Darwinist rhetorical tactics Functionally Specified Complex Information & Organization ID Foundations

On Active Information, search, Islands of Function and FSCO/I

Spread the love

A current rhetorical tack of objections to the design inference has two facets:

(a) suggesting or implying that by moving research focus to Active Information needle in haystack search-challenge linked Specified Complexity has been “dispensed with” [thus,too, related concepts such as FSCO/I]; and

(b) setting out to dismiss Active Information, now considered in isolation.

Both of these rhetorical gambits are in error.

However, just because a rhetorical assertion or strategy is erroneous does not mean that it is unpersuasive; especially for those inclined that way in the first place.

So, there is a necessity for a corrective.

First, let us observe how Marks and Dembski began their 2010 paper, in its abstract:

Needle-in-the-haystack problems look for small targets in large spaces. In such cases, blind search stands no hope of success. Conservation of information dictates any search technique will work, on average, as well as blind search. Success requires an assisted search. But whence the assistance required for a search to be successful? To pose the question this way suggests that successful searches do not emerge spontaneously but need themselves to be discovered via a search. The question then naturally arises whether such a higher-level “search for a search” is any easier than the original search. We prove two results: (1) The Horizontal No Free Lunch Theorem, which shows that average relative performance of searches never exceeds unassisted or blind searches, and (2) The Vertical No Free Lunch Theorem, which shows that the difficulty of searching for a successful search increases exponentially with respect to the minimum allowable active information being sought.

That is, the context of active information and associated search for a good search, is exactly that of finding isolated targets Ti in large configuration spaces W, that then pose a needle in haystack search challenge. Or, as I have represented this so often here at UD:

csi_defnUpdating to reflect the bridge to the origin of life challenge:

islands_of_func_chall

In this model, we see how researchers on evolutionary computing typically confine their work to tractable cases where a dust of random walk searches with drift due to a presumably gentle slope on what looks like a fairly flat surface is indeed likely to converge on multiple zones of sharply rising function, which then allows identification of likely local peaks of function. The researcher in view then has a second tier search across peaks to achieve a global maximum.

This of course contrasts with the FSCO/I [= functionally specific, complex organisation and/or associated information] case where

a: due to a need for multiple well-matched parts that

b: must be correctly arranged and coupled together

c: per a functionally specific wiring diagram

d: to attain the particular interactions that achieve function, and so

e: will be tied to an information-rich wiring diagram that

f: may be described and quantified informationally by using

g: a structured list of y/n q’s forming a descriptive bit string

. . . we naturally see instead isolated zones of function Ti amidst a much larger sea of non-functional clustered or scattered arrangements of parts.

This may be illustrated by an Abu 6500 C3 fishing reel exploded view assembly diagram:

abu_6500c3mag

. . . which may be compared to the organisation of a petroleum refinery:

Petroleum refinery block diagram illustrating FSCO/I in a process-flow system
Petroleum refinery block diagram illustrating FSCO/I in a process-flow system

. . . and to that of the cellular protein synthesis system:

Protein Synthesis (HT: Wiki Media)
Protein Synthesis (HT: Wiki Media)

. . . and onward the cellular metabolic process network (with the above being the small corner top left):

cell_metabolism

(NB: I insist on presenting this cluster of illustrations to demonstrate to all but the willfully obtuse, that FSCO/I is real, unavoidably familiar and pivotally relevant to origin of cell based life discussions, with implications onward for body plans that must unfold from an embryo or the like, OOL and OOBP.)

Now, in their 2013 paper on generalising their analysis, Marks, Dembski and Ewert begin:

All but the most trivial searches are needle-in-the-haystack problems. Yet many searches successfully locate needles in haystacks. How is this possible? A success-ful search locates a target in a manageable number of steps. According to conserva-tion of information, nontrivial searches can be successful only by drawing on existing external information, outputting no more information than was inputted [1]. In previous work, we made assumptions that limited the generality of conservation of information, such as assuming that the baseline against which search perfor-mance is evaluated must be a uniform probability distribution or that any query of the search space yields full knowledge of whether the candidate queried is inside or outside the target. In this paper, we remove such constraints and show that | conservation of information holds quite generally. We continue to assume that tar-gets are fixed. Search for fuzzy and moveable targets will be the topic of future research by the Evolutionary Informatics Lab.

In generalizing conservation of information, we first generalize what we mean by targeted search. The first three sections of this paper therefore develop a general approach to targeted search. The upshot of this approach is that any search may be represented as a probability distribution on the space being searched. Readers who are prepared to accept that searches may be represented in this way can skip to section 4 and regard the first three sections as stage-setting. Nonetheless, we sug-gest that readers study these first three sections, if only to appreciate the full gen-erality of the approach to search we are proposing and also to understand why attempts to circumvent conservation of information via certain types of searches fail. Indeed, as we shall see, such attempts to bypass conservation of information look to searches that fall under the general approach outlined here; moreover, conservation of information, as formalized here, applies to all these cases . . .

So, again, the direct relevance of FSCO/I and linked needle in haystack search challenge continues.

Going further, we may now focus:

is_ o_func2_activ_info

In short, active information is a bridge that allows us to pass to relevant zones of FSCO/I, Ti, and to cross plateaus and intervening valleys in an island of function that does not exhibit a neatly behaved objective function. And, it is reasonable to measure it’s impact based on search improvement, in informational terms. (Where, it may only need to give a hint, try here and scratch around a bit: warmer/colder/hot-hot-hot. AI itself does not have to give the sort of detailed wiring diagram description associated with FSCO/I.)

It must be deeply understood, that the dominant aspect of the situation is resource sparseness confronting a blind needle in haystack search. A reasonably random blind search will not credibly outperform the overwhelmingly likely failure of the yardstick, flat random search. Too much stack, too few search resources, too little time. And a drastically improved search, a golden search if you will, itself has to be found before it becomes relevant.

That means, searching for a good search.

Where, a search on a configuration space W, is a sample of its subsets. That is, it is a member of the power set of W, which has cardinality 2^W. Thus it is plausible that such a search will be much harder than a direct fairly random search.  (And yes, one may elaborate an analysis to address that point, but it is going to come back to much the same conclusion.)

Further, consider the case where the pictured zones are like sandy barrier islands, shape-shifting and able to move. That is, they are dynamic.

This will not affect the dominant challenge, which is to get to an initial Ti for OOL then onwards to get to further islands Tj etc for OOBP.  That is doubtless a work in progress over at the Evolutionary Informatics Lab, but is already patent from the challenge in the main.

To give an outline idea, let me clip a summary of the needle-to-stack challenge:

Our observed cosmos has in it some 10^80 atoms, and a good atomic-level clock-tick is a fast chem rxn rate of perhaps 10^-14 s. 13.7 bn y ~10^17 s. The number of atom-scale events in that span in the observed cosmos is thus of order 10^111.

The number of configs for 1,000 coins (or, bits) is 2^1,000 ~ 1.07*10^301.

That is, if we were to give each atom of the observed cosmos a tray of 1,000 coins, and toss and observe then process 10^14 times per second, the resources of the observed cosmos would sample up to 1 in 10^190 of the set of possibilities.

It is reasonable to deem such a blind search, whether contiguous or a dust, as far too sparse to have any reasonable likelihood of finding any reasonably isolated “needles” in the haystack of possibilities. A rough calc suggests that the ratio is comparable to a single straw drawn from a cubical haystack ~ 2 * 10^45 LY across. (Our observed cosmos may be ~ 10^11 LY across, i.e. the imaginary haystack would swallow up our observed cosmos.)

Of course, as posts in this thread amply demonstrate the “miracle” of intelligently directed configuration allows us to routinely produce cases of functionally specific complex organisation and/or associated information well beyond such a threshold. For an ASCII text string 1,000 bits is about 143 characters, the length of a Twitter post.

As just genomes for OOL  start out at 100 – 1,000 k bases and those for OOBP credibly run like 10 – 100+ mn bases, this is a toy illustration of the true magnitude of the problem.

The context and challenge addressed by the active information concept is blind needle in haystack search challenge, and so also FSCO/I. The only actually observed adequate cause of FSCO/I is intelligently directed configuration, aka design. And per further experience, design works by injecting active information coming from a self-moved agent cause capable of rational contemplation and creative synthesis.

So, FSCO/I remains as best explained on design. In fact, per a trillion member base of observations, it is a reliable sign of it. Which has very direct implications for our thought on OOL and OOBP.

Or, it should. END

272 Replies to “On Active Information, search, Islands of Function and FSCO/I

  1. 1
    kairosfocus says:

    On Active Information, search, Islands of Function and FSCO/I

  2. 2
    Zachriel says:

    kairosfocus: Conservation of information …

    No such theorem. You probably mean Wolpert & Macready’s No Free Lunch Theorem.

    kairosfocus: … any search technique will work, on average, as well as blind search.

    However, biological evolution is a specific ‘search algorithm’, not the universal set of search algorithms; and the natural environment is a specific ‘fitness landscape’, not the universal set of fitness landscapes.

    kairosfocus: {snip all that follows from faulty premises}

  3. 3
    cantor says:

    2 Zachriel May 2, 2015 at 7:27 am

    kairosfocus: … any search technique will work, on average, as well as blind search.

    However, biological evolution is a specific ‘search algorithm’,

    What part of “any” don’t you understand?

    .

  4. 4
    kairosfocus says:

    Z:

    1: I do not “mean” anything, I am explicitly (and cf the onward linked papers of some dozens of pp altogether) citing Marks, Dembski et al who do have an extensive discussion of conservation of info in search-like phenomena, including several theorems. So, no, conservation of information theorems and associated proofs exist, first restricted cases then more general ones, with yet more in prospect.

    2: before we get to OOBP and discuss biological evolution, we have to first get to biology, i.e. OOL.

    3: This means that we have to find an empirically grounded means of getting to the FSCO/I in cell based life including that of its embedded von Neumann code using kinematic self replication facility. Where, the only empirically warranted source of bridging active information is intelligently directed configuration, aka design.

    4: Having got to first cell based life with the only empirically warranted serious candidate on the table being design, that puts the next issue, that design sits at the table as of right in explaining OOBP.

    5: Where, it then emerges that such OOBP must be an account of origin of even more FSCO/I than for OOL dozens of times over, which then is under the same point, that FSCO/I is credibly a reliable signature of design as cause.

    6: In addition, the hoped for all-powerful chance variation plus differential reproductive success subtracting less successful varieties leading to incremental descent with modification extrapolated across the tree of life icon, is in the relevant sense a search mechanism, with chance variation creating novel varieties and with differential success subtracting the less successful.

    7: This mechanism is known to work somewhat within islands of function, e.g. finch beak lengths, and the variations of red deer that move from North American Elk to the deer of Europe . . . which were found to interbreed in the forests of New Zealand. And of course the different finches of the Galapagos are found to interbreed too.

    8: Varieties of dogs show that artificial selection can create similar variations.

    9: However, given the FSCO/I challenge, leading to the need to cross gaps between islands, which are of order 10 – 100+mn bases as a first estimate, OOBP by this mechanism is not credible.

    10: Further to this, there is no good evidence of a vast continent of incrementally improved varieties from microbes to Mozart, mice, molluscs and mango trees. Not in a fossil record of sudden appearance and disappearance, stasis of form and gaps in morphology. Not in the molecular structures, especially protein families and AA sequence space gaps (a capital case of islands of function), not in the Cambrian Revolution, not in bio-geography, not in industrial melanism, not in homology, not in real embryology, and so forth. Yes, it is demanded ideologically by a priori evolutionary materialism and it is imposed via the iconic tree of life, but the evidence for it, of incremental and cumulative broad, branching variation that starts with closely linked populations then diverges step by step is just not there.

    11: On the contrary the evidence just of deeply isolated protein fold domains in AA sequence space and the abundance of very small domains that often appear distinctively among taxonomically close species, itself points to a strong case of islands of function.

    12: So, once we move away from an initial island, we do face the issue of crossing a sea of non-functional forms to achieve another island of function, and this will make the very same Darwinist mechanism very likely indeed to meet the same challenge that it is unlikely to find the needles in the haystack given available atomic resources and time.

    _____________

    Nope, Darwinist chance variation and differential reproductive success leading to culling out of less successful varieties, is not a credible free lunch, golden search.

    Unsurprising, as the engines of variation it appeals to boil down to chance processes uncorrelated with future possible success, it can only reward so to speak immediate incremental success embedded in a population. (Notice, so-called natural selection, is an information SUBTRACTION process, culling out less successful varieties.)

    As in, your hoped for snip, find a point to dismiss, then sweep away all else, fails.

    KF

  5. 5
    kairosfocus says:

    Cantor, thanks for watching my 6. Indeed, we may ask, what part of “any” and what part of the need for OOBP to cross intervening arms of the ocean of non-functional forms on a small planet of is it 10^43 atoms within available time, is it that is not understood. But, for ideological reasons, there is a common perception that “natural selection” answers to all problems once we get to an initial self replicating entity. It seems to be very hard for the adherents to see that hey are facing serious issues with that broad brush extrapolation from minor adaptation to OOBP. And the bigger problem yet of accounting for OOL involving gated metabolic automata using sophisticated nanotech machinery, and embedding a code using von Neumann self replication facility is even more a-begging. But then it’s what, coming on three years we have had no comers to seriously answer to the UD essay challenge to provide an adequately warranted basis for the ToL from the root up. KF

  6. 6
    Carpathian says:

    kairosfocus:

    The information for a software replicator is not very improbable.

    That is something I intend to pursue in parallel with an ID model.

    I think that ID has bigger problems than improbability to deal with.

    For instance, what kind of body plan should I design to fit into a current environment without knowing what that environment will be like 200 years in the future?

    Information about the future is necessary in order for ID to work.

  7. 7
    NetResearchGuy says:

    Zachriel: First off, OOL precedes the existence of biological evolution, which starts when the first self replicator with heritable random variation comes into existence. So your entire point is moot regarding OOL.

    Second, what makes you think biological evolution is such a great search? It suffers from a lot of limitations. The most critical limitation is that each step in the search process must be “functional”, i.e. a living organism. The second limitation is that the random variation source is extremely poor quality — most mutations are neutral or harmful, and mutations are rare making the search process slow. Third, the selection function (what reproduces or not) is poor — it can only operate as a binary function on an entire organism, and can’t apply fine tuning to a specific feature at a time. Also there is a ton of randomness involved — for example 98% of tadpoles die before becoming adult frogs, and it’s mostly blind luck. What if one of those that died had a beneficial mutation?

    Those are just limitations specific to biological evolution. Evolutionary algorithms in general have additional challenges. There is a lot of fine tuning required to make one work. If selection is too strong, an algorithm gets stuck on local maxima, whereas if it’s too weak, it takes too long or drifts away from fitness peaks. If the mutation rate is too low, novelty requiring coordinated mutations can never arise, and if it’s too high, you get too many deleterious mutations fixed, leading to error catastrophe. Even with perfect tuning, it’s possible for fitness landscapes to exist that are impossible to traverse given finite resources.

    So far, the argument of evolutionists is question begging: because life exists, we know the search algorithm and fitness landscape are masterfully fine tuned to produce complex life. They have never demonstrated this fact though, and all experimental attempts to do so have failed. Their materialist philosophy requires them to believe this on faith.

    I will now predict your response: you will say that my description of evolution “leaves out some details”, and those magic golden details fix all the problems I listed (for example, to one of my previous comments, you threw out the buzzword “recombination” — I could pick that apart if you want). Or changing fitness landscapes, or deep time, or drift, or whatever. No evidence will be given to mathematically support the position that those details matter. Before you mention a detail, consider how much it really helps.

    The problem is that adding lots of details makes it impossible to rigorously model something. That’s what evolutionists hide behind: ID researchers haven’t modeled evolution completely — well neither have evolutionists! At least ID researchers try to make forward progress on the math (such as studying Active Information), whereas evolutionists are happy with their just so stories.

  8. 8
    velikovskys says:

    NRG:
    Second, what makes you think biological evolution is such a great search? It suffers from a lot of limitations. The most critical limitation is that each step in the search process must be “functional”, i.e. a living organism. The second limitation is that the random variation source is extremely poor quality — most mutations are neutral or harmful, and mutations are rare making the search process slow. Third, the selection function (what reproduces or not) is poor — it can only operate as a binary function on an entire organism, and can’t apply fine tuning to a specific feature at a time. Also there is a ton of randomness involved — for example 98% of tadpoles die before becoming adult frogs, and it’s mostly blind luck. What if one of those that died had a beneficial mutation?

    It makes you wonder what kind of designer creates such an inefficient design.

  9. 9
    velikovskys says:

    KF:

    A question, part 21200 seems to be a gear of some kind, if one loses a tooth does the FSCO/i decrease? if so does the broken tooth have the missing Fsco?

    The gear might be functional with the loss of one tooth but say we lose 10, and the part is no longer functional as per the design, would that be any ,part or total loss of FSco ? Thanks

    Just curious

  10. 10
    Zachriel says:

    cantor: What part of “any” don’t you understand?

    What part of “on average” don’t you understand. The term refers to the universe of possible search algorithms and the universe of possible fitness landscapes. However, specific search algorithms may do better on specific fitness landscapes. The question, then, isn’t mathematical, but empirical.

    kairosfocus: So, no, conservation of information theorems and associated proofs exist, first restricted cases then more general ones, with yet more in prospect.

    They only have validity as restatements of No Free Lunch theorems. Notably, you called it “conservation of information”, but made a statement from the No Free Lunch theorem.

    NetResearchGuy: First off, OOL precedes the existence of biological evolution, which starts when the first self replicator with heritable random variation comes into existence.

    The original post is discussing “evolutionary computing”.

    NetResearchGuy: Second, what makes you think biological evolution is such a great search? It suffers from a lot of limitations.

    Quite so! The vast majority of structures will be forever outside the reach of evolutionary search.

    NetResearchGuy: If selection is too strong, an algorithm gets stuck on local maxima, whereas if it’s too weak, it takes too long or drifts away from fitness peaks.

    Recombination resolves most of the problems associated with being stuck on local maxima.

    NetResearchGuy: They have never demonstrated this fact though, and all experimental attempts to do so have failed.

    There is substantial scientific evidence supporting evolution. You might start with the historical record.

  11. 11
    niwrad says:

    Zachriel

    There is substantial scientific evidence supporting evolution. You might start with the historical record.

    If “evolution” is the creation by unguided processes from a common ancestor of all the different kinds of living beings ever existed, its evidence is exactly zero.
    The historical record, being a collection of static findings, cannot prove such evolution (= dynamic) by definition.

  12. 12
    Bob O'H says:

    In this model, we see how researchers on evolutionary computing typically confine their work to tractable cases where a dust of random walk searches with drift due to a presumably gentle slope on what looks like a fairly flat surface is indeed likely to converge on multiple zones of sharply rising function, which then allows identification of likely local peaks of function.

    Huh? I thought GAs were used in a lot of horrible cases when, for example, the fitness function wasn’t (a) tractable (in which case by definition we can use methods which find the maximum), and (b) are multi-modal, so occasional large-scale changes are needed to find the maximum. Multi-modal problems are usually pretty intractable.

  13. 13
    Joe says:

    If unguided evolution is any clue then publishing is meaningless. 😛

  14. 14
    NetResearchGuy says:

    For me, the most fundamental piece of evidence that is missing to support evolution is the scarcity of phenotypic vestigial features in extant organisms. (Evolutionists will claim the genome contains vestigial junk in the form of duplicate or non coding DNA, which is debatable, but let’s ignore that).

    For example, the just so story for eye evolution suggests that some specialized refractive cells happened to spontaneously appear at the front of the eye to become a cornea. How did the cells know where to appear in the body? Why didn’t they appear in the animal’s mouth or spleen or elsewhere?

    In order for evolution to function, it must have tried a ton of failed experiments placing refractive cells in random places all over the body, before it finally had that success. So if we were to look at living organisms, we should see bizarre cases in biology of animals with parts in random places that do nothing useful. It should be the rule, rather an exception. I know there are mutations that cause extra wings or legs or toes, but those are just the same genes triggered extra times, not novel vestigial features. And those are never selectively beneficial.

    Regarding the historical record, I want to see a fossil with half a wing. Show me that.

  15. 15
    Zachriel says:

    niwrad: The historical record, being a collection of static findings, cannot prove such evolution (= dynamic) by definition.

    A temporal sequence can certainly support a scientific hypothesis when the sequence is entailed in that hypothesis.

    NetResearchGuy: or example, the just so story for eye evolution suggests that some specialized refractive cells happened to spontaneously appear at the front of the eye to become a cornea.

    Refraction is actually a complex process, so it wouldn’t appear randomly, but only after a number of other potentiating changes. The basic process is thought to be an eyepatch, then invagination, then a transparent membrane to protect the organ, then this creates a situation which is conducive to further specialization and refraction. Each of these steps is incremental and selectively advantageous, meaning it constitutes a plausible pathway. There are examples of each of these intermediate forms in nature.

    NetResearchGuy: Regarding the historical record, I want to see a fossil with half a wing.

    There are many intermediate wings in nature.
    http://iliketowastemytime.com/.....irrel1.jpg

  16. 16
    kairosfocus says:

    AS,

    isn’t it interesting how you neatly omitted the context in which I highlighted how Orgel and Wicken went on record since the 1970;s as well as Dembski in NFL pp 148 – 9 and Meyer in his response to Falk, from here on:

    http://www.uncommondescent.com.....ent-562213

    Do you see why you force me to conclude that you are playing a snip out of context and snipe, strawman caricature game?

    Besides, you are here plainly refusing to attend to something so blatantly obvious in front of you that it is at the level of A is A.

    Namely, that there is a relevant subset of specific, complex information in the biological and technological etc worlds, where the specification is functional. That is, per a wiring diagram configuration, parts put together in a certain cluster of ways will work to do a job that emerges from their interaction. This is shown in the OP above in several ways — and don’t overlook the text strings that because of how they are put together constitute a message in English.

    Let me clip Dembski from NFL, to help rivet the point, as he is discussing something that should not even be remarkable, just a description of patent facts — the debate should be on where functionally specific complex organisation and/or associated information [FSCO/I] as a relevant subset of the more general CSI comes from, not whether it exists. That, is patent, it is undeniably all around us, it is in the text of this comment and yours too, in the PCs we are using, and in our bodies beginning with DNA code and organisation of the living cell.

    So, Dembski:

    p. 148:“The great myth of contemporary evolutionary biology is that the information needed to explain complex biological structures can be purchased without intelligence. My aim throughout this book is to dispel that myth . . . . Eigen and his colleagues must have something else in mind besides information simpliciter when they describe the origin of information as the central problem of biology [–> which is precisely a functional context].

    I submit that what they have in mind is specified complexity, or what equivalently we have been calling in this Chapter Complex Specified information or CSI . . . .

    Biological specification always refers to function. An organism is a functional system comprising many functional subsystems. . . . In virtue of their function [[a living organism’s subsystems] embody patterns that are objectively given and can be identified independently of the systems that embody them. Hence these systems are specified in the sense required by the complexity-specificity criterion . . . the specification can be cashed out in any number of ways [[through observing the requisites of functional organisation within the cell, or in organs and tissues or at the level of the organism as a whole.

    {Dembski cites:}

    Wouters, p. 148: “globally in terms of the viability of whole organisms,”

    Behe, p. 148: “minimal function of biochemical systems,”

    Dawkins, pp. 148 – 9: “Complicated things have some quality, specifiable in advance, that is highly unlikely to have been acquired by ran-| dom chance alone. In the case of living things, the quality that is specified in advance is . . . the ability to propagate genes in reproduction.”

    On p. 149, he roughly cites Orgel’s famous remark from 1973, which exactly cited reads:

    In brief, living organisms are distinguished by their specified complexity. Crystals are usually taken as the prototypes of simple well-specified structures, because they consist of a very large number of identical molecules packed together in a uniform way. Lumps of granite or random mixtures of polymers are examples of structures that are complex but not specified. The crystals fail to qualify as living because they lack complexity; the mixtures of polymers fail to qualify because they lack specificity . . .

    And, p. 149, he highlights Paul Davis in The Fifth Miracle: “Living organisms are mysterious not for their complexity per se, but for their tightly specified complexity.”] . . .”

    p. 144: [[Specified complexity can be more formally defined:] “. . . since a universal probability bound of 1 [[chance] in 10^150 corresponds to a universal complexity bound of 500 bits of information, [[the cluster] (T, E) constitutes CSI because T [[ effectively the target hot zone in the field of possibilities] subsumes E [[ effectively the observed event from that field], T is detachable from E, and and T measures at least 500 bits of information . . . ”

    Meyer, backs this up:

    http://www.signatureinthecell......l-falk.php

    . . . [[W]e now have a wealth of experience showing that what I call specified or functional information (especially if encoded in digital form) does not arise from purely physical or chemical antecedents[[–> i.e. by blind, undirected forces of chance and necessity]. Indeed, the ribozyme engineering and pre-biotic simulation experiments that Professor Falk commends to my attention actually lend additional inductive support to this generalization. On the other hand, we do know of a cause—a type of cause—that has demonstrated the power to produce functionally-specified information. That cause is intelligence or conscious rational deliberation. As the pioneering information theorist Henry Quastler once observed, “the creation of information is habitually associated with conscious activity.” And, of course, he was right. Whenever we find information—whether embedded in a radio signal, carved in a stone monument, written in a book or etched on a magnetic disc—and we trace it back to its source, invariably we come to mind, not merely a material process. Thus, the discovery of functionally specified, digitally encoded information along the spine of DNA, provides compelling positive evidence of the activity of a prior designing intelligence. This conclusion is not based upon what we don’t know. It is based upon what we do know from our uniform experience about the cause and effect structure of the world—specifically, what we know about what does, and does not, have the power to produce large amounts of specified information . . .

