particularly informative to this subject. The simulation quite obviously imports substantial information and also has a target defined by the designer. It is interesting in that it illustrates the narrative regarding selection very well but it does not in any way demonstrate that natural selection is actually capable of generating what it purports to generate. A better description of it is documentary/propaganda.

[iNow]

Actually, I'd love to get the code of this simulator running for fun, and perhaps work on improving it...

Ask, and ye shall receive:

 

 

One thing I should add. The program does not draw the clocks. It maintains, mates, and simulates them, but the drawing must be done manually from the genome matrix.

 

The program is written in MatLab.

 

The hand rotations that begin with 86 are 86,000 not 86.000. When YouTube compressed the video it becam hard to tell a comma from a period.

 

To download this video go to:

http://www.mediafire.com/?9e1zz000mq7

 

To download the program go to:

http://www.mediafire.com/?1umdtnwayyp

Alternate link to download Clock Evolution Program:

http://www.mediafire.com/?yygr5xhmzdm

 

If you wish to translate this video you can download the PowerPoint file from:

http://www.mediafire.com/?z33mymkjmzn

 

Learn the facts, spread the truth, and most importantly, Think About It.

Edited July 26, 2010 by iNow

[Cap'n Refsmmat]

I don't find the slide show or the code particularly informative to this subject. The simulation quite obviously imports substantial information and also has a target defined by the designer. It is interesting in that it illustrates the narrative regarding selection very well but it does not in any way demonstrate that natural selection is actually capable of generating what it purports to generate. A better description of it is documentary/propaganda.

It is interesting that you seem more interested in appeals to ridicule than in explaining your counter-claims.

 

There is minimal information imported into the simulation. It has building block -- gears, pendulums, springs, and so on -- and basic physics. Similarly, it could have had chemical "building blocks" and basic chemistry. A "target" is indeed in this case defined by a "designer", but that target could equally well have been "survive in the environment" and be defined by nature itself.

 

At no point in the simulation did the "designer" help the parts achieve the "goal" of being clock-like. He did not nudge the parts and say "here, this'll work better to make a clock." He didn't even give the clocks an initial condition or setup.

 

One can also observe numerous steps in the process from worthless clock to good clock. At no point did the "designer" give the system information on how to achieve any of these states. He did not "teach" it how to achieve proto-clock status or to develop a dial. This occurred through random mutation and selection.

 

The selection function could just have easily been a result of the natural environment; there is no constraint that it must be designed by a human. A natural selection function would likely be less efficient, but that is a matter of degree and time required.

 

Ask, and ye shall receive:

Indeed. I've downloaded the code but I'm not a MATLAB fan. If I get time to work on it, I may rewrite it in Python; it's surprisingly short code.

[Mr Skeptic]

Show me a genetic algorithm that does not import active information and I might agree with your claim.

 

This is what is called a strawman, and also called moving the goalposts. I never claimed that a genetic algorithm has no information, so asking me to prove it does not is misleading. Also, I never claimed that you can make information without information, which is what you are now asking. Finally, the fact that an algorithm contains information does not in any way impair it from generating new information -- an arbitrarily large amount of information, to be exact.

 

Show me the moon is made of green cheese, and I might agree with your claim.

 

Sure I don't see that I implied that the experiment previously contained the derived mutations, but the mutations were an alteration of a pre-existing capability to metabolize citrate in deoxygenated conditions. This is clearly an adaptation. Furthermore the mutations involved fewer than four steps and was well within the reach of a blind random search.

 

Yet it is one step. Now it is a capability. What is to stop another step from being taken based on this new ability?

[cypress]

It is interesting that you seem more interested in appeals to ridicule than in explaining your counter-claims.

 

Hmm, I reread my post and find it quite informative regarding the reasons the example is not relevant to this discussion. Was there something I said that requires elaboration?

 

There is minimal information imported into the simulation. It has building block -- gears, pendulums, springs, and so on -- and basic physics.

 

The designer only included functional parts. This amounts to an incredible amount of active information since the algorithm is guaranteed to apply only functional components. The designer also rigged an oracle to guarantee that correct fits are better than incorrect fits, more information.

 

Similarly, it could have had chemical "building blocks" and basic chemistry. A "target" is indeed in this case defined by a "designer", but that target could equally well have been "survive in the environment" and be defined by nature itself.

 

Survive in the wild is not a target as there are countless ways to survive in the wild, they are overlapping, codependent and no configuration is clearly "best".

