Uncommon Descent Serving The Intelligent Design Community

# The “quine dilemma” of evolution

Share
Flipboard
Print
Email

Sorry if this post is a bit for computer programmers, anyway I trust that also the others can grasp the overall picture.

Evolutionists claim that what it takes to evolution to work is simply “a populations of replicators, random variations on them, and a competition for survival or resources”.
Today we will try to partially layout how to simulate on computer such process. First off, we need the replicators, i.e. digital programs able to self-reproduce. In informatics jargon, a computer program able to self-reproduce, i.e. to produce as output a copy of its source code is called a “quine”. Therefore in a sense a quine is a little, minimal digital “bio-cell”. You can write the code of a quine in any programming language. Also with the same language you can do that in various ways. Here I will examine a quine written in Perl language by Tushar Samant, it is but one of many examples.

Its source code is the following:

\$a=’X‘; print map “\\$a=’\$a’; \$_, q(\$_)”, q(print map “\\$a=’\$a’;\$_, q(\$_)”)

If you have Perl installed on your computer you can easily verify that if you run this script, it prints itself on the screen. Eventually if you redirect the output to a file, such file will be a perfect copy of its generator file.

Von Neumann mathematically proved that a self-reproducing automaton must contain a symbolic description or representation of itself and a constructor (see my previous post). Also our quine contains a symbolic description of itself (the code on the right, the quoted “q(print map…)”), while the code on the left (the first “print map”) is the operation on the description (the constructor), to output the whole quine. Some say that, this way, the quine necessarily works somehow in “auto-referential” mode. (About “quines”, self-reference, automata, meta-languages and artificial intelligence I suggest reading the book by Douglas Hofstadter, “Gödel, Escher, Bach”, 1979.)

Why in the source code I have highlighted in blue color a “neutral” zone, in red color a “critical zone”? This distinction holds not only for this quine, rather for almost all quines (and in a sense even for almost all computer programs or any system in general). Random variations in the red zone destroy the self-reproducing function. Differently, most variations in the neutral zone don’t cause malfunctioning. If, for example, we change the value of the \$a variable from “X” to, say, “fb_M+hF6.oia7-jj” we get a bigger script, but it continues to self-replicate.

Now, let’s imagine that we want to develop a small evolution simulator on our computer. We could set an initial number of those quines and make them self-reproduce to obtain a growing population. Eventually we could apply random variations, generation by generation, on their neutral zones. Then we have to write in our evolution simulator a “fitness function” working in this somewhat digital environment. A first simple idea could be to establish that only the bigger quines survive. However such evolution simulation would be very poor. In fact, the variations inside the digital organisms would be trivial, sure no new organization arises. Moreover the fitness function is poorly specified, because what matters is only the quantitative size of the quines, how much they are “fat” so to speak. Certainly no really different organism arise.

Therefore, if we want to test the above evolutionist claim, we could imagine a more complicated fitness function, based on the concept of predation, just a suggestion. The organisms that are somehow able to “eat” parts of other organisms are more fit to survive. They are the “predators”, while the organisms eaten are the “victims”, who necessarily die. This would be similar to what happens in nature per Darwinian selection. Also we could think of a selection based on a competition for resources.

At this point the question is: what variations are necessary to transform our initial quines into evolved predators or resource seekers? No random variation can produce such increase in organization, because, as seen above, almost all random variations in the red zone are fatal and the variations on the blue zone are neutral. However to transform our quines in predators or resource seekers is not impossible. But one has to increase the organization of the critical zone in substantial manner. New source code has to be written in the red zone. Changes in the blue zone are useless. The predation macro function needs sub-functions: movement, enemy detection, fight… Analogously, the resource seeker function needs: movement, resource detection, import of resources…

To keep our discourse simple, as an example, I modified the initial quine with a simple, very rudimental, movement sub-function (which serves to both the higher functions): now the replicator can perform a random walk on a grid with steps of 1 unit in 8 different directions. To do that I used the \$p variable to store the X/Y information (where the replicator stays on the grid at a given time). The result could be something like this:

\$a=’X‘;\$p=q(500_500);\$e=q((\$x, \$y) = split /_/, \$p;\$x+=int(rand(2))*(-1)**int(rand(2)); \$y+=int(rand(2))*(-1)**int(rand(2));\$p =~ s/\d+_\d+/\${x}_\${y}/;);eval \$e;print map “\\$a=’\$a’; \\$p=’\$p’; \\$e=’\$e’; eval ‘\$e’;\$_, q(\$_)”, q(print map ” \\$a=’\$a’;\\$p=’\$p’;\\$e=’\$e’; eval ‘\$e’; \$_, q(\$_)”)

With this modification the automaton continues to be able to self-replicate, and — if introduced in a suitable evolution program simulator (which I have not programmed thus far) — it moves on a grid. Notice however that both the constructor and the symbolic description are changed.

All that leads us directly to what I call the “quine dilemma” of unguided evolution. If random variations are harmless or neutral (blue zone) they create no new organization. If evolution has to create complex functional novelties, new organization, it must operate in the red zone and necessarily become potentially destructive. To speak of “dilemma” here is euphemistic. This dilemma is worse than Hamletic, because de facto is a show-stopper for evolution. The quine dilemma holds in computer programming, as in biology. In fact, in the lab you can crash the cellular replication by introducing random variations in a cell. Needless to say, this dilemma has a lot to do with the experimental fact that unicellulars grown in the lab haven’t yet evolved in … frogs or butterflies (e.g. Lenski’s work).

I like to cite Larry Wall, the computer scientist who invented Perl, who sums it up best: “The potential for greater good goes right along with the potential for greater evil”. Larry said that in the context of software development, but mutatis mutandis it holds also in general, biology included. In short, no power without risk.

I said “biology included” because the objection by evolutionists might be that in biological replicators there is no “quine” problem, because the information for new organization (which random variation applies on) is decoupled from the information for construction. This claim is fully illogical because the information for new organization is the information for construction, what else. An organism is constructed according to assembly instructions. If you want a different organism you have to modify them. No decoupling is possible between instructions and organism because the latter is the direct product (bit by bit) of the former. No decoupling is possible between cause and effect.

To sum up, the initial claim that evolution needs only “a populations of replicators, random variations on them and a competition” is only an hope, because just in simple replicators it crashes against basic conceptual obstacles, one of which is indeed the “quine” dilemma.