    All I have done is to add complexity to emphasise that there is a threshold — typically, 500 – 1,000 bits will be more than enough — of complexity to get the phrase, functionally specific complex information, FSCI. As GP often emphasises, when this is openly digitally coded FSCI, we may abbreviate dFSCI. And to take in cases where wiring diagram organisation may imply rather than openly express the information [per, a structured string of Y/N Q’s that describe and specify the functional wiring diagram, cf. OP above], we have functionally specific complex organisation and/or associated information, FSCO/I.

    There is no need to go to a further publication in the professional literature to establish the meaning and reasonableness of the term.

    That, should be trivial.

    That it is not, is a sign that it has such strong basis that by any rhetorical means deemed necessary it is to be excluded, dismissed, expelled, locked out and mocked with schoolyard taunt tactics.

    That, is shameful.

    KF

    PS, someone above asked what will happen to the FSCO/I in a reel that has a damaged gear. With a wholly or partially missing tooth, it might work though not as well — still within the island of function — this sometimes happens with automobile transmissions. If it is bent and will jam the gearing, it might cause functionality to cease. In either case, the issue of FSCO/I beyond a threshold being a reliable marker of design remains.

    .

  17. 17
    kairosfocus says:

    Bob O’H: With a lot of fine tuning, GAs and kin work within islands of function. With the addition of a higher tier they may search out a global optimum, as the cited case gives. But when the dominating sea of non function is material, it takes bridging active information to get them to do the job. And that is intelligently provided. KF

  18. 18
    kairosfocus says:

    Carpathian:

    The information for a software replicator is not very improbable.

    The hardware and background software to get a software replicator to work is quite complex beyond 128 bytes worth of info, thank you.

    Unsurprisingly, it is designed.

    But, we are not interested in cellular automata and toy virtual worlds. That is why I have repeatedly pointed to a von Neumann, code using KINEMATIC self replicator. Whether with nanotech or a so-called clanking replicator, that is what is needed.

    And, that is no mean feat, we can conceptualise but a system such as a universal computer or the like that embeds an integrated vNSR is by no means trivial. And as I have pointed to, it is what would trigger Industrial Revolution 3.) and open the gateway to solar system colonisation.

    That is also what in effect Paley put on the table over 200 years ago when he amplified on just what sort of watch he was pondering in his thought exercise, in Ch 2 of Nat Theol:

    Suppose, in the next place, that the person who found the watch should after some time discover that, in addition to all the properties which he had hitherto observed in it, it possessed the unexpected property of producing in the course of its movement another watch like itself — the thing is conceivable; that it contained within it a mechanism, a system of parts — a mold, for instance, or a complex adjustment of lathes, baffles, and other tools — evidently and separately calculated for this purpose [==> update, vNSR, with tape [a bar of cams is a program, as was used in so many C18 automata], and constructor keyed in as an ADDITIONAL facility integrated with the main machine — of course, IIRC a full size clanking unit considered by NASA was many, many tons in scale] . . . .

    The first effect would be to increase his admiration of the contrivance, and his conviction of the consummate skill of the contriver. Whether he regarded the object of the contrivance, the distinct apparatus, the intricate, yet in many parts intelligible mechanism by which it was carried on, he would perceive in this new observation nothing but an additional reason for doing what he had already done — for referring the construction of the watch to design and to supreme art [–> notice, the impact of seeing ADDITIONAL FSCO/I] . . . . He would reflect, that though the watch before him were, in some sense, the maker of the watch, which, was fabricated in the course of its movements, yet it was in a very different sense from that in which a carpenter, for instance, is the maker of a chair — the author of its contrivance, the cause of the relation of its parts to their use.

    That is where we need to begin considerations.

    And I find it highly material to note that I have never seen a dismissal of Paley’s watch discussion that seriously engages with what is implied by this.

    KF

  19. 19
    niwrad says:

    Zachriel

    A temporal sequence can certainly support a scientific hypothesis when the sequence is entailed in that hypothesis.

    Also if, for hypothesis, we concede that transformations occurred no evolutionist can prove that such transformations were the result of *unguided* evolution (instead of intelligent design frontloading).

  20. 20
    Zachriel says:

    niwrad: Also if, for hypothesis, we concede that transformations occurred no evolutionist can prove that such transformations were the result of *unguided* evolution (instead of intelligent design frontloading).

    Baby steps. So you now agree that historical evidence can support a scientific hypothesis? Say, the hypothesis that dinosaurs once roamed the Earth?

  21. 21
    kairosfocus says:

    Niw, & Z:

    A serious survey of the molecular situation and the gross morphology will come up with an absence of actual demonstration of the gradualistic incrementalism of the dominant neo-darwinian school of thought.

    That is a demonstrable fact that led to the rise of a whole alternate school of thought, punctuated equilibria.

    And, I simply point to Meyer’s runaway continued bestseller, Darwin’s Doubt, for book length documentation despite the usual attempted brushoffs.

    The fact is, the issues on the Cambrian fossil revolution were on the table in Darwin’s day and he hoped future finds would fill in gaps. 150 years later, 250,000+ fossil species, millions in museums, billions in the ground and the pattern of gaps is still there.

    The dominant pattern is NOT gradualism from unicellular organisms across the branching tree of life. And with so large a sample in hand, if such gradualism were dominant, it should have long since been seen.

    The island of function issue is abundantly manifest with proteins in AA space on up to gross body plans.

    Which is what the logic tells us to expect.

    If you dispute this simply provide the dominant evidence of this gradualism.

    And we have not touched the biggest gap of all: OOL.

    KF

  22. 22
    kairosfocus says:

    Z, that there were animals that we now call dinosaurs (and which would have been termed dragons at other times and places) that have gone extinct is neither a scientific hypothesis nor a seriously disputed fact. What would be a relevant hyp with good grounding would be molecular and gross morphology observations documenting in detail their gradual arrival. That, you do not have, so please do not take something else and use it in a way that suggests what you do not have. KF

    PS: A gliding squirrel is nowhere near addressing the transformations to create a bat or a bird.

  23. 23
    Mung says:

    Evolutionary algorithms in general have additional challenges. There is a lot of fine tuning required to make one work.

    Indeed. They must be carefully designed.

  24. 24
    kairosfocus says:

    Mung, as must essentially be all software and the hardware for them to run on. Ponder what is really going on in a Java Hello World. Or even an HTML one. KF

  25. 25
    cantor says:

    10 ZachrielMay 2, 2015 at 10:50 am

    specific search algorithms may do better on specific fitness landscapes

    Only because the information is front-loaded into the landscape (or the search).

    You haven’t solved the problem of the origin of that information.

    All you’ve done is push the problem to another domain.

  26. 26
    Zachriel says:

    kairosfocus: A serious survey of the molecular situation and the gross morphology will come up with an absence of actual demonstration of the gradualistic incrementalism of the dominant neo-darwinian school of thought. That is a demonstrable fact that led to the rise of a whole alternate school of thought, punctuated equilibria.

    Stephen Gould: “Since we proposed punctuated equilibria to explain trends, it is infuriating to be quoted again and again by creationists—whether through design or stupidity, I do not know—as admitting that the fossil record includes no transitional forms. Transitional forms are generally lacking at the species level, but they are abundant between larger groups.”

    kairosfocus: If you dispute this simply provide the dominant evidence of this gradualism.

    We can start with the overall historical record. You will find an overall evolutionary pattern; from unicellular organisms to simple colonies to chordates to vertebrates to gnathostomes to land vertebrates to amniotes to dinosaurs to birds; to consider one lineage, showing a broad pattern of unfolding of new traits. We could then look in more details at the specifics, such as dinosaurian phylogeny.

    kairosfocus: Z, that there were animals that we now call dinosaurs (and which would have been termed dragons at other times and places) that have gone extinct is neither a scientific hypothesis nor a seriously disputed fact.

    Good. Then we can reject the contention that we can’t make some reasonable determinations of historical events.

    cantor: Only because the information is front-loaded into the landscape (or the search).

    That isn’t necessary. We can take a standard evolutionary algorithm and apply it to many different landscapes. Landscapes that exhibit certain types of local clustering are amenable to evolutionary search. The natural world exhibits local clustering, so is amenable to evolution.

  27. 27
    kairosfocus says:

    Z, you are well off on tangents. A sure sign of little to say on the focal matter: origin of FSCO/I by blind needle in haystack mechanisms that show an alternative to intelligent injection of active information; i/l/o the challenge of want of atomic and temporal resources to cross/bridge seas of non-function. That said, it is quite plain that you do in fact have the problem of the gaps and lack of observational evidence to document blind watchmaker, body plan level macro evolution, and even moreso, naturalistic origin of life. What has been going on for far too long is at root a priori evolutionary materialism, and then suggesting and extrapolating far more than the actual evidence warrants. That, BTW is the fundamental reason why the coming on three years standing essay challenge has stood without a solid response. KF

  28. 28
    Mung says:

    Zachriel is just wrong. Again.

  29. 29
    cantor says:

    27 Zachriel May 2, 2015 at 6:44 pm

    We can take a standard evolutionary algorithm and apply it to many different landscapes. Landscapes that exhibit certain types of local clustering are amenable to evolutionary search. The natural world exhibits local clustering, so is amenable to evolution.

    It’s hard to tell if you are serious or joking.

    Firstly, “certain types of local clustering” is information.

    Secondly, if by “evolution” you mean and finch beaks and antibiotic resistance then you are equivocating.

    ~

  30. 30
    Mung says:

    Zachriel:

    We can take a standard evolutionary algorithm and apply it to many different landscapes.

    Please post a link to this “standard evolutionary algorithm.” I’d love to see the source code.

    What Zachriels meant to say was that you can create different evolutionary algorithms and they will each work for a different landscape, a landscape designed to work for that specific algorithm.

    Give a different fitness landscape to an algorithm not designed for that landscape and you may not get anything useful from the algorithm.

  31. 31
    niwrad says:

    Zachriel

    So you now agree that historical evidence can support a scientific hypothesis? Say, the hypothesis that dinosaurs once roamed the Earth?

    No. Obviously I don’t agree that historical evidence supports unguided evolution. The issue is not if “dinosaurs once roamed the Earth”. You know well the issue is an incomparably different thing, but you prefer to digress as usual. The real issue is if all the bio-complexity was created by unguided natural forces. That is an impossible thing in principle because bio-organization cannot be created by the physical/chemical laws science states.

  32. 32
    kairosfocus says:

    VS, Following back up. If one knocks out a single main gear tooth, performance will likely degrade but plausibly will continue. Ironically if the tooth is instead bent or distorted so as to jam the gear train, function will obviously cease. Stripping out 10 or so teeth will almost certainly destroy the function of the main gear . . . which is the one you highlighted. The import of this is (a) there is a range of variation within an island of function and (b) it has a shoreline so to speak. The problem in view is to arrive at the shore going the other way, in the context of function of cell based life forms. FSCO/I is quantifiable on structured string of Y/N q’s to describe the wiring diagram arrangement, and the “complex” part of this descriptive acronym — in the direct context of functional, wiring diagram specificity — points to a threshold beyond which it is not plausible for such an entity to arise by blind chance and/or mechanically necessary processes down to atomic level on the gamut of our solar system or the observed cosmos; 500 – 1,000 bits being a useful value. That can be confirmed on a back of the envelope calculation for 10^80 atoms, doing 10^13 – 14 steps of action per s [a fast chem rxn rate], for 10^17 s, as the OP discusses briefly. KF

    PS: If you blow up scale on the sphere diagrams, you will see random walk squiggles across the baseline surface. They are random, in a phase space type situation, if a process is deterministic, once it crosses itself it must be in a loop . . . it cannot cross itself at an angle so to speak, only one trajectory of states can cross a given particular state, but injection of chance variability breaks that iron predictability, as does of course intervention by external action. Oddly, that looping is more or less what so often happens in long division where a recurring loop appears, in a pattern of deterministic algorithmic repeating steps. Notice, how we are taught to impose halting on the loop in Grade school — this is of course a major problem with algorithms. A random walk can of course have a superposed drift; much as wind superposes on the molecular motion of air molecules.

  33. 33
    kairosfocus says:

    Mung:

    Give a different fitness landscape to an algorithm not designed for that landscape and you may not get anything useful from the algorithm.

    Thus, we readily see fine tuning as a pattern with such algorithms.

    KF

  34. 34
    Zachriel says:

    kairosfocus: you are well off on tangents.

    You brought up punctuated equilibrium. The scientist who devised the theory would certainly have some idea what the theory states.

    kairosfocus: origin of FSCO/I by blind needle in haystack mechanisms that show an alternative to intelligent injection of active information

    Your claim is that the landscape is not traversable. Looking at the history of adaptation would surely be relevant.

    cantor: Firstly, “certain types of local clustering” is information.

    Depending on your definition of information. For instance, sunlight is highly organized. The strongest radiation comes from a single source, which moves on a regular cycle.

    Mung: Please post a link to this “standard evolutionary algorithm.”

    Generate the initial population of individuals randomly, then repeat the following steps:

    1. Evaluate the individual fitness of each member of population;
    2. Compare each indvidual to landscape to determine best fit individuals;
    3. Breed new individuals through mutation and crossover to give birth to offspring;
    4. Replace least-fit population with new individuals.

    Mung: What Zachriels meant to say was that you can create different evolutionary algorithms and they will each work for a different landscape, a landscape designed to work for that specific algorithm.

    Same algorithm.

    Zachriel: So you now agree that historical evidence can support a scientific hypothesis? Say, the hypothesis that dinosaurs once roamed the Earth?

    niwrad: No.

    Seriously? You don’t think we can evaluate historical evidence to determine that dinosaurs once roamed the Earth.

    niwrad: Obviously I don’t agree that historical evidence supports unguided evolution.

    That wasn’t the question.

    niwrad: The issue is not if “dinosaurs once roamed the Earth”.

    Your statement was that “The historical record, being a collection of static findings, cannot prove such evolution (= dynamic) by definition.” That claim isn’t just about evolution, but about a collection of static findings providing evidence of change over time. You even say it is “by definition”!

    There were once mega-dinosaurs roaming the Earth. Today mega-dinosaurs do not roam the Earth. That is change. The scientific evidence strongly supports that such a change has occurred.

  35. 35

    Cantor:

    Firstly, “certain types of local clustering” is information.

    If clustering is information (and in some senses it can be regarded as such), then you are using a definition of information that has nothing to do with design. Lots of processes result in clustering, and most of them are non-biological.

    Secondly, if by “evolution” you mean and finch beaks and antibiotic resistance then you are equivocating.

    No, he is not. Both are examples of fast adaptive evolution.

  36. 36
    cantor says:

    35 Zachriel May 3, 2015 at 7:07 am

    cantor: Firstly, “certain types of local clustering” is information.

    Depending on your definition of information. For instance, sunlight is highly organized. The strongest radiation comes from a single source, which moves on a regular cycle.

    It is very difficult to tell if you are intentionally equivocating or just very uninformed or confused.

    Firstly, if you don’t know what information means in the context of ID then you shouldn’t be posting about it until you take the time and effort to educate yourself.

    Secondly, you didn’t bother to clarify what you meant by “evolution” in your post #27.

    ~

  37. 37
    cantor says:

    36 Elizabeth Liddle May 3, 2015 at 7:16 am

    If clustering is information (and in some senses it can be regarded as such), then you are using a definition of information that has nothing to do with design. Lots of processes result in clustering, and most of them are non-biological.

    Now you are equivocating Elizabeth. Look at the context of his use of the word “clustering”.

    Both are examples of fast adaptive evolution.

    More equivocation. Adaptation is non-controversial. Finch beaks and antibiotic resistance is not what we are discussing here.

    ~

  38. 38
    kairosfocus says:

    Z, Pardon but you have given a general outline not an algorithm. KF

  39. 39
    Zachriel says:

    cantor: Firstly, if you don’t know what information means in the context of ID then you shouldn’t be posting about it until you take the time and effort to educate yourself.

    With regards to the environment, you had said, “certain types of local clustering” is information. The environment is highly ordered, and it ordered in a way that is amendable to evolutionary processes. Here’s your original statement:

    cantor: Only because the information is front-loaded into the landscape (or the search).

    In biology, the landscape is the natural world, including other organisms. The natural world isn’t random, but highly ordered. We can model the evolutionary process with an evolutionary algorithm.

    cantor: Secondly, you didn’t bother to clarify what you meant by “evolution” in your post #27.

    Depends on which statement. We referred to both biological evolution and evolutionary algorithms.

    cantor: Look at the context of his use of the word “clustering”.

    Elizabeth Liddle correctly read our comment, saying “Lots of processes result in clustering, and most of them are non-biological.”

    kairosfocus: Pardon but you have given a general outline not an algorithm.

    An algorithm is step-by-step description of a process, which was provided. The fitness landscape is external to the evolutionary algorithm.

  40. 40

    Cantor:

    If clustering is information (and in some senses it can be regarded as such), then you are using a definition of information that has nothing to do with design. Lots of processes result in clustering, and most of them are non-biological.

    Now you are equivocating Elizabeth. Look at the context of his use of the word “clustering”.

    I did. As I said, clustering does provide “information” in the sense in which Ewert, Dembski and Marks use the term, whether that is the clustering of fitnesses around phenotypes, or the clustering of environmental properties with similar environmental properties. It appears to be a ubiquitous property of our universe, from sub-atomic to intergalactic scales (as I said elsewhere recently) and it is that property that makes smooth fitness landscapes possible. To be specific: because similar genotypes have similar phenotypes (e.g. your children are like, but not identical to you), and because similar phenotypes have similar properties vis a vis the resources and threats of the environment (long legs are good for running, whether they are 30 cm or 29), then organisms that reproduce as biological organisms do will find themselves on a smooth fitness landscape.

    You could argue that the universe we live in, in which such things are possible, is itself unlikely but that’s not an argument about evolution.

    Both are examples of fast adaptive evolution.

    More equivocation. Adaptation is non-controversial. Finch beaks and antibiotic resistance is not what we are discussing here.

    I would hope adaptation is non-controversial. But if it is, then Darwinian evolution is non-controversial, as that is what it is – adaptation. There are other factors in evolution that Darwin did not consider, e.g. drift, but adaptive evolution is what Origin is about.

    So if you are discussing something other than Darwinian evolution, you might be in the wrong thread! That’s what “evolutionary search” is.

  41. 41
    kairosfocus says:

    Z, from an algorithm it should be possible to implement the process. From an outline, it will not. Likewise the actual history of adaptations we have observational evidence of — as opposed to gross extrapolations and inferences controlled by questionable evolutionary materialist a priori assumptions — is of minor changes [e.g. finch beaks], not body plan level origins [e.g. origin of flying birds]. We also have strong reason to understand that functionally specific complex interactive organisation will come in isolated islands in large configuration spaces of possible clumped or scattered arrangements of parts, and that such will not be amenable to incremental blind needle in haystack search. I find it highly significant to see how consistently this issue is ducked or diverted from. KF

  42. 42
    cantor says:

    35 Zachriel May 3, 2015 at 7:07 am

    1. Evaluate the individual fitness of each member of population;

    2. Compare each indvidual to landscape to determine best fit individuals;

    3. Breed new individuals through mutation and crossover to give birth to offspring;

    4. Replace least-fit population with new individuals.

    What’s the difference, if any, between what you posted above and Dawkins’ thoroughly discredited “Weasel” program?

    ~

  43. 43
    Joe says:

    What is the evidence that Darwinian evolution can provide adaptation? And no “evolutionary search” does not equal Darwinian evolution. Intelligent Design Evolution is an evolutionary search and it isn’t Darwinian.

    You have no idea what is being debated and your ignorance is amusing.

  44. 44
    cantor says:

    41 Elizabeth Liddle May 3, 2015 at 7:48 am

    I would hope adaptation is non-controversial. But if it is, then Darwinian evolution is non-controversial, as that is what it is – adaptation.

    What you just posted is the ultimate bait-and-switch equivocation: the claim that blind-watchmaker evolution is simply finch beaks writ large. And that claim, Elizabeth, is controversial.

    ~

  45. 45

    What’s the difference, if any, between what you posted above and Dawkins’ thoroughly discredited “Weasel” program?

    Weasel is not “discredited” Cantor. It was just a very simplistic model in which phenotype=genotype=fitness function..

    In real evolution these things are all separate, and, specifically, the phenotype and genotype with optimal fitness is not known in advance (in Weasel, it was, obviously).

    So when I use the system Zachriel describes to find a solution to a problem, I do not know the solution in advance (or I wouldn’t bother to use the system). The system finds an optimal solution to my problem.

    All I have to do is to make sure that candidate solutions that partly solve it breed with higher problems than candidate solutions that solve it less well.

    Which is exactly how finch-beaks work, and antibiotic resistance, and is exactly what Darwin proposed.

  46. 46
    Zachriel says:

    kairosfocus: from an algorithm it should be possible to implement the process.

    Where do you feel the algorithm is wanting?

    kairosfocus: Likewise the actual history of adaptations we have observational evidence of is of minor changes, not body plan level origins.

    Yes, you keep repeating that. We might start with the historical record, which shows a pattern of incremental adaptation.

    kairosfocus: What’s the difference, if any, between what you posted above and Dawkins’ thoroughly discredited “Weasel” program?

    Weasel is a simplified implementation of an evolutionary algorithm. It doesn’t have crossover, for one. It has a very simple fitness landscape for another.

    There nothing discredited about Weasel. It’s just highly simplified.

  47. 47

    Cantor:

    What you just posted is the ultimate bait-and-switch equivocation: the claim that blind-watchmaker evolution is simply finch beaks writ large. And that claim, Elizabeth, is controversial.

    Well, it shouldn’t be. Finch beak evolution is exactly what Dawkins described as “blind-watchmaker evolution”.

    Why do you think it isn’t?

  48. 48
    cantor says:

    41 Elizabeth Liddle May 3, 2015 at 7:48 am

    clustering does provide “information” in the sense in which Ewert, Dembski and Marks use the term, whether that is the clustering of fitnesses around phenotypes, or the clustering of environmental properties with similar environmental properties. It appears to be a ubiquitous property of our universe, from sub-atomic to intergalactic scales (as I said elsewhere recently) and it is that property that makes smooth fitness landscapes possible.

    What you are describing above sounds like front-loaded design, not Darwinian evolution. Was that your intent?

    ~

  49. 49
    Joe says:

    Weasel was discredited as evidence for natural selection. It has nothing to do with biological evolution. All evolutionary and genetic algorithms are in that category. Not one supports natural selection. Not supports Darwinian nor neo-Darwinian evolution.

  50. 50
    Joe says:

    There isn’t any evidence tat finch beak evolution occurred via accumulations of genetic accidents, errors and mistakes. Lizzie is equivocating, again.

  51. 51
    Zachriel says:

    cantor: What you are describing above sounds like front-loaded design, not Darwinian evolution.

    That sunlight emanates from a single source that moves in a regular cycle while plants grow upward towards the sky is due to Apollo crossing the sky in a fiery chariot and the plants’ desire to touch the gods.

  52. 52

    What you are describing above sounds like front-loaded design, not Darwinian evolution. Was that your intent?

    If you want to describe a system in which similar genotypes produce similar phenotypes which have similar properties vis a vis the environment, as front-loading, then feel free. But that was Darwin’s starting point – what happens when you have such a system.

    So it’s Darwinian.

    Ewert, Dembski and Marks of course are now arguing, as you are, that those provisions are themselves Active Information, which you might regard as having been “front-loaded” into the system – but in that case, the ID argument is not against the effectiveness of Darwinian evolution (which is effective) but against the case that only an ID could have designed a system in which genotypes would arise that produce phenotypes that have similar real-world properties.

    Which is not anti-Darwinian at all, and certainly not anti-blind-watchmaker. Such a system would be blind – but would still produce well-adapted populations of diverse organism, which was Darwin’s point.

  53. 53

    There isn’t any evidence tat finch beak evolution occurred via accumulations of genetic accidents, errors and mistakes. Lizzie is equivocating, again.

    Yes, there is, Joe. Scads of it.

  54. 54
    Joe says:

    That is your opinion, Lizzie and only an opinion. Present some so we can take a look.

  55. 55
    Joe says:

    Elizabeth- You don’t even have a mechanism capable of getting beyond populations of prokaryotes and that is given starting populations of prokaryotes.

  56. 56
    Zachriel says:

    Elizabeth Liddle: Ewert, Dembski and Marks of course are now arguing, as you are, that those provisions are themselves Active Information, which you might regard as having been “front-loaded” into the system

    While it’s hard to pin ID down, front-loading usually means that functional genes already exist in the ancestor in a dormant form. So, for instance, front-loaders point out that many of the proteins used in nerves predate nerves. Of course that’s entirely consistent with {evolutionary co-option}, but they take that as evidence of front-loading, that the designer preconceived and planned for nerves and brains.

    This is different from the “designed to evolve” claim, which just adds an extraneous entity to the process.


    Edited for clarity.

  57. 57
    Joe says:

    While it’s hard to pin ID down, front-loading usually means that functional genes already exist in the ancestor in a dormant form.