 

At no point in the simulation did the "designer" help the parts achieve the "goal" of being clock-like. He did not nudge the parts and say "here, this'll work better to make a clock." He didn't even give the clocks an initial condition or setup.

 

No, importing information is a more subtle way to help out.

 

One can also observe numerous steps in the process from worthless clock to good clock. At no point did the "designer" give the system information on how to achieve any of these states. He did not "teach" it how to achieve proto-clock status or to develop a dial. This occurred through random mutation and selection.

 

Incorrect. The information added by the designer accomplished this task by eliminating innumerable unworkable combinations and the designer coded in an artificial fitness function that guaranteed success.

 

Are you defending this person's simulation because you actually believe it demonstrates that natural processes without aid and without importing any information (we can quibble about how much information was imported into this simulation but nobody can deny that some was imported and that information was critical to success of the simulation) can and does generate coherent information? If so can you provide a demonstration? If you don't believe this or you can't demonstrate it then why are we discussing this?

 

This is what is called a strawman, and also called moving the goalposts. I never claimed that a genetic algorithm has no information, so asking me to prove it does not is misleading. Also, I never claimed that you can make information without information, which is what you are now asking. Finally, the fact that an algorithm contains information does not in any way impair it from generating new information -- an arbitrarily large amount of information, to be exact.

 

First off unless I am mistaken, you quite clearly claimed that genetic algorithms only need the same attributes that evolution has to function. Perhaps you are claiming that evolutionary processes contain active information imported by a designer, but if not, then I have not mischaracterized your claim, I have not moved the goalpost, and you have claimed by induction that information can be made without information. Have I misunderstood your position?

 

Yet it is one step. Now it is a capability. What is to stop another step from being taken based on this new ability?

 

If it did take several more steps in fewer generations then it would be the example I was looking for. Speculation is nice but it is not evidence.

Edited July 26, 2010 by cypress

[Cap'n Refsmmat]

Hmm, I reread my post and find it quite informative regarding the reasons the example is not relevant to this discussion. Was there something I said that requires elaboration?

One must remember during debates that one's opponent often takes opposing positions and therefore requires convincing.

 

The designer only included functional parts. This amounts to an incredible amount of active information since the algorithm is guaranteed to apply only functional components. The designer also rigged an oracle to guarantee that correct fits are better than incorrect fits, more information.

First, you're changing the "information" goalposts. The original definition of "information" was in the genome, or in this case the configuration of components. Now you're saying that the kind of component also counts as information, and that the fitness function is information as well. The fitness function is a direct result of the environment, not the organism's genome. Make up your mind. What does "information" encompass?

 

Furthermore, it's naive to call a pile of gears and pendulums "functional parts." What is that supposed to mean? That a gear sitting in a lonely heap on a table is somehow "functional"? No, their "function" comes from their "genome", or rather the arrangement of parts inside the proto-clock. Merely supplying a gear does not help arrange these parts in any way.

 

The choice of gears and pendulums is arbitrarily. If my fitness function is "telling time," anything from a pile of sand to a stick stuck in the ground would suffice as starting places. Numerous different parts could be supplied to similar effects.

 

(Hmm. It'd be fun to evolve an accurate hourglass with minute and hour chambers. Perhaps day of the week too.)

 

Survive in the wild is not a target as there are countless ways to survive in the wild, they are overlapping, codependent and no configuration is clearly "best".

There are countless ways to make clocks, they are overlapping, codependent and no configuration is clearly "best." Your point?

 

There are countless ways to survive as a wild clock in the vast steppes of Timeberia.

[Mr Skeptic]

First off unless I am mistaken, you quite clearly claimed that genetic algorithms only need the same attributes that evolution has to function. Perhaps you are claiming that evolutionary processes contain active information imported by a designer, but if not, then I have not mischaracterized your claim, I have not moved the goalpost, and you have claimed by induction that information can be made without information. Have I misunderstood your position?

 

It seems that rather than having misunderstood my position, you are employing the tactic known as "begging the question". I've said the algorithm has information (all algorithms do, by the way). And you assume that the presence of the information is proof for a designer -- the very thing under dispute. First prove that the presence of information proves a designer, and then you can speculate about evolution being designed.

 

Perhaps this is a language issue. I consider any random string of data to be information. Because it is. If that random string had value of some kind for any reason, then it would just as clearly be information as would a string of data that the value of could be seen easily. You seem to have a more restricted view of what information is, but I don't know what yours is. If this is indeed our point of disagreement, I could expand on it further, and I would ask that you define what you mean by "information".

