Open thread and responses to stuff
by KrauzeAny repressed thoughts? Unload them here. Some old pieces to get the conversation going:
Over at Middlebrow, John Mark Reynolds reviews Michelle Goldberg's Kingdom Coming.
At Right Reason, Graeme Hunter responds to Michael Shermer on confirmation bias.
And at Thinking Christian, Tom Gilson takes a closer look at an article in the Journal of Clinical Investigation with the ominous title "Defending science education against intelligent design: a call to action".
August 7th, 2006 at 3:06 pm
from Gutz:
Only in the sense of being motors. When you look at how they function internally, they are massively different. Man-made motors are made of rigid parts. Moving parts are lubricated so as not to stick. Biological "motors" are built of flexible proteins; they use "parts" that physically change shape. They don't require lubrication; in fact, sticking together of the moving parts is required for function as the molecules must go through complex cycles of binding, changing shape, and unbinding–a motive mechanism very different from that used in man-made machines. We are only beginning to understand how to design molecules to do these sorts of things–the best strategies we have at present are to copy from nature, or to use mutation and selection to evolve what we need.
Comment by trrll — August 7, 2006 @ 3:06 pm
August 7th, 2006 at 3:21 pm
trll:
Yeah thats the sense I'm talking about.
trll:
Actually proteins can act as rigid parts. Flexible proteins can act like levers and springs, for example. There are some differences, but they are trivial. There are many more similarities , enzyme catalysis, for example, has a very machine-like nature, with coordinated , specific moving parts, and other similarities I mention here . Some characteristics from another paper:Like macroscopic machines, molecular machines are characterized by(i) the kind of energy supplied to make them work, (ii) the kind of movement performed by their components, (iii) the way in which their operation can be controlled and monitored, (iv) the possibility to repeat the operation, (v) the time needed to complete a cycle of operation, and (vi) the function performed. (here ) There more than just characteristics that are shared between human design and molecular systems, the NPC , for example, even uses a polymer brush-like mechanism to keep unwanted material from passing through. Polymer brushes are also common in human technology.
Trll:
There are no demonstrations of mutation and selection alone evolving a complex machine.
Comment by Guts — August 7, 2006 @ 3:21 pm
August 7th, 2006 at 3:33 pm
RandomThoughts:
Salvador Cordova will be on International TV next week along with Caroline Crocker, Edward Sisson, and IDEA GMU president Christine Chenette. See Coral Ridge Hour. (Disclaimer: I was intereviewed by their reporters, anything else said during that show is Coral Ridge's business).
I think Cornell geneticist John Sanford's book Genetic Entropy would be a good topic of study and discussion.
I have proposed an exploration of front-loaded evolution even a YEC would be interested in: Marsupials and Placentals
Nachman's Paradox has not been pursued enough.
A large number of "science" blogs which Nature highlighted discuss ID, and thus the topic will not die easily.
I think it would be cool if Mike Gene, under the pseudoname Mike Gene, gets a peer-reviewed publication to his name. After all, scientific American mentioned him!
Comment by Salvador T. Cordova — August 7, 2006 @ 3:33 pm
August 7th, 2006 at 3:36 pm
Kewl, Sal! Too bad I don't have TV…
Comment by Joy — August 7, 2006 @ 3:36 pm
August 7th, 2006 at 6:35 pm
In other words, ignoring the actual design.
Note that the human designed "molecular machines" in the article you cite are very different from all macroscopic human machines and are explicitly based upon biological examples. As I said before, "[one of] the best strategies we have at present are to copy from nature." So it is not a matter of biological molecular machines resembling human-designed machines, it is the other way around: human designs for molecular machines are explicitly and intentionally designed to copy the way that biological "molecular machines" work.
Comment by trrll — August 7, 2006 @ 6:35 pm
August 7th, 2006 at 6:39 pm
Oh, I'd in the spirit of open inquiry, I like to show others I study other hypotheses. Here is some data on a YEC/Panspermia Hypothesis.
Salvador
Comment by Salvador T. Cordova — August 7, 2006 @ 6:39 pm
August 7th, 2006 at 7:13 pm
trll:
Actually it is demonstrable that each of those characteristics are indeed common between molecular machines, and man-made machines, not just man made molecular machines, just as the author stated. The author then applies these shared characteristics in the design of a minature machine. It's also why he says: "Nature provides living systems with complex molecules called motor proteins which work inside a cell like ordinary machines built for everyday needs. Because of these biological engines, we can walk, talk, and even think." It is indeed a matter of biological molecular machines resembling human-designed macroscopic machines. Nature just helps us fill in some of the blanks, when it comes to minaturizing our designs. There are piles of articles on this subject alone.
Comment by Guts — August 7, 2006 @ 7:13 pm
August 7th, 2006 at 7:35 pm
trrll:
Some of the most compelling results have been similarities in nonfunctional DNA sequences, such as pieces left over from viral infections and damaged, nonfunctional pseudogenes, where there is no "intelligent" justification for such duplication
Intelligence does not require an absence of malfunction IOW perfection. None of the above refutes imputing intelligence as a causal factor at the origin of an initial genome. There is no force of nature implicated in generating encoded conventions and systems functioning according to them. There is good reason to believe the generation of such was the product of intelligence.
Comment by Bradford — August 7, 2006 @ 7:35 pm
August 7th, 2006 at 7:39 pm
trrll: Note that the human designed "molecular machines" in the article you cite are very different from all macroscopic human machines and are explicitly based upon biological examples. As I said before, "[one of] the best strategies we have at present are to copy from nature." So it is not a matter of biological molecular machines resembling human-designed machines, it is the other way around: human designs for molecular machines are explicitly and intentionally designed to copy the way that biological "molecular machines" work.
Of what significance is this observation?
Comment by Bradford — August 7, 2006 @ 7:39 pm
August 8th, 2006 at 1:04 am
It certainly is the case that "not very intelligent design" (NVID) is more consistent with many aspects of biology–the backwards design of the eye, for example–than intelligent design. Biology does in fact resemble a certain type of human design–say if a team of human designers were brought in on a massive software project for which the documentation had been lost, each person given only a part of the project to work on, and not allowed to communicate with one another.
But there are some features that are hard to explain even in terms of NVID. For example, it is difficult to understand why we should share code remnants of viral retroviruses–of exactly the same kind left by viral infection, and in the same position in the genome–with evolutionarily related primates, except in terms of common descent. NVID is not sufficient to explain this–you'd need a designer who is intentionally trying to deceive.
Natural selection does not require such a force. In biology, encoding turns out to be a consequence of simple binding. If a protein or RNA happens to bind to a particular sequence (and biological molecules tend to be promiscuously "sticky"–things bind to one another just by chance) then that sequence can be said to "encode" for that protein or RNA. And selection will either enhance or eliminate that encoding, depending upon whether it hurts or helps fitness.