    Only in a gene-centric front-loading scenario.

    Of course that’s entirely consistent with evolution

    Front-loading is evolution you cowardly equivocator. Grow up already.

    This is different from the “designed to evolve” claim, which just adds an extraneous entity to the process.

    That is your uneducated opinion, anyway. However it should be noted that your brand of evolution can’t even be modeled and has no entailments beyond change, stasis, disease and deformities.

  58. 58
    cantor says:

    48 Elizabeth Liddle May 3, 2015 at 8:03 am

    Finch beak evolution is exactly what Dawkins described as “blind-watchmaker evolution”.

    Stop playing word games Elizabeth. Dawkins didn’t become a multi-millionaire by writing 13 books about finch beaks.

    ~

  59. 59

    Cantor:

    Stop playing word games Elizabeth. Dawkins did become a multi-millionaire by writing 13 books about finch beaks.

    I am not “playing word games”, cantor. By “blind-watchmaker evolution”, as I understand him, Dawkins was referring to precisely the process that in the short term leads to generation-by-generation changes to the mean depth of finch-beaks in the Galapagos depending on whether recent weather cycles have favored the abundance of large or small seeds, and, in the long term, generation-by-generation change that leads to the adaptation of wings into flippers, or scales into feathers, or fins into wrists, etc.

    You seem to think that somehow two different processes are postulated – they aren’t. Alternatively, you think that somehow what works for finch-beaks over a short time scale, can’t work for much more extensive adaptions over a longer time scale. Dawkins doesn’t think so, and nor do I. So, far from “equivocating” I simply don’t accept that there are two different meanings to the term “adaptation”.

    Where are you drawing the line? Or, alternatively, how are you defining the meaning of the term “evolution” in the context of the discussion? Because in the context of the math of evolutionary searches, the process is identical to the process postulated to account for finch-beaks. If you think that something else is required for larger changes, than that isn’t what people refer to as “evolutionary search”.

  60. 60
    cantor says:

    53 Elizabeth Liddle May 3, 2015 at 8:11 am

    So it’s Darwinian.

    Darwin is spinning in his grave. It’s futile trying to have a logical discussion with someone who has their own private definitions.

    ~

  61. 61
    Zachriel says:

    cantor: Darwin is spinning in his grave.

    Darwinian has several related definitions, but usually refers to evolution by natural selection. How are you using the term?

  62. 62
    cantor says:

    60 Elizabeth Liddle May 3, 2015 at 8:50 am

    Finch beaks and antibiotic resistance are not controversial.

    Molecules-to-man via an undirected non-teleological non-frontloaded purely material process is controversial.

    To facilitate dialog and avoid equivocation I call the former “adaptation” and the latter Darwinian evolution (or blind watchmaker evolution).

    If you don’t like that choice of words, then go ahead and pick two different words that you can live with and use them consistently. If you think the mechanisms underlying them are identical, then use third word (different) word for that mechanism.

    ~

  63. 63
    Zachriel says:

    cantor: To facilitate dialog and avoid equivocation I call the former “adaptation” and the latter Darwinian evolution

    Evolution by natural selection is what is normally meant by darwinian evolution. Darwin’s theory is two-fold, adaptation plus branching descent.

  64. 64
    cantor says:

    52 Zachriel May 3, 2015 at 8:08 am

    plants grow upward towards the sky is due to Apollo crossing the sky in a fiery chariot and the plants’ desire to touch the gods.

    I can’t have an adult conversation with someone who gets their jollies posting something like that. So you just dropped off my radar, Mr Z. Goodbye.

    ~

  65. 65
    cantor says:

    46 Elizabeth Liddle May 3, 2015 at 8:00 am

    Weasel is not “discredited” Cantor. It was just a very simplistic model

    Good choice of words, simplistic. Synonyms: Facile, jejune, naive.

    http://i.imgur.com/jfiWTKK.png

    Weasel is teleological, which is exactly what Darwin did *not* intend. So it’s a discredited example of Darwinism.

  66. 66
    cantor says:

    46 Elizabeth Liddle May 3, 2015 at 8:00 am

    when I use the system Zachriel describes to find a solution to a problem, I do not know the solution in advance (or I wouldn’t bother to use the system). The system finds an optimal solution to my problem.

    I use SCO and DEPS “evolutionary” search algorithms all the time in my work. I put “evolutionary” in quotes because such algorithms have nothing to do with “evolution” (as meant by the leading proponents of “evolution”): I have to front-load the search with a distant goal. That’s not Darwinian.

    ~

  67. 67
    Zachriel says:

    cantor: I can’t have an adult conversation with someone who gets their jollies posting something like that.

    Our statement had a point. You conflated different meanings of the word information when you said “Only because the information is front-loaded into the landscape (or the search).”

    cantor: What you are describing above sounds like front-loaded design, not Darwinian evolution.

    To return to your statement, that the natural world is highly ordered makes it amenable to evolution. For instance, sunlight emanates from a single source, a typical ordering that allows plants to orient themselves towards the light, and compete for that light. You can say this was “front-loaded”, but astronomers would say it’s because of gravity and accretion and hydrogen fusion.

    cantor: Weasel is teleological, which is exactly what Darwin did *not* intend. So it’s a discredited example of Darwinism.

    Weasel was not proposed as a model of biological evolution, but as a simple example of an evolutionary algorithm with a specific target showing that it would work much faster than random search.

  68. 68
    Mung says:

    Zachriel, you’re predictable to a Ti.

    Now code that up into something that can be tested on a computer.

    Zachriel said:

    The question, then, isn’t mathematical, but empirical.

    Indeed. And then when asked to put something to an empirical test you can’t come through.

  69. 69
    Zachriel says:

    Mung: Now code that up into something that can be tested on a computer.

    Been there. Done that. One of many we created used the dictionary as a fitness landscape. An IDer had claimed it couldn’t be traversed by an evolutionary algorithm, that it would stall after no more than seven letter words.

    Zachriel: The question, then, isn’t mathematical, but empirical.

    Mung: Indeed.

    The claim was of a mathematical proof. By agreeing that it is empirical, you are rejecting that original argument.

  70. 70
    Mung says:

    Elzabeth Liddle:

    All I have to do is to make sure that candidate solutions that partly solve it breed with higher problems than candidate solutions that solve it less well.

    And that’s where you’re wrong, and that’s where Zachriel is wrong.

    If you and Zachriel insist on hand-waving, please wave your hands over your keyboard and come up with some actual code that can be tested.

    Elizabeth Liddle:

    So if you are discussing something other than Darwinian evolution, you might be in the wrong thread! That’s what “evolutionary search” is.

    I can quote you on this over at TSZ? Darwinian evolution is a search?

  71. 71
    Zachriel says:

    Mung: If you and Zachriel insist on hand-waving, please wave your hands over your keyboard and come up with some actual code that can be tested.

    The literature is replete with studies of evolutionary algorithms. All you need is the basic process to test it for yourself. Try Weasel for a very simple example, and work your way up from there. None of this is rocket engineering.

  72. 72
    Mung says:

    And, Zachriel, I’ll bet your genotype is designed for your landscape. Yes, folks, the candidate solutions must be designed. (Well, they don’t have to be designed. But good luck going down that road.)

  73. 73
    Zachriel says:

    Mung: I’ll bet your genotype is designed for your landscape.

    No. Genotypes evolve by random mutation of letters and random recombination of snippets. If and when they form a word, they are added to the population. Like this:

    o
    ox
    box
    fox
    for
    fore
    fort

    And so on.

  74. 74

    Mung:

    And, Zachriel, I’ll bet your genotype is designed for your landscape. Yes, folks, the candidate solutions must be designed. (Well, they don’t have to be designed. But good luck going down that road.)

    This just isn’t correct, Mung. You keep saying it, but it isn’t true. The only thing you need to “design in” to your starting genotype is the ability to replicate with variation, in other words, you need to do the “OoL” part.

    Then the population evolves a solution. If you knew the optimum solution before you started, you wouldn’t need to run the system.

    In AVIDA, for instance, none of the starting population need be able to perform any functions at all, and most mutations are not advantageous – many are disadvantageous. Yet functions evolve, and they manner in which the successful organisms perform the functions is different every time (or can be – not sure how many solutions there are to each function).

    You DON’T design in the solution. What you “design in” (apart from your starter population of minimally-reproducing virtual organism) is the problem. The problem you want solved is the fitness function. In order to survive and breed better than their fellows, individual virtual organisms have to solve your problem better than their fellows – you don’t tell them how to do it.

    Eventually, you get a population of virtual organisms that can solve the problem really efficiently – but by a method that you did not know beforehand, nor did you program in.

    In nature, of course, you don’t need a designer to “design in” the problem. It’s right there, in the environment: “use these resources and avoid these threats and you will survive to breed and breed again”.

  75. 75

    Mung:

    I can quote you on this over at TSZ? Darwinian evolution is a search?

    As long as you retain my scare quotes, sure. It can be thought of as a “search” as long as we are clear about the ways in which it is, and is not, like a literal search. Just as you can describe my post as “food for thought”, as long as you do not complain about being hungry afterwards.

  76. 76
    Joe says:

    AVIDA is not a genetic algorithm and it does not simulate Darwinian evolution.

    With evolutionism the “target” is already reached- that “target” being survival and reproduction. Whatever else happens is contingent serendipity.

  77. 77
    Joe says:

    The literature is replete with studies of evolutionary algorithms.

    Yes and they all support evolution by intelligent design. Is that your point?

  78. 78
    Mung says:

    Mung: I’ll bet your genotype is designed for your landscape.

    Zachriel: No. Genotypes evolve by random mutation of letters and random recombination of snippets.

    And how does that even begin to address my claim?

    Zachriel: If and when they form a word, they are added to the population.

    What do words (or letters for that matter) have to do with anything? Why are you weeding out potential solutions in advance?

    Oh, that’s right. You’re using a dictionary. Dictionaries contain words. Words are constructed from letters. Your genotype is designed to work with your landscape.

    http://www.merriam-webster.com.....dictionary

    Why would you even deny this?

    Now take your example genome and use a Chinese or Arabic dictionary in your application without modification to the design of your genome.

    Elizabeth Liddle:

    This just isn’t correct, Mung. You keep saying it, but it isn’t true.

    I keep saying it because it is true. What will happen to your NSCSI program if we change the design of the genotype you used? Will it still show that Natural Selection can create Complex Specified Information?

    That fact that you can sit here and claim that you did not design the genotype you used in that program is just mind boggling. Did someone else design it? Walk us through it.

    You know what I am saying is true and any standard text on GA’s will say the same thing I am saying.

  79. 79
    Mung says:

    Sometimes I think I must be speaking a different language from Zachriel and Elizabeth.

    When I say that the genotype in a GA is designed, what do the two of you think I mean by that?

    You’re trying to solve a problem. You have to figure out some way to represent your problem as a genotype. There has to be a mapping of the genotype to the solution domain. This all requires intelligent design.

    Do you really want to argue this?

    Is there some general purpose genotype that works for all domains that you’d care to share with me?

  80. 80
    Zachriel says:

    Mung: And how does that even begin to address my claim? {your genotype is designed for your landscape.}

    Because the genotype isn’t designed, but evolved.

    Mung: What do words (or letters for that matter) have to do with anything?

    It was the fitness landscape chosen by an IDer to supposedly represent unconnected islands of function.

    Mung: Your genotype is designed to work with your landscape.

    You seem to be using the term genotype in some unusual manner. The genotypes are the sequences of letters that enter the population because they meet the minimum requirement of being perfectly spelled words. If you mean the dictionary is designed, that’s the fitness landscape, not the genotype.

  81. 81

    Mung:

    Sometimes I think I must be speaking a different language from Zachriel and Elizabeth.

    Yes, that has struck me as well.

    When I say that the genotype in a GA is designed, what do the two of you think I mean by that?

    I guess you mean that the genotype in a GA is designed. Whereas it isn’t – the starter genotype is designed, for sure, so that it is capable of replication, but it typically has no more than chance success of solving your problem.

    The winning genotype – the one that solves your problem, is not designed.

    You’re trying to solve a problem. You have to figure out some way to represent your problem as a genotype.

    Nope. You have to find some way to represent your problem as a fitness function. The genotype sorts itself out.

    There has to be a mapping of the genotype to the solution domain. This all requires intelligent design.

    What has to happen is that the genotype has to entail a phenotype. This mapping is indeed designed – it’s there in the starter population as I said. But it needs only to be designed enough to replicate – the equivalent of OOL.

    From there on, the population evolves to solve your problem – which is represented in the fitness function, not the genotype.

    Do you really want to argue this?

    No. I just want you to try to understand the language we are using. The designer does not design a genotype that will produce a phenotype that will solve the problem. The system does that. What the designer does is design the original population of virtual organisms (OOL), in which genotypes are varied, and map on to phenotypes. When the genotype varies, so will the phenotype. The designer does NOT design the solution – the genotype that solves her problem. The system does that. And it does so because the designer presents her problem as a fitness function.

    Is there some general purpose genotype that works for all domains that you’d care to share with me?

    Of course not. An equation won’t work in a primordial soup.

    As we’ve said, repeatedly, Mung: to evolve a solution, you need a starter population of virtual organisms, and the problem represented as a fitness function.

    Clearly you can’t shove worms into your computer and expect them to evolve into solutions as to how to discriminate patients with schizophrenia from patients with bipolar on the basis of their brain scans.

    That’s why the system presupposes OOL – a system (“domain” if you like) in which a population of virtual organisms can reproduce with variance, and have their probability of breeding dependent on their capacity to exploit the resources and avoid the threats of the environment you provide.

    And these are the parts you design.

    You do NOT, repeat NOT, design the genotype that specifies a phenotype that solves your problem. You do not even KNOW the phenotype that solves your probem, so you CAN’T design its genotype.

    THAT is what I mean by “the genotype in a GA is NOT designed”.

    OK? Are we speaking the same language now?

  82. 82
    Mung says:

    Let’s say we want to kick off a run of an evolutionary algorithm and we want to initialize the starting population with random candidate solutions.

    initial_population = Population.new(GenotypeFactory)

    Now in this code snippet the GenotypeFactory is not used to create the random candidate solutions, but rather it constructs a template (a Genotype) from which the random candidate solutions can be generated. The random candidate solutions are then generated from that template, so that they all share the same genotype design.

    Now say our GenotypeFactory can produce many different Genotypes of different design. When constructing our starting population we would not want to just grab any old Genotype from the factory and hope it works for the problem we are trying to solve.

    Nor would we want our initial starting population of candidate solutions to be generated using different Genotypes. We want them all to share the same Genotype, and thus to all have in a common the same genotype design.

    We want a Genotype designed for our domain.

    There is nothing at all controversial about this.

  83. 83
    Mung says:

    Elizabeth, looks like we cross-posted 🙂

    My latest @83) is not a response to your 82, rather it is a continuation of my 79 and 80.

  84. 84
    Zachriel says:

    Mung: Now in this code snippet the GenotypeFactory is not used to create the random candidate solutions

    It can be, and often is. Random sequences are frequently used as the seed for an evolutionary algorithm. Or it might be just a very simple sequence. See #35 above.

  85. 85
    Mung says:

    Zachriel, anyone familiar with programming ought to understand that there is a difference between a type and an instance of a type. Even non-programmers can grasp that.

    And just because you can change the properties of an object it doesn’t follow that the object is not designed.

  86. 86
    Zachriel says:

    Mung: And just because you can change the properties of an object it doesn’t follow that the object is not designed.

    You seem to be confusing the model with the thing being modeled. If the sequence is random, the sequence is not designed.

  87. 87
    Mung says:

    Mung: Now in this code snippet the GenotypeFactory is not used to create the random candidate solutions

    Zachriel: It can be, and often is.

    LoL. Yet I just told you that it isn’t. It’s my code, I can design it the way I want it 🙂

    Zachriel:

    Random sequences are frequently used as the seed for an evolutionary algorithm. Or it might be just a very simple sequence. See #35 above.

    Irrelevant or redundant.

    The point of the GenotypeFactory is separation of concerns and to clarify and illustrate.

    The GenotypeFactory generates Genotypes which are themselves then used to create candidate solutions.

    So first and foremost, when I say your genotype is designed it is this Genotype that I am talking about, for it defines what your potential candidate solutions can look like.

    You have some questions to answer still.

  88. 88
    Joe says:

    Elizabeth:

    The winning genotype – the one that solves your problem, is not designed.

    Of course it is. Computers don’t do anything they are not intelligently designed to do. And if you write a program to solve a problem, and it does so, then it did so by intelligent design. That was the purpose of the program.

    The following comment explains why you are wrong: a lesson on computers.

  89. 89

    Joe: if you want to model biological evolution, then you need to use a computer, which is designed by human beings. If we used your logic, and computer model of any natural phenomenon whatsoever would be “designed” because the computer is designed.

    In a computer model of evolution, the computer itself, its operating system, and the code that runs the simulation are models of the physical environment – the physical and chemical laws of the natural world.

    The population of virtual organisms, is a model of a population that has not yet evolved to thrive within the new environment we are just about to present them. So they represent proto-biological population, if you like, sometimes after OOL.

    The fitness function represents the resources and hazards of the natural environment. We can design it so that they represent a specific problem we want to solve, or a natural environment we want to model.

    And what happens is that within that MODEL system, in which all the DESIGNED parts (computer, operating system, EV code, virtual critters, virtual environment) stand-in for the natural world. What is NOT designed are the evolved critters themselves. We do not know what genotypes/phenotypes will emerge from the process as the optimal solution to survival and reproduction within the environment we have provided from them.

    Therefore, such systems are a MODEL of how, GIVEN: a population of self-replicating organisms; a physics and chemistry; an environment of resources and threats – that initial of simple, minimally functional virtual organisms will evolve adaptively to that environment, EVEN THOUGH we do not specify what they will look like, what features they will have, what surprising properties they will evince.

    And they do.

    Therefore, in principle, we know that Darwin’s mechanism works.

    It is also very useful for solving problems that human designers aren’t very good at solving.

  90. 90
    Mung says:

    Elizabeth Liddle:

    I guess you mean that the genotype in a GA is designed. Whereas it isn’t – the starter genotype is designed, for sure, so that it is capable of replication, but it typically has no more than chance success of solving your problem.

    And you have just said both that the genotype is and is not designed. Unless you are equivocating, you’re contradicting yourself.

    So what is the starter genotype and what is the relevant difference between the starter genotype and the genotype in a GA you mention such that the starter genotype is designed and the genotype in a GA is not designed?

    Are you working with a language other than MATLAB now? I would not mind working some specific examples so that we both understand specifically what we are referring to.

  91. 91
    Zachriel says:

    Mung: Yet I just told you that it isn’t.

    Okay. You can set it up anyway you want, even if it doesn’t meet the outline provided above. You might as well dispense with the algorithm entirely if you already know what you want.

    Mung: So first and foremost, when I say your genotype is designed it is this Genotype that I am talking about, for it defines what your potential candidate solutions can look like.

    So now it’s our genotype. Wish you’d make up your mind. According to the definition provided above, the original population is made up of random sequences.

  92. 92
    Mung says:

    Mung: You’re trying to solve a problem. You have to figure out some way to represent your problem as a genotype.

    Elizabeth Liddle: Nope. You have to find some way to represent your problem as a fitness function. The genotype sorts itself out.

    Another annoying trait you and Zachriels share.

    It may in fact be the case that you have to represent your problem as a fitness function, but that is not a rebuttal to what I wrote. So your rebuttal consists of “Nope.”

    In computer programming, genetic representation is a way of representing solutions/individuals in evolutionary computation methods.

    http://en.wikipedia.org/wiki/G.....esentation

    If I say potential solution instead of problem will that change at all the fact that this is a part of the design?


    A typical genetic algorithm requires:

    a genetic representation of the solution domain,
    a fitness function to evaluate the solution domain.

    http://en.wikipedia.org/wiki/G.....esentation

  93. 93

    And you have just said both that the genotype is and is not designed. Unless you are equivocating, you’re contradicting yourself.

    I’ve made myself very clear, Mung: the winning genotype is not designed. And none of the modifications to the initial population are designed.

    In other words, the very thing that is at issue – the thing evolves – is NOT designed.

    I have always said, over and over, that the STARTING population of reproducing virtual organism IS designed – this is the equivalent of OOL. Darwinian evolution is predicated on the existence of a minimally functional starting population of self-replicators and we do not, as yet, know how those got going. Darwin’s theory starts AFTER that point. And AFTER that point, the genotype is NOT designed – OK? It EVOLVES. Nobody knows, in advance, what the winning or optimal design will be.

    So what is the starter genotype and what is the relevant difference between the starter genotype and the genotype in a GA you mention such that the starter genotype is designed and the genotype in a GA is not designed?

    The difference is that the starter genotype is very very simple. All it has to do is provide reproductive functionality. Often, the designer actually generates an initial set of random variants – all that is designed is their capacity to replicate with variance.

    They don’t doesn’t have to do anything else. Everything else – everything we actually want it to evolve to do, is NOT designed. It EVOLVES by replication with heritable variation in reproductive success in the current environment. The current environment IS designed – by the modeller, either to simulate a natural environment, or to express a problem that the modeler wants solved.

    And the genotype that produces a phenotype that solves the problem is NOT designed, even though its remote ancestor at the virtual OOL at the start of the run, was.

    Are you working with a language other than MATLAB now? I would not mind working some specific examples so that we both understand specifically what we are referring to.

    I’ve been using Eureqa, not writing them myself. I’ve been too busy.

  94. 94

    Mung:

    It may in fact be the case that you have to represent your problem as a fitness function, but that is not a rebuttal to what I wrote. So your rebuttal consists of “Nope.”

    Yes it does consist of “nope”. Nope as in: you do not have to represent your problem as a genotype. That doesn’t even make any sense. You might, conceivably, represent your solution as a genotype, but if you knew the solution beforehand you wouldn’t need the GA.

  95. 95
    Zachriel says:

    Mung: A typical genetic algorithm requires: …

    You do realize that genetic algorithms are a subset of evolutionary algorithms?

    Mung: You’re trying to solve a problem. You have to figure out some way to represent your problem as a genotype.

    The fitness landscape is the problem. The genotypes are the evolved solutions.

  96. 96
    Mung says:

    Zachriel: According to the definition provided above, the original population is made up of random sequences.

    So? Random sequences of what? Spaghetti? Jello?

    Why do your “genotypes” use letters rather than pictograms?

  97. 97
    Mung says:

    Zachriel: You do realize that genetic algorithms are a subset of evolutionary algorithms?

    Yes. So? Do you have a point?

  98. 98
    Joe says:

    Elizabeth:

    Joe: if you want to model biological evolution, then you need to use a computer, which is designed by human beings. If we used your logic, and computer model of any natural phenomenon whatsoever would be “designed” because the computer is designed.

    That doesn’t follow from what I posted. Also mere evolution is not being debated. I asked how to model UNGUIDED evolution. Natural selection is eliminative. The “goal” is to survive and reproduce. And that is a given.

    The only way we would say that evolution is unguided is if unguided processes produced life in the first place. If the OoL is intelligently designed then we would say that organisms were designed to evolve and evolved by design.

    EV doesn’t model unguided evolution.

  99. 99
    Zachriel says:

    Mung: Random sequences of what? Spaghetti? Jello?

    If you are solving a complex equation, for instance, they might be the terms of the equation.

    Mung: Why do your “genotypes” use letters rather than pictograms?

    The genotypes have to correlate to the landscape. If they don’t, then an evolutionary algorithm will not work. So, in biological evolution, if genotypes didn’t map through phenotypes to the environment somehow, then biological evolution wouldn’t occur. In addition, the fitness landscape must exhibit positive ordering. Turns out that nature clumps, light and gravity and water and all sorts of things!

  100. 100
    Mung says:

    Since we’re being all pedantic and stuff:

    Zachriel: The fitness landscape is the problem.

    No, it isn’t.

    Zachriel: The genotypes are the evolved solutions.

    You mean potential solution or candidate solution.

    And the behaviors and properties that all your “genotypes” share in common? For example, what is the longest word that you allow for, and why? Do you allow your potential words to consist of non-word characters, and if not why not?

  101. 101
    Mung says:

    Zachriel: If you are solving a complex equation, for instance, they might be the terms of the equation.

    Precisely. And that would be by design. Took forever, but we got there in the end. thank you.

  102. 102
    Mung says:

    Zachriel: The genotypes have to correlate to the landscape. If they don’t, then an evolutionary algorithm will not work.

    And now you’re just repeating what I have been saying all along. 🙂

  103. 103
    Joe says:

    Elizabeth, Intelligent Design and evolution are not mutually exclusive. Intelligently designing something to evolve is still Intelligent Design.

  104. 104
    Mung says:

    Elizabeth, if you are going to make a pretense of disputing my argument you need to address my argument. It’s like you’re not paying attention to what I actually say going off on some tangent about things I’m not even talking about.

    In other words, the very thing that is at issue – the thing evolves – is NOT designed.

    Probably another reason we’re having difficulty communicating. That’s not what’s at issue for me. My focus in this thread has been very narrow and specific.

    I have always said, over and over, that the STARTING population of reproducing virtual organism IS designed – this is the equivalent of OOL.

    I wasn’t even going that far. But ok.

    The difference is that the starter genotype is very very simple. All it has to do is provide reproductive functionality. Often, the designer actually generates an initial set of random variants – all that is designed is their capacity to replicate with variance.