In evolutionary algorithms, the environment and its relationship to the replicators is considered separate from the replicators.
Evolutionary algorithms are examples of Intelligent design evolution. No one knows how to model unguided/ blind watchmaker evolution. That is just another reason why the concept isn't scientific. Joe
Zachriel continues to expose the problem with evolutionism:
You have to distinguish your hypothesis empirically from mere ignorance, something you can’t do until you, well, you know, state a testable hypothesis.
Evolutionism, ie unguided/ blind watchmaker evolution, doesn't have any testable hypotheses and relies on our ignorance. Thanks again, Zachriel Joe
Gary S Gaulin says The problem is that none that I know of are in the ID movement I say Well good luck with that then. If that is what you want. I honestly don't wish you any ill will. I just don't get what you are trying to say. Can you not even consider the possibility that a reason your idea is being ignored by ID enthusiasts is that we find it to be mostly incoherent and completely irrelevant to the question of Intelligent Design? you say, I cannot afford more years of trying to explain it I say, I'm not asking for years. Just the minutes it takes to write a decent synopsis explaining why your project is in any way relevant. Apparently you don't have time for that either. peace PS This is not at all about your simulation I think your program is reasonably cool. I just can't connect the dots between it and ID perhaps other than an example of why machines are not intelligent. fifthmonarchyman
fifthmonarchyman, I already found those who accept the model and theory. The model even won a superior coding award. The problem is that none that I know of are in the ID movement, and I cannot afford more years of trying to explain it to those who tell me that they cannot understand any of it at all and I need to rewrite the whole thing all over again, and again, and again, and again. Gary S. Gaulin
Gary S. Gaulin says, it is obvious that all in biology is fully covered by only four requirements. I say, I've read your paper and I would disagree. I think there is a lot in biology that is not explained by "behaviors" possessing “sensory sensors” and "motor muscles" and "guess new memory action" and "confidence (central hedonic, homeostasis) system". I don't think you've demonstrated otherwise. I don't think you've demonstrated those things are even coherent. I'm not even sure I know what it means for all of biology to be "covered" by a "behavior" that has 4 “circuit requirements”. But I could be completely misunderstanding your point. I suppose I probably am. I guess we will never know since you are unwilling to summarize and I can't make heads or tails of what you are trying to say while "studying the rest of the theory". Oh well I gave it a shot. Sorry you feel my inability to begin to grasp your argument is a sign I want nothing to do with science. Good luck with with finding someone to accept your project. peace fifthmonarchyman
fifthmonarchyman:
What is a “circuit requirement” and why is your fourfold list both necessary and comprehensive? You don’t say.
The text that I quoted is from the Introduction section explaining the illustration of the "circuit" in which all four of the "circuit requirements" are also shown algorithmically. Why the four "requirements" are necessary and comprehensive will require studying the rest of the theory where the same four requirements are explained a good number of times in relation to all levels of intelligence that the theory covers, so that it is obvious that all in biology is fully covered by only four requirements. Gary S. Gaulin
Hey Gary G, Your quote is a perfect example of what I'm talking about here. What is a "circuit requirement" and why is your fourfold list both necessary and comprehensive? You don't say. I would say for a behavior to qualify as intelligent it should be non-algorithmic and non-random. Why are my requirements unnecessary and yours self evident? Another thing, whatever "motor muscles" and "sensory sensors" are I don't see how a behavior could possibly possess them. A behavior is what an intelligent agent exhibits. I am not a behavior I am a agent. Even if a "behavior" could possess these things you give no evidence or argument why they are necessary or sufficient for intelligence. The rest of your project is equally as confusing as this paragraph. At least from my perspective. If I overlook these things I still don't see the point. you define intelligence in such a way that It appears that you trying to explain a phenomena that does not need explaining. IMHO you need to at least layout clearly what the problem is you are trying to address. If your project is just about describing what makes the computer Watson tick you wasting your time as far as I can tell No one I know cares very much what makes Watson tick . If you were explaining what makes IBM scientists tick it might be more interesting. What you would need to show is that the intelligence in Watson is necessarily the same as the intelligence in the scientists that created it Anyway all these comments assume that I understand what you are getting at and I'm not at all sure I do. peace fifthmonarchyman
This is one example of what fifthmonarchyman is unable to understand:
A behavior qualifies as intelligent behavior by meeting all four circuit requirements for this ability, which are: [1] body (or modeling platform) with motor muscles (proteins, electric speaker, electronic “write” to a screen) to control, [2] memory addressed by sensory sensors where each motor action and its associated confidence value are separate data elements, [3] confidence (central hedonic, homeostasis) system that increments (stored in memory) confidence value of a successful motor action else decrements the confidence value, [4] guess new memory action when associated confidence level sufficiently decreases (and if not prerandomized motor data then when first addressed). For flagella powered cells reversing motor direction can produce a tumble to a new heading direction, guess where to go. The IBM Watson system that won at Jeopardy qualifies as intelligent. Word combinations for hypotheses were guessed then tested against memory for confidence in each being a hypothesis that is true, and whether confident enough in its best answer to “push the buzzer”. The Watson platform simply had a speaker (for vocal muscles) and muscles guiding a pen was simulated by an electric powered writing device.
After several years of the same old excuses it's obvious that several more years of rewriting the theory (after already having added several illustrations algorithmically showing the above information as well as computer models showing this in code) is a waste of time on those who want nothing to do with science. Gary S. Gaulin
Fifthmonarchyman it would be helpful for you to explain what you did understand, as opposed to suggesting that you could not understand anything at all then try to get rid of me. According to the theory: How is intelligence systematically qualified? In the computer model how is intelligence detected? What is the primary mechanism responsible for the Cambrian Explosion? Gary S. Gaulin
fifthmonarchyman:
It’s obvious (I think) that what you’ve come up with has not as of yet been shown to be relevant at all to the questions that ID folks are interested in.
I have to add that the theory is included with the ID Lab computer models. It's expected to be difficult for someone who is used to an easy read that uses metaphors and such. My purpose is to explain the scientific details that show why the premise of the Theory of Intelligent Design holds true when tested. Regardless of how I phrase the paragraphs it will never be an easy read. I also have to make sure all the right words are in the right places or else experienced scientists and programmers who study it would find the theory to be logically incoherent. That level of detail makes it even harder for someone with no experience in the given areas of science, but I can't help that. My main concern has to be that when understood the model and theory remain standing on its own scientific merit. Simply attacking another theory (as other ID literature does) might be an easy read but it's not able to stand on its own scientific merit. I have to start with a model that requires experimenting with to fully understand how it works, what it does, and what happens when intelligence is present. It's otherwise just more of the usual ID literature that the majority of experienced scientists, programmers and science teachers have no interest in. Gary S. Gaulin
fifthmonarchyman: Nothing specific can be proposed while the very idea of an oracle is ruled out of bounds as somehow not science. You're more than welcome to speculate, however, you would have to distinguish empirically any proposed scientific hypothesis from mere ignorance. fifthmonarchyman: When I get some time I expect maybe perhaps to organize my thoughts Good idea. Zachriel
Gary S. Gaulin says The problem is that I know of no better way to initially present the material I say, If I were you I'd start be composing a paragraph that explains in detail the problems you want to address and why it's important to address them. Then a paragraph that lists the reasons you feel that previous attempts have failed to adequately address those problems Finally one that explains exactly how your work addresses those problems in a better way, You need to be specific. Simply asserting that yours is the first Theory of Intelligent design is not going to cut it. It's obvious (I think) that what you've come up with has not as of yet been shown to be relevant at all to the questions that ID folks are interested in. You need to explain why it is relevant with something more than it has the words Intelligent Design in the title. That is just three paragraphs. Surely you can do that. peace fifthmonarchyman
Fifthmonarchyman that's why I'm at UD, to explain the theory. The problem is that I know of no better way to initially present the material. I don't want it to turn into a 500 page book that will take you a month to get through so I kept it as brief and to the point as possible, which maybe causes another problem that is caused by my not knowing where others need more info. I'm good in this area of science but I'm not a professional writer, which is why I admit to needing help writing up this project! When I read what I wrote it's as clear and scientifically precise as I can make it, but since I'm not you I can't see where you are having trouble. It's also something that could easily become a four+ year college subject but I know you don't want to get into that much detail. How to cover the most important detail in around 50 pages is a very frustrating problem for me. Any ideas to help overcome this are very welcomed. Gary S. Gaulin
Gary S. Gaulin says, I now wonder whether fifthmonarchyman is a culprit, or just another victim of the sloppy science that the Discovery Institute has become famous for. I say, I can assure you I'm not a victim of the Discovery Institute. I don't pay them much mind whatsoever. I am totally confused by your "theory" however. Granted I'm pretty slow but I can't make heads or tails of it for the most part. If you want folks like me to look at it. I would suggest you go back and see if you could express yourself with a little more clarity peace fifthmonarchyman
Zac says, So someone on the Internet is standing in your way from unlocking the secrets of the universe! I say, Who said anything about unlocking the secrets of science? I have a "game" that I feel is interesting and has something to say about the limits of algorithms like evolution. I'm trying to work through the implications that is all. I'm curious what others who don't share my worldview would think about the conclusions I'm drawing and predictions that flow from those conclusions. That is all this discussion is about. When I get some time I expect maybe perhaps to organize my thoughts and pass then off to someone who is in a position to tell me how all wet I am. If you are looking to unlock the "secrets of science" in one fell swoop you are wasting your time with me. peace fifthmonarchyman
And Planet Source Code is back online! Both of the above links are working again. That's sure a relief.. Gary S. Gaulin
Excellent point Zachriel:
You have to distinguish your hypothesis empirically from mere ignorance, something you can’t do until you, well, you know, state a testable hypothesis.
I think this is a good time to repost a few links so that hopefully at least some of the adults at UD might bother to learn what 3-6 year old US children learn by watching educational TV: Buddy has a hypothesis - YouTube https://www.youtube.com/watch?v=r0CGhy6cNJE Dinosaur Train. A Colorful Hypothesis Activity | PBS Parents http://www.pbs.org/parents/dinosaurtrain/activities/a-colorful-hypothesis/ “What’s a hypothesis?” https://tosciencewithlove.wordpress.com/tag/dinosaur-train/ The good news is that thanks to the work of creative educators it's getting harder to sell the nonsense that is still being promoting to the general public. I now wonder whether fifthmonarchyman is a culprit, or just another victim of the sloppy science that the Discovery Institute has become famous for. Gary S. Gaulin