Simply pointing out how easy it is to become confused about causality. There is a tendency to interpret aspects of the natural world that appear to resemble products of human design as indicative of a human-like intelligence at work. But one of the greatest inspirations for human design has always been nature. So it is perhaps more reasonable to see such similarity as indicative of the ability of humans to imitate the products of natural selection.
Comment by trrll — August 8, 2006 @ 1:04 am
August 8th, 2006 at 10:53 am
There is no force of nature implicated in generating encoded conventions and systems functioning according to them. There is good reason to believe the generation of such was the product of intelligence.
Natural selection does not require such a force. In biology, encoding turns out to be a consequence of simple binding.
Baloney. Encoded messages result from the ink in the pen right? Binding means nothing when there is no replicating organism within which binding occurs.
If a protein or RNA happens to bind to a particular sequence (and biological molecules tend to be promiscuously "sticky""“things bind to one another just by chance) then that sequence can be said to "encode" for that protein or RNA.
That's the most absurd statement I've come across in this forum. The encoded nature of a nucleic acid lies in the content and sequential order of its codons as well as a biological capacity to translate the information contained to amino acid chains which also must be properly sequenced to confer function. Chance interaction of molecules does not result in codes. And of course the issue of nucleic acid replication is ignored.
Comment by Bradford — August 8, 2006 @ 10:53 am
August 8th, 2006 at 11:30 am
A few interesting sections on the argument from design can be found at the link below (especially 24-30):
Don't worry Salvador, such a scenario of giving up 'Design' for 'General Laws' would seem suitable for your talents. That is, if you give up your scientistic defense of theism. Glad to see you reading someYEC stuff again - how about some mature evolutionary creationism?
Comment by g arago — August 8, 2006 @ 11:30 am
August 8th, 2006 at 11:47 am
Further with the same author:
This use of 'intelligent design' dates to 1889. Have other earlier uses been found also?
g.A.
Comment by g arago — August 8, 2006 @ 11:47 am
August 8th, 2006 at 11:56 am
Since you are a biologist I'm baffled at your way of interpreting the inverted retina. There are some scientific evidences that the inverted retina of vertebrates is a necessity not simply bad design. OTOH it sounds like an argument from ignorance: "Since we currently can not identify its function, it serves no function and it is definitely bad design."
It seems the myth of inverted retina is only popular among evolutionists. I asked about role of the inevrted retina from an ophthalmologist friend and he replied that it is speculated that the inverted retina serves vital functions that some of them are already identified. However I forgot to ask him if he is a creationist or not. Next time I'll ask about it.
Comment by Farshad — August 8, 2006 @ 11:56 am
August 8th, 2006 at 12:06 pm
trrll said,
So because trrll would have designed the human eye differently, the eye must not be a very intelligent design?!?! There's a word for this! LOL!!
Consider that a single rod can detect a single photon. If trrll was the most intelligent human in history and tried to design a mechanism of similar capability, he would make a rod out of an avalanche photodiode cooled with liquid nitrogen (the current best human design).
If trrll was the designer, we might have to carry around liquid nitrogen tanks, - but DAMMIT, those retinas would be RIGHT SIDE UP!
Comment by chunkdz — August 8, 2006 @ 12:06 pm
August 8th, 2006 at 1:54 pm
How much intelligence does it take to realize that it is not a smart design to run the wiring in front of the photosensors where they will scatter the light? Or if you do make that dumb mistake, to run the wires around the edges to the back instead of punching through near the middle where it will make a blind spot in the visual field? Sure, the eye has enough sensitivity to capture a single photon, so running the nerves in front probably doesn't matter if all you want to do is tell light from dark (which is probably all that the earliest eyes did). On the other hand, if you are interested in visual acuity, it's downright stupid. Of course natural selection doesn't have the foresight to realize that a design that works fine for simple light detection might not be the best basis on which to build a high resolution camera, but natural selection pretty much has to work by optimizing what it already has–it doesn't have the luxury of an intelligent designer who can throw out the original concept when design goals change and go back and redesign the system from the scratch.
If trrll was the most intelligent human in history and tried to design a mechanism of similar capability, he would make a rod out of an avalanche photodiode cooled with liquid nitrogen (the current best human design). If trrll was the designer, we might have to carry around liquid nitrogen tanks, - but DAMMIT, those retinas would be RIGHT SIDE UP!
No, just a design more like the octopus: photosensors in front, nerves in back, no blind spot. Pretty elementary–a child could figure it out. But natural selection doesn't have even the foresight of a child. If it is lucky enough to start down the best path, all well and good. If it is unlucky to start with an inferior approach, it will make it work as well as it can–so we ends up with odd kluges like the mammalian retina or the panda's "thumb."
Comment by trrll — August 8, 2006 @ 1:54 pm
August 8th, 2006 at 2:28 pm
Stupid? It seems to me that humans have been pretty successful on planet Earth. Obviously, the design of the eye must have been pretty good in order to put humans at the top of the food chain. If you want to prove bad design, you have to prove it's ineffectiveness or it's inefficiency. In other words, rate the eye on it's performance, not the way it's wired. The human eye has a proven succesful track record, and it's not enough for you to say that a child would have assembled it differently.
So the octopus eye is intelligently designed, but mammalian eyes are not?
And why stop there? Why not add a telephoto option, infra-red sensors, UV shielding, and a macro-focus mode? Hey, that designer really left out a lot of gadgets that should have been in there. When would the eye be good enough to be considered an intelligent design? Do we all need little Hubble Telescopes in our eye sockets? If your argument is that "no true designer" would design a mammalian eye, then I'll just have to ask you what else you know about the 'true' designer that you're not telling us.
Comment by chunkdz — August 8, 2006 @ 2:28 pm
August 8th, 2006 at 4:37 pm
Of course, considering that all of our vertebrate competitors have the same suboptimal eye design that we do, that hardly seems like a great endorsement of our eye design. But I don't want to minimize the power of natural selection to wring the last bit of performance out of even a suboptimal design. The argument is not that the eye is bad–it is that it is not designed in the way that an intelligent designer would do it, with knowledge of ultimate goals making it possible to choose an optimal approach at the outset, rather than coming up with clever work-arounds. ID advocates like to point at the similarities between the design of nature and the design of man. But if you are going to do that, then you also are obliged to make note of the differences. Do you know of any type of camera designed by man that runs the wiring in front of the light sensor? Sure, we could do it if we had to, and probably end up with decent performance–wiring can be made almost transparent–but a human designer would anticipate the problem early in the design process and avoid the need for such a work-around.