    They don’t doesn’t have to do anything else.

    See Zachriel @ 100.

    You might, conceivably, represent your solution as a genotype, but if you knew the solution beforehand you wouldn’t need the GA.

    But you’re not representing the solution [that’s a straw man]. You are creating a context that generates potential or candidate solutions and then allows them to be tested.

    If your fitness function expects words and your candidate solutions are representations of pictures you can have replication with variance all day long for all the good it will do you.

  105. 105
    Joe says:

    Genotypes map to phenotypes? Except if you are a vole, apparently:

    Voles- A lot of micro but no macro

    The study focuses on 60 species within the vole genus Microtus, which has evolved in the last 500,000 to 2 million years. This means voles are evolving 60-100 times faster than the average vertebrate in terms of creating different species. Within the genus (the level of taxonomic classification above species), the number of chromosomes in voles ranges from 17-64. DeWoody said that this is an unusual finding, since species within a single genus often have the same chromosome number.  

    Among the vole’s other bizarre genetic traits:  

    •In one species, the X chromosome, one of the two sex-determining chromosomes (the other being the Y), contains about 20 percent of the entire genome. Sex chromosomes normally contain much less genetic information.
    •In another species, females possess large portions of the Y (male) chromosome.
    •In yet another species, males and females have different chromosome numbers, which is uncommon in animals. 

    A final “counterintuitive oddity” is that despite genetic variation, all voles look alike, said DeWoody’s former graduate student and study co-author Deb Triant. 

    “All voles look very similar, and many species are completely indistinguishable,” DeWoody said.  

    In one particular instance, DeWoody was unable to differentiate between two species even after close examination and analysis of their cranial structure; only genetic tests could reveal the difference.  

  106. 106

    Mung

    Precisely. And that would be by design. Took forever, but we got there in the end. thank you.

    This is just weird, Mung. But yes, you specify the physics and chemistry of your virtual world. So you don’t hope to evolve equations in alphabet world, any more than you would expect equations to evolve in DNA world.

  107. 107

    Mung:

    You are creating a context that generates potential or candidate solutions and then allows them to be tested.

    Yes, precisely. And they are generated by replication with random variation, just as Darwin proposed. And it works. The solutions are not designed. The virtual population is “directed” by nothing more than feedback from tne fitness function, just as in nature.

    Thus Darwin’s proposed mechanism is modeled, and shown to work: complex and unplanned solutions to the problem of surviving in an environment full of resources and threats emerge simply from the mechanism.

    The fact that we design the physics chemistry, the environment, and the initial population are irrelevant; Darwin assumed all these were in place, and proposed his mechanism simply to account for the adaptive evolution of the reproducing population.

    Glad that’s sorted.

  108. 108

    Elizabeth, Intelligent Design and evolution are not mutually exclusive. Intelligently designing something to evolve is still Intelligent Design.

    I entirely agree, Joe.

    The question is whether a system needs to be intelligently designed in order to result in evolution.

    But I’m glad we seem to agree that as long as we have the initial population of self-replicators (OOL), and an environment full of resources and hazard, adaptive evolution will tend to occur without any steering from a designer, cool.

  109. 109
    mike1962 says:

    EL: The fact that we design the physics chemistry, the environment, and the initial population are irrelevant;

    Of course they are relevant. The nature of the initial population (the systems, processes and control information they contain) determines to some extent what kinds of variations are even possible for any putative selection to act on. If I program a GA that searches for optimized antennas, it is not going to output any optimizations for airplane propellers. How would I even specify an “initial propeller” in a system designed to find optimized antennas?

    What the designers program as the initial conditions and the initial objects, and their potential properties, limits and constraints is determinative. Hardly “blind” evolution as Darwin envisioned.

  110. 110

    Joe:

    The only way we would say that evolution is unguided is if unguided processes produced life in the first place. If the OoL is intelligently designed then we would say that organisms were designed to evolve and evolved by design.

    Well, that’s not usually what people mean by “guided” in this context, Joe, which is why they sharpen their knives for Darwin, not the OoL people.

    Darwin did not have anything to say about OoL. In fact, in Origin, he famously, he assumes that life was originally “breathed” into the starting population of living things:

    There is grandeur in this view of life, with its several powers, having been originally breathed into a few forms or into one; and that, whilst this planet has gone cycling on according to the fixed law of gravity, from so simple a beginning endless forms most beautiful and most wonderful have been, and are being, evolved.

    So if all you are saying is that OoL had to be Intelligently Designed, your own complaint, anyway, is not against Darwin, nor against the “Blind Watchmaker” concept. These computer models demonstrate clearly that, given self-replicators that replicate with heritable variance in reproductive success in the current environment, complex solutions to the problems of survival in that environment will evolve, unguided by anything except the environment itself.

    Now, it may take a Designer to design the set-up – but the evolution of the solutions is indeed unguided.

  111. 111
    Joe says:

    Elizabeth:

    The question is whether a system needs to be intelligently designed in order to result in evolution.

    The evidence says that it does as basic biological reproduction is irreducibly complex.

    But I’m glad we seem to agree that as long as we have the initial population of self-replicators (OOL), and an environment full of resources and hazard, adaptive evolution will tend to occur without any steering from a designer, cool.

    Unguided evolution tends towards the more simple so it would never get beyond simple molecular replicators.

  112. 112
    mike1962 says:

    Elizabeth Liddle: unguided by anything except the environment itself.

    Incorrect. They are also guided by the processes, systems, information, limits and constraints inherent in the replicators themselves.

  113. 113
    Joe says:

    Elizabeth:

    Well, that’s not usually what people mean by “guided” in this context, Joe, which is why they sharpen their knives for Darwin, not the OoL people.

    That is what Creationists and IDists have been saying- ie that is what is meant by “guided”.

    Darwin did not have anything to say about OoL

    And that is why his idea was doomed to fail as the OoL holds the key.

    So if all you are saying is that OoL had to be Intelligently Designed, your own complaint, anyway, is not against Darwin, nor against the “Blind Watchmaker” concept.

    Of course it is. If the OoL = ID then both of those explanations are non-starters.

    These computer models demonstrate clearly that, given self-replicators that replicate with heritable variance in reproductive success in the current environment, complex solutions to the problems of survival in that environment will evolve, unguided by anything except the environment itself.

    What computer models? EV is about binding sites. AVIDA does nothing but go towards the more simple when realistic parameters are entered.

  114. 114

    Mike:

    Of course they are relevant. The nature of the initial population (the systems, processes and control information they contain) determines to some extent what kinds of variations are even possible for any putative selection to act on. If I program a GA that searches for optimized antennas, it is not going to output any optimizations for airplane propellers. What the designers program as the initial conditions and the initial objects, and their potential properties and limits matters immensely. It’s hardly “blind” a-telic evolution, as Darwin conjectured.

    Blind is precisely what it is – it cannot “see” beyond the current generation.

    But sure, in a limited model, only a very limited range of kinds of evolved solution will be possible. Nonetheless, they evolve “blind” – nobody knows in advance what the solution will be, nobody steps in and picks the solutions they think are on the right track – they just sit back and watch the best solution evolve, blindly.

    It’s important (and tried to make this point a few times, and I’ll try again) to distinguish between the problem that the EA designer wants solving (a better antenna) and the problem the population is solving (a higher reproductive rate).

    We can (but need not) set up the environment so that BY solving the problem of how to survive within it, the virtual population also solves our problem.

    But you can also (and I’ve done this) set up a random environment – indeed one that changes over time. And what we observe is that the evolving population adapts to the environment you provide, even if it was generated from a random-environment-generator, and if you change it, the population then adapts to the new environment.

    In other words the only guidance is coming from the environment.

    We do not need to invoke a designer, in the natural world, to account for an environment of resources and hazards. It’s just there, and its constantly changing, so there’s every reason to expect that populations of organisms will evolve, blindly, to optimise their ability to thrive within the current environmental conditions. This is “a-telic”. The population is only “guided” in the sense that gravity “guides” mountain streams down hillsides via stream-beds that take them most efficiently to the sea. The raindrops do not need to know where the sea is – they just push blindly against the obstruction-of-least-resistance.

  115. 115
    Joe says:

    AGAIN- with unguided evolution the only goal is survival and reproduction. And guess what? It starts with that so the goal is already met. Everything else is just contingent serendipity doo-dah.

  116. 116
    Joe says:

    Elizabeth, Just because no one knows what the solution will be doesn’t mean the process is unguided. Obviously the solution is guided by the program. With unguided evolution the variation has to already exist when the environment changes. The less fit get eliminated.

    Computers programs use actual selection whereas natural selection is mere elimination. The two processes are very different as I, via Mayr, explained earlier.

  117. 117
    mike1962 says:

    Elizabeth Liddle: Blind is precisely what it is – it cannot “see” beyond the current generation.

    One could say that of a robot that’s programmed to make it’s way around a room. Would you call a robot that’s programmed to handle environmental challenges “blind” because it cannot see beyond it’s next move? Even “blind” humans are not blind in the Darwinian sense you seem to be asserting. There are sophisticated processes and informational systems in play that shape and constrain the outcome in addition to the environment.

    The nature of the object is part of the determining factor of any result. Not merely the environment.

  118. 118

    Joe:

    Elizabeth, Just because no one knows what the solution will be doesn’t mean the process is unguided. Obviously the solution is guided by the program.

    It’s guided by the element in the program that represents the environment. In the real world, that guidance is provided by the real environment. Not by a designer. And it is only “guidance” in the “hunt the thimble” sense. The environment, whether in nature or in the computer, simply says “warmer” or “cooler” – in other words only provides feedback as to how good the current state is, not whether the population is travelling towards some goal.

    With unguided evolution the variation has to already exist when the environment changes. The less fit get eliminated.

    Well, “fitness” in this context simply means “how well you breed”. The “less fit” are simply those who breed less than their peers or parents, whereas the “more fit” are those who breed more than their peers or parents. A “less fit” parent can still breed, and sometimes, its descendents prove to be extra fit.

    It is not the case that “natural selection” is simply an elimination process.

    Computers programs use actual selection whereas natural selection is mere elimination. The two processes are very different as I, via Mayr, explained earlier.

    No, both processes are the same. In both, those who breed most, leave most offspring (by definition) and those who breed less, leave fewer offspring (by definition). Therefore, each generation will be enriched with the genes of the parents who left the most children.

    In a computer, you can do it in various ways: eliminate those who score lowest, and breed to replace them from those who score better. But it’s usually best to make it probabilistic – if you are too rigid with your elimination, you will often lose slightly weird variants with the capacity to leave offspring with something novel and useful.

    In Nature, the probabilistic part is built in – sometimes a super-stud will get unlucky and die in a landslide; at other times a runt will just happen to mate with a superstud. But the end result is the same – those that breed most will leave more of their genes in future generations than those that breed least. Therefore those genes for those features that increase their bearer’s chance of breeding in the current environmetn will become more prevalent.

    There really is no difference between the computer version and the Nature version, except that in the computer version we have to design the chemistry, the environment, and the starting population, whereas in Nature we have them already there before the process starts.

    And once the process starts, no further designer-input is necessary.

  119. 119
    Mung says:

    Zachriel and I actually agree on something:

    The genotypes have to correlate to the landscape. If they don’t, then an evolutionary algorithm will not work.

    It’s a new day here at UD.

  120. 120

    Mike:

    Incorrect. They are also guided by the processes, systems, information, limits and constraints inherent in the replicators themselves.

    OK, accepted. But all those things have natural counterparts that do not require an intentional designer.

    Sure, you can set up a system in which the results are highly constrained. But many systems exist in which the results are highly constrained, but we do not say: aha! It must have been designed. “Guided” as in “constrained by high granite cliffs” is not the same meaing of “guided” as “led by someone who knows the way and will take you to where she wants you to go”.

    If all people mean by “guided evolution” is “evolution constained by the laws of physics and chemistry” then, sure, all evolution is “guided”. But that tells us nothing about whether a designer is involved, and I can certainly tell you that in computer evolution, once the thing is set up, you sit back and wait for the result. No Designer Intervention required.

    So if all IDers are saying is that a Designer must have been required to set up the evolutionary system that produced us, then, fine.

    But in that case, stop beating up on poor old Darwin!

  121. 121
    Mung says:

    The fact that the physics, the chemistry, the environment, and the initial population are all designed is irrelevant.

    They aren’t just “all designed.” They are designed for a purpose and they are designed to serve the same purpose. The different aspects are all designed to work together.

    See @ 120.

    And that’s all irrelevant? You’re a hoot.

  122. 122
    Joe says:

    Elizabeth:

    It’s guided by the element in the program that represents the environment.

    It is actively guided towards a solution. It does exactly what I said earlier, in the other thread.

    In the real world, that guidance is provided by the real environment.

    In the real world a diverse group of organisms share the same environment. In the real world it is a process of elimination in which whatever is good enough survives and reproduces.

    Well, “fitness” in this context simply means “how well you breed”. The “less fit” are simply those who breed less than their peers or parents, whereas the “more fit” are those who breed more than their peers or parents. A “less fit” parent can still breed, and sometimes, its descendents prove to be extra fit.

    Exactly. That you cannot see how impotent such a process is is beyond me.

    It is not the case that “natural selection” is simply an elimination process.

    Sure it is. It’s just that what gets eliminated and how quickly, can change.


    Computers programs use actual selection whereas natural selection is mere elimination. The two processes are very different as I, via Mayr, explained earlier.

    No, both processes are the same.

    You are absolutely loonie. Mayr went over the differences in “What Evolution Is”.

    Only selection can produce dog breeds. Without selection there would never have been any.

    In selection only a SELECT few get to survive whereas with elimination only a select few are eliminated. Which selection there is an actual objective to the survival and reproduction of the selected. With elimination the objective is met with survival and reproduction. Whatever is good enough survives, and that changes and can be many different existing variants.

    And once the process starts, no further designer-input is necessary.

    Exactly. That is the whole purpose. However we have to have a mechanism capable of explaining the diversity of life. Changes to genetics doesn’t seem to be capable.

    Everything we observe says that genetic change offers limited physiological change. There aren’t any cases of small changes that we can extrapolate into large changes. Anti-biotic resistance and beak changes are about existing variation. Peppered moths- existing variation.

  123. 123
    Joe says:

    Elizabeth:

    But that tells us nothing about whether a designer is involved, and I can certainly tell you that in computer evolution, once the thing is set up, you sit back and wait for the result. No Designer Intervention required.

    Umm a designer is involved if one set the whole thing up and provided life with the programming required to help us adapt.

    We don’t know how intense that initial set up had to be. It could very well be that a special creation-type start is required. Darwinian-type evolution would work well with that and it would explain the extinction rate.

  124. 124
    Mung says:

    You see. mike1962 get’s it.

    In Elizabeth’s [in]famous NS can generate CSI program she chose a genome size of 500 “bases” and populated each “base” with a zero or a one for each member of the initial population, with the size of the initial population set at 100 individuals.

    Why 100 individuals? Why 500 “bases”? Why zero and one? Why not different numbers? Why not letters? Why not url’s to interesting travel sites on the internet?

    Why did she select the particular design she used if in fact any population of self-replicators would have worked?

    The answer is glaringly obvious.

    And it seems Elizabeth has selective amnesia. Her program didn’t work, at first. She had to go back and make some design tweaks. Right Elizabeth?

    It’s ok. You’re human. Fallible. The end product of an infallible process.

  125. 125
    Joe says:

    Selection vs elimination:

    From “What Evolution Is”, Ernst Mayr, page 117:

    What Darwin called natural selection is actually a process of elimination.

    Page 118:

    Do selection and elimination differ in their evolutionary consequences? This question never seems to have been raised in the evolutionary literature. A process of selection would have a concrete objective, the determination of the “best” or “fittest” phenotype. Only a relatively few individuals in a given generation would qualify and survive the selection procedure. That small sample would be only to be able to preserve only a small amount of the whole variance of the parent population. Such survival selection would be highly restrained.

    By contrast, mere elimination of the less fit might permit the survival of a rather large number of individuals because they have no obvious deficiencies in fitness. Such a large sample would provide, for instance, the needed material for the exercise of sexual selection. This also explains why survival is so uneven from season to season. The percentage of the less fit would depend on the severity of each year’s environmental conditions. (bold added)

  126. 126
    Mung says:

    Blind is precisely what it is – it cannot “see” beyond the current generation.

    I can’t see beyond my immediate surroundings, and I can’t see through walls, but I’m not blind. Some people can’t see what is right in front of their face, but they are blind.

    Perhaps it’s time the evolutionists came up with a better metaphor.

  127. 127
    Mung says:

    Elizabeth Liddle:

    Yes, precisely. And they [candidate solutions] are generated by replication with random variation

    No one ever said otherwise.

    Unless I’ve misunderstood you [a not insignificant possibility], the replication mechanism is designed, and you admit this.

    What about the “random variation” mechanism? Also designed?

    In your NS can generate CSI program did you design the mutation (random variation) mechanism? [Say yes. Save us all the trouble. Thank you.]

  128. 128

    No doubt, Mung, no doubt. Metaphors can be very misleading. Best thing is simply to describe it directly. My best description of evolution theory is: in a population of self-replicators that reproduce with heritable reproductive success, those features that best promote reproductive success in the current environment will tend to become more prevalent.

    That’s all it is. And it works, as, logically, it must do.

  129. 129
    Mung says:

    Elizabeth Liddle:

    So if all IDers are saying is that a Designer must have been required to set up the evolutionary system that produced us, then, fine.

    But in that case, stop beating up on poor old Darwin!

    But Darwin was misguided and his conclusions were false. His intent was to absolve the designer of responsibility. If the designer designed the Darwinian process then Darwin failed. There is no design without a designer.

    Losers get beat, by definition.

  130. 130
    Mung says:

    Hi Elizabeth,

    ok, trying to get back on topic. 🙂

    Zachriel:

    However, biological evolution is a specific ‘search algorithm’, not the universal set of search algorithms; and the natural environment is a specific ‘fitness landscape’, not the universal set of fitness landscapes.

    Biological evolution is not a specific search algorithm and the natural environment is not a specific fitness landscape. Granted, that’s Zachriel spouting their usual nonsense.

    Zachriel in a more lucid moment:

    However, specific search algorithms may do better on specific fitness landscapes.

    Who ever thought otherwise?

    NetResearchGuy:

    Evolutionary algorithms in general have additional challenges. There is a lot of fine tuning required to make one work.

    Mung:

    Indeed. They must be carefully designed.

    That sets the general context for the discussion which followed.

  131. 131
    Mung says:

    Elizabeth Liddle @ 129:

    That’s all it is. And it works, as, logically, it must do.

    Indeed. It’s a tautology. Welcome to the dark side. 😀

  132. 132
    Zachriel says:

    Mung: You mean potential solution or candidate solution.

    Typically, they’re approximate solutions.

    Mung: For example, what is the longest word that you allow for, and why?

    No specified limit.

    Mung: Do you allow your potential words to consist of non-word characters

    If a mutation results in a sequence not found in the dictionary, it is still-born, that is, doesn’t enter the population.

    Mung: and if not why not?

    Because that was the fitness landscape specified by the IDer.

    Mung: And that would be by design.

    The use of the genome to represent terms of an equation is the defined relationship between the genome and the fitness landscape, i.e. the ‘chemistry’.

    Mung: And now you’re just repeating what I have been saying all along.

    And what Darwin pointed out in 1859. Natural selection works on heritable traits (genotypes) that provide a difference in reproductive fitness (landscape).

    mike1962: The nature of the initial population (the systems, processes and control information they contain) determines to some extent what kinds of variations are even possible for any putative selection to act on

    That’s not generally a function of the initial population, but of the fitness landscape.

  133. 133
    mike1962 says:

    Elizabeth Liddle: OK, accepted.

    Here comes the “ubiquitious ‘but'”…

    But all those things have natural counterparts that do not require an intentional designer.

    Whoa Nellie. Says who? Are you telling me you’ve got your head around the nature of the replicators of earth so well that you know that they didn’t require a designer?

    By all means, do tell.

    Sure, you can set up a system in which the results are highly constrained. But many systems exist in which the results are highly constrained, but we do not say: aha! It must have been designed.

    Such as?

    “Guided” as in “constrained by high granite cliffs” is not the same meaing of “guided” as “led by someone who knows the way and will take you to where she wants you to go”.

    Granite cliffs are not replicators. I thought we were talking about how replicators, and how only the environment matters with regards to their evolution. You claim only the environment is determinative of the outcome. I claim it’s both the properties of the replicators and the environment.

    This should be trivially obvious.

    If all people mean by “guided evolution” is “evolution constained by the laws of physics and chemistry” then, sure, all evolution is “guided”.

    No, that’s not all we mean. We mean the entire bio process which includes the properties of the replicators. The properties of the replicators are determinative in any outcome. Not just the environment.

    But that tells us nothing about whether a designer is involved, and I can certainly tell you that in computer evolution, once the thing is set up, you sit back and wait for the result. No Designer Intervention required.

    Who said anything about intervention after the properties of the replicators are determined? That’s another topic.

    So if all IDers are saying is that a Designer must have been required to set up the evolutionary system that produced us, then, fine.

    Fine? Well, that’s a hell of a concession. But by “system” we mean the replicators themselves. Nor just the environment. At least, I do.

    But in that case, stop beating up on poor old Darwin!

    The problem with Darwin (and his faithful followers) is that you assume that the replicators have no engineered constraints that led to certain outcomes as they related to the environment.

    That’s not demonstrable. Imaginations of the faithful notwithstanding.

  134. 134
    Zachriel says:

    Mung: Zachriel in a more lucid moment: However, specific search algorithms may do better on specific fitness landscapes. Who ever thought otherwise?

    From the original post: “Conservation of information dictates any search technique will work, on average, as well as blind search. Success requires an assisted search.”

    In fact, evolution only requires a fitness landscape that is positively ordered and not chaotic.

  135. 135
    Zachriel says:

    mike1962: I claim it’s both the properties of the replicators and the environment.

    The notion of a fitness landscape entails that there is a defined relationship between the replicators and the landscape, the ‘chemistry’ of the artificial world.

  136. 136
    mike1962 says:

    Zachriel: That’s not generally a function of the initial population, but of the fitness landscape.

    Give me an example with regard to software implemented replicator objects interacting with their environment.

  137. 137
    mike1962 says:

    Zachriel: The notion of a fitness landscape entails that there is a defined relationship between the replicators and the landscape, the ‘chemistry’ of the artificial world.

    Then you agree with me, and not Elizabeth. In order for a replicator to have a relationship with it’s environment it has to have certain properties that will necessarily be determinative of any future outcome.

    This is trivially obvious.

    In fact, evolution only requires a fitness landscape that is positively ordered and not chaotic.

    No, it also requires that the replicators have particular properties that allow it to successfully replicate in the environment.

  138. 138
    Joe says:

    Mung and mike1962- over on TSZ Elizabeth posted the following (parapsychology thread- Randi challenge):

    Well, I don’t see any reason why psi effects can’t be investigated by normal scientific methodology.

    Well, we don’t see any reason why macroevolution can’t be investigated by normal scientific methodology. 😎

    Intelligent Design can be investigated by normal scientific methodology.

  139. 139
    Mung says:

    Mung: You mean potential solution or candidate solution.

    Zachriel: Typically, they’re approximate solutions.

    We’re being pedantic, remember?

    Approximate:

    1: located close together
    2: nearly correct or exact

    So no, you lose again.

    Let’s recall that the initial population is randomly generated. That’s because we don’t want them located close together. And that is because we have no idea whether they will be a nearly correct or exact solution.

    All part of the design.

  140. 140
    Mung says:

    Zachriel: The notion of a fitness landscape entails that there is a defined relationship between the replicators and the landscape…

    Isn’t that what I have been saying all along?

    Isn’t that what Elizabeth denies?

    How is this relationship defined in an EA?

    I’m guessing it is designed.

  141. 141
    NetResearchGuy says:

    EL/Z: I think Mung’s point is that known functional evolutionary algorithms start with a fixed set of alleles that are designed. For example Weasel starts with letters, the antenna evolving algorithm starts with a working antenna made out of metal and a list of allowable mutations, the nozzle evolving algorithm starts with a working nozzle, etc. In other words, the initial genotype is designed to start on an island of function, and the allowable variations to the genotype are designed to remain on that island of function.

    For the nozzle evolving example, let’s say you started with a spherical piece of material, topologically lacking a hole for water to flow through. How would an evolutionary algorithm modify that into a starting point of a functional nozzle? The starting point is irreducibly complex, either it has a hole of the right shape to interface with a hose it needs to connect to or it doesn’t. How would an evolutionary algorithm create that irreducibly complex structure? Given infinite time it could stumble onto that initial island of function, but not finite time.

    Let’s say you wanted the nozzle algorithm to generate a sprayer, with multiple holes instead of one. It couldn’t do that because the initial allowed range of variation in the genotype of the nozzle algorithm doesn’t allow it to generate topological shapes with multiple holes.

    To consider another dimension of the problem, let’s say you gave your nozzle evolving program just the laws of physics (i.e. the physics of individual water molecules), and ran it using that. It would be too slow! Even the world’s most powerful supercomputer would take too long to evaluate the fitness function at a molecular level for a single nozzle shape in a practical amount of time.