fifthmonarchyman: The “you” in this case is Alan Turing who defined the term. Turing's oracle is not a machine. However, in mathematics, an oracle can be of any complexity class. It can even have a random output. We provided a standard use of the term from Dembski. http://en.wikipedia.org/wiki/Oracle_machine If you want to specify that the oracle you have in mind can't be a machine, that's fine, but you can't simply say it has to be so just because you name it an oracle. fifthmonarchyman: Nothing specific can be proposed while the very idea of an oracle is ruled out of bounds as somehow not science. So someone on the Internet is standing in your way from unlocking the secrets of the universe! We told you the scientific requirements. You have to distinguish your hypothesis empirically from mere ignorance, something you can't do until you, well, you know, state a testable hypothesis. Zachriel
Note: The entire Planet Source Code site unexpectedly went offline soon after I posted the above links. If they did not work for you then please give PSC time to get everything back online again. This has happened before. Sorry for the inconvenience. Gary S. Gaulin
fifthmonarchyman:
Apparently you don’t get what is trying to be explained in ID. Hint it’s not intelligence itself.
I am very aware of the theory having been premised to explain how our Creator (Allah, God, etc.) works. In fact I'm the only one who has had such a model (theory included) online for experimenters to experiment with. This was the first ID model, from 2011. Reading the comments to it will explain that its 5 globe rating was voted down by the usual protesters (for political not scientific reasons): http://www.planetsourcecode.com/vb/scripts/ShowCode.asp?txtCodeId=74175&lngWId=1 And more: http://www.planetsourcecode.com/vb/scripts/BrowseCategoryOrSearchResults.asp?txtCriteria=Gary+Gaulin&lngWId=1 The only thing you would accomplish by submitting a model and theory that does not exist to Planet Source Code is waste your time and theirs (having to reject it). The Theory of Intelligent Design already exists. But it's not presented at UD or is available from the Discovery Institute. Therefore the game (of pretending to have a model and theory) is already over. Checkmate... Gary S. Gaulin
Zac said, So when someone wants to explore your idea, you say no, you won’t go there. I say, The "you" in this case is Alan Turing who defined the term. You say You haven’t proposed anything yet. I say, Nothing specific can be proposed while the very idea of an oracle is ruled out of bounds as somehow not science. That is the point you choose to exclude for purely metaphysical reasons the only idea that will ever yield a complete theory. you say, Scientific theories are always incomplete, whether algorithmic or not. I say We will never get to proposals till that metaphysical muddle in your thinking is cleared up. peace fifthmonarchyman
fifthmonarchyman: You have already acknowledged that an algorithmic theory is necessarily incomplete. Scientific theories are always incomplete, whether algorithmic or not. fifthmonarchyman: The addition of the oracle makes the “theory of evolution” complete therefore the oracle must be non-algorithmic The addition of a ill-defined oracle doesn't make a theory complete. You provided an example, the axiom of random mutation in neo-Darwinian theory. fifthmonarchyman: We will not go any further into the nature of this oracle than to say that it cannot be a machine. So when someone wants to explore your idea, you say no, you won't go there. fifthmonarchyman: Infinity verses finitude is an important topic I’d like to explore sometime with someone who is willing to look at this objectively So when someone wants to explore your idea, you say no, you won't go there. fifthmonarchyman: When you exclude the oracle you are left with a theory with less explanatory power. It is as simple as that You haven't proposed anything yet. Zachriel
Hey Zac, in the interest of time and clarity I'm going to be very selective in my responses to you here. Just addressing things that may not have been sufficiently addressed before. If a lurker feels that I am missing something I will hope they point it out to me zac says What makes you think the oracle is non-algorithmic? I say, Isn't it obvious? You have already acknowledged that an algorithmic theory is necessarily incomplete. The addition of the oracle makes the "theory of evolution" complete therefore the oracle must be non-algorithmic Then there is this----- quote: We will not go any further into the nature of this oracle than to say that it cannot be a machine. end quote: Alan Turing you say, Gödel’s theory does not apply to finite sets. I say, Infinity verses finitude is an important topic I'd like to explore sometime with someone who is willing to look at this objectively for now this should be sufficient from here http://plato.stanford.edu/entries/goedel-incompleteness/ quote: Any consistent formal system F within which a certain amount of elementary arithmetic can be carried out is incomplete; i.e., there are statements of the language of F which can neither be proved nor disproved in F end quote: you say, In any case, you don’t necessarily have to take the unknown as a given. For instance, the continuum hypothesis can be taken as either true or false with equal consistency. I say, And you can include or exclude the Oracle with equal consistency. It's not a question of consistency it's a question of utility. When you exclude the oracle you are left with a theory with less explanatory power. It is as simple as that peace fifthmonarchyman
fifthmonarchyman: A complete theory of how my cheese burger was made does not have to explain the dreams of the fry cook. It’s not a complete theory if an *explanatory entity* is outside of investigation. A recipe is not a scientific theory. fifthmonarchyman: Remember Gould’s Drunkard’s walk. We could hypothesize that that all things being equal the Oracle tends to push the drunk toward the curb or visa versa. You can speculate all you want, but a hypothesis has to have testable entailments. fifthmonarchyman: That the oracle is a black box is itself a necessary part of an explanation. You can’t explain the Oracle with out invalidating the explanation it facilitates. You haven't explained how that differs from simple ignorance. You can always section off the unknown, say it’s an oracle, and call it a day, but it’s scientifically sterile. fifthmonarchyman: Think of it as an axiom. like all axioms it can’t be described within the system it describes but is instead must be taken as a given. This is simply Godel’s incompleteness theorem. Gödel's theory does not apply to finite sets. In any case, you don't necessarily have to take the unknown as a given. For instance, the continuum hypothesis can be taken as either true or false with equal consistency. fifthmonarchyman: An axiom of the theory of evolution is that random variation occurs. There is no requirement that the inner-workings of randomness in a deterministic universe be explained before the theory of evolution can progress. We can observe mutations, and can show that they are random with respect to fitness, so the axiom is empirically supported. Furthermore, we can investigate the process of mutation to determine its causes. fifthmonarchyman: You can scientifically test whether the addition of an Oracle increases the explanatory power of “evolution” and you can scientifically probe the characteristics of the Oracle. Well, we're waiting. fifthmonarchyman: Prohibition of non-algorithmic entities is not science it’s merely materialistic metaphysics. What makes you think the oracle is non-algorithmic? Zachriel
Gary S. Gaulin says, Then explain how the system you call an “oracle” works. I say, That the oracle is a black box is itself a necessary part of an explanation. You can't explain the Oracle with out invalidating the explanation it facilitates. Think of it as an axiom. like all axioms it can't be described within the system it describes but is instead must be taken as a given. This is simply Godel's incompleteness theorem. It is high time that the science made peace with this mathematical reality. There will always be things that are necessarily true but must remain unexplained. You say, If what you are attempting to ultimately explain is “evolution” then the logical name of your proposed theory would be something like “Oracle Theory of Evolution” ....not the proposed “Theory of Intelligent Design” I say, Apparently you don't get what is trying to be explained in ID. Hint it's not intelligence itself. let me try and help you out The theory of "evolution" is offered as an explanation of the panorama of life not as an explanation of change. change is treated a given An axiom of the theory of evolution is that random variation occurs. There is no requirement that the inner-workings of randomness in a deterministic universe be explained before the theory of evolution can progress. The theory simply specifies that variation is random with respect to fitness as a starting point and moves on to the implications of that axiom. All I'm doing is introducing another axiom that increases the explanatory power of the theory. You say A scientific hypothesis is an idea you can scientifically test. A philosophical hypothesis is an idea you can philosophically test, I say That is why the Oracle idea is a scientific hypothesis and not a philosophical one. You can scientifically test whether the addition of an Oracle increases the explanatory power of "evolution" and you can scientifically probe the characteristics of the Oracle. You say, but even where the hypothesis holds true after you tested it that does not make it a scientific hypothesis or a theory I say, Again with the "it's not science unless it's algorithmic" meme. Prohibition of non-algorithmic entities is not science it's merely materialistic metaphysics. You are entitled to your own metaphysics but don't confuse them with science peace fifthmonarchyman
Also, if what you are attempting to ultimately explain is "evolution" then the logical name of your proposed theory would be something like "Oracle Theory of Evolution" or "Oracle Evolutionary Theory" but not the proposed "Theory of Intelligent Design" due to its premise specifically requiring you to explain how "intelligent cause" works. And this statement is a philosophical/religious "hypothesis" (not theory):
The oracle helps to explain how something works in this case “evolution”.
A scientific hypothesis is an idea you can scientifically test. A philosophical hypothesis is an idea you can philosophically test, but even where the hypothesis holds true after you tested it that does not make it a scientific hypothesis or a theory. Gary S. Gaulin
The oracle helps to explain how something works in this case “evolution”.
Then explain how the system you call an "oracle" works. Gary S. Gaulin
Gary S. Gaulin says A scientific theory explains how something works. I say. The oracle helps to explain how something works in this case "evolution". You are confusing explaining something with explaining everything. A Scientific Theory does not require us to explain everything. The problem arises when folks choose to resist explaining anything completely unless they can explain everything. That is not science that is simply materialistic metaphysics. you say, Therefore shrugging of shoulders (instead of explaining how something works) only explains why they have no scientific theory. I say, It's the other side who is shrugging their shoulders on this one. What do you think is going on when they say no complete theory is possible peace fifthmonarchyman
Maybe this might help fifthmonarchyman and others understand the commonsense logic of theories: A scientific theory explains how something works. Therefore shrugging of shoulders (instead of explaining how something works) only explains why they have no scientific theory. Also see: http://en.wikipedia.org/wiki/Theory_of_operation Gary S. Gaulin
Zac says By definition. I say Then you don't understand what I mean by a complete theory. A complete theory of how my cheese burger was made does not have to explain the dreams of the fry cook. It merely has to explain completely how my cheese burger came to be. The consciousness of the fry-cook is beside the point. You say. Great! What’s the next step? I say, quote: Do you really not know? I have an oracle I want to understand it as much as I can. It's simple. The problem is you rule out the oracle and choose to remain ignorant of it's characteristics. You say, Well, that’s very nice. But you haven’t proposed a testable model that explains anything. I say, Do you really have this limited an imagination. Must I spoon feed this stuff too you. Just think about all the different possible Oracles there could be. Each of these would have testable entailments. Again off the very top of my head. Remember Gould's Drunkard's walk. We could hypothesize that that all things being equal the Oracle tends to push the drunk toward the curb or visa versa. So when we come to gaps in the algorithm we would expect less complexity or visa versa. Surely you can think of tests of your own similar to this one. The important thing is we would not be stymied in our discovery efforts until we could explain every single step materialistically as we are now. I'm sensing that further discussion is not going to be productive. So unless you have something new other than the tired old "it can't possibly be Science" stuff, Then you can have the last word. peace peace fifthmonarchyman