If you think that you can make a case for the absence of these features being evidence against intelligent design, feel free to do so. Just as long as you are clear that it is your argument rather than mine.
Comment by trrll — August 8, 2006 @ 4:37 pm
August 8th, 2006 at 5:58 pm
Why would a designer build a plane that was inherently aerodynamically unstable? One might argue that to make a plane that couldn't glide was poor design - that even a child knows that a plane must be able to glide. One could point to perfectly fine jet designs that were inherently stable, and then wonder at the stupidity of the designer who would make a jet that can't even glide or "fly-by-wire".
An uninformed layman's opinion may be that the US Air Force was stupid for designing the F-16 to be mechanically unstable. The logic might not seem evident until you discover that the jet is stabilized by flight control computers, and that "relaxed static stability" is actually a design 'feature', not a flaw. What seemed "stupid" is actually discovered to be "brilliant" as evidenced by the superior performance characteristics of the F-16 Fighting Falcon.
Could it be that the eye exhibits a similar design that seems counterintuitive on the surface, but reveals a superior design upon the understanding of the entire system?
We know that the brain acts as the control computer for visual processing. We know that the brain corrects for the "blind spot" created by the optic nerve, as well as the inverted image on the retina. What we find is a design that seems inherently flawed, but when controlled by the brain, becomes a highly efficient visual sensor. Other advantages of the reversed retina design are that rods and cones can come in contact with the retinal pigment epithelial cells, distributing nutrients to the retina, and recycling photopigments. The reversed orientation absorbs excessive UV radiation, and light loss is compensated for in the central fovea by displacing the cells sideways to allow light regulation at the most sensitive part of the retina.
I wouldn't call these design features "clever workarounds" anymore than I would call the F-16's Flight Control Computer a "clever workaround". The 'instability' was designed into the system from the beginning. That's why I pointed out that for you to comment on the poor design of the marvelously complex human eye is 'hubris'. I would say the same of anyone who in their ignorance complained about the 'stupidity' of a plane that can't glide.
Comment by chunkdz — August 8, 2006 @ 5:58 pm
August 8th, 2006 at 6:17 pm
POLITICS!!! YIKES!!!
In all this back-and-forth on the Armstrong deal, I get the feeling that others don't appreciate the irony as much as I do. Yeah, irony is part of the whole clown college curriculum, so maybe I've got more formal training on that than others. So because that may be so, I offer a short front-page post over at a website so overrun today you may or may not be able to access, about the may or may not be able to access silliness involved in an important primary challenge we 'progressives' are keeping track of today…
From Markos Moulitsas, owner of "DailyKos.com" and author of "Crashing the Gate," and a very interesting young man with some very interesting ideas. About a very funny couple of days' worth of charges and counter-charges related to the fact that Joe Lieberman's campaign website is down right at the culmination of a desperate primary battle for his political life. The site says Joe didn't pay the bill, the shills claim it's been hacked. Do read the comments, they're hilarious, and have pertinence to the Armstrong/Crowther issue we've been playing with this week…
MyDD is down. It's Lieberman's fault.
Comment by Joy — August 8, 2006 @ 6:17 pm
August 8th, 2006 at 7:19 pm
No the brain does not "correct" for the blind spot. It can't correct for what it cannot see; the best it can do is conceal it by interpolating, sometimes incorrectly, from the surroundings. This is basically sweeping the problem under the rug. It leaves vertebrates with a vulnerability, in that they cannot perceive a threat to the eye whose image falls into the blind spot, and will not blink to protect the eye.
I'm not sure what you mean by "correcting for the inverted image." There is not a right way or a wrong way for an image to be displayed on the retina, and (in contrast to the blind spot) no information is lost by displaying the image upside down. The retina is an information gathering device that enables the brain to build up a model of the outside world. The brain is very flexible in using whatever information it can gather for this purpose, although it sometimes makes mistakes.
Nutrients can be distributed to the retina from the back whichever way it is oriented. The reversed retina design actually makes it more difficult to get nutrients to the nerve cells of the retina, because they are in front of the photoreceptors and further from the blood vessels, which have to be kept behind the retina to avoid obstructing vision (there are illnesses in which blood vessels grow in front of the photoreceptors, destroying vision).
There is no need to use nerve cells to absorb excessive UV radiation. They are not particularly well-suited for that purpose (having other business). The cornea and lens are much better suited and situated for that purpose, and provide almost all of the UV filtration. And displacing the cells sideways is just a kluge to get around a problem that wouldn't exist if the eye were designed the right way around to begin with.
Comment by trrll — August 8, 2006 @ 7:19 pm
August 9th, 2006 at 11:52 am
Actually, binding works even without an organism. For example, random peptides can catalyze chemical reactions by binding to the transition state. The sequence of random peptide that catalyzes a particular reaction can reasonably be said to code for that reaction.
All of which occurs based upon binding–what sticks to what.
There is no reason why a chance interaction, optimized by selection, could not result in the kind of pattern of binding specificity that we refer to as a "code."
Replication also occurs by binding. Nucleotides align themselves in a particular order by binding to a complementary sequence on the other strand. Alignment of reactants is the simplest form of catalysis, reducing the entropy barrier to a chemical reaction. Once you have catalysis, it can be optimized by selection to achieve greater efficiency, such as by binding to and stabilizing a transition state.
Comment by trrll — August 9, 2006 @ 11:52 am
August 9th, 2006 at 12:52 pm
trrll,
You previously pointed to the octopus eye as an example of a better design than the human eye. Obviously, you are not basing this opinion on overall performance, because that would clearly point to the human eye as being superior in function. If not function, then what is your main reason for assuming that cepholopods have better eye design?
Is it the placement of the wiring? The inversion of the retina? You wrote:
This is what would happen in a Canon Powershot SD600, but not in the human eye. While counterintuitive for a camera designer, it doesn't seem to be a problem for the eye. Even Dawkins admits this.
Is the problem the blind spot? What our eyes don't correct with stereo vision, and stochastic mapping, the brain sufficiently interpolates. Doesn't seem to be much of a problem for human survival, and with stereo vision we get added depth perception! Might seem like bad design for a Nikon D200, but works great for us humans.
Is the problem with visual acuity? You wrote:
But if you want to compare verted eyes like those in the octopus, with inverted eyes like, say, an eagle, then I can't understand the 'stupid' assertion.
So the human eye is superior in overall function to the cephalopod eye, yet you claim that it is poorly designed. Your earlier posts offer an insight into your reasoning.
and…
It seems that you think that if eyes are not designed to approximate cameras, then they are poorly designed. But cameras are weak approximations of the total visual information processing system in our heads. So it is simply your intuition that tells you that the eye should be designed more like a camera.