 

If it did take several more steps in fewer generations then it would be the example I was looking for. Speculation is nice but it is not evidence.

 

Right, but what would it prove? Even were I to give an example, where 4 consecutive new abilities were evolved within a reasonable time while being observed, you would simply change your demand. Look at it this way: just because something has not been directly observed to happen, does not mean it didn't/doesn't. I demand that you show me someone being raised from the dead by holiness/faith, or I won't believe it ever did. I demand that you show me Napoleon conquering even just 1 country, or I won't believe that he conquered several. Not only does your demand prove nothing, but also you will change it were an example given.

[cypress]

First, you're changing the "information" goalposts. The original definition of "information" was in the genome, or in this case the configuration of components. Now you're saying that the kind of component also counts as information, and that the fitness function is information as well. The fitness function is a direct result of the environment, not the organism's genome. Make up your mind. What does "information" encompass?

 

Coherent Information trims away alternatives. When I say "It will rain", I have conveyed very little information but when I say "It will rain between 6 and 7 this morning in New Orleans" I have eliminated a great deal of alternative scenarios and have conveyed a great deal of information. In information theory the quantity of information is measured by the quantity of alternatives it eliminates. So when I said that the designer eliminated a huge number of alternative configurations that person inserted large amounts of information into the simulation.

 

Furthermore, it's naive to call a pile of gears and pendulums "functional parts." What is that supposed to mean? That a gear sitting in a lonely heap on a table is somehow "functional"? No, their "function" comes from their "genome", or rather the arrangement of parts inside the proto-clock. Merely supplying a gear does not help arrange these parts in any way.

 

I find your claim vacuous. Something is functional if it has a configuration that can be employed for any useful task. I am surprised you would deny this reality since these parts can and are used for multiple purposes with the same configuration. Likewise protein components have configurations that are functional and a great deal more that are disorganized blobs with no shape conformance, no stability and therefore no function because we know that protein components require shape stability, shape conformance and matched binding sites to be functional. Sequence conformance is generally not required.

 

The choice of gears and pendulums is arbitrarily. If my fitness function is "telling time," anything from a pile of sand to a stick stuck in the ground would suffice as starting places. Numerous different parts could be supplied to similar effects.

 

How can you demonstrate this to be true. I don't see any way for you to say they are arbitrary. If anything could have been used then the simulation would require these anythings to be included. They were not included because the simulation would have failed.

 

 

There are countless ways to make clocks, they are overlapping, codependent and no configuration is clearly "best." Your point?

 

There are countless ways to survive as a wild clock in the vast steppes of Timeberia.

 

I don't know if there are or not, but the designer of the simulator programmed it with a definition of best as opposed to "survive". Only those that run closest to the same speed as time survive. This is artificial, it is active information and it guarantees the simulation will suceed.

 

Do you actually believe this simulation demonstrates that natural processes are capable of generating the information required to generate functional systems?

Edited July 26, 2010 by cypress

[cypress]

It seems that rather than having misunderstood my position, you are employing the tactic known as "begging the question". I've said the algorithm has information (all algorithms do, by the way). And you assume that the presence of the information is proof for a designer -- the very thing under dispute. First prove that the presence of information proves a designer, and then you can speculate about evolution being designed.

 

From the beginning I have been speaking of evidence as opposed to proof and if you read my posts you would note that I am not assuming this is proved. The questions I asked were rhetorical to help illustrate how I believe your complaints were misplaced. Uniform experience and observation confirms that intelligent agents generate coherent encoded information but there are no confirmed cases of natural processes accomplishing the same. This is evidence but since alternatives are still in play, it is not proof. If you believe that my statement is incorrect and that natural processes do generate coherent information then let’s discuss it. Otherwise the minor complaints you raise are side issues that I don’t think need to be discussed any further.

 

Perhaps this is a language issue. I consider any random string of data to be information. Because it is. If that random string had value of some kind for any reason, then it would just as clearly be information as would a string of data that the value of could be seen easily.

 

Random strings of data is .... well.... data so there is no reason to redefine it as information. If random strings “had value” in the sense that it could be processed to produce functional systems, then it indeed would be information in the sense I mean. Do you have any example of random data containing large amounts of this kind of information capable of being processed to produce functional systems?

 

You seem to have a more restricted view of what information is, but I don't know what yours is. If this is indeed our point of disagreement, I could expand on it further, and I would ask that you define what you mean by "information".