fifthmonarchyman: Again is this an unwritten law or just a personal preference? By definition. fifthmonarchyman: Is archeology not science unless we can crack the mystery of consciousness? Archaeology is a science, but it's not a theory, much less a complete theory. fifthmonarchyman: We will not go any further into the nature of this oracle than to say that it cannot be a machine. Why not? fifthmonarchyman: You are going to have to provide evidence for this one From the definition, "The problem can be of any complexity class." http://en.wikipedia.org/wiki/Oracle_machine fifthmonarchyman: The replicators are part of the algroythym itself. Quibbling. Here's a reference to a computer oracle from Dembski et al. http://evoinfo.org/publications/efficient-per-query-information-extraction-from-a-hamming-oracle/ fifthmonarchyman: The replicators are part of the algroythym itself. In evolutionary algorithms, the environment and its relationship to the replicators is considered separate from the replicators. The only information that flows between them has to be limited to differential reproductive success. fifthmonarchyman: The recognition of the oracle is not the end of the process it’s just the beginning. Great! What's the next step? fifthmonarchyman: 1) What attributes must the oracle have 2) What other kinds of problems might be solvable by this particular Oracle 3) Are there tendency in the products of the Oracle? Is one outcome more likely than another. Well, that's very nice. But you haven't proposed a testable model that explains anything. Zachriel
Zac says, It’s not a complete theory if an explanatory entity is outside of investigation. I say, Again is this an unwritten law or just a personal preference? Is archeology not science unless we can crack the mystery of consciousness? you say, Computable as in the usual mathematical definition of computable, Turing computability. I say, quote: We will not go any further into the nature of this oracle than to say that it cannot be a machine. end quote: Alan Turing you say, In an evolutionary algorithm, it’s called an oracle because information about the workings of the oracle are not available to the replicators. I say, You are going to have to provide evidence for this one It's not an oracle unless it's workings are not available to us the observer. The replicators are part of the algroythym itself. you say, You can always section off the unknown, say it’s an oracle, and call it a day, but it’s scientifically sterile. I say, The recognition of the oracle is not the end of the process it's just the beginning. The alternative is to continue to ignore a perfectly viable solution because you don't like the implications. Talk about scientifically sterile You say, Propose an oracle theory that is testable and we can discuss. I say, There all all kinds of testable hypothesis we can explore. off the top of my head 1) What attributes must the oracle have 2) What other kinds of problems might be solvable by this particular Oracle 3) Are there tendency in the products of the Oracle? Is one outcome more likely than another. etc etc etc peace fifthmonarchyman
fifthmonarchyman: the purpose of the oracle is to yield a complete theory of course It's not a complete theory if an explanatory entity is outside of investigation. fifthmonarchyman: When you say computable you mean in theory right? Computable as in the usual mathematical definition of computable, Turing computability. fifthmonarchyman: If the Oracle is not a black box it is not an Oracle In an evolutionary algorithm, it's called an oracle because information about the workings of the oracle are not available to the replicators. fifthmonarchyman: here is a quote for you: That's lovely. You forgot the application. fifthmonarchyman: I’m frustrated by the exclusion of entities with explanatory utility for purely metaphysical reasons. You can always section off the unknown, say it's an oracle, and call it a day, but it's scientifically sterile. Propose an oracle theory that is testable and we can discuss. Zachriel
Suppose you have an oracle that only answers yes or no. Is there a question too complex to be answered by chipping away with yes or no questions? Petrushka
fifthmonarchyman says:
I’m frustrated by the exclusion of entities with explanatory utility for purely metaphysical reasons.
I’m frustrated by all the hype over a "better theory" that UD does not even want to discuss or further develop. The only interest I found was in promoting an imaginary scientific theory that leaves a void in science to fill with scripture. As far as science is concerned that is very wrong. Gary S. Gaulin
Zac says, It depends on your purpose, of course. I say, the purpose of the oracle is to yield a complete theory of course you say, Computable oracles are used in many applications, including evolutionary algorithms. I say, When you say computable you mean in theory right? If the Oracle is not a black box it is not an Oracle you say, You’re taking a mathematical model and apparently trying to apply it to science somehow. I say, here is a quote for you: quote: The miracle of the appropriateness of the language of mathematics for the formulation of the laws of physics is a wonderful gift which we neither understand nor deserve. We should be grateful for it and hope that it will remain valid in future research and that it will extend, for better or for worse, to our pleasure, even though perhaps also to our bafflement, to wide branches of learning. end quote: Eugene Wigner you say, It just seems you prefer to fill your gaps with Oracle-brand metaphysical paste. I say, No I want a better theory. I'm frustrated by the exclusion of entities with explanatory utility for purely metaphysical reasons. It seems you prefer an inferior theory simply because is does not offend your metaphysical sensibilities. peace fifthmonarchyman
fifthmonarchyman: Right, the benefits of an Oracle lie in it being incomputable. No. An oracle may or may not be a computable function. fifthmonarchyman: If we could break it down to a step by step process we would defeat the whole purpose that it serves in the first place. It depends on your purpose, of course. Computable oracles are used in many applications, including evolutionary algorithms. You're taking a mathematical model and apparently trying to apply it to science somehow. It just seems you prefer to fill your gaps with Oracle-brand metaphysical paste. Zachriel
Zac says, Just to clarify, the machine isn’t incomputable because there’s an oracle, but because the oracle is defined as incomputable. I say Right, the benefits of an Oracle lie in it being incomputable. If we could break it down to a step by step process we would defeat the whole purpose that it serves in the first place. This obvious fact has nothing to do with what is going on behind the curtain. What is going on back there is inaccessible to us if we want an oracle. peace fifthmonarchyman
fifthmonarchyman, that was helpful information. I still though don't understand how that applies to how intelligence and intelligent cause works. I know of no black-box or ever found internal workings that are not (technology willing) assessable to scientific study. If a model for what is inside the black-box is incomputable then there is no model to base a theory upon. Gary S. Gaulin
fifthmonarchyman: Is this an unwritten rule or is it just a preference on your part? It's an observation. Theories are formed by simplifying a messy reality, primarily by generalizing about a limited domain. fifthmonarchyman: Even though the values of functions based on oracle machines cannot be computed (since they are by definition incomputable), it is still possible to reason about which problems are reducible to oracles and which oracles they are reducible to. Posed another way, if a programmer had an oracle for a given problem, what other problems could be solved? Just to clarify, the machine isn't incomputable because there's an oracle, but because the oracle is defined as incomputable. Zachriel
zac says All scientific theories are incomplete because they are, at best, approximations. I say, Is this an unwritten rule or is it just a preference on your part? Gary S Gaulin says, Please elaborate. I say, check it out from here http://www.blythinstitute.org/images/data/attachments/0000/0041/bartlett1.pdf Quote: A Turing Oracle (hereafter oracle) is a black-box function (ie., no implementation description is given) which solves an incomputable function and yields its answer in a single step. An oracle machine is a combination of a normal computational system which also has access to an oracle. If the oracle is well-defined in its abilities,it can be used to reason about the process even if the process as a whole is incomputable. An oracle machine, then, is a regular machine (i.e.,a normal computable function) which is connected to an oracle (i.e., the function has access to an operation which is incomputable). Alan Turing describes the oracle machine as follows: Let us suppose that we are supplied with some unspecified means of solving number theoretic problems; a kind of oracle as it were. We will not go any further into the nature of this oracle than to say that it cannot be a machine. With the help of the oracle we could form a new kind of machine (call them o-machines), having as one of its fundamental processes that of solving a given number theoretic problem. (Turing, 1939) Even though the values of functions based on oracle machines cannot be computed (since they are by definition incomputable), it is still possible to reason about which problems are reducible to oracles and which oracles they are reducible to. Posed another way, if a programmer had an oracle for a given problem, what other problems could be solved? :end quote peace fifthmonarchyman
fifthmonarchyman:
Blackbox is not defined as untestable. It means the internal workings are not assessable to us.
Please elaborate. What exactly is "not assessable to us" and how did you scientifically test that conclusion? Gary S. Gaulin
fifthmonarchyman: It means the internal workings are not assessable to us. We might be able to surmise some aspects of the internal workings from its output, or make reasonable predictions about the output based on past performance. Please restate your point, if you had one. fifthmonarchyman: “repeating a famous quote more than three times is a sure sign your argument is bankrupt”- We asked whether you understood the quote, but you didn't answer. Meanwhile, you ask questions that are answered by reference to the quote. fifthmonarchyman: was it a complete explanation? All scientific theories are incomplete because they are, at best, approximations. fifthmonarchyman: The Point is that to be complete a theory of planetary formation does not need to explain all phenomena just those it purports to explain. All scientific theories are incomplete because they are, at best, approximations. fifthmonarchyman: Are you actually claiming that in order to be scientific a theory must fail to explain what it sets out to explain? All scientific theories are incomplete because they are, at best, approximations. Zachriel
Zac said, If you using the term oracle to mean a blackbox, then it is defined as untestable. I say, Blackbox is not defined as untestable. It means the internal workings are not assessable to us. We test lots things without seeing their inner-workings. Are you claiming that psychology and sociology and archeology are not science because human conscientiousness is a black-box? you say, Is this a case of you redefining terminology again? I say, Exploring implications is not the same as redefining terminology. If we can't use terms in new contexts then language is useless You say, You’re testing the output, not the oracle. I say, Just like in psychology and sociology and archeology among other things. you say, “All models are wrong, but some are useful.” — George E. P. Box I say "repeating a famous quote more than three times is a sure sign your argument is bankrupt"- Fifthmonarchyman ;-) peace fifthmonarchyman
fifthmonarchyman: was it a complete explanation? All models are wrong ... fifthmonarchyman: Who said it is beyond testing? If you using the term oracle to mean a blackbox, then it is defined as untestable. Is this a case of you redefining terminology again? It's as if we have to keep a list of your idiosyncratic vocabulary. fifthmonarchyman: Hypothesis: the oracle is not Zac. Hypothesis: the oracle does not violate natural law No way to tell. fifthmonarchyman: Hypothesis: results explained by the oracle can not confused for random noise. You're testing the output, not the oracle. fifthmonarchyman: The Point is that to be complete a theory of planetary formation does not need to explain all phenomena just those it purports to explain. All models are wrong ... fifthmonarchyman: Are you actually claiming that in order to be scientific a theory must fail to explain what it sets out to explain? “All models are wrong, but some are useful.” — George E. P. Box Zachriel
Zac said, Lots of things were once unexplained, then scientific explanations were later found. I say, was it a complete explanation? you say, An oracle is ad hoc and beyond testing, so has no scientific utility. I say, Who said it is beyond testing? I can think of all kinds of tests. For example off the top of my head. Hypothesis: the oracle is not Zac Hypothesis: the oracle does not violate natural law Hypothesis: results explained by the oracle can not confused for random noise. I'm just getting started Who said the oracle is ad hoc? It seems perfectly generalizable to other problems from physics to AI. Zac says, While there are robust theories of planetary formation, they are not complete. I say, Agreed but you are apparently missing the point. The Point is that to be complete a theory of planetary formation does not need to explain all phenomena just those it purports to explain. Are you actually claiming that in order to be scientific a theory must fail to explain what it sets out to explain? That is quite a claim There is no rule that a scientific theory must be incomplete unless you define Scientific as algorithmic. A theory that "saves more of the phenomena" is to be preferred. This should not be controversial peace fifthmonarchyman