That's as silly as saying that the F-16 is inferior to the Wright Flyer because the F-16 was designed with inherently instability.
Comment by chunkdz — August 9, 2006 @ 12:52 pm
August 9th, 2006 at 1:45 pm
No, I pointed to it as an example of a more intelligent design, in that it obviates the need to work around a number of unnecessary obstacles. It also demonstrates that there is no fundamental biological constraint that makes it impossible to put the photoreceptors in front.
The key word here being "most." Once again, the argument isn't that it is not possible to get acceptable performance with the nerve cells in the way, the argument is that it is not an intelligent way to do it. Since when does "not as silly as it sounds" equate to "intelligent?" Why have any light scattered by nerve cells that could just as well be situated behind, with the additional advantage of eliminating the blind spot?
So one might predict, based on the hypothesis of intelligent design, that the designer might use a backwards eye design for animals with stereo vision, and use a photosensors-in-front design for animals without stereo vision, such as those with eyes on opposite sides of their heads. Does reality agree with this prediction? And by the way, "stochastic mapping" isn't going to help you blink when something is flying at your eye, right in your blind spot.
Anything unnecessarily scattering light is going to interfere with acuity, so it is hard to see how the eyes of an eagle wouldn't be improved with the photosensors in front. Yes, the eyes of the eagle are good enough–that is one of the signal differences between natural selection and intelligent design. Natural selection will settle for good enough; an intelligent designer will seek to do things the best way.
I would say that it is a bit more than weak intuition that says that any kind of optical sensor will work better without unnecessary stuff in the light path.
Stability and agility are opposing goals, so any vehicle must be a trade off between those two, and different vehicles will have different trade offs. But there is no plausible trade off of design goals that would make it reasonable to put unnecessary stuff in the light path of an optical sensor.
Comment by trrll — August 9, 2006 @ 1:45 pm
August 9th, 2006 at 3:27 pm
trrll,
You said
and
I found some.
And this…
An inverted retina seems to be a very intelligent compromise, and probably optimal for vertebrates.
You have made your judgement about the retina without considering the system as a whole. Kind of like looking at the Hubble telescope and saying "what kind of stupid designer put a hole in the primary mirror?!?!?"
Learn about compound telescopes as a whole system and the understanding becomes clear.
Learn about the eye as a whole system and the understanding becomes clear.
Comment by chunkdz — August 9, 2006 @ 3:27 pm
August 9th, 2006 at 6:41 pm
Correct. We know this from the octopus. This is important because it eliminates silly nonsense such as what you quote next regarding imaginary problems with the retinal pigment layer and blood vessels. You may find an illustration of octopus retinal anatomy in Yamamoto et al., J. Cell Biol 25:345-59. You'll note that the photoreceptor cells are tightly packed, comparable to vertebrate retina. The support function of the vertebrate retinal pigment epithelium is presumably provided by the tiny processes of the support cells, which are so thin that they do not reduce photoreceptor packing density. Note also the presence of abundant capillaries immediately below the outer segments of the photoreceptor cells–indeed, it appears that every photoreceptor cell is in direct contact with a capillary. If anything, the vascular support for the octopus retina is even better than the vertebrate retina, and without reducing the packing density (which Denton, who apparently had never bothered to actually look at an octopus retina, found "hard to imagine")
Comment by trrll — August 9, 2006 @ 6:41 pm
August 10th, 2006 at 1:18 am
Bradford:
Imagination is fundamental to science. Much of ID seems to derive from failure of imagination: "I can't imagine how this could have happened, so it must be magic." But one thing science does is train our imagination. Of course, hypotheses derived from imagination must be tested experimentally. This is currently ongoing with various hypotheses of the origin of life.
Considering how many ways people have found that membrane-like vesicles can self-assemble, it would not be surprising if the earliest forms of life were spontaneously formed membrane vesicles that happened to trap catalytically active peptides or nucleic acid polymers. For example, a fatty acid vesicle that contained a nucleic acid or peptide capable of catalyzing the synthesis of a proton ionophore (a not very demanding function that many random peptides could satisfy) would be able to grow by incorporating free fatty acids, and even divide (although division would be a passive rather than an active process, driven by hydrodynamic sheer forces in the environment, not molecular motors). Those protocells that were fortunate enough to trap more useful catalytically active molecules (ideally, a catalytic set, in which the synthesis of each member of the set is catalyzed by some other member) would have a selective advantage. This would be a form of protolife without either ribosomes or chromosomes.
The prior function would be a catalytic activity that synthesizes something useful to the protocell. Experiments have shown that catalytic activity can be found even in short random nucleotide polymers or peptides.
Comment by trrll — August 10, 2006 @ 1:18 am
August 10th, 2006 at 4:14 am
trrll: Imagination is fundamental to science. Much of ID seems to derive from failure of imagination: "I can't imagine how this could have happened, so it must be magic."
You've got it backwards trrrll. It is really you who believes in magic and then calls it science. There is nothing known about organic chemistry predicting that nucleic acid sequences would form in functionally sequential ways.
Of what significance would a functional sequence have in an extra-cellular envirnment. Would you care to imagine why such a sequence would be "selected" and on what basis prior to the existence of a cell?
Considering how many ways people have found that membrane-like vesicles can self-assemble, it would not be surprising if the earliest forms of life were spontaneously formed membrane vesicles that happened to trap catalytically active peptides or nucleic acid polymers.
These "membrane-like vesicles" are not to be confused with actual membranes whose proteins enable a flow of nutrients and waste materials in and out of cells. In addition you never answered the question posed which asks why nucleic acid sequences having functional utility would be selected in a prebiotic environment. If through an incredibly unlikely event a series of codons formed having protein coding function why would such a sequence be preserved in a selection process? You are aware of what is required to enable the expression of a single gene are you not trrll? Do I have to drag this out step by step to illustrate the silliness of a natural selection concept in this contrived prebiotic soup?
For example, a fatty acid vesicle that contained a nucleic acid or peptide capable of catalyzing the synthesis of a proton ionophore (a not very demanding function that many random peptides could satisfy) would be able to grow by incorporating free fatty acids, and even divide (although division would be a passive rather than an active process, driven by hydrodynamic sheer forces in the environment, not molecular motors). Those protocells that were fortunate enough to trap more useful catalytically active molecules (ideally, a catalytic set, in which the synthesis of each member of the set is catalyzed by some other member) would have a selective advantage. This would be a form of protolife without either ribosomes or chromosomes.
Where does this "protolife" exist? Cite the study indicating a continuously self-replicating protocell like the one you described.
Improved function? What is the prior function and what gives rise to ribozymes? On what basis are you concluding that any selection occurs?