 

I am speaking specifically of encoded digital information such that when processed generates functional systems. Computer code is an example. Instruction manuals are another. The information encoded into DNA is a third example.

 

 

Right, but what would it prove? Even were I to give an example, where 4 consecutive new abilities were evolved within a reasonable time while being observed, you would simply change your demand.

 

Would I? Do you read minds? I have asked for this example because the inability to produce one is confirmation of the claim I have made. If you produce an example then I will have to retract my claim that random error and selection does not produce novel function and body plans within the available geologic timeframe.

 

Look at it this way: just because something has not been directly observed to happen, does not mean it didn't/doesn't.

 

Perhaps not, but unless the claim cites a causally adequate process currently in operation and known to be capable of generating the claimed effects it is not a scientific claim, it is metaphysical. Are you making an appeal to the unknown?

 

I demand that you show me someone being raised from the dead by holiness/faith, or I won't believe it ever did. I demand that you show me Napoleon conquering even just 1 country, or I won't believe that he conquered several. Not only does your demand prove nothing, but also you will change it were an example given.

 

Demand what you like but I don’t make these claims. For someone who objects to logical fallacies you seem awfully fond of them.

 

So what we have at this point is the fact that the physical constants in this universe are fine-tuned and we know that mind is capable of fine-tuning while we are unaware of any natural process capable of fine-tuning. We also have coherent encoded digital information as fundamental and necessary component of biological systems and we know that mind is capable of generating encoded digital information while once again we are unaware of any natural process capable of generating this specific kind of information.

 

These are two pieces of evidence that suggests this universe and life in it was created. Furthermore since there is no causally adequate natural explanation for attributes of the universe or life in it, creation is currently the best explanation for both. This may change with new data but it is silly to deny the current reality.

[Mr Skeptic]

Very well then, it seems "information" seems to be our main point of contention. If you can give even one example of a random string of data that has no value of information, while allowing me to provide context for said data, I shall change my mind. On the other hand, if I can demonstrate that any arbitrary string of data will have value as information with a given context, then I would expect you would change your mind. Sound fair?

 

On my part, I apologize for some of the ridiculous "demands" I have made, although I am sure you realize those were more of a rhetorical device, examples of why your demand could be seen as unreasonable and most definitely didn't prove anything. While I cannot read minds, I am 99.9% sure that you would change your demand were I to give you one example fitting that description that I can think of.

 

Also, it really makes no sense to claim that natural processes cannot generate information, since the information produced by evolution is both observed and encoded and digital, as you requested. Similarly with fine-tuning, evolution is a natural process that can produce organisms fine-tuned to an environment (which has also been observed). Of course, since I'd also consider random data to be information, I think it best we work that out first.

[Klaplunk]

This post was by mistake. Posted in the wrong thread.

Edited July 26, 2010 by Klaplunk

[cypress]

Very well then, it seems "information" seems to be our main point of contention. If you can give even one example of a random string of data that has no value of information, while allowing me to provide context for said data, I shall change my mind. On the other hand, if I can demonstrate that any arbitrary string of data will have value as information with a given context, then I would expect you would change your mind. Sound fair?

 

No I don't think so. Your use of the term "value of information" seems altogether different than "coherent digitally encoded information that when processed generates functional systems". Did you intend to change the meaning?

 

On my part, I apologize for some of the ridiculous "demands" I have made, although I am sure you realize those were more of a rhetorical device, examples of why your demand could be seen as unreasonable and most definitely didn't prove anything. While I cannot read minds, I am 99.9% sure that you would change your demand were I to give you one example fitting that description that I can think of.

 

I recognize that you were attempting to make analogies, but I consider them to be misapplied. I am also somewhat disappointed in your lack of trust. Unfortunately we will likely not face this issue because there are no such examples. The neo-Darwinian narrative of gradual change over time predicts millions upon millions of these pathways and yet we can identify none. How odd.

 

Also, it really makes no sense to claim that natural processes cannot generate information, since the information produced by evolution is both observed and encoded and digital, as you requested.

 

It makes a great deal of sense because we don't know of any evolutionary process in operation today that is observed to have generated any significant quantity of new information required for new form or function beyond what little can be shown to be derived by a blind search plus the probabilistic resources available to search shape and chemical affinity space using population genetics modeling. It is a tautology for you to presuppose that there exists such a process. If observations of known processes existed then you would be able to provide an example of one of these multistep evolutionary pathways we discussed. Again let me remind you that research has not yet identified any pathway greater than three steps but we need pathways in the millions of steps.