    These types of issues are the point of ID. Evolution doesn’t always work in every case — it’s quite easy to construct examples where evolution can’t work, at least with finite resources. I.e. irreducible complexity, uncrossable maladaptive holes in the fitness landscape, insufficient time resources, etc.

    I’ve never seen EL or Z even tangentially address these issues. These are unimportant and nonexistent in their minds. As long as there is a non zero probability evolution could work, it doesn’t matter how many zeroes there are in the probability exponent.

  142. 142
    Mung says:

    NetResearchGuy:

    EL/Z: I think Mung’s point is that known functional evolutionary algorithms start with a fixed set of alleles that are designed.

    E/Z would probably disagree. They would argue that the content of any given allele is generated randomly and thus is not designed. But that would miss the point.

    That a ‘0’ in the genome represents a ‘T’ in the phenome is a design decision. That a ‘1’ in the genome represents an ‘H’ in the phenome is a design decision. That ‘T’ and ‘H’ have relevance to a potential solution is a design decision.

    We can of course create in software a Genotype that has no correlation with the fitness function.

    The Darwinists need to answer the question, why is it that your genotypes are correlated to your fitness functions.

  143. 143
    sparc says:

    How often have we seen this very thread before? I am not interested in fishing but even I realize that I’ve seen the Abu 6500 C3 reel before (according to Google it appears 42 times on this site). Just opening another thread will not bring the stillborn FSCO/I to life. Didn’t you read what WE had to say about it? And what about Dembski, Meyer, Behe, Marks et al.? Do you think they even consider FSCO/I? FSCO/I just dead and never lived.

  144. 144
    Upright BiPed says:

    Mung: That a ’0′ in the genome represents a ‘T’ in the phenome is a design decision. That a ’1′ in the genome represents an ‘H’ in the phenome is a design decision.

    Net Research: I’ve never seen EL or Z even tangentially address these issues. These are unimportant and nonexistent in their minds.

    Actually, they’ve both addressed them clearly. Dr Liddle has accepted that she cannot re-create the process from virtual “critters” and Brownian motion; and Zachriel hides behind an alternate process that does not accomplish what must be accomplished, and demands that everyone ignore that fact.

  145. 145
    Mung says:

    Upright BiPed:

    Actually, they’ve both addressed them clearly. Dr Liddle has accepted that she cannot re-create the process from virtual “critters” and Brownian motion;

    Elizabeth now admits that the initial population of virtual critters is in fact designed. Progress?

  146. 146
    kairosfocus says:

    Z, 74 (attn EL & Mung):

    I quote:

    Genotypes evolve by random mutation of letters and random recombination of snippets. If and when they form a word, they are added to the population. Like this:

    o
    ox
    box
    fox
    for
    fore
    fort

    And so on.

    This inadvertently illustrates the huge list of begged questions behind the darwinist tree of life, incrementalist icon and its failure to cogently address the message that the FSCO/I required for OOL and OOBP must be adequately accounted for. Which of course, is the precise context of active, bridging information that crosses the gaps to and between islands of function, as is discussed and illustrated in the OP.

    The same OP, that has been assiduously snipped, wrenched into strawman caricatures such as the just cited, and sniped at without truly cogently facing the pivotal questions.

    What has been done here is to take a simplistic — thus strawmannish — short skip distance case, 7 bits apart, and then grossly extrapolate to cases of such higher degrees of complexity and skip distance. The result of such gross oversimplification, is that the result is a fallacy. Strawman caricature, per Nizkor:

    The Straw Man fallacy is committed when a person simply ignores a person’s actual position and substitutes a distorted, exaggerated or misrepresented version of that position. This sort of “reasoning” has the following pattern:

    Person A has position X.
    Person B presents position Y (which is a distorted version of X).
    Person B attacks position Y.
    Therefore X is false/incorrect/flawed.

    This sort of “reasoning” is fallacious because attacking a distorted version of a position simply does not constitute an attack on the position itself . . .

    As a first answer, we can challenge you, Z, to provide an onward incrementalist functional at each step chance variation and selection based progress to something like:

    The quick brown fox jumps over the lazy dog,

    . . . with the underscored requisite that at each step the result is a functional sentence, created in reasonable time by chance variation on the ASCII code by known chance variation.

    Say, known to be chance variation, as tracing to whitened Zener noise or sky noise. (Such are used in modern random number generators.)

    Of course, we already know the answer to a realistic monkeys at keyboards exercise. As Wikipedia notes on random sentence generation by such means:

    The infinite monkey theorem states that a monkey hitting keys at random on a typewriter keyboard for an infinite amount of time will almost surely type a given text, such as the complete works of William Shakespeare.

    In this context, “almost surely” is a mathematical term with a precise meaning, and the “monkey” is not an actual monkey, but a metaphor for an abstract device that produces an endless random sequence of letters and symbols. One of the earliest instances of the use of the “monkey metaphor” is that of French mathematician Émile Borel in 1913,[1] but the earliest instance may be even earlier. The relevance of the theorem is questionable—the probability of a universe full of monkeys typing a complete work such as Shakespeare’s Hamlet is so tiny that the chance of it occurring during a period of time hundreds of thousands of orders of magnitude longer than the age of the universe is extremely low (but technically not zero) . . . .

    The theorem concerns a thought experiment which cannot be fully carried out in practice, since it is predicted to require prohibitive amounts of time and resources. Nonetheless, it has inspired efforts in finite random text generation.

    One computer program run by Dan Oliver of Scottsdale, Arizona, according to an article in The New Yorker, came up with a result on August 4, 2004: After the group had worked for 42,162,500,000 billion billion monkey-years, one of the “monkeys” typed, “VALENTINE. Cease toIdor:eFLP0FRjWK78aXzVOwm)-‘;8.t” The first 19 letters of this sequence can be found in “The Two Gentlemen of Verona”. Other teams have reproduced 18 characters from “Timon of Athens”, 17 from “Troilus and Cressida”, and 16 from “Richard II”.[24]

    A website entitled The Monkey Shakespeare Simulator, launched on July 1, 2003, contained a Java applet that simulates a large population of monkeys typing randomly, with the stated intention of seeing how long it takes the virtual monkeys to produce a complete Shakespearean play from beginning to end. For example, it produced this partial line from Henry IV, Part 2, reporting that it took “2,737,850 million billion billion billion monkey-years” to reach 24 matching characters:

    RUMOUR. Open your ears; 9r”5j5&?OWTY Z0d…

    That infinite monkeys theorem case [equivalent to the issue of needle in haystack search that you are hoping to dismiss without actually cogently addressing . . . ] has been pointed out here at UD many, many times, for instance at #4 in the now longstanding ID Foundations series, where the link between CSI and functional organisation was discussed. (FSCO/I was discussed at #5 with roots tracing to Orgel et al, and Borel’s underlying analysis appears at #11, together with a heuristic for the Chi_500 expression seen in the OP.)

    As Wiki was clipped in #11:

    In one of the forms in which probabilists now know this theorem, with its “dactylographic” [i.e., typewriting] monkeys (French: singes dactylographes; the French word singe covers both the monkeys and the apes), appeared in Émile Borel‘s 1913 article “Mécanique Statistique et Irréversibilité” (Statistical mechanics and irreversibility),[3] and in his book “Le Hasard” in 1914. His “monkeys” are not actual monkeys; rather, they are a metaphor for an imaginary way to produce a large, random sequence of letters. Borel said that if a million monkeys typed ten hours a day, it was extremely unlikely that their output would exactly equal all the books of the richest libraries of the world; and yet, in comparison, it was even more unlikely that the laws of statistical mechanics would ever be violated, even briefly.

    The physicist Arthur Eddington drew on Borel’s image further in The Nature of the Physical World (1928), writing:

    If I let my fingers wander idly over the keys of a typewriter it might happen that my screed made an intelligible sentence. If an army of monkeys were strumming on typewriters they might write all the books in the British Museum. The chance of their doing so is decidedly more favourable than the chance of the molecules returning to one half of the vessel.[4]

    These images invite the reader to consider the incredible improbability of a large but finite number of monkeys working for a large but finite amount of time producing a significant work, and compare this with the even greater improbability of certain physical events. Any physical process that is even less likely than such monkeys’ success is effectively impossible, and it may safely be said that such a process will never happen.

    In short, this is a discussion of fluctuations from the thermodynamic equilibrium that is to be expected with some fluctuation but with such a dominance of the bulk cluster that deeply isolated special zones are empirically unobservable even on an experiment on the scale of the observed cosmos of 10^80 or thereabouts atoms and 10^17 s duration.

    To see the point, consider each of those atoms to be given a tray of 1,000 coins. Or equivalently, a paramagnetic substance with 1,000 atoms in a weak B-field with possible orientations, N up or N down. Flip and test every 10^-14 s. As the OP discusses — but which you, Z, plainly insist on ignoring — the experiment maxes out at 10^111 possibilities, about 1 in 10^190 of the full config space W of 1.07*10^301. Where, in that wider space, we may find every possible ASCII code string of 143 characters.

    But, English sense-making strings or computer programs etc will be so deeply isolated and will be so small a fraction of the possibilities that the bulk, predominant group of near 50-50 outcomes in no particular order, within several standard deviations of the central peak, will utterly dominate.

    Such, is well known, and/or readily accessible.

    Indeed, its logic is foundational to the molecular statistics analysis that undergirds the second law of thermodynamics, 2LOT. For over 100 years now.

    So it is no wonder that we see something like picking up a scale of complexity that comes from a space of about 10^50, a factor of 10^150 short of the 500 bit/72 ASCII character Sol system limit, and a factor of 10^250 short of the observed cosmos limit.

    All this has been pointed out over and over, just studiously ignored and strawmannised, as above.

    Likewise, just to get to a typical, 300 AA protein, we are dealing with DNA code of 900 bases, or 1800 bits of basic info carrying capacity. For a simplistic first cell based life of 100 such proteins, that is 180,000 bits, or 90,000 bases. This shows the lower end of the range, 100,000 – 1 mn bases for the genome of a reasonable first cell based life. Remember, such would have to be a gated, encapsulated metabolic automaton with integrated, code using von Neumann self replicator.

    I safely conclude, the FSCO/I required for such first life is not a plausible product of any cluster of actually observed spontaneously occurring processes to be found in a Darwin warm, salty lightning struck pond or any other seriously proposed OOL scenario.

    The only empirically, observationally warranted source of the active bridging information to get us to the first island of function, the tap root of Darwin’s tree of life, is intelligently directed configuration. AKA, Design.

    Design, therefore, sits at the table from the root up, as of right, not grudging sufferance. (Which sufferance, thanks to ideological commitment to a priori evolutionary materialism, is scarce.)

    Going on, a simple calculation for informational requisites for origin of body plans, OOBP, will readily come up with the range, 10 – 100+ mn bases. As I noted some time back in discussing OOBP in the IOSE:

    . . . the sort of novel body plans observed in the Cambrian fossil life revolution reasonably required 10 – 100+ millions of functional four-state DNA bases. This is more than 100,000 times the 500 – 1,000 bit threshold at which the undirected search resources of the observed cosmos would be inadequate to carry out a credible search of the relevant configuration spaces.

    Some would doubt such a range, so let us do a fresh calculation: 50 new tissue types to make up the organs for a new body plan would easily take up probably 10 – 100 proteins [[including enzymes etc] per type, i.e we are looking at 500 – 5,000 proteins as a reasonable/ conservative estimate — VERY conservative at the low end. 500 * 300 = 1.5 *10^5 codons, or 4.5 *10^5 bases, plus regulatory, let’s say about 10% more, 1/2 mn bases.At the upper end, we would arrive at 4.5 *10^6 bases.

    But this estimate is too low:

    Arabidopsis thaliana [[a flowering plant] 115,409,949 DNA bases

    Anopheles gambiae [[a mosquito] 278,244,063 bases

    Sea urchin 8.14 x 10^8 bases

    Amphibians 10^9–10^11

    Tetraodon nigroviridis (a pufferfish) 3.42 x 10^8

    In short, 10 – 100 million bases for a novel body plan is reasonable, even generous. And in any case the config space of 500 k bases is: 9.9 *10^301,029 possibilities.

    To cross the intervening sea of non-function to arrive at such deeply isolated islands of function, points strongly to intelligently directed, active configuration as source of relevant information and functional organisation. Especially, as design already sits at the table of candidate explanations of FSCO/I as of right.

    But, it will be predictably asserted — just see above — that all that is needed is incremental chance variation and natural selection leading to descent with modification across a branching tree structure.

    This is tantamount to the assertion or implication that there is a vast continent of living forms, from first common ancestral microbes, to Mozart, molluscs and mango trees, etc.

    How is such observationally grounded?

    In the end by gliding over the systematic pattern of missing intermediate forms across the fossil record, highlighting the ever-changing set of icons held to show what is otherwise concealed by the imperfections of the record and the like.

    After 250,000+ fossil species from all eras and across the world, with millions of samples in museums and billions seen in the ground, I don’t buy that argument. The Cambrian fossil life revolution is emblematic of the actual dominant pattern of gaps at exactly the points where transitional forms should utterly dominate the record. And that has been so since Darwin’s day.

    More to the point, such incrementalism runs cross-grain to the known, natural logic of how FSCO/I works.

    Take the fishing reel in the OP as an example. We see that many correct parts must be properly oriented, aligned, arranged and coupled per a wiring diagram for interactive function to emerge. That extends to the petroleum refinery and ever so many other familiar cases. it also applies to the many cases in the living cell, as protein synthesis and wider metabolism show. This also extends to higher level organisation of complex, multicellular life forms.

    In short, FSCO/I required for OOBP, implies a drastic limitation of acceptable arrangement of correct parts to achieve function, from the space of possible clumped or scattered arrangements of the sane parts. That obtains whether the scale is nm in the cell or mm to cm in a fishing reel. Put the parts of a 6500 C3 reel in a bait bucket and shake all you want. Unlike the simplistic case EL suggested of sorting stones, it is highly reliably predictable that a functional reel will not result.

    Because of the search space challenge highlighted in the OP but never cogently addressed above.

    The only empirically warranted adequate cause of the FSCO/I required to explain OOL and OOBP across the tree of life from the root up, is intelligently directed configuration. AKA, design.

    KF

  147. 147
    kairosfocus says:

    sparc, 144:

    How often have we seen this very thread before? I am not interested in fishing but even I realize that I’ve seen the Abu 6500 C3 reel before (according to Google it appears 42 times on this site). Just opening another thread will not bring the stillborn FSCO/I to life. Didn’t you read what WE had to say about it? And what about Dembski, Meyer, Behe, Marks et al.? Do you think they even consider FSCO/I? FSCO/I just dead and never lived.

    How eager — suspiciously eager — you are to not look at what is in front of you: concrete cases in point illustrating the reality and characteristics of functionally specific complex organisation and associated information, FSCO/I; and of course, to write a stillbirth certificate.

    Utterly revealing of the underlying problem.

    For, this is a blatant case of knocking over a strawman and pretending the strawman was all that ever was there. Backed up, by a hoped for negative appeal to authority in the teeth of the actual demonstration of the reality of FSCO/I before your eyes that you want to avert your eyes and mind from, complaining that you have seen these cases before.

    Exactly.

    (That is, with all due respect, you inadvertently showed how your root problem is the fallacy of the closed, indoctrinated mind. A particularly virulent form of selective hyperskepticism that has led you to disbelieve the testimony of your eyes. As the just linked explains. [I add: those who would caricature that other phrase I have often used as a descriptive phrase for what Simon Greenleaf long ago termed the error of the skeptic, should note that it describes a real problem encountered live also: inconsistent, double standards of required warrant that project a demand on what one is inclined to reject that one would not accept for a comparable case one is inclined to accept. And, where typically, if the hyperskeptical standard were to be generally applied whole fields of learning or common sense useful or even vital knowledge would vanish.])

    Since you are trying to suggest that functionally specific complex organisation and associated information, FSCO/I, is a non-issue . . . predictable, given a recent exchange in another thread alluded to above, I will AGAIN cite the remark made by Dembski in No Free Lunch, where he highlighted that the functional subset of complex specified information is the relevant one for the biological world. That is, the concept just described and abbreviated, FYI, does appear at a crucial point in the writings of Dr Dembski:

    p. 148:“The great myth of contemporary evolutionary biology is that the information needed to explain complex biological structures can be purchased without intelligence. My aim throughout this book is to dispel that myth . . . . Eigen and his colleagues must have something else in mind besides information simpliciter when they describe the origin of information as the central problem of biology.

    I submit that what they have in mind is specified complexity [[cf. here below], or what equivalently we have been calling in this Chapter Complex Specified information or CSI . . . .

    Biological specification always refers to function. An organism is a functional system comprising many functional subsystems. . . . In virtue of their function [[a living organism’s subsystems] embody patterns that are objectively given and can be identified independently of the systems that embody them. Hence these systems are specified in the sense required by the complexity-specificity criterion . . . the specification can be cashed out in any number of ways [[through observing the requisites of functional organisation within the cell, or in organs and tissues or at the level of the organism as a whole. {Dembski cites:}

    Wouters, p. 148: “globally in terms of the viability of whole organisms,”

    Behe, p. 148: “minimal function of biochemical systems,”

    Dawkins, pp. 148 – 9: “Complicated things have some quality, specifiable in advance, that is highly unlikely to have been acquired by ran-| dom chance alone. In the case of living things, the quality that is specified in advance is . . . the ability to propagate genes in reproduction.”

    On p. 149, he roughly cites Orgel’s famous remark from 1973, which exactly cited reads:

    In brief, living organisms are distinguished by their specified complexity. Crystals are usually taken as the prototypes of simple well-specified structures, because they consist of a very large number of identical molecules packed together in a uniform way. Lumps of granite or random mixtures of polymers are examples of structures that are complex but not specified. The crystals fail to qualify as living because they lack complexity; the mixtures of polymers fail to qualify because they lack specificity . . .

    And, p. 149, he highlights Paul Davis in The Fifth Miracle: “Living organisms are mysterious not for their complexity per se, but for their tightly specified complexity.”] . . .”

    p. 144: [[Specified complexity can be more formally defined:] “. . . since a universal probability bound of 1 [[chance] in 10^150 corresponds to a universal complexity bound of 500 bits of information, [[the cluster] (T, E) constitutes CSI because T [[ effectively the target hot zone in the field of possibilities] subsumes E [[ effectively the observed event from that field], T is detachable from E, and and T measures at least 500 bits of information . . . ”

    Nor is this a novelty introduced by Dembski. When he would have been in High School, in 1973, FYFI, this is what leading OOL researcher, Leslie Orgel, had to say:

    . . . In brief, living organisms [–> bio-functional context] are distinguished by their specified complexity. Crystals are usually taken as the prototypes of simple well-specified structures, because they consist of a very large number of identical molecules packed together in a uniform way. Lumps of granite or random mixtures of polymers are examples of structures that are complex but not specified. The crystals fail to qualify as living because they lack complexity; the mixtures of polymers fail to qualify because they lack specificity . . . .

    [HT, Mung, fr. p. 190 & 196:] These vague idea can be made more precise by introducing the idea of information. Roughly speaking, the information content of a structure is the minimum number of instructions needed to specify the structure. [–> this is of course equivalent to the string of yes/no questions required to specify the relevant “wiring diagram” for the set of functional states, T, in the much larger space of possible clumped or scattered configurations, W, as Dembski would go on to define in NFL in 2002 . . . ] One can see intuitively that many instructions are needed to specify a complex structure. [–> so if the q’s to be answered are Y/N, the chain length is an information measure that indicates complexity in bits . . . ] On the other hand a simple repeating structure can be specified in rather few instructions. [–> do once and repeat over and over in a loop . . . ] Complex but random structures, by definition, need hardly be specified at all . . . . Paley was right to emphasize the need for special explanations of the existence of objects with high information content, for they cannot be formed in nonevolutionary, inorganic processes. [The Origins of Life (John Wiley, 1973), p. 189, p. 190, p. 196.]

    And FYSYFI a few years later, J S Wicken in 1979, probably while Dembski was in College as an undergrad:

    ‘Organized’ systems are to be carefully distinguished from ‘ordered’ systems. Neither kind of system is ‘random,’ but whereas ordered systems are generated according to simple algorithms [[i.e. “simple” force laws acting on objects starting from arbitrary and common- place initial conditions] and therefore lack complexity, organized systems must be assembled element by element according to an [[originally . . . ] external ‘wiring diagram’ with a high information content . . . Organization, then, is functional complexity and carries information. It is non-random by design or by selection, rather than by the a priori necessity of crystallographic ‘order.’ [[“The Generation of Complexity in Evolution: A Thermodynamic and Information-Theoretical Discussion,” Journal of Theoretical Biology, 77 (April 1979): p. 353, of pp. 349-65. (Emphases and notes added. Nb: “originally” is added to highlight that for self-replicating systems, the blue print can be built-in.)]

    The descriptive phrase, functionally specific complex organisation and associated information, FSCO/I for short, FYI, is directly based on Wicken’s remarks and is informed by a much wider circle of considerations.

    You go on to call up Meyer, obviously having failed to note what he was already noted to have stated in answer to Falk’s hostile review of the 2009 book, Signature in the Cell:

    For nearly sixty years origin-of-life researchers have attempted to use pre-biotic simulation experiments to find a plausible pathway by which life might have arisen from simpler non-living chemicals, thereby providing support for chemical evolutionary theory. While these experiments have occasionally yielded interesting insights about the conditions under which certain reactions will or won’t produce the various small molecule constituents of larger bio-macromolecules, they have shed no light on how the information in these larger macromolecules (particularly in DNA and RNA) could have arisen. Nor should this be surprising in light of what we have long known about the chemical structure of DNA and RNA. As I show in Signature in the Cell, the chemical structures of DNA and RNA allow them to store information precisely because chemical affinities between their smaller molecular subunits do not determine the specific arrangements of the bases in the DNA and RNA molecules. Instead, the same type of chemical bond (an N-glycosidic bond) forms between the backbone and each one of the four bases, allowing any one of the bases to attach at any site along the backbone, in turn allowing an innumerable variety of different sequences. This chemical indeterminacy is precisely what permits DNA and RNA to function as information carriers. It also dooms attempts to account for the origin of the information—the precise sequencing of the bases—in these molecules as the result of deterministic chemical interactions . . . .

    [[W]e now have a wealth of experience showing that what I call specified or functional information (especially if encoded in digital form) does not arise from purely physical or chemical antecedents [[–> i.e. by blind, undirected forces of chance and necessity]. Indeed, the ribozyme engineering and pre-biotic simulation experiments that Professor Falk commends to my attention actually lend additional inductive support to this generalization. On the other hand, we do know of a cause—a type of cause—that has demonstrated the power to produce functionally-specified information. That cause is intelligence or conscious rational deliberation. As the pioneering information theorist Henry Quastler once observed, “the creation of information is habitually associated with conscious activity.” And, of course, he was right. Whenever we find information—whether embedded in a radio signal, carved in a stone monument, written in a book or etched on a magnetic disc—and we trace it back to its source, invariably we come to mind, not merely a material process. Thus, the discovery of functionally specified, digitally encoded information along the spine of DNA, provides compelling positive evidence of the activity of a prior designing intelligence. This conclusion is not based upon what we don’t know. It is based upon what we do know from our uniform experience about the cause and effect structure of the world—specifically, what we know about what does, and does not, have the power to produce large amounts of specified information . . .

    When you can cogently address such, you will rise above strawman tactics.

    But in fact, your strawman errors go beyond that.

    For, the focal issue addressed in the OP is not “old hat.” It is specifically the role of active bridging information in the origin of FSCO/I, required to bridge seas of non-function to arrive at OOL and OOBP in the relevant configuration spaces.

    Worse, this is a case of refusing to notice what is in front of you.

    The 6500 fishing reel is emblematic of literally trillions of cases of FSCO/I all around you, including in the PC you used to compose your dismissive comment and to read this one in reply. FSCO/I is real.

    The same phenomenon is exhibited in the digitally coded text s-t-r-i-n-g in your dismissive statement, which directly manifests the kind of specific functional organisation found in DNA, mRNA and onward in proteins and enzymes assembled through the FSCO/I rich process and systems in the ribosome. All, backed up by the further FSCO/I in the metabolic networks of the living cell.

    All of which are illustrated in the OP for this thread, and all of which you are ever so eager to sweep away by resorting to strawman tactics.

    Worse, the FSCO/I involved in the embedded integrated von Neumann kinematic self replicating facility of the living cell also cries out for adequate causal explanation.

    And, in case you want to revert to the longstanding strawman tactic of dismissing Paley’s watch argument as a flawed analogy, let me cite what Paley directly went on to say regarding self-replication in Ch 2 of his Nat Theol, his actual main watch argument . . . an argument I find suspiciously missing in far too many dismissals of what he had to say back in 1804.

    Let me clip 19 above (yes, you seem to be commenting dismissively without having first interacted seriously with the thread of discussion).

    Paley:

    Suppose, in the next place [–> immediately following a short C1], that the person who found the watch should after some time discover that, in addition to all the properties which he had hitherto observed in it, it possessed the unexpected property of producing in the course of its movement another watch like itself — the thing is conceivable; that it contained within it a mechanism, a system of parts — a mold, for instance, or a complex adjustment of lathes, baffles, and other tools — evidently and separately calculated for this purpose [==> update, vNSR, with tape [a bar of cams is a program, as was used in so many C18 automata], and constructor keyed in as an ADDITIONAL facility integrated with the main machine — of course, IIRC a full size clanking unit considered by NASA was many, many tons in scale] . . . .