The prior function would be a catalytic activity that synthesizes something useful to the protocell. Experiments have shown that catalytic activity can be found even in short random nucleotide polymers or peptides.
Cite the study indicating that catalysts evolve according to a selection process in a prebiotic environment. What is the selection criteria and what experimental evidence documents your claim?
Comment by Bradford — August 10, 2006 @ 4:14 am
August 10th, 2006 at 5:50 am
Evidently, with all new discoveries, the myth of inverted retina will be placed amongst Junk DNA and other evolutionary stories of "bad design".
Comment by Farshad — August 10, 2006 @ 5:50 am
August 10th, 2006 at 9:07 am
There is now experimental evidence that short random RNA sequences exhibit enzymatic activity and that novel activities can be derived and optimized by selection. This is just was was predicted based on the "RNA world" hypothesis. Nothing magical about it.
Actually, experiments have shown that even simple fatty acid membranes can grow by taking up free fatty acids (nutrients) and establish ion gradients. This supports the prediction that protomembrane vesicles can have useful functions that could be optimized by selection.
I did. Protomembrane vesicles containing sequences that add function (e.g. ion transport, catalysis of synthesis of other useful sequences) that reinforces vesicle growth will reproduce and constitute selectable units.
As I said, this is an area of ongoing research, so such experiments are a promising direction. Of course, there are many challenges, since nobody knows exactly what the environment was like where life initially formed, so there are a large number of possibilities to test.
Nobody knows exactly what the prebiotic environment was, and there are many hypotheses as to the chemical structure of the first biopolymers. What has been established thus far is that a variety of catalytic activities can be found in random populations of short polymer sequences of a variety of monomer types.
Comment by trrll — August 10, 2006 @ 9:07 am
August 10th, 2006 at 10:55 am
trrll: There is now experimental evidence that short random RNA sequences exhibit enzymatic activity and that novel activities can be derived and optimized by selection. This is just was was predicted based on the "RNA world" hypothesis. Nothing magical about it.
"Optimized by selection?" I'm not referring to how RNA can be manipulated. Where is the evidence that in reasonably hypothesized prebiotic conditions RNA sequences with enzymatic activity evolve according to natural selection? Cite the reference and the selection basis.
These "membrane-like vesicles" are not to be confused with actual membranes whose proteins enable a flow of nutrients and waste materials in and out of cells.
Actually, experiments have shown that even simple fatty acid membranes can grow by taking up free fatty acids (nutrients) and establish ion gradients. This supports the prediction that protomembrane vesicles can have useful functions that could be optimized by selection.
It does not show what is a necessary prerequisite for life- the capacity for a membrane to enclose functional biomaterial that includes functionally sequenced nucleic acid capable of coding for the biomolecules that make replication possible.
In addition you never answered the question posed which asks why nucleic acid sequences having functional utility would be selected in a prebiotic environment.
I did. Protomembrane vesicles containing sequences that add function (e.g. ion transport, catalysis of synthesis of other useful sequences) that reinforces vesicle growth will reproduce and constitute selectable units.
Wrong answer trrll. Protomembrane vesicles do not contain the inheritable information contained in nucleic acids. Keep playing trrll but while doing so kindly get around to answering the question.
Where does this "protolife" exist? Cite the study indicating a continuously self-replicating protocell like the one you described.
As I said, this is an area of ongoing research, so such experiments are a promising direction. Of course, there are many challenges, since nobody knows exactly what the environment was like where life initially formed, so there are a large number of possibilities to test.
In other words the answer to my question is: Self-replicating protocells exist in trrll's imagination.
Cite the study indicating that catalysts evolve according to a selection process in a prebiotic environment.
Nobody knows exactly what the prebiotic environment was, and there are many hypotheses as to the chemical structure of the first biopolymers. What has been established thus far is that a variety of catalytic activities can be found in random populations of short polymer sequences of a variety of monomer types.
Writing that catalytic activities are observed in short polymer sequences is not a dramatic revelation nor one that makes your case. Back to the drawing boards
Comment by Bradford — August 10, 2006 @ 10:55 am
August 10th, 2006 at 6:28 pm
RNA evolution has been demonstrated under laboratory conditions. Nobody knows what prebiotic conditions were exactly. Moreover, initiation of life might require very specific conditions of temperature, flow, etc. Nor is it clear whether it is best to start with RNA, some other nucleic acid derivative, amino acids, or something completely different. So I wouldn't anticipate that recreation of the origin of life will be achieved anytime in the near future. I'm not sure what your point is here. If all you are saying is that we don't know the origin of life, I already stated that. However, so far there is no evidence that it could not occur by natural processes, and a number of predictions of origin of life models–such as the fact that it is fairly easy to find short polymers with catalytic activity, or that protomembranous structures self assemble, grow and reproduce under a variety of conditions–have been confirmed.
I'm not sure what you mean by "functionally sequenced." However, there is certainly evidence that short RNA sequences can exhibit catalytic activity, including synthesis of RNA with specific sequence. And it is easy to get things inside of membranes. So there does not seem to be any major obstacle at that level.
Selection does not even require that inheritable information be in the form of nucleic acids. All that is required for natural selection to work is the ability to reproduce and mutate–it does not require that inheritable information be encoded in a particular type of molecule. It is even possible that the earliest forms of life did not use nucleic acids at all, although they do have some attractive properties. But some models use short peptides rather than nucleic acids. So if some protomembrane vesicles happen to enclose biopolymers that enhance the membranes ability to grow and reproduce, you have a selectable unit.
They exist under laboratory conditions. Again, if your point is merely that we don't know the origin of life and it has not been demonstrated in the laboratory, I said that at the outset. So yes, replicating protocells are hypothetical. But unlike certain other things that have not been shown to exist other than in the imagination of people–nonhuman intelligent designers, for example–origin of life theory has yielded testable predictions and given rise to important discoveries about membranes, self-assembly of biomolecules, the nature of catalysis, laboratory evolution of novel enzymatic activity, etc. It's a young field. If after a few decades, nobody has yet managed to generate some kind of protolife under plausible prebiotic conditions, people will begin looking for other models. But even if the current models turn out to be wrong, they will end up having made a valuable contribution to scientific knowledge. As usual, it is the people who are looking for natural explanations for natural phenomena who end up making the discoveries.
Comment by trrll — August 10, 2006 @ 6:28 pm
August 10th, 2006 at 7:07 pm
"Optimized by selection?" I'm not referring to how RNA can be manipulated. Where is the evidence that in reasonably hypothesized prebiotic conditions RNA sequences with enzymatic activity evolve according to natural selection? Cite the reference and the selection basis
trrll: RNA evolution has been demonstrated under laboratory conditions.