 

Similarly with fine-tuning, evolution is a natural process that can produce organisms fine-tuned to an environment (which has also been observed).

 

Once again I must correct you. We have a metaphysical belief but we do not know if natural process produce new organisms. We do know that evolution can generate adaptations of pre-existing function but these adaptations on average destroy as much or more information as they generate. Adaptations tweak the dials within the range minimal range of survival but they must be in that range to begin with for adaptation to work.

 

Of course, since I'd also consider random data to be information, I think it best we work that out first.

 

Yes we should, but first we need to adjust the criteria for what constitutes the kind of information we are talking about. If we relax the criteria enough we can describe nearly anything as information.

[Mr Skeptic]

No I don't think so. Your use of the term "value of information" seems altogether different than "coherent digitally encoded information that when processed generates functional systems". Did you intend to change the meaning?

 

If I understand your sentence, you are saying that everything you have just told me contains no information, since it does not generate a functional system. My measure of quantity of information is approximately the size of the data when it is as compressed as possible. Function of it or lack thereof has nothing to do with any particular string of information, since there is no difference between useful and useless information, not without looking at the context.

 

My invitation to suggest a string of data that you believe contains no information is still standing.

 

I recognize that you were attempting to make analogies, but I consider them to be misapplied. I am also somewhat disappointed in your lack of trust. Unfortunately we will likely not face this issue because there are no such examples. The neo-Darwinian narrative of gradual change over time predicts millions upon millions of these pathways and yet we can identify none. How odd.

 

On the contrary, we have identified several. The metabolic pathway of glucose is a good example. No, what you are asking is that we observe an event which the theory predicts to be unlikely to happen within the confines of a lab within a reasonable amount of time, and then pretending that is evidence against the theory, or that it is a reasonable suggestion and you remain open-minded. In this your demand is far more close-minded than my demand to see someone raised from the dead, since the latter is something that not only has (according to the Bible) happened about 5 times in a few thousand years but also that only requires a little bit of faith the size of a mustard seed for anyone to do.

 

It makes a great deal of sense because we don't know of any evolutionary process in operation today that is observed to have generated any significant quantity of new information required for new form or function beyond what little can be shown to be derived by a blind search plus the probabilistic resources available to search shape and chemical affinity space using population genetics modeling. It is a tautology for you to presuppose that there exists such a process. If observations of known processes existed then you would be able to provide an example of one of these multistep evolutionary pathways we discussed. Again let me remind you that research has not yet identified any pathway greater than three steps but we need pathways in the millions of steps.

 

But why the need to do it the difficult way? Why the requirement to observe something which the theory itself predicts you could have to wait millions of years to observe (really, the odds of an ability sufficiently novel to satisfy you, and on top of that for it to happen to the same species limited to a minuscule population (1 ton would be like 0.00000001% of earth's biomass) within a literal blink of an eye (0.000001%) of geologic time, and if you wanted them to build off each other than it has to relate to 1 out of thousands of genes three times, which would be a chance of 0.0000001% of a chance if you wanted 4 such examples and the four did happen anyways within the lab population and experiment runtime.

 

Why such an unlikely thing, when you could simply look at the historical and DNA evidence? Like the Bible, evolution makes specific predictions as to what DNA should look like. According to the Bible, DNA should look like the quantity of DNA that fits on a big wooden boat, and the maximum number of alleles depending on the type of animal -- 4 alleles per unclean animal, 14 alleles per clean animal or bird, and only 1 allele in the Y chromosome and only one type of mitochrondrial DNA for unclean animals. Evolution predicts that DNA will look like it was copied and mutated, with less mutations to the more vital genes due to natural selection and probability of improving an already well-refined gene, and more mutations in neutral DNA. Guess which one turns out to be true?

 

Once again I must correct you. We have a metaphysical belief but we do not know if natural process produce new organisms. We do know that evolution can generate adaptations of pre-existing function but these adaptations on average destroy as much or more information as they generate. Adaptations tweak the dials within the range minimal range of survival but they must be in that range to begin with for adaptation to work.

 

This is quite obviously false. Many times in evolutionary history genes have been copied wholesale, and then one of the copies modified to a new purpose. This does not destroy the original information.

 

Yes we should, but first we need to adjust the criteria for what constitutes the kind of information we are talking about. If we relax the criteria enough we can describe nearly anything as information.

 

When there is no difference between two things it is reasonable to say they are identical. So yes, I do consider any string of data to contain information, though you are welcome to try to show a string of data with no information (again though, I shall be the one to put some context to said string).