fifthmonarchyman: Thus far no scientific theory has been able to explain consciousness does that mean that no scientific theory can do so? No. Lots of things were once unexplained, then scientific explanations were later found. fifthmonarchyman: Thus far no scientific theory has been able to unify relativity and QM does that mean that no scientific theory can do so? No. Lots of things were once unexplained, then scientific explanations were later found. Good example, though, of how theories are limited in their domains. In this case, where they would overlap, they are in contradiction. fifthmonarchyman: A theory does not have to explain every phenomena to be complete. It only has to explain what it attempts to explain completely. “All models are wrong, but some are useful.” — George E. P. Box fifthmonarchyman: A complete theory of planetary formation does not have to explain protein folding. Good example. While there are robust theories of planetary formation, they are not complete. fifthmonarchyman: the Oracle increases the fit to the evidence. As such it is not extraneous by definition. An oracle is ad hoc and beyond testing, so has no scientific utility. Zachriel
Zac says. Every scientific model ever proposed is incomplete. I say, Thus far no scientific theory has been able to explain consciousness does that mean that no scientific theory can do so? Thus far no scientific theory has been able to unify relativity and QM does that mean that no scientific theory can do so? come on Zac use you head Zac says, The “Theory of Everything” refers to the unification of physics, but would not, in fact, be a theory of every phenomena. I Say. A theory does not have to explain every phenomena to be complete. It only has to explain what it attempts to explain completely. A complete theory of planetary formation does not have to explain protein folding. use your head You say, Theories are judged by their fit to the evidence. Extraneous entities are vacuous. I say, the Oracle increases the fit to the evidence. As such it is not extraneous by definition. peace fifthmonarchyman
fifthmonarchyman: Evidence please? Every scientific model ever proposed is incomplete. fifthmonarchyman: Were the physicists who looked for a “theory of everything” not doing science? The "Theory of Everything" refers to the unification of physics, but would not, in fact, be a theory of every phenomena. fifthmonarchyman: There is no scientific reason that a theory must be incomplete is there? Models are always incomplete because they only concern certain aspects of the world with limitations on their domain of applicability. fifthmonarchyman: A theory can be wrong and complete or correct and incomplete. Do you not understand “All models are wrong, but some are useful”? fifthmonarchyman: Is the theory that explains less more likely to be true in Zac’s world? Theories are judged by their fit to the evidence. Extraneous entities are vacuous. Zachriel
Fifth: I recommend Wagner's Arrival of the Fittest. Perhaps you can borrow Joe's copy. He doesn't seem to be using it. Petrushka
Zac says, Scientific models are always incomplete. I say, Evidence please? Were the physicists who looked for a "theory of everything" not doing science? zac quotes All models are wrong I say. Wrong and complete are not antonyms. A theory can be wrong and complete or correct and incomplete. There is no scientific reason that a theory must be incomplete is there? Is the theory that explains less more likely to be true in Zac's world? peace fifthmonarchyman
Hey Petrushka, Do you know about my game? It's a long story! It's been my experience that a "solution" usually can be found for any particular feature in a string. Such solutions often come with a cost that shows up in other places however. So it's possible that we can narrow the search in one area only to expand it in another. Bottom line is we don't ever get rid of the Oracle we just move it around. peace fifthmonarchyman
fifthmonarchyman: An Algorithmic theory can provide detail but not completeness. Scientific models are always incomplete. “All models are wrong, but some are useful.” — George E. P. Box fifthmonarchyman: A non-Algorithmic theory can provide completeness but not detail. Scientific models are always incomplete. “All models are wrong, but some are useful.” — George E. P. Box Zachriel
A hybrid model like ID that has an Oracle in addition to an algorithmic process can provide the detail of an algorithm plus completeness.
The interesting thing in biology is whether the "oracle" is connected. That is, can you search functional space by making only small steps (mutations. If there are always functional neighbors, then positions are not isolated islands. In other words, for any given gene sequence, there will always be functionally equivalent variants (alleles). If so, Dembski's search is not 10^150 or 10^500. it is just the space one mutation away. And if a neighbor sequence is reached, it will have new neighbors. Petrushka
Me_Thinks No. I am saying we can’t model all the variables of evolution as we can’t capture the inputs of all the variables. I say, Suppose we quantified all the information that can be modeled with evolution and subtracted that information from the the information in life as a whole. This should be a relatively easy task if that what can be modeled has already been modeled Would what is left be random noise or would it contain interesting patterns? Would there be any way to even in principle to know these un-capture-able inputs were not the result of Design decisions? Peace fifthmonarchyman
Zac says That doesn’t mean a non-algorithmic model wouldn’t be a better fit. I say An Algorithmic theory can provide detail but not completeness. A non-Algorithmic theory can provide completeness but not detail. A hybrid model like ID that has an Oracle in addition to an algorithmic process can provide the detail of an algorithm plus completeness. Such a model is obviously a better fit than an algroythym alone peace fifthmonarchyman
Let me phrase my point a different way. Can you name or list something about evolution that Michael Behe says is possible that you say is impossible. Behe says common descent is possible. Petrushka
I understand your position. It seems to be that OOL is not possible without intervention. You are not addressing evolution, or if you are, you are simply wrong. I do not need to appeal to materialists on this. Behe will do. Petrushka
Petrushka #76
Chemistry contains the potential to self organize.
Chemistry contains the potential to self organize an organism as a set of Lego pieces contains the potential to self organize the Internet network, ie zero. The potential to create organization (with all its formalisms, information processing, functions, tasks, signal communication...) can come only from intelligence, the source of any formalism. Open any textbook of chemistry (inorganic and organic), you find molecules, reactions and basic laws to combine them. All the above organization formalisms are missing, both potentially and actually. Here EugeneS has just explained that better than me. niwrad
I'm not familiar with Kauffman. What I'm referring to is an experiment that could be replicated by anyone having a computer and able to write a simple program emulating logic gates. The question isn't about the origin of complexity, but about the connectedness of configurations. The challenge has been to see whether random testing of the immediate neighborhood can proceed indefinitely. Petrushka
Petrushka, Are you referring to what Kauffman did? He did not succeed with his 'edge of chaos' because logic comes first. There must be something that makes choices which gate to use, which one to throw away. Even if it is connected, this does not move you one bit further. You still need intelligence of some form. Rules =/= constraints. The crux of intelligence is decision making. In chemistry or physics, there is NO decision making ability hidden whatsoever. To state the opposite leads to a contradiction because the laws of nature cannot decide. At the level of physics/chemistry there is no propensity to making choices. Tilt the chessboard to see if gravity can play. EugeneS
fifthmonarchyman: In simple terms, the Church–Turing thesis states that a function on the natural numbers is computable in an informal sense Yes, and there's reason to believe that natural evolution is sufficiently granular that algorithmic evolution approximates natural evolution well enough to provide insights into the process. That doesn't mean a non-algorithmic model wouldn't be a better fit. fifthmonarchyman: If a process is algorithmic it can be modeled It doesn't mean it can be modeled as a practical matter. As already pointed out, we can't even model protein folding. fifthmonarchyman: Therefore “evolution” is a necessarily incomplete explanation for the grand scope of life. And gravity theory is an incomplete explanation. So? That doesn't mean it's not a valid theory. "All models are wrong, but some are useful." — George E. P. Box Zachriel
fifthmonarchyman @ 75
Tell that to Zac
Zac will certainly not agree with the truncated statement !
Are you conceding that any materialistic explanation will always and necessarily be incomplete??
No. I am saying we can't model all the variables of evolution as we can't capture the inputs of all the variables. What can be modeled has been modeled. Me_Think
Chemistry contains the potential to self organize. The question of whether this is designed is above my pay grade. Regarding the questionof whether fuctional sequences are endlessly connectable by single nucleotide substitutions is a decidable property of chemistry. Wagner says yes; Behe says no. Oddly enough, it's aproperty shared by logic circuits. Connect 16 or more logic gates -- AND, OR and NOT -- together in an arbitrary wiring, and you can replace them all without changing the output, by changing just one element at a time. The function space is connected. Petrushka
me_thinks says, Yes. Of-course, I say, Tell that to Zac you say but only if you are able to bring in all variables affecting evolution (or any process, for that matter). I say. Are you conceding that any materialistic explanation will always and necessarily be incomplete?? You say, How can you possibly compute all long ranging temporal environmental variables affecting evolution ? I say, You can't IMO. Therefore "evolution" is a necessarily incomplete explanation for the grand scope of life. What percentage of life would you think can be deemed to forever elude explanation by an "evolutionarily" model 10% 50%? peace fifthmonarchyman
fifthmonarchyman @ 73,
If a process is algorithmic it can be modeled
Yes. Of-course, but only if you are able to bring in all variables affecting evolution (or any process, for that matter). How can you possibly compute all long ranging temporal environmental variables affecting evolution ? Whatever could be modeled has been modeled (Eg: Check out Avida) Me_Think
me_think asks How is this relevant to modelling all/some aspects of evolution ? I say, If a process is algorithmic it can be modeled Peace fifthmonarchyman
Me_Think, You are greatly mistaken. It just means that installed molecular logic is at play. Chemotaxis is a wonderful example of preloaded intelligence. It is more complex than you think and can't be causally reduced just to chemistry. Although it utilizes it. Just like a television set utilizes the properties of electro-margnetic fields but in the entirety of its function cannot be reduced to those properties. Genuine decisions are made between multiple physico-chemically indeterminate states and by definition a bona fide decision is not reduced to chemistry. Whatever is explained by physics or chemistry alone cannot be classified as 'decision' as it is determined by law. In contrast, if you see in your experiments that something truly makes decisions, it merely means that this phenomenon causally traces back to intelligence. You are not seeing the woods for the trees. EugeneS
69 Andre @ 69
So chemicals do not obey the laws of nature, instead they yearn for truth?
It is the opposite. You don't need brains for complicated pathways and 'decisions'. Simple chemical and pressure based taxis will do. It doesn't bode well for 'mind is outside the body' ID science Me_Think
fifthmonarchyman @ 68
In simple terms, the Church–Turing thesis states that a function on the natural numbers is computable in an informal sense (i.e., computable by a human being using a pencil-and-paper method, ignoring resource limitations) if and only if it is computable by a Turing machine.
How is this relevant to modelling all/some aspects of evolution ? Me_Think
Me_Think
There is thus amazing control and ‘decision’ making just by chemicals.
So chemicals do not obey the laws of nature, instead they yearn for truth? That's not good for your position as a materialist. Just sayin..... Andre
zac says. Saying not all aspects of evolution can be modeled doesn’t mean they are or are not algorithmic. I say. quote: In simple terms, the Church–Turing thesis states that a function on the natural numbers is computable in an informal sense (i.e., computable by a human being using a pencil-and-paper method, ignoring resource limitations) if and only if it is computable by a Turing machine. and, To establish that a function is computable by Turing machine, it is usually considered sufficient to give an informal English description of how the function can be effectively computed, and then conclude "By the Church–Turing thesis" that the function is Turing computable (equivalently partial recursive). end quote: from here http://en.wikipedia.org/wiki/Church%E2%80%93Turing_thesis peace fifthmonarchyman