Evolving to what? Not a genome. That is what must take place before your belief has a semblance of plausibility.
Nobody knows what prebiotic conditions were exactly. Moreover, initiation of life might require very specific conditions of temperature, flow, etc. Nor is it clear whether it is best to start with RNA, some other nucleic acid derivative, amino acids, or something completely different. So I wouldn't anticipate that recreation of the origin of life will be achieved anytime in the near future. I'm not sure what your point is here. If all you are saying is that we don't know the origin of life, I already stated that. However, so far there is no evidence that it could not occur by natural processes,
There is no evidence that would ever convince you because your belief is not grounded in scientific plausibility but rather some extra-scientific concerns.
and a number of predictions of origin of life models"“such as the fact that it is fairly easy to find short polymers with catalytic activity, or that protomembranous structures self assemble, grow and reproduce under a variety of conditions"“have been confirmed.
These short polymers you allude to have as much relevance to cellular generation as do a pile of steel girders to a skyscraper construction. The disconnect between the reality and the conceived end result is evident and the degree of extrapolation is obscene.
It does not show what is a necessary prerequisite for life- the capacity for a membrane to enclose functional biomaterial that includes functionally sequenced nucleic acid capable of coding for the biomolecules that make replication possible.
I'm not sure what you mean by "functionally sequenced."
I think you do. If you have any doubts head on over to a databank and get a readout on an RNA polymerase or a gene that codes for it. Scramble the order of the AAs or the codons and you get nonsense instead of a functional sequence.
Protomembrane vesicles do not contain the inheritable information contained in nucleic acids.
Selection does not even require that inheritable information be in the form of nucleic acids. All that is required for natural selection to work is the ability to reproduce and mutate"“it does not require that inheritable information be encoded in a particular type of molecule.
Cite a study indicating that lipids pass on information filtered by natural selection.
It is even possible that the earliest forms of life did not use nucleic acids at all, although they do have some attractive properties. But some models use short peptides rather than nucleic acids. So if some protomembrane vesicles happen to enclose biopolymers that enhance the membranes ability to grow and reproduce, you have a selectable unit.
You can cite models all night long. Models come and go. The ones that stick around are those confirmed by empirical evidence. If you have some indicating that biochemicals, other than nucleic acids, substitute for DNA function in a life form then out with it.
In other words the answer to my question is: Self-replicating protocells exist in trrll's imagination.
They exist under laboratory conditions. Again, if your point is merely that we don't know the origin of life and it has not been demonstrated in the laboratory, I said that at the outset. So yes, replicating protocells are hypothetical.
LOL. I'll take the last sentence rather than the first as the true meaning.
Comment by Bradford — August 10, 2006 @ 7:07 pm
August 11th, 2006 at 12:30 am
Where is the evidence that they can't? Yes, this has not yet been demonstrated, and based on the progress of the field, I'd be surprised to see such a demonstration any time soon even if the hypothesis is correct. There's just too many possibilities and too much uncertainty about the proper conditions
No, and I'd be surprised if that is ever demonstrated directly in the laboratory, even if the hypothesis is correct. Evolution protolife into a genome as we understand it today could well take thousands of years.
History has taught scientists to be suspicious of claims that something cannot work. Historically, most such alleged proofs–the impossibility of heavier than air flight, the impossibility of breaking the sound barrier, the impossibility of the sun continuing to burn for millions of years, etc., etc. have turned out to be wrong. Coming up with objections is easy. Progress is always made by the people who say, "Let's suppose that it can happen. How might it work?" Certainly if it were found that short random peptides or nucleic acids did not exhibit catalytic activity, or if spontaneous self assembly of membranes had not been demonstrated under a variety of conditions, I would have a problem with the hypothesis. But it seems to me that the really serious hurdles have already been surmounted.
Piles of steel girders do not spontaneously start building things. But random peptides or nucleic acids do start catalyzing chemical reactions. As to the relevance to cellular generation, spontaneous catalysis is the earliest step in the hypothesized pathway to life, so discovering that it works as hypothesized is significant support for the theory.
The information would not be passed on by the lipids, but by the biopolymers enclosed within (which might or might not be nucleic acid). The membranes are important because for selection to work, you have to have separate selectable units–i.e. protomembrane vesicles that happen to enclose catalytic biopolymers that enhance the rate of membrane growth will reproduce more rapidly than those that do not. Those in which the biopolymers constitute an autocatalytic set (i.e. they collectively catalyze their own synthesis) will be able to sustain such reproduction without being diluted out.
The evidence is that other types of polymers, such as peptides, can also exhibit sequence-specific catalytic activity, so it is possible that there could be an autocatalytic set comprised entirely of peptides. Confined within a membrane, such a set would constitute a selectable unit–i.e. a protolife form. The appealing nature of nucleic acids is that complementary binding assists fidelity of synthesis. And of course, we need to end up with nucleic acids eventually, so starting with random sequence ribozymes is attractive. The two hypotheses are not mutually exclusive–it is easy to imagine a protolife form in which simple ribozymes and short catalytic peptides catalyze one another's synthesis.
Comment by trrll — August 11, 2006 @ 12:30 am
August 11th, 2006 at 1:24 am
"Optimized by selection?" I'm not referring to how RNA can be manipulated. Where is the evidence that in reasonably hypothesized prebiotic conditions RNA sequences with enzymatic activity evolve according to natural selection?
trrll: Where is the evidence that they can't?
Why is selection invoked at all? There is no organism at this hypothetical point. Invoking selection does not appear motivated by an objective assessment of physical realities.
Evolving to what? Not a genome.
No, and I'd be surprised if that is ever demonstrated directly in the laboratory, even if the hypothesis is correct. Evolution protolife into a genome as we understand it today could well take thousands of years.
It's not a matter of time. The issue is cause. There is simply no reason to believe random chemical reactions lead to encoding molecules unless you convince yourself there are no other options.
There is no evidence that would ever convince you because your belief is not grounded in scientific plausibility but rather some extra-scientific concerns.
History has taught scientists to be suspicious of claims that something cannot work. Historically, most such alleged proofs"“the impossibility of heavier than air flight, the impossibility of breaking the sound barrier, the impossibility of the sun continuing to burn for millions of years, etc., etc. have turned out to be wrong. Coming up with objections is easy. Progress is always made by the people who say, "Let's suppose that it can happen. How might it work?"
Let's be clear about my objection. I have no doubt that genomes can be generated. The solution is the same one used to create air flight etc.- application of intelligence. When examining cells and how they function however, there is nothing in evidence that indicates a self-generating process is plausible.