    The first effect would be to increase his admiration of the contrivance, and his conviction of the consummate skill of the contriver. Whether he regarded the object of the contrivance, the distinct apparatus, the intricate, yet in many parts intelligible mechanism by which it was carried on, he would perceive in this new observation nothing but an additional reason for doing what he had already done — for referring the construction of the watch to design and to supreme art [–> notice, the impact of seeing ADDITIONAL FSCO/I] . . . . He would reflect, that though the watch before him were, in some sense, the maker of the watch, which, was fabricated in the course of its movements, yet it was in a very different sense from that in which a carpenter, for instance, is the maker of a chair — the author of its contrivance, the cause of the relation of its parts to their use.

    In short, Paley anticipated Darwin et al by 50 years, and von Neumann by 150 years.

    The issue he put on the table cries out for adequate, empirically grounded causal explanation.

    And, going back to the focal issue in the OP, the only adequate, empirically grounded cause of the required FSCO/I — emphatically not a dead or ignorable matter — and for the required, active bridging information to navigate the seas of non-function to hit on deeply isolated islands of function, is intelligently directed configuration. AKA, design.

    KF

  148. 148
    kairosfocus says:

    F/N: FSCO/I is BTW a genuine, legitimately accounted for case of the emergent behaviour of systems comprising interacting parts. But, of course, while it readily gets you to mechanical GIGO limited computation, it will not allow you to indulge the fantasy of poof, we get North to rational self-aware contemplation by insistently heading West to blindly mechanical computation. KF

  149. 149
    timothya says:

    Kairosfocus: have you submitted a paper covering a theoretical or experimental justification of your FSCO/I proposition to any peer-reviewed journal? If you have, can you provide a citation?

    Thanks in advance.

  150. 150
    Zachriel says:

    mike1962: Give me an example with regard to software implemented replicator objects interacting with their environment.

    A simple example is Weasel, but for a scientific example, see Krupp & Taylor, Social evolution in the shadow of asymmetrical relatedness, Proceedings of the Royal Society B: Biological Sciences 2015.

    mike1962: In order for a replicator to have a relationship with it’s environment it has to have certain properties that will necessarily be determinative of any future outcome.

    It’s not determinative other than in the trivial sense.

    mike1962: No, it also requires that the replicators have particular properties that allow it to successfully replicate in the environment.

    That’s what is meant by a fitness landscape. If the fitness is always one or always zero, then the landscape is flat, and evolution would be no different than a random walk.

    Mung: Let’s recall that the initial population is randomly generated.

    That’s what is commonly done. Consider them conjectured solutions.

    Mung: How is this relationship defined in an EA?

    It’s entailed in the concept of a fitness landscape. Given a sequence, we determine its fitness. And if the fitness landscape is non-chaotic, then replicators will tend to find higher levels of fitness. If we can’t determine fitness or if fitness doesn’t vary, then evolution would be no different than a random walk.

    NetResearchGuy: I think Mung’s point is that known functional evolutionary algorithms start with a fixed set of alleles that are designed. For example Weasel starts with letters, the antenna evolving algorithm starts with a working antenna made out of metal and a list of allowable mutations, the nozzle evolving algorithm starts with a working nozzle, etc.

    It’s entailed in the notion of a fitness landscape, otherwise there would be no relationship between the replicators and the fitness landscape. The fitness landscape would be random or flat with respect to sequence. Evolution only works when the fitness landscape is positively ordered.

    NetResearchGuy: In other words, the initial genotype is designed to start on an island of function, and the allowable variations to the genotype are designed to remain on that island of function.

    In many evolutionary algorithms, most variations are not viable.

    NetResearchGuy: Evolution doesn’t always work in every case — it’s quite easy to construct examples where evolution can’t work, at least with finite resources.

    Absolutely. There are chaotic landscapes, as well as perverse landscapes, that are not amenable to evolutionary processes. Evolution won’t work on the vast majority of conceivable fitness landscapes, even with resources short of exhaustive sampling. However, biological evolution works in a highly ordered, albeit complex, fitness landscape.

    NetResearchGuy: I’ve never seen EL or Z even tangentially address these issues.

    Many times. The reason evolution works in rational fitness landscapes is because it doesn’t have to sample the vast majority of space, just like you don’t have to cover every bit of ground to reach the top of a hill.

    kairosfocus: As a first answer, we can challenge you, Z, to provide an onward incrementalist functional at each step chance variation and selection based progress to something like

    Provide us the fitness landscape, a way to determine whether something is a valid phrase in English, then we will provide the algorithm. A phrase book should contain these among other snips: “quick dog” “lazy fox”, “jumps over”, “the quick”, etc.

  151. 151
    velikovskys says:

    Kf, thanks for the info

  152. 152
    kairosfocus says:

    TA, did you actually take time to read what is in front of you that shows that the concept and almost the exact phrasing in the acronym FSCO/I has long been in relevant literature from leading lights starting with Orgel and Wicken in the 1970’s? Apart from, that such a descriptive phrase describes things that are literally right in front of us as we participate in this discussion — text strings, PCs that process such, ever so many artifacts, etc? Or, are you just trying to pile on a dismissive talking point that evades addressing a patent issue? KF

  153. 153
    kairosfocus says:

    Z:

    kairosfocus: As a first answer, we can challenge you, Z, to provide an onward incrementalist functional at each step chance variation and selection based progress to something like

    [Z:} Provide us the fitness landscape, a way to determine whether something is a valid phrase in English, then we will provide the algorithm. A phrase book should contain these among other snips: “quick dog” “lazy fox”, “jumps over”, “the quick”, etc.

    Z, it is you who by suggesting a stepwise task suggested such. Therefore, it is part of your own burden of proof.

    And had you observed the Million Monkeys task, you would see the obvious point, matching a large corpus of literature stored electronically as a way to address it.

    All of which goes to the absence of actual algorithms specifying step by step procedure that computationally implemented on a reasonable machine solve the problem in hand.

    KF

  154. 154
    kairosfocus says:

    VS, welcome. KF

  155. 155
    sparc says:

    kairosfocus @ 148
    You mean Demsbki, Behe, Meyer et al. were referring to FSCO/I before you even coined the term? Dembski’s and Marks’s co-worker and co-author Winston Ewert was obviously reluctant to even consider FSCO/I. When asked directly

    Is FSCO/I something you’ve heard of? If you have, do you (and as spokesman for DEM) endorse it?

    he replied

    I’ve seen posts about it. I’m not inclined to take it seriously until I see it published some place more serious then a blog.

    BTW, nobody is mentioning Lewontin in this thread.

  156. 156
    kairosfocus says:

    sparc:

    You obviously are refusing to look at and take due note of the evidence right in front of your eyes.

    FSCO/I is a descriptive acronym, that actually builds on Orgel, Wicken and Thaxton et al. Dembski and Meyer speak to the same cluster of facts, phenomena, concepts and issues.

    That has been on record for decades.

    All around you the phenomenon is manifest.

    Whatever we may wish for by way of more research, those facts are already more than adequate to address seriously the functionally specific subset of complex specified information that is as blatantly in your face as the text s-t-r-i-n-g-s we are using to communicate with in this thread. It is as commonplace as the electrical and electronic circuits in the PCs we are using.

    It is in the coded info in the programs in those PCs.

    It is in the exploded view wiring diagrams behind ever so many machines, whether PCs or fishing reels or automobiles or airplanes.

    It is in the DNA of the cells, the ribosomes, the proteins, the von Neumann self replicating facilities in the cells.

    And, consistently, when we see the cause, it is intelligently directed configuration. With a very good reason as discussed in the OP above you seem to be studiously avoiding, namely that wiring diagram interactive functionality sharply constrains effective arrangements of parts, creating isolated islands of function in large config spaces.

    So, the means to bridge seas of non-function to reach such islands is by injection of active information that moves us beyond what we may reasonably expect of blind chance and mechanical necessity.

    KF

  157. 157
    mike1962 says:

    Zechrial: A simple example is Weasel

    Let’s stick with Weasel for now:

    What part of the software and/or data is the replicator object and what part is the environment?

  158. 158
    Zachriel says:

    mike1962: What part of the software and/or data is the replicator object and what part is the environment?

    The population of sequences are the replicates. The environment is the target string, “Methinks it is like a weasel”.

    kairosfocus: Therefore, it is part of your own burden of proof.

    Let’s start with the simpler case. Using a standard dictionary for the fitness landscape, are seven-letter words isolated islands? Ten-letter words?

  159. 159
    kairosfocus says:

    Z, it is you who still have that burden to show. And the target zone — assuming use of the 128 ASCII character set, is 72 characters of coherently meaningful text strings. So far, as cited above and long since headlined here at UD, 19 – 24 characters is what’s been done with W ~ 10^50, a factor of 10^100 short of the scope where FSCO/I is a material factor. KF

  160. 160
    Zachriel says:

    kairosfocus: it is you who still have that burden to show.

    We could use a book of every possible phrase in the English language, but that would require quite a bit of memory. How about a partial phrase book? That would result in even larger gaps between valid sequences, and make the job of navigating it with an evolutionary algorithm considerably more difficult. The phrase book would include words and phrases, a phrase defined as any contiguous words found in a valid sentence. Does that work for a fitness landscape to show islands of function?

  161. 161
    kairosfocus says:

    Z, how you do it is your problem. By hoping to get away with extrapolating on short words that will be closer in Hamming space, you set yourself that burden. On the cases I showed, they obviously compared a corpus of literature. KF

  162. 162
    Zachriel says:

    kairosfocus: how you do it is your problem.

    You won’t stake out a position so that you can disavow it later. There’s no point testing your position when you won’t cooperate in devising a suitable test.

    kairosfocus: By hoping to get away with extrapolating on short words that will be closer in Hamming space, you set yourself that burden.

    Then let’s consider longer words. Do you think 12-letter words are isolated islands?

  163. 163
    mike1962 says:

    Zechriel: The population of sequences are the replicates. The environment is the target string, “Methinks it is like a weasel”.

    So then, is all the code and data except the target string part of the replicators’ set of properties?

  164. 164
    Zachriel says:

    mike1962: So then, is all the code and data except the target string part of the replicators’ set of properties?

    Here’s the usual steps:

    * Evaluate the individual fitness of each member of population;
    * Determine best fit individuals to become parents;
    * Breed new individuals through mutation and crossover to give birth to offspring;

  165. 165
    Carpathian says:

    Mung:

    That a ’0? in the genome represents a ‘T’ in the phenome is a design decision. That a ’1? in the genome represents an ‘H’ in the phenome is a design decision. That ‘T’ and ‘H’ have relevance to a potential solution is a design decision.

    This sounds to me more of a case of labeling, not design.

    You seem to be confusing the “lab” with the “Device Under Test”.

    If I type that I prefer to use a pen instead of using a computer, I have not demonstrated that I am a hypocrite by typing this sentence. It is simply the case that this is the best and probably only way to get that information to you.

  166. 166
    Zachriel says:

    mike1962: So then, is all the code and data except the target string part of the replicators’ set of properties?

    There’s a relationship between the sequences and the fitness landscape. That’s what it means to evaluate the fitness. In natural biology, when a mutated sequence is expressed, there may be a variety of very complex changes in the organism. In terms of reproductive potential, this can be beneficial, detrimental, or neutral. We don’t have to know the details of this process to observe whether a variation is advantageous or not in a given environment.

    This process is abstracted in an evolutionary algorithm.

  167. 167
    Joe says:

    Zachriel is a projectionist:

    You won’t stake out a position so that you can disavow it later. There’s no point testing your position when you won’t cooperate in devising a suitable test.

    Talk about not knowing one’s place.

    Then let’s consider longer words. Do you think 12-letter words are isolated islands?

    Not for intelligently designed processes like word mutagen.

  168. 168
    Joe says:

    Zachriel:

    This process is abstracted in an evolutionary algorithm.

    Evolutionary algorithms model intelligent design evolution. Nice own goal.

  169. 169
    Joe says:

    sparc- Dembski wrote that biological specification refers to function. All FSCO/I do0es it make it much clearer, ie more specific, what is being discussed.

  170. 170
    Carpathian says:

    Joe:

    Evolutionary algorithms model intelligent design evolution. Nice own goal.

    You are confusing the lab and the test.

    The tools we use may be designed, but that should not have an effect on what we are measuring.

    Sometimes the effects can be diminished and sometimes we just have to live with them.

    As an example, there is no reason we could not use designed math to model random probabilities resulting in a dealt hand in a card game.

  171. 171
    Joe says:

    No Carpathian, YOU are confused. GAs are intelligently designed and so are their solutions. GAs use selection- actual intelligent selection, to drive variants towards a goal.

  172. 172
    Carpathian says:

    Joe:

    A designed algorithm does not mean the problems it solves are designed.

    Take “X + Y = Z”. Given that numbers are infinite, there is no possible way that all solutions are known beforehand.

    In the case of Dawkin’s Weasel program, the algorithm doesn’t know the answer.

    (pointer to solution) = WeaselFunction((pointer to string));

    The weasel function requires a string as input since the algorithm doesn’t have a built-in solution.

  173. 173
    Joe says:

    They solve the problem by design. They are no different than thousands of qualified people all working independently to solve the problem.

    In Dawkins weasel if the answer wasn’t provided the sentence would never be found. In the antenna GA if the specification for the antenna wasn’t provided the solution would never be found. GAs do what humans do, they just do it faster and with virtual resources.

  174. 174
    Carpathian says:

    In Dawkin’s Weasel algorithm, the input sentence is the “specification”.

    You could replace the input string with a short “DNA sequence” and take the next randomly altered by Weasel “DNA sequence” and use it as a GMO modifier.

    If the corn or green beans turn out to be a better crop, you feed that short “DNA sequence” back into Weasel to reproduce with random changes.

    You would never actually compare any strings. The test would be better crops.

    This way there is no reason to say the solution is known since we would never compare it to any specific sequence.

  175. 175
    Joe says:

    The solution isn’t known but what is wanted is known. The specification for the antenna was known but the solution, ie the actual antenna design, was not. Without the specification the antenna would never have been realized.

  176. 176
    Carpathian says:

    Yes, the “specification” was the “environment” the antenna would be operating in.

    Sometimes we don’t even know that.

    In the case of a Weasel type program generating “DNA snippets”, we don’t have a clue what they might be and we don’t know we will like the result even if it survives.

    The point is, that the criteria for allowing that “DNA snippet” to survive is unknown before the organism is actually tested.

    This is a blind test to the farmer, the seed company and Dawkin’s Weasel program.

    No one knows the “solution” which will simply be something the farmer likes.

  177. 177
    Joe says:

    No, Carpathian, the specification was the wavelength and type of wave- you know typical antenna stuff that engineers use. The GA emulated engineers.

    They all use actual selection as opposed to natural selection which is a process of elimination. Selecting the top 1% is much different from eliminating the bottom 1%. Selection drives towards a target.

    I know no one knows the solution. The point is GAs actively drive towards that solution.

  178. 178
    Winston Ewert says:

    Didn’t you read what WE had to say about it? And what about Dembski, Meyer, Behe, Marks et al.? Do you think they even consider FSCO/I? FSCO/I just dead and never lived.

    What I said was that I wasn’t familiar with it, and suggested that it should be pursued via publication in a venue like a paper or conference rather than in blog posts. I have no opinion on the merits of FSCO/I. I don’t know enough about it to have an opioion.

  179. 179
    Mung says:

    No word is an island.

  180. 180
    Mung says:

    In the case of Dawkin’s Weasel program, the algorithm doesn’t know the answer.

    So?

    The weasel function requires a string as input since the algorithm doesn’t have a built-in solution.

    So?

    WeaselFunction(pointer to string)

    Why a string? Why not a pointer to a peanut butter and jelly sandwich?

    Let me tell you why. Because the programmer knew that the desired solution was a string of a certain length composed of specific letters in a specific order and found a way to code that into the program. The program didn’t magically conceive that on it’s own.

    We say GA’s are designed because they are designed. And then people say that’s irrelevant. Har.

  181. 181
    kairosfocus says:

    Z, you are simply showing just how difficult the real problem is. KF

  182. 182
    sparc says:

    179
    Winston EwertMay 4, 2015 at 1:56 pm

    Didn’t you read what WE had to say about it? And what about Dembski, Meyer, Behe, Marks et al.? Do you think they even consider FSCO/I? FSCO/I just dead and never lived.

    What I said was that I wasn’t familiar with it, and suggested that it should be pursued via publication in a venue like a paper or conference rather than in blog posts. I have no opinion on the merits of FSCO/I. I don’t know enough about it to have an opioion.

    According to Google FSCO/I has been metioned at UD more than 1000 times. I.e., more often than Dembski’s original CSI which although much older was mentioned 918 times according to Google. Maybe it’s my English skills but your statements

    I’ve seen posts about it. I’m not inclined to take it seriously until I see it published some place more serious then a blog.

    You really think a comment on a blog that quotes other people and calls them idea-roots for FSCO/I qualifies as a serious presentation of the idea of FSCO/I?

    don’t really sound as if you don’t have an opinion on FSCO/I.
    BTW, is the founder of UD even aware of FSCO/I? Did you ever discuss FSCO/I with Dr. Dembski or Dr. Marks in the past or in the since it was mentioned in the previous thread? If not I guess everybody here would be happy if you discuss FSCO/I with him and Dr. Marks and publish a summary of the discussion here.

  183. 183
    kairosfocus says:

    sparc:

    Why are you unresponsive to what Dembski stated in one of his main works, in discussing that in the biological context the specification in question in CSI is functional: “Biological specification always refers to function?

    Similarly, why are you unresponsive to the specific citation from Meyer that refers to “specified or functional information (especially if encoded in digital form)”?

    Especially, as both have been specifically cited to you above?

    Let me cite Dembski again, from the cite that appears most recently above at comment no. 148 just yesterday morning (rather coincidental with it’s coming from p. 148 f in NFL . . . ); for your information:

    p. 148:“The great myth of contemporary evolutionary biology is that the information needed to explain complex biological structures can be purchased without intelligence. My aim throughout this book is to dispel that myth . . . . Eigen and his colleagues must have something else in mind besides information simpliciter when they describe the origin of information as the central problem of biology.

    I submit that what they have in mind is specified complexity [[cf. here below], or what equivalently we have been calling in this Chapter Complex Specified information or CSI . . . .

    Biological specification always refers to function. An organism is a functional system comprising many functional subsystems. . . . In virtue of their function [[a living organism’s subsystems] embody patterns that are objectively given and can be identified independently of the systems that embody them. Hence these systems are specified in the sense required by the complexity-specificity criterion . . . the specification can be cashed out in any number of ways [[through observing the requisites of functional organisation within the cell, or in organs and tissues or at the level of the organism as a whole. {Dembski cites:}

    Wouters, p. 148: “globally in terms of the viability of whole organisms,”

    Behe, p. 148: “minimal function of biochemical systems,”

    Dawkins, pp. 148 – 9: “Complicated things have some quality, specifiable in advance, that is highly unlikely to have been acquired by ran-| dom chance alone. In the case of living things, the quality that is specified in advance is . . . the ability to propagate genes in reproduction.”

    On p. 149, he roughly cites Orgel’s famous remark from 1973, which exactly cited reads:

    In brief, living organisms are distinguished by their specified complexity. Crystals are usually taken as the prototypes of simple well-specified structures, because they consist of a very large number of identical molecules packed together in a uniform way. Lumps of granite or random mixtures of polymers are examples of structures that are complex but not specified. The crystals fail to qualify as living because they lack complexity; the mixtures of polymers fail to qualify because they lack specificity . . .

    And, p. 149, he highlights Paul Davis in The Fifth Miracle: “Living organisms are mysterious not for their complexity per se, but for their tightly specified complexity.”] . . .”

    p. 144: [[Specified complexity can be more formally defined:] “. . . since a universal probability bound of 1 [[chance] in 10^150 corresponds to a universal complexity bound of 500 bits of information, [[the cluster] (T, E) constitutes CSI because T [[ effectively the target hot zone in the field of possibilities] subsumes E [[ effectively the observed event from that field], T is detachable from E, and and T measures at least 500 bits of information . . . ”

    (It would be appreciated if you were to explain to us why this is apparently unacceptable to you as pointing out that in the biological world, we address the functionally specified subset of complex, specified information, instantly making it reasonable to use an acronym for a stock descriptive phrase that describes the explicit and implicit cases of such biologically functional, information-rich specified complexity. Namely, again, FSCO/I = functionally specific complex organisation and/or associated information. BTW, why have you been consistently unresponsive to the repeated explanation or expansion of the acronym?)

    Likewise, FYFI, here is Meyer as cited to you just yesterday at 148, in his response to Falk’s hostile review of Signature in the Cell, 2009:

    For nearly sixty years origin-of-life researchers have attempted to use pre-biotic simulation experiments to find a plausible pathway by which life might have arisen from simpler non-living chemicals, thereby providing support for chemical evolutionary theory. While these experiments have occasionally yielded interesting insights about the conditions under which certain reactions will or won’t produce the various small molecule constituents of larger bio-macromolecules, they have shed no light on how the information in these larger macromolecules (particularly in DNA and RNA) could have arisen. Nor should this be surprising in light of what we have long known about the chemical structure of DNA and RNA. As I show in Signature in the Cell, the chemical structures of DNA and RNA allow them to store information precisely because chemical affinities between their smaller molecular subunits do not determine the specific arrangements of the bases in the DNA and RNA molecules. Instead, the same type of chemical bond (an N-glycosidic bond) forms between the backbone and each one of the four bases, allowing any one of the bases to attach at any site along the backbone, in turn allowing an innumerable variety of different sequences. This chemical indeterminacy is precisely what permits DNA and RNA to function as information carriers. It also dooms attempts to account for the origin of the information—the precise sequencing of the bases—in these molecules as the result of deterministic chemical interactions . . . .

    [[W]e now have a wealth of experience showing that what I call specified or functional information (especially if encoded in digital form) does not arise from purely physical or chemical antecedents [[–> i.e. by blind, undirected forces of chance and necessity]. Indeed, the ribozyme engineering and pre-biotic simulation experiments that Professor Falk commends to my attention actually lend additional inductive support to this generalization. On the other hand, we do know of a cause—a type of cause—that has demonstrated the power to produce functionally-specified information. That cause is intelligence or conscious rational deliberation. As the pioneering information theorist Henry Quastler once observed, “the creation of information is habitually associated with conscious activity.” And, of course, he was right. Whenever we find information—whether embedded in a radio signal, carved in a stone monument, written in a book or etched on a magnetic disc—and we trace it back to its source, invariably we come to mind, not merely a material process. Thus, the discovery of functionally specified, digitally encoded information along the spine of DNA, provides compelling positive evidence of the activity of a prior designing intelligence. This conclusion is not based upon what we don’t know. It is based upon what we do know from our uniform experience about the cause and effect structure of the world—specifically, what we know about what does, and does not, have the power to produce large amounts of specified information . . .

    Why are you further unresponsive to Orgel’s focus on specified complexity as a characteristic of cell based life that distinguishes it — in a specifically OOL context — from crystals and random polymers?

    Also, to Wicken’s apt point that wiring diagram, information rich function is involved in the organisation found in that general context?

    (Both of these were also cited to you specifically.)

    Further to this, why have you been so unresponsive to the commonplace fact that simply the s-t-r-i-n-g-s of text in this discussion thread and more generally (surely you know how important this type of data structure is . . . ) are a form of wiring diagram organisation?

    Likewise, why have you been unresponsive to the patent fact that, it is an utter commonplace for parts to work together to create a functional result based on how they are arranged and coupled together, e.g. in electrical/ electronic circuits, or the sort of engineering diagrams commonly called exploded views and also wireframes?

    Why have you been unresponsive to the widespread fact that such a nodes and arcs mesh may be reduced informationally to a structured sequence of Y/N q’s, i.e. to a binary digit (= bit) based description, such as one may see with AutoCAD etc, or more mathematically with the approach Orgel took of speaking to length of the descriptive string as quantifying the information involved.

    Have you taken note that since 2005 Trevors and Abel have distinguished and discussed in the peer reviewed literature for strings — directly involved in D/RNA and implied by wiring diagram functional organisation — ordered, random and functional sequence complexity. Which in turn reflects the following contrast from Thaxton et al in The Mystery of Life’s Origin, 1984 — the very first technical work that sparked the rise of design theory, which builds on Orgel and Wicken:

    1. [Class 1:] An ordered (periodic) and therefore specified arrangement:
    THE END THE END THE END THE END

    Example: Nylon, or a crystal . . . .

    2. [Class 2:] A complex (aperiodic) unspecified arrangement:
    AGDCBFE GBCAFED ACEDFBG

    Example: Random polymers (polypeptides).

    3. [Class 3:] A complex (aperiodic) specified arrangement:
    THIS SEQUENCE OF LETTERS CONTAINS A MESSAGE!

    Example: DNA, protein.

    Where, the same authors go on to say:

    Yockey7 and Wickens5 develop the same distinction, that “order” is a statistical concept referring to regularity such as could might characterize a series of digits in a number, or the ions of an inorganic crystal. On the other hand, “organization” refers to physical systems and the specific set of spatio-temporal and functional relationships among their parts. Yockey and Wickens note that informational macromolecules have a low degree of order but a high degree of specified complexity. In short, the redundant order of crystals cannot give rise to specified complexity of the kind or magnitude found in biological organization; attempts to relate the two have little future. [NB: The name should be Wicken.]