EugeneS @ 66,
If you mean the emergence of true novelty i.e. new function, control, you are wrong.
A single cell Slime Mold (Physarum polycephalum) can take decisions , solve mazes, replicate Japan rail network etc based on slime trails, by chemo-taxis and protoplasmic streaming. There is thus amazing control and 'decision' making just by chemicals. Me_Think
Petrushka, "Chemistry self-organizes". If you mean ordering phenomena, I am with you. However, ordering is trivial and it gives absolutely no novelty. If you mean the emergence of true novelty i.e. new function, control, you are wrong. Control presupposes measurement of state as part of working out control signals to steer the system towards a goal state with highest pragmatic utility. Control needs memory to store/retrieve representations of measurements of previous states. Measurement cannot be described by the laws of nature because in the centre of it lies semiosis, i.e. the problem of instantiation of the logical relation between the symbol and what it denotes. Consequently, it cannot be described by the laws of the dynamics of matter (whether in their physical or chemical form). Self-organization in the latter sense is a great misnomer regardless of the field, be in chemistry, physics, IT, biology, etc. Organization in the sense of function, control, formalism, pragmatic utility is impossible without intelligence. Physicality (the laws of nature) is inert to all that, it does not care if anything processes information to produce utility. It only provides some basic properties, the ground layer upon which intelligence can work further. Nothing non-trivial in the above sense is capable of emerging by itself just from physicality. EugeneS
Petrushka #60
Chemistry self-organizes. Software does not.
Here we must be very clear. A system can develop organization only if such system contains a previous related potentiality of organization inside itself. A bird egg develops a bird because the egg contains the embryologic potentiality of developing the bird. This potentiality is something of very precise, bit by bit, as a computer program. This is the reason why a bird egg develops a bird, not a... crocodile and viceversa. If "chemistry self-organizes" -- as you say -- and must account for the organization of life, then chemistry must contain the correspondent potentiality of such organization. The science of chemistry, as known today, contains nothing of that. It is a science of laws. These laws explain many basic things, but per se don't explain the organization of life. Analogy: arithmetic alone doesn't explain the organization of a computer, because we must add countless layers of formalisms to pass from arithmetic to computers. Analogously, if chemistry has to explain life it must embodies all the countless layers of formalisms that are necessary to pass from such basic laws to life's organization. In other words, chemistry should potentially embody all the designs of life. In short, chemistry would become, not simply a science of laws, rather a science of designs. This would mean that the Designer frontloaded in "chemistry" (in my above advanced design sense) just from the beginning all the intelligence needed for developing all the life forms in time. This way the difference between "chemistry" and human software would be that the latter is developed step by step in time, while the former is a giant "bio-software" all installed just from the beginning, before time, by the Great Designer of the universe. That is logically necessary, because nothing comes from nothing and more doesn't come from less (in a previous comment I had spoken of "matter of principles"...). niwrad
EugeneS #62 Thanks for your encouragements. It is persons like you who give us the strength to continue in our commitment to defend the truth, despite of all and until the end. niwrad
This should help with definitions:
Algorithm From Wikipedia, http://en.wikipedia.org/wiki/Algorithm In mathematics and computer science, an algorithm is a self-contained step-by-step set of operations to be performed. Algorithms exist that perform calculation, data processing, and automated reasoning.
The algorithm for demonstrating intelligence was found to be a simple four step loop. Biological brains have multiple memory systems working in parallel, each modeled using the same algorithm. (1) Sensory bits Address Memory, which sends Action Data to Motors. (2) If Confidence=0 then Guess new Memory Action. (3) Run Motor Contol one more Time Step in time. (4) Adust and Save new Confidence in Memory Action. If action was successful its confidence increments, else decrements. GoTo (1) To be objective as possible I qualify intelligence by these four steps required to model it. It's then not a fuzzy subjective opinion that uses one of the various dictionary type generalizations, like the algorithm for intelligence the system has all needed for it to self-learn. You then know it's there, by the model doing awesome things that does in fact take real intelligence. It's not like with Darwinian models where what it models is not real chemistry, in this case biology most easily accomplishes the storing of long term information in a memory made of molecules and/or neurons instead of digital RAM chips. Either will learn to control a muscle system, which results in the same thing happening at the same time in either intelligence even where one is being algorithmically modeled in a computer. Whether or not consciousness exists does not matter, either way intelligence is still being demonstrated. It's one advantage to focusing on what makes it possible for even molecular systems to learn over time. Focsing on something else greatly complicates the modeling challenge with an algorithm that becomes ever greater in size and detail by needing to model chemistry. The algorithm for intelligence remains the same four steps, regardless of how complex the system is or what its components are made from. What makes the virtual living things virtually alive (and like we are) are the multiple levels of intelligence at work at the same time in the model. To demonstrate intelligence any one level can be modeled by itself, while what happens in biology is emergent from the behavior of matter that powers genetic level intelligence that all by itself causes emergence of cellular intelligence that all by itself causes emergence of multicellular intelligence. It would take more than a PC to model behavior of matter on up but (when available) modeling that way is made easy by only two of the steps of the algorithm required for modeling matter. Darwinian models attempt to model everything that is the result of a relatively simple underlying intelligent process, by using a Selection variable and others intelligence does not have. Dawinian models are then stuck being an outside view generalization, while intelligence based models are stuck being the inside view that greatly simplifies understanding of what is most important to know about how living things develop, how they work. Gary S. Gaulin
Niwrad, Great stuff. Thanks. Absurdity raised to the Nth power. This is it. It's difficult to look for a black cat in a dark room, especially if it's not there. I like you posts very much. EugeneS
Sorry but for me is fully incomprehensible that even a professional programmer believes that organization can arise without intelligent design Seems like some of the assumptions of your worldview need to be updated then. wd400
The problem is you have no definition of self-organization. Chemistry self-organizes. Software does not. It is evident that you are not modelling chemistry. Petrushka
wd400 #58
Your doubts are misplaced. I’m a computational biologist, and use python, R, and C++ in my work, pretty much every day.
Sorry but for me is fully incomprehensible that even a professional programmer believes that organization can arise without intelligent design. In my mind a software developer Darwinist is a living contradiction. (I don't want to be offensive though). niwrad
Why? It helps to understand that increase of organization is never gratis because if you want to add a function into it you have to program new code representing the organization of the function. How does that help us understand biology? And in fact, how is that not trivially obvious without or without the quine. It is a simple concept but… something makes me doubt you are a programmer. Your doubts are misplaced. I'm a computational biologist, and use python, R, and C++ in my work, pretty much every day. FWIW, I'm aksing these questions because I really don't understand why certain programmers and engineers think they can talk about biology without referring to any actual biology (or indeed, knowing much biology). It's a curious phenomenon. wd400
niwrad: It helps to understand that increase of organization is never gratis because if you want to add a function into it you have to program new code representing the organization of the function. The demonstration only applies to the one instance. You can't extrapolate from your particular example because it is not a model exhibiting general properties, but an instance specific to the particular non-biological environment. Zachriel
wd400 #54
niwrad: "I continue to think that a quine, despite of its extreme simplicity and all its defects, can help to understand that increase of organization is never gratis." wd400: "Why? Nothing in your post seems to lead to this point."
Why? It helps to understand that increase of organization is never gratis because if you want to add a function into it you have to program new code representing the organization of the function. It is a simple concept but... something makes me doubt you are a programmer. niwrad
niwrad: No reasonable person conflates the model with the thing being modeled, like no one conflates the symbol with the symbolized thing. Reasonable people can be confused. A computer simulation is a model which should be independent of the programming language. Your example is not a model, but an instance that is tied intimately to the particular computer environment, an environment which is very brittle to mutation. Zachriel
My aim was only to show how a software replicator (an ultra simplified model/symbol of a bio replicator) behaves under random variations. So no biology. I showed that such model cannot get new functions without intelligent design Where? You evolutionists disagree and say that the biological reality behaves differently. Yup. I continue to think that a quine, despite of its extreme simplicity and all its defects, can help to understand that increase of organization is never gratis. Why? Nothing in your post seems to lead to this point. wd400
Zachariel #52
However, you had claimed to be simulating the process on a computer. There are problems with your simulation, most important, you conflate the model with the thing being modeled. In other words, the algorithm should be independent of the programming environment.
No reasonable person conflates the model with the thing being modeled, like no one conflates the symbol with the symbolized thing. Computer simulations can be bad or good but obviously they never are the reality itself. Also -- say -- flight simulators, which actually are very good, have not the same reality of an airplane, by definition. I think that on this concept we all agree. My aim was only to show how a software replicator (an ultra simplified model/symbol of a bio replicator) behaves under random variations. I showed that such model cannot get new functions without intelligent design. You evolutionists disagree and say that the biological reality behaves differently. I continue to think that a quine, despite of its extreme simplicity and all its defects, can help to understand that increase of organization is never gratis. niwrad
niwrad: I have no need of writing computer programs to know that random creation of organization is absurdity raised to the Nth power. Nowhere organization arises without intelligent design. It is a matter of principles. Glad you have principles. However, you had claimed to be simulating the process on a computer. There are problems with your simulation, most important, you conflate the model with the thing being modeled. In other words, the algorithm should be independent of the programming environment. Zachriel
DATCG at 49: I've read over Behe's responses -- several of them -- and haven't found any argument that K76T is deleterious. He doesn't address, for example, why it remains in populations after chloroquine is discontinued, and even remains in populations that regain sensitivity to quinine. I think Behe's entire Edge argument requires limited or narrow pathways, or some obstacle (the grand cnyon is his metaphor). In short, Behe and Wagner are on a collision course. Behe argues that pathways are narrow and full of roadblocks, and Wagner argues that there are innumerable paths. I would also point out that extinction is always an option. If bacteria could quickly evolve resistance to drugs then antibiotics wouldn't work at all. As it is , it takes years for populations to become resistant. Petrushka
Petrushka + Zachariel
You can’t disprove evolution by writing an incompetent computer program.