Certainly if it were found that short random peptides or nucleic acids did not exhibit catalytic activity, or if spontaneous self assembly of membranes had not been demonstrated under a variety of conditions, I would have a problem with the hypothesis. But it seems to me that the really serious hurdles have already been surmounted.
The serious hurdle is generating genetic information where none exists without intelligence.
These short polymers you allude to have as much relevance to cellular generation as do a pile of steel girders to a skyscraper construction.
Piles of steel girders do not spontaneously start building things. But random peptides or nucleic acids do start catalyzing chemical reactions. As to the relevance to cellular generation, spontaneous catalysis is the earliest step in the hypothesized pathway to life, so discovering that it works as hypothesized is significant support for the theory.
The problem is analysed strictly from a physical perspective. The ommited dimension is a means of generating information, storing, transmitting and translating it into functional molecular tools. There is more to this than catalysis.
Comment by Bradford — August 11, 2006 @ 1:24 am
August 11th, 2006 at 9:08 am
That depends upon your definition of "organism." To a biologist, an organism is anything that reproduces with heritable variation. That's all you need to invoke selection.
There is no reason to believe that they cannot. You seem to take this as an article of faith, but science requires evidence. The problem with an ID hypothesis for the origin of life is that it doesn't seem to make predictions other than by exclusion–"there are no conditions in which random chemical reactions can self organize to produce something that reproduces with heritable variation". Since there is essentially an infinite number of possible conditions, this is a prediction that can never be tested in a finite amount of time. And hypotheses that are not practicably testable are always, always scientific dead ends. On the other hand the various spontaneous self-assembly hypotheses make detailed, testable predictions, many of which seem to be holding up to experimental tests.
And nothing to indicate that it is not. Both are simply hypotheses. But one hypothesis leads to predictions and discoveries, and the other does not.
The notion that information cannot be generated without intelligence has been disproved experimentally in a variety of computer simulations. Now I am aware of the traditional ID out: "It takes intelligence to set up the simulation; that's where the information comes from." The problem with that dodge is that it is possible to measure the amount of information coming out, and it can be far more than the total amount of information required to define the simulation. So where does the extra information come from?
And even if we accept that some "seed" of information derived from intelligence is required to start the process off (there seems to be no reason why that should be true, but never mind), why does it have to be at this particular point (the origin of life). Why not at an earlier point? The origin of the universe, perhaps? Or the definition of the laws of physics (which seems more analogous to setting up a simulation)? Or the origin of logic itself (if that has any meaning at all)?
Comment by trrll — August 11, 2006 @ 9:08 am
August 11th, 2006 at 9:42 am
About two months ago there was a thread at UD regarding this issue and here is one of my comments:
Comment by Farshad — August 11, 2006 @ 9:42 am
August 11th, 2006 at 9:45 am
trrll: That depends upon your definition of "organism." To a biologist, an organism is anything that reproduces with heritable variation. That's all you need to invoke selection.
The peptides and truncated RNAs cited are observed to "reproduce" under very limiting conditions which includes availability of AAs and nucleotides and no means for the "organism" to synthesize them in a hostile environment.
It's not a matter of time. The issue is cause. There is simply no reason to believe random chemical reactions lead to encoding molecules unless you convince yourself there are no other options.
There is no reason to believe that they cannot. You seem to take this as an article of faith, but science requires evidence.
Science involves observing. Encoding entities are not observed resulting from random chemical reactions but rather from applied intelligence. Faith is believing in what is unseen. In this case that applies to your position.
The problem with an ID hypothesis for the origin of life is that it doesn't seem to make predictions other than by exclusion"“"there are no conditions in which random chemical reactions can self organize to produce something that reproduces with heritable variation".
I'll give you a reason that is testable. Found universally are error detection and repair mechanisms that maintain genomic integrity. Causes of genomic mal-function include factors that were present on earth at life's origins. The prediction is that such mechanisms are absolutely essential to the viability of a genome and a test involves observing replication without such mechanisms. If the rate of information loss exceeds gains a generation mechanism is fatally compromised.
The serious hurdle is generating genetic information where none exists without intelligence.
The notion that information cannot be generated without intelligence has been disproved experimentally in a variety of computer simulations. Now I am aware of the traditional ID out: "It takes intelligence to set up the simulation; that's where the information comes from." The problem with that dodge is that it is possible to measure the amount of information coming out, and it can be far more than the total amount of information required to define the simulation. So where does the extra information come from?
I've encountered enough computer simulations to realize that unless they are examined in detail it is foolish for me to accept a claim on face value. I do not doubt your sincerity but need a test drive.
And even if we accept that some "seed" of information derived from intelligence is required to start the process off (there seems to be no reason why that should be true, but never mind), why does it have to be at this particular point (the origin of life). Why not at an earlier point? The origin of the universe, perhaps? Or the definition of the laws of physics (which seems more analogous to setting up a simulation)? Or the origin of logic itself (if that has any meaning at all)?
The argument that information is loaded into the origin of the universe has been advanced. While the topic is fascinating I'll leave it to others to hash that one out
Comment by Bradford — August 11, 2006 @ 9:45 am
August 11th, 2006 at 9:59 am
Many months ago at ARN, I showed in about half a dozen different ways (all the ways were related, of course, as they were grounded in mathematics and logic and mathematical theory) why — and how — the Avidistas are deluding themselves in their fervent desire and need to believe that 'information' can self-originate or come-from-nowhere.
And how did the Avidistas respond? How did they attempt to disprove my proof that their claims/assertions (and beliefs) are contrary both to reason and to mathematics? Oh, it was the easiest thing in the world — they just declared that I'm ignorant (if not a 'liar'), they just declared that I "don't understand 'science'" and that I certainly don't understand "biology." (Biology!? 'Avida' is a computer program, it's not "biology.")
Comment by Ilion — August 11, 2006 @ 9:59 am
August 11th, 2006 at 7:41 pm
Except that there is no continuous "energy" feed. The programmer sets up the conditions and lets it go. Perhaps you could argue that the intelligence that the programmer puts into designing a fitness function is some kind of "intelligence battery," but you are then left with the problem, "Why doesn't it run down?" For example, a genetic algorithm can be used to solve traveling salesman problems. The fitness function is a simple "Shorter is better" criterion. How much intelligence can be "banked" in the few bits that it takes to define that criterion? Yet this algorithm will derive a solution for any traveling salesman problem, generating an unlimited number of bits of information.
Comment by trrll — August 11, 2006 @ 7:41 pm
August 11th, 2006 at 8:00 pm
How hostile is the environment? There are a lot of microenvironments on earth (and some have even hypothesized that some of the reactions might take place in outer space). Many different conditions of temperature, pressure, availability of nutrients, solar radiation, etc.