    They go on to comment: “the redundant order of crystals cannot give rise to specified complexity of the kind or magnitude found in biological organization; attempts to relate the two have little future.”

    Thirty years later this has been fully justified.

    Now, sparc, you have been a critic of design thought in and around UD for years. I cannot believe that the above has escaped your notice all this time; your purpose is plainly rhetorical, and the facts as repeated to you are obviously inconvenient to where you want to go.

    I will simply point out, therefore, that the above conclusively shows that the concepts and context addressed by the acronym FSCO/I have been at the heart of design thought for 30 years, were originally built on considerations brought to the table by Orgel, Wicken and Yockey et al, and have in fact been highlighted by both Dembski and Meyer as noted.

    Further to all this, the phrase is descriptive of a common, easily observed pattern. One that appears in text of comments of this thread of discussion, and also in things like fishing reels, petroleum refineries, watches, engines, electronic circuits and software. All of which are abundantly familiar in a high tech age. Some of which are exemplified in the OP above. Trillions of cases altogether.

    FSCO/I is objectively real and not a mere fairy-tale figment of some dismissible IDiot’s imagination.

    That you find it so hard to acknowledge that patent reality that is literally staring you in the face when you read the text of comments or compose comments of your own simply underscores its cogency and inductive force that obviously strongly points where you desperately do not want to go.

    Consistently, reliably — and for reasons connected to the need for actively, intelligently inserted functional information and linked organisation (as the OP you are also largely unresponsive to discusses) — such FSCO/I is a reliable sign of intelligently directed configuration as cause.

    AKA, design.

    Now, if you want to challenge such an inductively grounded conclusion, the path is obvious. Provide a counter example that credibly shows that nope FSCO/I is also adequately caused by blind chance and/or mechanical necessity. Which, in any case is needed to warrant claims commonly made from an evolutionary materialist perspective regarding OOL and OOBP. If, such are to meet the vera causa test of causal adequacy.

    The problem of course is patent. In the teeth of a trillion member base of cases, you cannot provide such cases. Dozens of attempts to do so in and around UD for years have consistently come up as instead inadvertently showing intelligently directed configuration with requisite complex specified information and/or linked organisation — typically functionally specific — thus being again and again showed to be strong and reliable signs of design.

    But, equally obviously, for reasons connected to the ideological dominance of evolutionary materialist thought on origins science — and this descriptive term traces ultimately to concepts discussed by Plato in The Laws Bk X, FYI — that is not a welcome result.

    Hence, the current semantics games, unresponsiveness to evidence and refusal to acknowledge that, on the record, these concepts have long been a part of the design theory discussions.

    So, while it is not a little annoying to have to deal with such drumbeat unresponsiveness to cogent evidence, the very fact of such rhetorical patterns on the part of objectors to design theory inadvertently underscores the substantial force of the design inference on FSCO/I and broader CSI. (Which of course has in its turn come in for endless drumbeat repetition of talking points in objection. And, it is not coincidental that the best way to ground the reality of the general concept is through functionally specific cases, especially those based on digital code strings and obvious wiring diagrams.)

    I am confident that the astute onlooker will be able to see the balance on the merits, thus why it is reasonable and empirically warranted to speak of FSCO/I as a patent empirical reality and strong sign of design as cause.

    KF

  184. 184
    kairosfocus says:

    PS: Just as a fresh perspective on complex biological, coded functional information, we may wish to ponder:

    http://www.ncbi.nlm.nih.gov/pm.....MC3319427/

    Journal ListTheor Biol Med Modelv.9; 2012PMC3319427

    Theor Biol Med Model. 2012; 9: 8.

    Published online 2012 Mar 14. doi: 10.1186/1742-4682-9-8
    PMCID: PMC3319427

    Dichotomy in the definition of prescriptive information suggests both prescribed data and prescribed algorithms: biosemiotics applications in genomic systems

    David J D’Onofrio,corresponding author1 David L Abel,corresponding author2 and Donald E Johnson3

    Bioinformatics has opened up the field of molecular biology through the use of computer science and statistics. Data mining of genetic information includes discovering relationships between individual DNA sequences and variability in disease [1]. More importantly, the application of computer science will contribute to identifying intricate complex data and algorithmic structures that are part of the biological processes that manage and maintain metabolic functions of the cell.

    Biological organisms are considered to be controlled and regulated by Functional Information (FI) [2-8]. FI comes closer to expressing the intuitive and semantic sense of the word “information” than mere Shannon combinatorial uncertainty or reduced uncertainty (poorly termed “mutual entropy”). The innumerable attempts that have been made to reduce the functional information of genomics and molecular biology to nothing more than physical combinatorics and/or thermodynamics will fail for reasons best summarized in the peer-reviewed anthology entitled The First Gene: The Birth of Programming, Messaging and Formal Control [9].

    “Functional Information (FI)” has now been formalized into two subsets: Descriptive Information (DI) [7] and Prescriptive Information (PI) [7,10,11]. This formalization of definitions precludes the prevailing confusion of informational terms in the literature. The more specific and accurate term “Prescriptive Information (PI)” has been championed by Abel [12-16] to define the sources and nature of programming controls, regulation and algorithmic processing. Such prescriptions are ubiquitously instantiated into all known living cells [13]. PI either instructs or produces formal function [12] in such a way as to organize and institute a prescribed set of logic-gate programming choices. Without such steering of physicochemical interactions by “Choice-Contingent Causation and Control” (CCCC) [17-19], metabolic pathways and cycles would be impossible to integrate into a cooperative and holistic metabolism. The Organization (O) Principle [19] states that nontrivial formal organization can only be produced by CCCC. [–> Cf OP]

    Maynard Smith [20] argued that bioinformation is both specific and intentional. Maynard Smith also pointed out in this same paper the irreversibility of information transfer. Information moves only from signal to response, not in the reverse direction. He argued that genetic information implies the possibility of misinterpretation or error. Maynard Smith also considered genetic information to be undetermined by cause-and-effect necessity. But he considered genetic information to be gratuitous (not called for by the circumstances: unwarranted) [20].

    Jablonka [21] argues that life is dependent upon semantic information, and that Shannon “information” is insufficient to explain life. She emphasizes, as does Adami [22], the importance of “aboutness.” Aboutness relates to meaning which in biology relates to biofunction.

    Jablonka [21] also argues that semantic information can only exist with living or designed systems. “Only a living system can make a source into an informational input.” On page 588 Jablonka emphasizes the function of bioinformation. Thus the joint authors of this paper are not alone in our emphasis on the formal nature of life’s many control mechanisms.

    A closer examination of Prescriptive Information (PI) has led to a dichotomy in its definition to differentiate between 1) what are prescribed data, and 2) what are prescribed algorithms. As the concepts of computer science are applied to the cell, it is necessary to deconstruct information structures to identify and differentiate data from algorithms. The DNA polynucleotide molecule consists of a linear sequence of nucleotides, each representing a biological placeholder of adenine (A), cytosine (C), thymine (T) and guanine (G). This quaternary system is analogous to the base two binary scheme native to computational systems. As such, the polynucleotide sequence represents the lowest level of coded information expressed as a form of machine code . . .

    And, the beat goes on and on and on . . .

  185. 185
    kairosfocus says:

    PPS: And some more of the same,

    http://www.ncbi.nlm.nih.gov/pm.....MC4092032/

    C R Biol. 2011 Jan; 334(1): 1–5.
    Published online 2010 Dec 30. doi: 10.1016/j.crvi.2010.11.008
    PMCID: PMC4092032
    NIHMSID: NIHMS561755

    Thermodynamic perspectives on genetic instructions, the laws of biology, diseased states and human population control

    J. T. Trevors corresponding author and M. H. Saier, Jr.

    This article examines in a broad perspective entropy and some examples of its relationship to evolution, genetic instructions and how we view diseases. Many knowledge gaps abound, hence our understanding is still fragmented and incomplete. Living organisms are programmed by functional genetic instructions (FGI), through cellular communication pathways, to grow and reproduce by maintaining a variety of hemistable, ordered structures (low entropy). Living organisms are far from equilibrium with their surrounding environmental systems, which tends towards increasing disorder (increasing entropy). Organisms must free themselves from high entropy (high disorder) to maintain their cellular structures [–> i.e. cellular, functional organisation] for a period of time sufficient enough to allow reproduction and the resultant offspring to reach reproductive ages. This time interval varies for different species. Bacteria, for example need no sexual parents; dividing cells are nearly identical to the previous generation of cells, and can begin a new cell cycle without delay under appropriate conditions. By contrast, human infants require years of care before they can reproduce. Living organisms maintain order in spite of their changing surrounding environment, that decreases order according to the second law of thermodynamics. These events actually work together since living organisms create ordered biological structures by increasing local entropy. From a disease perspective, viruses and other disease agents interrupt the normal functioning of cells. The pressure for survival may result in mechanisms that allow organisms to resist attacks by viruses, other pathogens, destructive chemicals and physical agents such as radiation. However, when the attack is successful, the organism can be damaged until the cell, tissue, organ or entire organism is no longer functional and entropy increases. . . . .

    Atoms are ancient relics of the hypothesized Big Bang (Matsuno, 2008) and can be used to construct life forms under the control of FGIs (functional genetic instructions). Living organisms are programmed by FGIs, which flow through a biochemical communication pathway involving DNA–> RNA–> proteins, to instruct cells how to assemble into living organisms. They are programmed to grow and reproduce by maintaining a variety of hemistable, ordered structures (low entropy state) (Schrodinger, 1944). They are far from equilibrium with their surrounding environment, which tends towards increasing disorder (Dolev & Elitzur, 1998). This is achieved by absorption of energy, from our thermonuclear sun, which provides the energy for the conversion of inanimate material into living organisms. This occurs on our planet with conditions commensurate with the maintenance of the life forms that comprise our singular biosphere system (Dolev & Elitzur, 1998; Gatenby & Frieden, 2007).

    Researchers have devoted time and effort to defining and understanding the characteristics of life, from the atomic to the biospherical levels of organization (Penzlin, 2009; Schrodinger 1944) and in more recent years the possibility of synthetic single-celled life. Biology can therefore be viewed as the study of life (and death) at all levels of biological organization . . . Science relies on the fundamental laws of thermodynamics in addition to the knowledge that: (1) the cell is the basic unit of life; (2) life arises only from life; (3) a cell is the only living structure that can grow and divide (Trevors, 2004), and (4) functional genetic instructions flow along a cellular communication pathway to provide the instructions for the challenges from entropy, with reproduction as the normal outcome. Although natural selection prevents many individual organisms from reproducing, others must succeed if a species is to survive, even though all individuals within a species die, generally just not at the same time.

    The presence of the pattern of thought captured in the acronym FSCO/I should be abundantly apparent.

    But, the beat goes on and on and on . . .

  186. 186
    kairosfocus says:

    PPS: And despite posting hiccups and web vanishing, let us continue:

    http://www.ncbi.nlm.nih.gov/pubmed/18762901/

    Naturwissenschaften. 2009 Jan;96(1):1-23. doi: 10.1007/s00114-008-0422-8. Epub 2008 Sep 2.

    The riddle of “life,” a biologist’s critical view.

    Penzlin H1.
    Author information
    Abstract

    To approach the question of what life is, we first have to state that life exists exclusively as the “being-alive” of discrete spatio-temporal entities. The simplest “unit” that can legitimately be considered to be alive is an intact prokaryotic cell as a whole. In this review, I discuss critically various aspects of the nature and singularity of living beings from the biologist’s point of view. In spite of the enormous richness of forms and performances in the biotic realm, there is a considerable uniformity in the chemical “machinery of life,” which powers all organisms. Life represents a dynamic state; it is performance of a system of singular kind: “life-as-action” approach. All “life-as-things” hypotheses are wrong from the beginning. Life is conditioned by certain substances but not defined by them. Living systems are endowed with a power to maintain their inherent functional order (organization) permanently against disruptive influences. The term organization inherently involves the aspect of functionality, the teleonomic, purposeful cooperation of structural and functional elements. Structures in turn require information for their specification, and information presupposes a source. This source is constituted in living systems by the nucleic acids. Organisms are unique in having a capacity to use, maintain, and replicate internal information, which yields the basis for their specific organization in its perpetuation. The existence of a genome is a necessary condition for life and one of the absolute differences between living and non-living matter. Organization includes both what makes life possible and what is determined by it. It is not something “implanted” into the living beings but has its origin and capacity for maintenance within the system itself. It is the essence of life. The property of being alive we can consider as an emergent property of cells that corresponds to a certain level of self-maintained complex order or organization.

    But, the beat must go on and on . . .

  187. 187
    kairosfocus says:

    PPPPS: Just for hammering the point home:

    http://www.ncbi.nlm.nih.gov/pubmed/18202877/

    Theory Biosci. 2004 Jun;123(1):3-15. doi: 10.1016/j.thbio.2004.03.001.

    Evolution of cell division in bacteria.

    Trevors JT1.

    Author information
    Abstract

    Molecular evolution in bacteria is examined with an emphasis on cell division. For a bacterial cell to assemble and then divide required an immense amount of integrated cell and molecular biology structures/functions to be present, such as a stable cellular structure, enzyme catalysis, minimal genome, septum formation at mid-cell and mechanisms to take up nutrients and produce and use energy, as well as store it. The first bacterial cell(s) capable of division must have had complex cell and molecular biology functions. At this stage of evolution, they would not have been primitive cells but would have reached a threshold in evolution where cell division occurred in a regulated manner.

    But, but, but the drum-beat of objections and dismissals MUST go on . . .

  188. 188
    Zachriel says:

    Carpathian: In the case of Dawkin’s Weasel program, the algorithm doesn’t know the answer.

    Mung: So?

    An evolutionary algorithm typically consists of two modules; the evolving population and an oracle. The oracle can be simple, such as in Weasel, or it can be complex, such as a simulation of the natural environment. The key is that the only contact between the two parts is the return of a fitness value from the oracle when evaluating the elements of the population.

    In other words, the population module knows nothing about the workings of the oracle, and the only information it receives is through the fitness function. The oracle can be simple or exceedingly complex.

    Mung: Why a string?

    Because an evolutionary algorithm is a simple model of genomes, which can be represented by an evolving population of strings.

    Mung: Let me tell you why. Because the programmer knew that the desired solution was a string of a certain length composed of specific letters in a specific order and found a way to code that into the program.

    That’s not necessarily the case. While Weasel uses strings of a given length, a population can be composed of strings of arbitrary length.

  189. 189
    Carpathian says:

    Mung:

    Carpathian: In the case of Dawkin’s Weasel program, the algorithm doesn’t know the answer.

    Mung: So?

    It means that despite not knowing what the “target” is, the algorithm can find it.

    The feedback from the “environment”, (string comparison), for a string or “DNA snippet” is simply die versus reproduce.

    The string that is evaluated is not a part of the algorithm itself.

    Zachriel explains it well in #189.

  190. 190
    Joe says:

    What utter drivel- GAs ACTIVELY search for the solution they are designed to find. They use actual selection which is contrary to evolutionism’s mechanisms.

  191. 191
    Mung says:

    Zachriel is a hoot.

    A population module and an oracle and “the population module knows nothing about the workings of the oracle.” So what?

    This is ground already covered and Zachriel wants to cover it again.

    The population module is generator for potential solutions (candidate solutions) to the problem. IOW, it helps traverse the search space.

    But the candidate solutions are not just randomly generated objects of no particular design.

    Zachriel:

    Zachriel: The notion of a fitness landscape entails that there is a defined relationship between the replicators and the landscape…

    And just how and where is this relationship defined?

    Zachriel:

    An evolutionary algorithm typically consists of two modules; the evolving population and an oracle. The oracle can be simple, such as in Weasel, or it can be complex, such as a simulation of the natural environment. The key is that the only contact between the two parts is the return of a fitness value from the oracle when evaluating the elements of the population.

    And that’s nothing more than a red herring and is only part of the story.

  192. 192
    kairosfocus says:

    P^5S: Let’s try a sample chapter:

    6
    SYSTEMS BIOLOGY OF
    CELL ORGANIZATION

    The first few chapters of this textbook laid the foundation for understanding cell structure and function. We learned that life depends on organic molecules, which form the building blocks for macromolecules such as proteins, nucleic acids, and carbohydrates. In addition, we considered cell organ-ization at a higher level. Cells contain complex structures such as membranes, chromosomes, ribosomes, and a cytoskeleton. Eukaryotic cells have organelles that provide specialized com-partments to carry out various cellular functions.

    In short, complex specifically functional organisation of the cell based on interaction of correct, correctly arranged parts is a basic fact of life. Just as the OP describes. And just as the acronym FSCO/I describes.

    FSCO/I is real and it is relevant to biology from the molecular nanotech of the cell on up.

    Where there is a debate, is on the design inference on this form of CSI.

    Fine, if you want to debate that do so.

    Don’t try to pretend that FSCO/I is a dubious notion cooked up by some dismissible IDiot we can cyberstalk or stalk on the ground and smear with all sorts of false accusations, year after year, or enable such outrages and pretend that nothing is going on or that it is to object to such that is wrong.

    If you try to debate it, recognise that FSCO/I can be reduced informationally by creating a descriptive language based on addressing the nodes-arcs pattern with a structured string of Y/N q’s. So, we can reasonably assign a value for functionally specific info to it.

    Further, recognise that once you deal with the constraints of multipart interactive, specific function, you are clustering a relatively narrow range of possible configs, as opposed to any arbitrary clumped or scattered arrangement. Thus, if you are trying to suggest arriving at FSCO/I via blind chance and mechanical necessity, you are dealing with blind needle in haystack search. Where when the info content hits or exceeds 500 – 1,000 bits, you are looking at searches of the type in the OP: at 1,000 bits, 1 straw sampled from a haystack that would dwarf the observed cosmos.

    No wonder we have a trillion member inductive sample where FSCO/I reliably comes about by intelligently directed configuration, design.

    If you doubt me, start with the Internet, then go on to nuts, bolts, gear trains etc.

    So, design is an inductively warranted best current causal explanation of FSCO/I.

    WHERE — ABSENT A PRIORI IMPOSITION OF EVOLUTIONARY MATERIALISM — THERE IS NO SERIOUS COMPETING EXPLANATION.

    (My caps lock stuck by accident, but I’ll leave that.)

    That is what is on the table.

  193. 193
    Carpathian says:

    Mung:

    But the candidate solutions are not just randomly generated objects of no particular design.

    In Weasel, random is exactly what they are.

    The mutation component may make one string less acceptable and another more acceptable to the “environment”. This is not a problem since the less acceptable one will not be allowed to reproduce and mutate.

    Despite mutational setbacks, the solution string eventually appears in the “environment” despite the fact that the mutating component does not know what the target string should be.

  194. 194
    Joe says:

    The designer of the program knew what the target was and that knowledge was programmed in. The program was intelligently designed to reach that target. Remove the intelligently programmed knowledge and weasel would still be running unable to find the target.

    Also “random” wrt GAs (like weasel) just means an equal probability of occurring. It does not mean accidental or unplanned. The debate is about the latter, ie random as in accidental or unplanned.

  195. 195
    Carpathian says:

    Joe:

    The designer of the program included all components of Weasel into one program but nothing stops someone from removing the target string and passing it as a parameter from a console or a client on the web.

    The mutation component does not know what the target string is and thus does not know which character should be modified.

    It modifies any character without regard to the target.

  196. 196
    Joe says:

    The mutation component does not know what the target string is and thus does not know which character should be modified.

    You say that as if it means something. The mutation component is always actively guided towards a solution. That is the very important part. The mutations happen when programmed to happen and then the variants guided as programmed. Selection actively guides the variants.

  197. 197
    Carpathian says:

    Joe:

    Selection actively guides the variants.

    Very true.

    The mutation component however, does not know what the target string is.

    During a run, a character matching the target string sometimes gets turned into a non-matching character.

    This shows that mutation is completely random.

    Your are looking at Dawkins program as one single unit, but it isn’t.

    Programmers sometimes get slapped on the wrist for putting too much functionality in one file when it really should be spread out across multiple components.

    Dawkin’s Weasel incorporates all functionality into one file but that was simply a case of quickly getting it done.

    There should have been three separate components:
    Reproduction, Mutation and Environmental Acceptance.

  198. 198
    Joe says:

    The mutation component however, does not know what the target string is.

    All knowledge comes from the programmer. The mutation component is part of the program.

    This shows that mutation is completely random.

    Yes, as in equally probable. I said that already.

    Your are looking at Dawkins program as one single unit, but it isn’t.

    I am not doing that so lay off your false accusations.

    There should have been three separate components:
    Reproduction, Mutation and Environmental Acceptance.

    Reproduction is by intelligent Design. Mutation is part of the program and it is artificial selection, not environmental acceptance.

    Selection actively guiding the variants = Intelligent Design.

  199. 199
    Carpathian says:

    Joe:

    All knowledge comes from the programmer. The mutation component is part of the program.

    There is no point in debating this as you could simply look at the program and see that the functions are separate and could be in three different modules.

    e.g. Mutate( char *Population[]);

    Reproduction( char *Population[]);

    char *Environment( char *Target, char *Population[]);

    Notice that the populations are passed to the modules which means we don’t know what they are.

    Notice also that we don’t know how Environment rates the Population. It might compare the strings as Dawkin’s did or splice the “DNA snippets” into frogs and report the string that caused the frog’s skin to turn the bluest.

    The modules don’t know what is going on outside of them.

    They simply do their specific job.

  200. 200
    Joe says:

    The steering component of my car doesn’t know where it is going. It does’t know where my car is going. The car doesn’t know either- no matter how many times I reproduce the route.

    And yet my car and all of its components are intelligently designed.

  201. 201
    Joe says:

    There is no point in debating this as you could simply look at the program and see that the functions are separate and could be in three different modules.

    e.g. Mutate( char *Population[]);

    Reproduction( char *Population[]);

    char *Environment( char *Target, char *Population[]);

    Reproduction is by intelligent Design.

    Mutation is part of the program with random = equally probable; and

    it is artificial selection, not environmental acceptance.

    Selection actively guiding the variants = Intelligent Design.

    Computers emulate people. They can just do it faster. Everything they do traces back to humans. There isn’t anything about evolutionary and genetic algorithms that simulates natural selection.

    And one part Intelligent Design Evolution is what they model.

  202. 202
    sparc says:

    That is what is on the table.

    Too bad that obviously Dembski, Behe and Meyer are sitting at another one and don’t even look over to see what’s on your table.

  203. 203
    Mung says:

    And the red herrings continue.

    Let’s say that the program mutates using the uppercase characters A-Z and the space character but the target is all lower case characters. good luck with that.

    Let’s say that the length of the candidate solutions is 28 characters and I pass in a string of 128 characters [via the command line to this magical program] as the target. good luck with that.

    I would like to see Carpathian code this wonderful weasel that can find any target.

  204. 204
    Mung says:

    Yes, Carpathian, any half-wit coder can string together a bunch of modules that don’t work. But that’s hardly the point.

  205. 205
    Carpathian says:

    Mung:

    I would like to see Carpathian code this wonderful weasel that can find any target.

    Wonderful Weasel cannot find any target anymore than a chicken egg could contain any bird species.

  206. 206
    Zachriel says:

    Mung: But the candidate solutions are not just randomly generated objects of no particular design.

    They are random mutations of (roughly) “the best so far”.

    Mung: Let’s say that the program mutates using the uppercase characters A-Z and the space character but the target is all lower case characters. good luck with that.

    That’s actually not a problem. It would only decrease the effectiveness somewhat, but it would still be far faster than searching by random trial.

    Mung: Let’s say that the length of the candidate solutions is 28 characters and I pass in a string of 128 characters [via the command line to this magical program] as the target. good luck with that.

    Again, not a problem, as long as mutation includes deletions and insertions.

  207. 207
    kairosfocus says:

    sparc, you show no evidence of even seriously reading what is on the thread right in front of you. That unresponsiveness speaks sad volumes. KF

  208. 208
    Mung says:

    Mung: I would like to see Carpathian code this wonderful weasel that can find any target.

    Carpathian: Wonderful Weasel cannot find any target anymore than a chicken egg could contain any bird species.

    I know it. And you know it. Yet you claimed:

    The designer of the program included all components of Weasel into one program but nothing stops someone from removing the target string and passing it as a parameter from a console or a client on the web.

    But what would be the point of that, given that the program must be provided with information about the nature of the string. And if you mean they could manually enter the exact same string that the program had been designed to find and the program would still find it, that merits a big so what, lol.

    I thought you had something more interesting to say. Silly me.

    So if someone enter’s “methinks it is like a weasel” on the command line the search would still find the target?

    What if they entered: “methinks it is a cloud with the appearance of a Mustela nivalis”?

    What then?

    In how many places would your modular program need to be changed?

  209. 209
    Zachriel says:

    Mung: And if you mean they could manually enter the exact same string that the program had been designed to find and the program would still find it, that merits a big so what, lol.

    It’s not merely that it finds it, but that it finds it much more rapidly than random trial.

  210. 210
    Mung says:

    Zachriel: It’s not merely that it finds it, but that it finds it much more rapidly than random trial.

    Yes, a search guided by intelligent input will find a target more rapidly than a search not guided by intelligent input. I didn’t realize that was even a point of dispute.

  211. 211
    Zachriel says:

    Mung: Yes, a search guided by intelligent input will find a target more rapidly than a search not guided by intelligent input.