I have no need of writing computer programs to know that random creation of organization is absurdity raised to the Nth power. Nowhere organization arises without intelligent design. It is a matter of principles. If sometimes at UD I write notes from an informatics point of view it is because usually the readers who work in informatics are particularly apt to grasp the above principles. If I am "incompetent" at simulating evolution, well I wait for competent evolutionists to provide simulations proving their theory of unguided macroevolution of all species from unicellular. No wonder nothing of the sort has been produced by them thus far. It will never be, given the above impossibility. niwrad
thanks Petrushka, was curious if that was it. Aware of Behe's responses to Miller's claims. I'm sure UD will cover it, so I'll await that post and not discuss it more on Niwrad's post. DATCG
Pure contradiction. “Robustness (=resistance to variations) is the enabler for… variations.” Likewise, for you, brake is enabler of speed!
Ha! You have a completely different notion of 'Robustness'. I request you to read the latest literature Eg: Andreas Wagner's papers or you could read his book - 'Arrival of the Fittest'. Me_Think
Put simply, Behe describes a certain mutation -- K76T -- as deleterious. A grand canyon that must be jumped over before it can participate in a beneficial combination. Everyone else seems to disagree. Since K76T shows up wherever malaria is treated with cloroquinine, there are a lot of grand canyon jumpers. I suspect this is a case where Wagner's concept of multiple equivalent metabolic pathways might be relevant. I look forward to the exchange. Petrushka
Datcg at 41: http://www.millerandlevine.com/evolution/behe-2014/Behe-2014.pdf Behe has recently responded. Stay tuned. Petrushka
niwrad: Nowhere I said that a quine is a complete evolution simulator. As pointed out above, a quine, which is a replicating program, is a poor model because it conflates the model and what is being modeled. A proper simulation will be a process independent of the computational system. There will be a replicator, an environment, and a relationship between the two. "You can’t disprove evolution by writing an incompetent computer program." — Petrushka fifthmonarchyman: For convention Let’s call those aspects that can’t be modeled algorithmically “design functions”. Saying not all aspects of evolution can be modeled doesn't mean they are or are not algorithmic. For instance, a sequence folds into a complex three-dimensional shape with unevenly distributed charges. This can't be currently modeled, but it doesn't mean that the folding of enzymes is being done by an external agent. The evidence indicates it's a natural consequence of the molecular structure of the sequence. Indeed, random sequences can fold into structures that are designed to bind to specific substrates. Zachriel
Mapou #26
The only way for evolution to create new genes is via mutations. However, since bacteria could not survive without the ability to exchange genes, evolution never gets started.
Both statements are incorrect. New genes can be generated by recombination - from donor sequence both external and internal to the current cell - as well as by mutation. The boundary of a functional product is no more 'visible' to recombinational mechanisms than it is to mutation. And bacteria don't need to exchange genes to survive. They only need to replicate what they have. Of course getting that going is a tough problem. But such genes as are exchanged are replicated ones - copies of another sequence, not de novo sequences. Hangonasec
Me_Think #38
Do you really think Quine represents these aspects of evolution?"
Nowhere I said that a quine is a complete evolution simulator. However it is a first tile of the puzzle. And this tile just doesn't square with the pretension of unguided evolution.
Robustness is the enabler for searching new pathways and genetic variations.
Pure contradiction. "Robustness (=resistance to variations) is the enabler for... variations." Likewise, for you, brake is enabler of speed! niwrad
Robustness doesn't create anything. Robustness is just a word indicating that equivalent coding strings are always nearby. In terms of evolution, it also helps that there are many functionally equivalent metabolisms. The recent and ongoing research into robustness has changed th terms of the debate. If you want to claim that islands of function are isolated and unreachable, you are going to have to do some actual research. Actually, a lot of research. Petrushka
22 Petrushka said,
"Behe asserted it would take two simultaneous fortuitous mutations, but after the experiments are run, there are at least seven paths involving one mutation at a time."
Can you cite paper and experiments please? DATCG
Me_think says, Did I say RM/NS is abandoned ? Read again: No model can simulate all aspects of evolution. I say, "all aspects" is what I meant when I said "what ever". Read carefully I'm not saying you are abandoning RM/NS. I'm saying your side is abandoning the idea that evolution is algorithmic. If I understand you, you are saying that there are at least some parts of evolution that are non-algorithmic. These are what I'm calling "design functions". What parts of evolution are "design functions"? Is it more that 10%? 50%? peace fifthmonarchyman
fifthmonarchyman @ 37
I’m amazed that the other side is now abandoning the idea that RM/NS plus what ever is algorithmic. If we can’t model evolution what good is it as a materialistic theory?
Did I say RM/NS is abandoned ? Read again: No model can simulate all aspects of evolution. Me_Think
A quine is a model with possibility of replication, memory, inheritance, variation, all things related to evolution of unicellulars
Quine just prints it's own simple source code, so unless you 'pre-load' memory allocation, inheritance and variation in code, you won't get the output. Evolution depends on environment, random genetic mutation and competition pressure which are not just random but temporal too. Do you really think Quine represents these aspects of evolution? Why would anyone create Avida and Mendel's Accountant if all those could be done with a Quine ?!!
(Don’t say “biocells are robust” for I just explained that robustness opposes evolution.)
Why not? Robustness is what creates the genetic and metabolic hyper-dimension network (Wagner et al.). It is the enabler for searching new pathways and genetic variations. Me_Think
I'm amazed that the other side is now abandoning the idea that RM/NS plus what ever is algorithmic. If we can't model evolution what good is it as a materialistic theory? me_thinks says No model can simulate all aspects of evolution. I say, For convention Let's call those aspects that can't be modeled algorithmically "design functions". So now according to ME_think we can say that evolution equals "algorithmic functions" plus "design functions" Can you tell us precisely what aspects of evolution are "design functions"? Thanks in advance peace fifthmonarchyman
Me_Think #33
Quine as an evolutionary model is ‘not even wrong’. You need to raise the level of discussion to atleast Avida Vs Mendel’s Accountant.
A quine is a model with possibility of replication, memory, inheritance, variation, all things related to evolution of unicellulars. So it is not the crap you say. Macroevolution, as meant by all Darwinists, is "unicellular2man" by unintelligent processes. Before "raising the level of discussion" explains why a quine crashes when you apply unintelligent variations to get a new function. (Don't say "biocells are robust" for I just explained that robustness opposes evolution.) If you fail to get a single new function go figure if your blind evolution is able to create organisms, which are giant hierarchies of countless correlated functions. niwrad
Mung @ 34 Computer models do help us in simulating some aspects of evolution - like studying robustness, random walk, creating and testing multidimensional genome networks, studying fitness landscapes etc. Of course none of them can ever match Nature, but they do help in advancing our understanding of the evolutionary process. Me_Think
Me_Think: No model can simulate all aspects of evolution. You're missing the point. We're talking about computer models. Unless some significant progress has been made that I am unaware of, there are no biological computer models. Computer models can tell us nothing about evolution. Mung
No model can simulate all aspects of evolution. Quine as an evolutionary model is 'not even wrong'. You need to raise the level of discussion to atleast Avida Vs Mendel's Accountant. Me_Think
Petrushka: You can’t disprove evolution by writing an incompetent computer program. L Oh freaking L. Mung
Neil Rickert:
Maybe some evolutionists claim that. To me, that looks too simplistic. I see a need for the biology, not just the replication.
ok, but we're talking about computers. You disagree with those who think they can model biology in a computer?
And, as I see it, change in the environment is what drives a lot of evolutionary change.
So change in the environment causes genetic change? How does that work? Mung
fifthmonarchyman: Guaranteeing “fixity” while programing “nonfixity” is a problem that will not be solved by a clever computer program. As there exist evolutionary algorithms that don't crash, it's clear that crashing is not inevitable. fifthmonarchyman: If the weather is algorithmic what I have in my computer can be equivalent in every way to what I see float by on a summer day. What you have in your computer is a model, not clouds. If you model evolution, then you don't have organisms and environments, but models of organisms and environments. Zachriel
This is an excellent topic. From what I know Avida is now like the Holy Grail of Darwinian evolution education. This is what it has for variables:
The basic components of the Darwinian evolutionary mechanism are variation (V), inheritance (I), natural selection (S) and time (T).
http://avida-ed.msu.edu/ http://avida-ed.msu.edu/curricula/UnderstandIntroGenVar_Handout.pdf I bolded the words that precisely describe what it demonstrates, the "Darwinian evolutionary mechanism". What matters to ID is that the Darwinian model is not to demonstrate "intelligent cause" or even "intelligence" it's limited to "Darwinian evolution" only. The winning strategy is to provide the models required to demonstrate what ID theory is premised to explain. That's why the multiple level model explained by the theory was so vital to the success of ID in science. Only needed this, which is the same url my name links to: http://theoryofid.blogspot.com/ It's good to know how the Darwinian evolution model works. And for the ID movement it's even better to know how the ID model works. So does anyone want to code a simple as possible model to demonstrate intelligence? Which of the three intelligence levels to model up from is all up to you. Gary S. Gaulin
zac said, The k-complexity of a sequence is never much more than the length of the sequence itself. I say, yep that is the point. My guess and it's only a guess is that the required sequence will remain just out of reach. you say, Why would the simulation crash? I say, Did you not read the OP? Huh? I say, Guaranteeing "fixity" while programing "nonfixity" is a problem that will not be solved by a clever computer program. X+F-F=X no matter how you slice it You say, It’s like modeling the weather. You don’t have actual clouds inside your computer. I say, If the weather is algorithmic what I have in my computer can be equivalent in every way to what I see float by on a summer day. That is the point. peace fifthmonarchyman
fifthmonarchyman: 1) We could measure the complexity of the model. I have a guess that it will quickly become apparent that any successful mutating quine will have unbounded k-complexity. The k-complexity of a sequence is never much more than the length of the sequence itself. fifthmonarchyman: 2) I agree that what we really need to do is model the environment that contains the quine in addition to the quines themselves. This multiplies the required complexity exponentially. It multiplies the complexity of the environmental simulation, but not of the replicators, which is the fundamental error made in the original post. fifthmonarchyman: 3) Increased computing power will not provide any relief from the dilemma. All it will do is allow the stimulation to crash at an increased rate. Why would the simulation crash? fifthmonarchyman: 4) A better designed program will not provide relief because the problem is not lack of efficiency. You can’t make 2+2=5 simply by developing a more clever arithmetic. Huh? fifthmonarchyman: 5) I’m amazed that the other side is now claiming that computer programs are a poor analog for evolution. A quine, which is a replicating program, is a poor model because it conflates the model and what is being modeled. A proper simulation will be a process independent of the computational system. There will be a replicator, an environment, and a relationship between the two. It's like modeling the weather. You don't have actual clouds inside your computer. Mapou: Bacteria’s ability to exchange genes is not an evolutionary mechanism for the simple reason that it does not create new genes. Of course it's an evolutionary mechanism. Zachriel
Petrushka got it all wrong:
Does complexity harm the evolutionary hypothesis? Not at all. For one thing, microbes do not have to invent everything by themselves. They exchange genes rather promiscuously. Anything new and useful spreads.