They are not seen resulting from anything other than human intelligence. But we know from the existence of life that that is not the only way that codes can come into being. So we must hypothesize something different from what we have seen. Whether one is hypothesizing a mechanism of stepwise generation of codes by selection, or an unseen nonhuman designer, one necessarily is hypothesizing about something that cannot be observed. What science teaches is that hypotheses about the unobserved are worthless unless they can be used to derive predictions that can be used to test the hypothesis.
Science is about theorizing about the unseen, and then testing those theories by experiment. Faith is about believing in that which is both unseen and untestable, such as the hypothetical Designer.
So once again, the best that you can come up with is a negative prediction: reproduction without separate error correction mechanisms will never be observed. Obviously, such a prediction cannot be confirmed in a finite amount of time.
Comment by trrll — August 11, 2006 @ 8:00 pm
August 11th, 2006 at 9:31 pm
Science involves observing. Encoding entities are not observed resulting from random chemical reactions but rather from applied intelligence.
They are not seen resulting from anything other than human intelligence.
Anti-IDers try to imply there is something unique about this point. Yes, we reason abstractly but any being similarly equiped would have the same encoding capacities. That's why intelligence is where the inference ends.
Faith is believing in what is unseen. In this case that applies to your position.
Science is about theorizing about the unseen, and then testing those theories by experiment. Faith is about believing in that which is both unseen and untestable, such as the hypothetical Designer.
The testing would focus on what is observable. Test results would have to be consistent with the hypothesis that intelligence better explains the origins of life.
I'll give you a reason that is testable. Found universally are error detection and repair mechanisms that maintain genomic integrity. Causes of genomic mal-function include factors that were present on earth at life's origins. The prediction is that such mechanisms are absolutely essential to the viability of a genome and a test involves observing replication without such mechanisms. If the rate of information loss exceeds gains a generation mechanism is fatally compromised.
So once again, the best that you can come up with is a negative prediction: reproduction without separate error correction mechanisms will never be observed.
Predictions entailing negations are useful. Physical realities can reflect one option or its opposite. Why assume one is valid without testing? It is an irrational approach.
Obviously, such a prediction cannot be confirmed in a finite amount of time.
We can get a very good indication of what occurs under varying conditions and if the data is overwhelmingly lopsided after much testing a conclusion can be drawn. Confirming events of natural history may be intrinsically difficult but the best fit will win out in the end.
Comment by Bradford — August 11, 2006 @ 9:31 pm
August 11th, 2006 at 11:48 pm
As a rule they are not useful. Scientifically, they are almost always dead ends, because people who make negative predictions never actually seem to try to test them. It is just human nature–people find it hard to motivate themselves to attempt something that they believe to be impossible. Essentially all scientific progress is made on the basis of positive predictions. For example, even though you have tried to frame this as a prediction of ID, there are no ID advocates attempting to test it. All of the work on testing origin of life scenarios is being done by people who hypothesize that such a mechanism exists.
Comment by trrll — August 11, 2006 @ 11:48 pm
August 12th, 2006 at 12:03 am
Predictions entailing negations are useful. Physical realities can reflect one option or its opposite. Why assume one is valid without testing?
trrll: As a rule they are not useful. Scientifically, they are almost always dead ends, because people who make negative predictions never actually seem to try to test them. It is just human nature"“people find it hard to motivate themselves to attempt something that they believe to be impossible.
In this case testing genomic corruption as well as biological mechanisms designed to prevent it yields very useful data related to disease and cancer. It is no doubt true that funding could be problematic for an ID oriented study and anti-IDers would want to fry any publisher of adverse test results. But that's the political dimension of it.
Comment by Bradford — August 12, 2006 @ 12:03 am
August 12th, 2006 at 12:33 am
Perhaps the Discovery Institute should stop spending its money trying to get ID into the high schools, but instead fund studies by ID scientists. After all, the biggest reason why ID has such a poor reputation among scientists is the lack of major discoveries by ID advocates.
Comment by trrll — August 12, 2006 @ 12:33 am
August 13th, 2006 at 1:07 pm
Regarding "Genetic Algorithms" -
I know. Or maybe not.
Yes, they "emerged from evolutionary theory" - I never said they didn't. But you are absolutely incorrect in claiming that they "do 'exactly parallel evolutionary theory" (note I had said "theory", not "mechanisms", in my original claim). I sat in on a class on "Biomorphic Computing" last summer, and several lectures dealt with "Genetic Algorithms" and "Genetic Programming". I noted quite quickly that unless the desired goal or "fitness" was directly, specifically, and INTELLIGENTLY included in the program/algorithm from the start, all the searching the "genetic program/algorithm" might do would be futile. I asked the professor about this, and asked him what would happen if not only the search was allowed to proceed "randomly" according to genetic models, but also if the very "target" itself was allowed to be randomly varied (as would have to happen in Nature if no "target" was actually intended). His response, after several seconds of careful thought, was an assured, "It would spit out garbage" (in essence).
Comment by Douglas — August 13, 2006 @ 1:07 pm
August 13th, 2006 at 1:11 pm
Continuing…
There's the problem with your claim that genetic algorithms "EXACTLY" mimic/model/analogize evolutionary theory - evolutionary theory implies (unless it is a version which accepts an active Designer involved in the results and mechanisms) that even the "fitness landscape" itself must "evolve" (or, rather, be allowed to change over time), and must do so "randomly". I'm afraid genetic algorithms would produce just so much "mush" if they incorporated that aspect of evolutionary theory into their design.
Comment by Douglas — August 13, 2006 @ 1:11 pm
August 13th, 2006 at 1:19 pm
Yes, they "emerged from evolutionary theory" - I never said they didn't. But you are absolutely incorrect in claiming that they "do 'exactly parallel evolutionary theory" (note I had said "theory", not "mechanisms", in my original claim).
Bradford: I assume the term mechanisms references mainly selected biological genetic mutations which do not in fact exactly parallel genetic algorithms. Analogous is a better description.
Comment by Bradford — August 13, 2006 @ 1:19 pm
August 13th, 2006 at 2:54 pm
Recall that Bradford said, "Evolution is irrelevant to technological progress."
I replied that this was not so, given the widespread use of genetic algorithms (derived from evolutionary theory) in many fields, and provided several real-world examples. I don't know about Bradford, but I gather even Douglas now (reluctantly) assents to this.
Now Douglas seems to have some novel idea of what is meant by "fitness landscapes." Douglas, what exactly do you think biologists are talking about when they refer to fitness landscapes? Are you suggesting that environments don't change over time, often drastically? Or that the dynamics within a given ecosystem where various species live remain static? Are we talking about angels pushing the clouds around, or what?
Comment by takuan — August 13, 2006 @ 2:54 pm