Bacteria's ability to exchange genes is not an evolutionary mechanism for the simple reason that it does not create new genes. It is an adaptive mechanism that presupposes the existence of a huge variety of existing functional genes. The only way for evolution to create new genes is via mutations. However, since bacteria could not survive without the ability to exchange genes, evolution never gets started.
Does it make origin of life a difficult problem? Yes. but we already knew that.
A difficult problem? That's a laugh. It shows that random processes are insufficient to create living, self-replicating, adaptive organisms. The gap between molecules and the complexity of life is impossible for natural forces to breach. The only solution is design. It's simple, really, but have fun drinking your kool-aid. Looking at the rings on your hands does not make you a hell of a man. Mapou
Hey niwrad, This is an excellent and fascinating thread a couple of observations if I may 1) We could measure the complexity of the model. I have a guess that it will quickly become apparent that any successful mutating quine will have unbounded k-complexity. 2) I agree that what we really need to do is model the environment that contains the quine in addition to the quines themselves. This multiplies the required complexity exponentially. 3) Increased computing power will not provide any relief from the dilemma. All it will do is allow the stimulation to crash at an increased rate. 4) A better designed program will not provide relief because the problem is not lack of efficiency. You can't make 2+2=5 simply by developing a more clever arithmetic. 5) I'm amazed that the other side is now claiming that computer programs are a poor analog for evolution. The whole point of Darwin's contribution was to frame biology in an algorithmic light. RM/NS is simply an algorithm. Algorithms are what computers were made for. peace fifthmonarchyman
Very well, show me a textbook that says amoebas evolved into men. It appears that you are confusing modern microbes with common ancestors. This is wrong on two counts. The firs, and least important, is that modern microbes are every bit as evolved as multi-celled organisms. the second, is that evolution is a walk along the network of chemistry that allows survival. It doesn't have a direction toward anything. If the Cambrian were re-run, we would expect to see entirely different outcomes. Populations change, but they do not evolve toward targets or goals. As for Behe's Edge, it is the best reasoned objection to evolution. Behe accepts common descent. His argument is that there are places on the map with no roads connecting towns. this is a testable proposition, but the technology to do the testing is recent. Petrushka
Petrushka #22 "Amoebas have not evolved to men ... if you want to test evolution, take something like Behe’s chloroquine resistance in malaria". You seem to sell a low-profile definition of evolution. But in school for "evolution" they teach children Darwin's macroevolution, not Behe’s chloroquine resistance. niwrad
Amoebas have not evolved to men, and there is no biologist who thinks they could in the remaining lifetime of the earth. That is the kind of thing creationists think evolution means, and that confusion is why folks like Doug Axe can make a living proving it doesn't happen. If you want to test evolution, take something like Behe's chloroquine resistance in malaria, and do the chemistry to see how possible it is. Behe asserted it would take two simultaneous fortuitous mutations, but after the experiments are run, there are at least seven paths involving one mutation at a time. The multiplicity of paths is part of what I mean by robustness. The universality of genomic robustness could not have been tested ten years ago. The manufacture of biological molecules was not possible at a reasonable cost. Now it can be tested, and it turns out that the possibility of silent mutations and multiple pathways is the rule, rather than the exception. So if you are going to model evolution, you need to model actual chemistry, which is much more permissive than computer programming. Petrushka
Petrushka #19 If ameba has to evolve to man something has to change, robustness or non-robustness. You claim robustness (=fixity) contra ID computer simulation and in the same time you need much variation (=non-fixity) to produce evolution. You cannot have both ways. niwrad
In other words, you’re saying that evolution requires a lot more complexity to work properly than we thought. How does that help Darwinism?
It makes evolution possible. This has been known for a long time, but it has only recently been possible to manufacture molecules to order at a cost that allows testing this hypothesis. Does complexity harm the evolutionary hypothesis? Not at all. For one thing, microbes do not have to invent everything by themselves. They exchange genes rather promiscuously. Anything new and useful spreads. Does it make origin of life a difficult problem? Yes. but we already knew that. Petrushka
Programs are good analogies of biological developments. You cannot deny the development of any organism is a programmed construction, a task governed by instructions (like computer programs).
Programs are a terrible analogy if they are not robust in the face of mutation. If you cannot make small changes to program code without changing the output of the program, you are not modeling chemistry. Not every change needs to be silent, but some silent changes must exist. To model biochemistry, you must be able to walk a program stepwise -- without changing its function -- to an equivalent program with little of the original code. This can be done with genes, so if you are modelling genetics, your programming environment must incorporate this robustness. Petrushka
Petrushka:
When you can write a computer program that can not only replicate itself, but can tolerate mutation (can have silent mutations) and which exists in an environment in which silent mutations can be chained so that a chain of equivalent programs can exist leading to an equivalent program with none of the original code — then you will have a passable model of chemistry.
Interesting. In other words, you're saying that evolution requires a lot more complexity to work properly than we thought. How does that help Darwinism? Where did this complexity come from? How did it evolve? And here I was thinking that Darwinian evolution goes from the simple to the complex when, all along, things have to be complex to start with. How did this pile of blatant, in-your-face superstitions find its way into science? Inquiring minds and all that. Mapou
Petrushka #16 "The fact that genomes are robust and computer programs are not robust means that programs are not good analogies to genomes." Programs are good analogies of biological developments. You cannot deny the development of any organism is a programmed construction, a task governed by instructions (like computer programs). The relative robustness of genomes is an intelligent design feature. In a sense robustness of genomes disagrees with unguided evolution. In fact, evolution needs variations, if genomes resist variations then they oppose evolution. niwrad
You can't disprove evolution by writing an incompetent computer program. You can model some of the mathematical features of evolution on computer, but you cannot yet model the chemistry involved. The fact that genomes are robust and computer programs are not robust means that programs are not good analogies to genomes. Edit to add: When you can write a computer program that can not only replicate itself, but can tolerate mutation (can have silent mutations) and which exists in an environment in which silent mutations can be chained so that a chain of equivalent programs can exist leading to an equivalent program with none of the original code -- then you will have a passable model of chemistry. Petrushka
Nirward, I'm not sure if you are equivocating on the meaning of "many" or are just throwing out chaff. Just to be clear. When I said "many mutations" I mean a substantial proportion of spontaneously arising mutations have a "bad" effect. There is nothing in that fact to suggest evolution is impossible. wd400
…many mutations had the effect to morph ameba into man.
That is not something a biologist would claim. Petrushka
Petrushka #12 "That is not something a biologist would claim." That is exactly what Darwinian biologists do claim, random mutations + natural selection => all-species. niwrad
...many mutations had the effect to morph ameba into man.
That is not something a biologist would claim. Petrushka
wd400 #10 "...many mutations have a bad effect." You deny evolution, many mutations had the effect to morph ameba into man. niwrad
Evolutionary biologists simulate relevant parts of evolutionary theory through programs. Your example has no biology, and "proves" something everyone already knows: many mutations have a bad effect. That evolutionary biology still exists when everyone knows this suggest the "problem" is no a "show stopper". So, I'm left wondering what the point is, and what we learn from the program example that we can't learn from studying actual biology. wd400
wd400 #8 "The computer program seems to be completely unnecessary." Evolutionists offer computer programs of evolution. That is good. IDers offer computer programs of non-evolution. That is bad. Heads you lose, tails I win... niwrad
You think a program is like an organism? The computer program seems to be completely unnecessary. No one in the right mind thinks mutations can't be damaging, a poor analogy to biology doesn't get us any further. wd400
wd400 #6 "I think you forgot to explain why this “problem” is a show-stopper, or indeed a problem at all?" Why it is a problem? Yourself can verify on your computer. Apply random variations on the quine, open its source code in a text editor, and type into it with your eyes closed. This way you simulate blind evolution. You will see the program crashes. niwrad
All that leads us directly to what I call the “quine dilemma” of unguided evolution. If random variations are harmless or neutral (blue zone) they create no new organization. If evolution has to create complex functional novelties, new organization, it must operate in the red zone and necessarily become potentially destructive. To speak of “dilemma” here is euphemistic. This dilemma is worse than Hamletic, because de facto is a show-stopper for evolution.
I think you forgot to explain why this "problem" is a show-stopper, or indeed a problem at all? wd400
"Today we will try to partially layout how to simulate on computer such process." Thanks for interesting beginning. Did you review papers and discussions at Evolutionary Informatics Lab at Baylor by Marks, Dembski, et al? Their review and discussion on Avida programs and simulation? If so, curious how you see this in comparison to Avida. Avida claims
The claim of the paper was to demonstrate "how complex functions can originate by random mutation and natural selection."
DATCG
You seem to be conflating the model with the thing being modeled. A quine is a self-replicating program, not a simulated organism. Keep in mind that once you have a biological replicator, it entails the acquisition of the resources necessary for replication. It might be simple, such as an available pool of nucleotides, but the resources are necessarily limited, hence there is competition for those resources. Furthermore, in nature, there is a complex environment, as well as an intimate relationship between the structure of the replicator and the environment, which provides opportunities. A complete model of this would necessarily be extremely complex, indeed, intractable; so evolutionary algorithms use simplified environments. But even these can be quite complex. Arguing that providing a complex environment undermines the concept of an evolutionary algorithm is fallacious. Rather, the complexity of the environment and its relationship to the replicator is as much a given as is the ability to replicate. Zachriel
Evolutionists claim that what it takes to evolution to work is simply “a populations of replicators, random variations on them, and a competition for survival or resources”.
Maybe some evolutionists claim that. To me, that looks too simplistic. I see a need for the biology, not just the replication. And, as I see it, change in the environment is what drives a lot of evolutionary change.
At this point the question is: what variations are necessary to transform our initial quines into evolved predators or resource seekers?
I see that as a mistake. You are wanting evolution to be directional, to evolve toward particular outcomes. But I see evolution as a population doing its best to survive in changing circumstances. Without environmental change, I don't expect to see a lot of evolutionary change. Neil Rickert
If you have Perl installed on your computer you can easily verify that if you run this script, it prints itself on the screen. Eventually if you redirect the output to a file, such file will be a perfect copy of its generator file.
I think to parallel nature, your quine also needs to be running as a program simultaneously with all the other quines. Otherwise it's just an illustration of a snowflake. Which raises the question, what in nature is parallel to the application program running that allows for umpteen quines to run at once? Which raises another question what in nature parallels quines communicating with each other for the benefit of the whole and where did that protocol come from? awstar
Thought provoking, on what is required to set up a self-sustaining, replicating body plan, starting with the very first. kairosfocus