    If you mean that evolutionary processes are ‘intelligent’, then sure.

  212. 212
    Joe says:

    Yes, Zachriel, with Intelligent Design Evolution the evolutionary processes would be intelligent. With unguided evolution the evolutionary processes are blind and mindless, ie not intelligent.

  213. 213
    Carpathian says:

    Mung:

    What if they entered: “methinks it is a cloud with the appearance of a Mustela nivalis”?

    It would find that string or any you enter.

    You could type in this one for example:
    “ACTGGCTGCATTCATCCCAATGAGGATC”

    In how many places would your modular program need to be changed?

    None.

  214. 214
    Mung says:

    Mung: Yes, a search guided by intelligent input will find a target more rapidly than a search not guided by intelligent input.

    Zachriel: “If you mean that evolutionary processes are ‘intelligent’, then sure.”

    If you mean that you can write a program that finds a target with greater probability than blind search then yes, that is an intelligently guided search.

  215. 215
    Mung says:

    Carpathian, let us see the program.

    Now you’re back to claiming the Wonderful Weasel program can find any target after previously saying that it cannot.

  216. 216
    Carpathian says:

    Mung:
    It can find any ASCII string. That’s not the same as any target.

    As far as seeing the program, any programmer could write it.

    Here are the functions:
    Mutate( char *Population[]);

    Reproduction( char *Population[]);

    char *Environment( char *Target, char *Population[]);

    Notice that the target and population are parameters supplied from outside the function.

  217. 217
    Mung says:

    Carpathian: It can find any ASCII string. That’s not the same as any target.

    What “it” are you talking about?

    There are strings that are not ASCII strings. And Dawkins weasel program doesn’t even accept all ASCII strings. It is limited to uppercase A-Z and the space character.

    Carpathia: Notice that the target and population are parameters supplied from outside the function.

    So? That’s not the point.

    Let me try again.

    You take the Weasel program of Dawkins and modularize it, as you suggest.

    Now you want it to work with any ASCII string and not just the limited subset it currently works with.

    Which module or modules would you need to change?

  218. 218
    Zachriel says:

    Mung: If you mean that you can write a program that finds a target with greater probability than blind search then yes, that is an intelligently guided search.

    More particularly a program that utilizes random mutation and selection.

  219. 219
    Mung says:

    Zachriel: More particularly a program that utilizes random mutation and selection.

    Yes. Guided. by. Intelligent. Design.

    Zachriel seems to think that if you can make a “random” change to something that it follows that there’s no intelligent design involved.

    Zachriel seems to think that if you can “select” something that it follows that there’s no intelligent design involved.

  220. 220
    Zachriel says:

    Mung: Yes. Guided. by. Intelligent. Design.

    Biological evolution is posited to work by random mutation and selection. An evolutionary algorithm emulates this process showing the capability of such a system. If you consider it “intelligent”, the intelligence is a property of the process, not an external agent.

    Mung: Zachriel seems to think that if you can “select” something that it follows that there’s no intelligent design involved.

    Selection in biology is due to the natural and observable relationship of the population and the environment, again without an external agent.

  221. 221
    rhampton7 says:

    Mung,
    A pair of dice are intelligently designed, but the outcome is random. Likewise Zachriel’s program. And it follows that if the Universe is analogous to a program that generates ‘random’ results, then you have adopted a TE position on Creation: God knows how randomness (and free will) will unfold and need only intervene for individual revelation/salvation. I’m quite happy with that, but it’s not warmly embraced by most ID proponents.

  222. 222
    Mung says:

    rhampton7: A pair of dice are intelligently designed, but the outcome is random.

    If by random you mean equally likely then you’ve not said anything particularly interesting. The outcome is equally likely because the dice are designed. That’s never been in dispute.

  223. 223
    Joe says:

    Biological evolution is posited to work by random mutation and selection

    That is incorrect. Biological evolution is posited to work by random, as in accidental, mutation and elimination.

    An evolutionary algorithm emulates this process showing the capability of such a system.

    It emulates SELECTION, ie artificial selection. Selecting the top 1% is much different from eliminating the bottom 1%.

    Selection in biology is due to the natural and observable relationship of the population and the environment, again without an external agent.

    Natural selection is a process of elimination. You are cowardly in your attempt to obfuscate.

    Why are evolutionists such dishonest people?

    Lizzie does it. Petrushka is a master at it. Alan Fox does it. keiths is just pathetic. Zachriel is their clown/ jester.

    Very sad indeed.

  224. 224
    Joe says:

    Clueless and dishonest evos think nature selects because the word “selection” is part of natural selection.

    From “What Evolution Is”, by Ernst Mayr, page 117:

    What Darwin called natural selection is actually a process of elimination.

    Page 118:

    Do selection and elimination differ in their evolutionary consequences? This question never seems to have been raised in the evolutionary literature. A process of selection would have a concrete objective, the determination of the “best” or “fittest” phenotype. Only a relatively few individuals in a given generation would qualify and survive the selection procedure. That small sample would be only to be able to preserve only a small amount of the whole variance of the parent population. Such survival selection would be highly restrained.

    By contrast, mere elimination of the less fit might permit the survival of a rather large number of individuals because they have no obvious deficiencies in fitness. Such a large sample would provide, for instance, the needed material for the exercise of sexual selection. This also explains why survival is so uneven from season to season. The percentage of the less fit would depend on the severity of each year’s environmental conditions.

    It’s funny watching evos continue to misunderstand the very idea they are supposed to be defending.

  225. 225
    mike1962 says:

    Me: So then, is all the code and data except the target string part of the replicators’ set of properties?

    Zachriel: Here’s the usual steps: * Evaluate the individual fitness of each member of population; * Determine best fit individuals to become parents; * Breed new individuals through mutation and crossover to give birth to offspring;

    Non-responsive.

    I’ll ask again: is all the code and data except the target string part of the replicators’ set of properties?

  226. 226
    Zachriel says:

    Mung: The outcome is equally likely because the dice are designed.

    Many things in nature exhibit randomness.

    mike1962: is all the code and data except the target string part of the replicators’ set of properties?

    The genotypic traits of the members of the population are the individual sequences. The fitness landscape is separate. There is also the relationship between the members of the population and the fitness landscape.

  227. 227
    mike1962 says:

    Zachriel: The genotypic traits of the members of the population are the individual sequences. The fitness landscape is separate. There is also the relationship between the members of the population and the fitness landscape.

    Still non-responsive.

    Again: is all the code and data except the target string part of the replicators’ set of properties?

  228. 228
    Mung says:

    Zachriel’s started a second band: Masters of Non-Sequiturs.

  229. 229
    Carpathian says:

    Joe:

    Clueless and dishonest evos think nature selects because the word “selection” is part of natural selection.

    It’s funny watching evos continue to misunderstand the very idea they are supposed to be defending.

    The term “selection” is an analogy. There is no agent actually “selecting” either good or bad performance in an environment.

    If you believe that the analogy represents reality, you don’t understand what evos mean by the term evolution.

  230. 230
    Joe says:

    Carpathian:

    The term “selection” is an analogy.

    No, the word “selection” is wrong and misleading. Obviously you didn’t understand what Mayr said. Perhaps you need a better education.

    Evos don’t understand what they mean by the word “evolution”.

  231. 231
    Carpathian says:

    Joe:

    Since “evolution” is an “evos” word, it’s highly likely that “evos” know what it means.

    It’s probably a given that the French know what words in their own language mean.

  232. 232
    Joe says:

    Carpathian, evos misrepresent evolution on a daily basis. They think that because “selection” is part of natural selection that actual selecting takes place. Evos are nothing but fabricators and equivocators. Dishonest to the core.

  233. 233
    Carpathian says:

    Joe:

    Carpathian, evos misrepresent evolution on a daily basis. They think that because “selection” is part of natural selection that actual selecting takes place. Evos are nothing but fabricators and equivocators. Dishonest to the core.

    You are right that the term “selection” is a case of eliminating organisms that cannot survive in their environment but that is true by definition.

    There is nothing wrong with saying “selected for reproduction” and looking at it from a positive viewpoint.

    It is simply a different way of expressing something which makes it easier to work with.

    No one would think of having a podium for the 20 or so Olympic athletes who lose a certain event though that is a very accurate portrayal of what happens.

    No one would think of removing negative numbers from math simply because they don’t exist. They are used because they make describing or solving a product easier.

    In no case however, has anyone implied that there is a judge of evolution “selecting” anything, positively or negatively.

  234. 234
    Joe says:

    OK Carpathian, obviously you have no idea what you are talking about nor what Mayr wrote. The two processes are different. Eliminating the bottom 1% is not the same as selecting the top 1%.

    As I said evos misrepresent evolution on a daily basis and here you are.

  235. 235
    Carpathian says:

    Joe:

    OK Carpathian, obviously you have no idea what you are talking about nor what Mayr wrote. The two processes are different. Eliminating the bottom 1% is not the same as selecting the top 1%.

    You’ve just mispresented what I said.

    I gave you an example of 20 Olympic losers.

    There are only 3 winners.

    3 is not equal to 20.

    If you “select” 20 losers or “select” 3 winners the percentages are not equal.

    I never said they were.

    “Selection” is an analogy regardless of your intent to make it a real methodology you can then dispute.

  236. 236
    Joe says:

    Carpathian:

    You’ve just mispresented what I said.

    You mis(re)present what Mayr said. What he says follows this: The two processes are different. Eliminating the bottom 1% is not the same as selecting the top 1%.

    “Selection” is an analogy regardless of your intent to make it a real methodology you can then dispute.

    It isn’t an analogy. It is wrong and misleading.

    Mayr got it right. Until you can grasp what he said you will continue to get it wrong.

  237. 237
    Carpathian says:

    Joe:

    You mis(re)present what Mayr said. What he says follows this: The two processes are different. Eliminating the bottom 1% is not the same as selecting the top 1%.

    He is right about that.

    You are right that “selection” is elimination.

    I am right that you misunderstand the point I am making.

    Since you misunderstand me, you have misrepresented my position.

    When Mayr refers to a “selection” of the top 1%, he means a “elimination” of 99%.

    When he refers to a “elimination” of 1%, he means a “selection” of 99%.

    He can’t mean anything else since the math would not add up.

    “selection” is an analogy.

  238. 238
    Joe says:

    I understand you. I understand that you misrepresented the entire scenario.

    When Mayr refers to a “selection” of the top 1%, he means a “elimination” of 99%.

    When he refers to a “elimination” of 1%, he means a “selection” of 99%.

    He can’t mean anything else since the math would not add up.

    And the difference between the two are huge. One drives the survivors towards some objective and the other doesn’t.

  239. 239
    Zachriel says:

    mike1962: is all the code and data except the target string part of the replicators’ set of properties?

    As already explained, the code is not part of the replicators’ set of properties. The code is an implementation, not the thing being modeled. The properties of the replicators are represented by the sequence, i.e. traits.

  240. 240
    mike1962 says:

    As already explained, the code is not part of the replicators’ set of properties. The code is an implementation, not the thing being modeled. The properties of the replicators are represented by the sequence, i.e. traits.

    What you are positing is three categorical distinctions: replicators, environment, and algorithms to “implement” them. That’s a cheat because the replicators are not replicators without the code to implement them. Therefore such code is properly understood as necessary properties of the replicators. Any code associated with modifying the environment would be likewise understood as necessary properties of the environment. There is no middle ground. And there is no triple categories situation in nature.

    Therefore, your statement, “That’s not generally a function of the initial population, but of the fitness landscape” is false, since the initial population necessarily has properties in the form of algorithms that govern their evolution. That genomes in nature contain both code and data to replicate themselves changes nothing logically.

  241. 241
    Zachriel says:

    mike1962: What you are positing is three categorical distinctions: replicators, environment, and algorithms to “implement” them.

    The computer code is not part of what is being modeled. When we model weather, the parameters are pressure and moisture plus the physics of their interaction, not the programming language.

    mike1962: That’s a cheat because the replicators are not replicators without the code to implement them.

    There’s no actual rain in a weather simulation. You don’t soak your computer to model a rain storm. You’re conflating the model with what is being modeled.

  242. 242
    mike1962 says:

    Zachriel: The computer code is not part of the model.

    It is an essential element in the implementation, just like molecules and the physics of their interaction are an essential element in implementing real weather.

    When we model weather, the parameters are pressure and moisture plus the physics of their interaction, not the programming language.

    Not the programming language, but certainly the algorithms coded in the programming language.

    The analogy to real weather is merely an analogy. Real weather is “implemented” using molecules. It would be logically absurd to say that the molecules – the means of implementation – are not essential properties of real weather systems. Weather modelling programs do not model weather down to the molecular level, where real weather is implemented.

    In a real genome, the code and data and the processes are the means to replicate(genes, ribosomes) and are self-contained with the genome itself and perform any procedures necessary for replication. For GAs, the computer code is the means by which replications occurs. The computer code is necessarily analogous to the processes contained in a biological replicator. It would be absurd to say that the processes that do the replication are an essential property in natural replicators but not software based replicators. If the software code that makes replicators cannot be considered an essential property of the replicators then whatever objects are being created are not replicators.

  243. 243
    Zachriel says:

    mike1962: just like molecules and the physics of their interaction are an essential element in implementing real weather.

    The physics or chemistry, not the code, if that is what you mean. In an evolutionary model, we might have parameters such as fecundity, variation, and population limits; i.e. the knobs you adjust to test the behavior of the system. If you’re modeling a real-life organism, then you would tune the knobs to the parameters that apply to that organism.

    mike1962: Weather modelling programs do not model weather down to the molecular level, where real weather is implemented.

    That’s right. All models are simplified approximations. In weather simulations, they use pressure and moisture. The limitation isn’t so much molecular behavior, but that weather is a chaotic system. But even though we don’t model every molecule, we can still learn something from the models.

    Science is all about models, whether Newtonian mechanics or ideal gases.

    mike1962: For GAs, the computer code is the means by which replications occurs.

    We don’t have to model every detail to have a useful model. For instance, we can show how populations of deer rise and fall in a chaotic interaction with predators. We might set up the model so that replication depends on energy levels, which is typical in many organisms. Find food and replicate. Don’t find food and keep looking until out of juice.

  244. 244
    Mung says:

    Carpathian:

    There is nothing wrong with saying “selected for reproduction” and looking at it from a positive viewpoint.

    Not if you believe in teleology. Do you?

  245. 245
    Mung says:

    See how Zachriel weaves and dodges, stings like a butterfly, floats like a bee. Good luck pinning him down on anything.

    Zachriel, do you set yourself a goal each day to say some minimum number of irrelevant things on this site?

    Zachriel: We don’t have to model every detail to have a useful model.

    And no one suggested that we do have to.

    Zachriel: That’s right. All models are simplified approximations.

    Who said otherwise?

    Zachriel: The computer code is not part of what is being modeled.

    Who said it was? It’s obviously part of the model. But you won’t even admit that much.

    Zachriel: In an evolutionary model, we might have parameters such as fecundity, variation, and population limits;

    Good luck doing that without code.

    Zachriel: There’s no actual rain in a weather simulation.

    There could be a weather simulation that uses real rain. Think wind tunnels.

    Zachriel: You don’t soak your computer to model a rain storm.

    You might if you wanted to observe the effects so that you could better model them.

    Zachriel: … the code is not part of the replicators’ set of properties.

    It can be.

    In artificial intelligence, genetic programming (GP) is an evolutionary algorithm-based methodology inspired by biological evolution to find computer programs that perform a user-defined task.

    I’m pretty sure Zachriel knows this, which means he was just caught lying through his teeth.

  246. 246
    Zachriel says:

    Mung: do you set yourself a goal each day to say some minimum number of irrelevant things on this site?

    Actually, we try to only post on topic.

    Mung: Who said it was?

    mike1962: What you are positing is three categorical distinctions: replicators, environment, and algorithms to “implement” them.

    Mung: Good luck doing that without code.

    Algorithms are independent of code.

    Zachriel: You don’t soak your computer to model a rain storm.

    Mung: You might if you wanted to observe the effects so that you could better model them.

    Heh.

  247. 247
    Mung says:

    Zachriel: In an evolutionary model, we might have parameters such as fecundity, variation, and population limits

    Mung: Good luck doing that without code.

    Zachriel: Algorithms are independent of code.

    Parameters are not algorithms.

    Is that what you call being on topic? Posting an utterly irrelevant rejoinder is not what I would call posting on topic.

    Zachriel: The computer code is not part of what is being modeled.

    Mung: Who said it was? It’s obviously part of the model.

    Zachriel:

    mike1962: What you are positing is three categorical distinctions: replicators, environment, and algorithms to “implement” them.

    I missed where he wrote “computer code.”

    Maybe you are getting to it and maybe you just missed it, but just in case:

    Zachriel: …the code is not part of the replicators’ set of properties.

    Mung: It can be.

    In artificial intelligence, genetic programming (GP) is an evolutionary algorithm-based methodology inspired by biological evolution to find computer programs that perform a user-defined task.

    If it’s not the code that is evolving, what is?

    Maybe you just didn’t know. But somehow I doubt that. Maybe you just forgot for a bit.

  248. 248
    Mung says:

    mike1962, if you have coding experience, demand code. Require specific examples. It quickly exposes the bankruptcy of the critics.

  249. 249
    Mung says:

    Zachriel:

    biological evolution is a specific ‘search algorithm’, not the universal set of search algorithms; and the natural environment is a specific ‘fitness landscape’, not the universal set of fitness landscapes.

    Not true.

  250. 250
    Mung says:

    mike1962, and if you have the will for it, go back to previous posts. You’ll often find the critics contradicting themselves or failing to respond to pertinent questions, hoping it will go unnoticed.

  251. 251
    mike1962 says:

    All of your irrelevancies aside…

    Zachriel: Algorithms are independent of code.

    Not in a computer program. No code, no algorithms.

    Zachriel: …the code is not part of the replicators’ set of properties.

    Of course it is. That’s the part of the replicator that actually performs the replicating.

    Show me some source code of a program, where the replicator replicates without any algorithmic code to implement its replications. If the so-called replicator itself is not performing the replications, it’s not a replicator. You’ll have to call it something else.

  252. 252
    Steve says:

    Thank you mike1962 and Mung for taking the time and effort to tackle Zachriel’s (the many person poster) tedious pedantic tomfoolery.

    It believes as long as it replies, it survives to tweet another day.

    You SIRS are valuable assets to the logical and rational blog world.

    I (and I’m sure many other onlookers) enjoy cyber-watching the dismantling of non-sensical evo-devo storytelling.

  253. 253
    Zachriel says:

    mike1962: Not in a computer program. No code, no algorithms.

    That’s right. A computer program is an implementation of an algorithm.

    mike1962: Of course it is. That’s the part of the replicator that actually performs the replicating.

    Yes, and in a weather simulation, it’s the code that ‘moves’ the air around. You’re still confusing the model with the thing being modeled.

    mike1962: If the so-called replicator itself is not performing the replications, it’s not a replicator.

    In an evolutionary algorithm, elements of the population have the innate ability to replicate. That ability is due to what is called fitness, which is determined by the fitness landscape.

    If you prefer, you may call them individuals, elements of the population, candidate solutions, or Fred and Wilma.

    Mung: Good luck doing that without code.

    In the old days, they used paper and pencil; for example, see Fisher, The Genetical Theory of Natural Selection, Oxford University Press 1930.

  254. 254
    Carpathian says:

    Mung, mike1962:

    Zachriel: The computer code is not part of what is being modeled.

    Mung:Who said it was? It’s obviously part of the model. But you won’t even admit that much.

    The computer code is not a part of the thing being modeled any more than a meter is a part of the battery whose voltage you are measuring.

    The computer code that performs the simulation or model can be considered to be the lab where the test is being performed.

    All scientists make this distinction.

    You will never see a technician pick up a spark plug with his hand and then put that hand over a flame to see how well the spark plug responds to heat.

    The technician is not a part of the thing being tested.

  255. 255
    Carpathian says:

    Mung:

    mike1962, if you have coding experience, demand code. Require specific examples. It quickly exposes the bankruptcy of the critics.

    It is pointless to ask for code if you can’t understand it.

    I showed you code that demonstrated that a Weasel implementation could be written that was independent of the information required for the search which appeared to be hard for you to understand.

    What chance does someone have who has little software background to understand a demo?

  256. 256
    Joe says:

    Carpathian, We can only simulate that which we understand. That means we cannot simulate biological evolution.

  257. 257
    Carpathian says:

    Joe:

    Carpathian, We can only simulate that which we understand. That means we cannot simulate biological evolution.

    Simulations are used to gain knowledge.

    You make one simulation model and find it doesn’t seem to work well. You then refine it over and over until you get close to seeing its behaviour resembling the actual thing you are modeling.

    Simulation is a tool, not a replacement for the thing being modeled.

  258. 258
    Joe says:

    Nice non-sequitur, Carpathian.

  259. 259
    Carpathian says:

    Nice self-denial Joe.

  260. 260
    Joe says:

    Nice self-delusion, Carp.

  261. 261
    mike1962 says:

    mike1962: Of course it is. That’s the part of the replicator that actually performs the replicating.

    Zachriel: Yes, and in a weather simulation, it’s the code that ‘moves’ the air around. You’re still confusing the model with the thing being modeled.

    We’re not even anywhere near “modeling” yet. One can create a software program that generates “objects” that interact with an “environent” that do not model anything real. For now, I’m talking about replicators. (And, I was talking about Dawkins’s Weasel program, but you never bothered to answer my question.)

    In a software implementation that is actually generating replicators it makes no sense to say that the code that actually does the replicating is not a property of the replicators. Take the code away, and what do you have? Whatever it is, it cannot be a replicator since replicators by definition contain processes that reproduce themselves.

    M: If the so-called replicator itself is not performing the replications, it’s not a replicator.

    Z: In an evolutionary algorithm, elements of the population have the innate ability to replicate. That ability is due to what is called fitness, which is determined by the fitness landscape.

    And it is determined by the processes of the replicators themselves, which you deny. Remember, here’s what we are discussing:

    M: The nature of the initial population (the systems, processes and control information they contain) determines to some extent what kinds of variations are even possible for any putative selection to act on.

    Z: That’s not generally a function of the initial population, but of the fitness landscape.

    Your reply is patently false.

    M: Give me an example with regard to software implemented replicator objects interacting with their environment.

    Z: A simple example is Weasel

    What is a “replicator” object in Weasel?

  262. 262
    mike1962 says:

    Carpathian: The computer code is not a part of the thing being modeled any more than a meter is a part of the battery whose voltage you are measuring.

    It is in the case of software replicators. It makes no sense to say a software object is a “replicator” if it has no processes that can replicate. Those processes are necessarily an essential property of the replicator object. Otherwise it’s not a replicator.

    Replicators by definition contain processes that reproduce themselves.

    Think about it.

  263. 263
    Zachriel says:

    mike1962: One can create a software program that generates “objects” that interact with an “environent” that do not model anything real.

    Quite so. For instance, Weasel is abstracted.

    mike1962: In a software implementation that is actually generating replicators it makes no sense to say that the code that actually does the replicating is not a property of the replicators.

    The code is not what is being modeled. Sequences are modeled with the ability to replicate. That’s why its called an evolutionary algorithm. There’s a number of ways to flesh out the details, such as requiring the acquisition of resources before reproduction.

    mike1962: Take the code away, and what do you have?

    The same you would have if you take away the code for a computer weather simulation. Not much.

    mike1962: Whatever it is, it cannot be a replicator since replicators by definition contain processes that reproduce themselves.

    In an evolutionary algorithm, the sequences are assumed to have the ability to be replicated. You can call them Fred and Wilma, if you prefer.

    mike1962: The nature of the initial population (the systems, processes and control information they contain) determines to some extent what kinds of variations are even possible for any putative selection to act on.

    We misread your statement originally. The available variations are a function of the properties of the sequences as defined by the parameters of the evolutionary algorithm, the ‘chemistry’ of the world, if you prefer. It’s analogous to the rules of how pressure system work in a weather simulation. It’s not a trait of the atmosphere itself, but the physics of the abstracted world.

    Here’s a simplified evolutionary algorithm: Take a sequence of letters. If the sequence spells a word, then assume it can replicate. Allow word sequences to mutate and recombine. And so on.

  264. 264
    Mung says:

    Carpathian: It is pointless to ask for code if you can’t understand it.

    It’s pointeless to write code that doesn’t work.

    Take the following for example:

    int CountMaxCompares(char *Target, char *PopMember)
    {
    int MatchingCount, Position;

    MatchingCount, Position = 0;
    while( Target[Position] && PopMember[Position] )

    There are some issues here.

    1. What if the length of the target string is not the same as the length of the population member string?

    2. How does the code determine that the end of the target string has been reached?

    3. Your while loop will terminate the first time a zero is encountered in either the target or the population member. Is that what you designed it to do or was that just an oversight?

    Maybe you should get your code to compile and run before asking us to.

  265. 265
    Mung says:

    Carpathian:
    What chance does someone have who has little software background to understand a demo?

    Personally, I have been writing code since I learned BASIC in the mid 1970’s. Do the math.

    Maybe you mean mike1962.

    I also know how C allows you to go off into memory locations that have nothing to do with your code. That’s apparently a lesson you haven’t learned yet.

  266. 266
    Mung says:

    Zachriel: If the sequence spells a word, then assume it can replicate.

    Apparently this happens by magic.

    And for the most part, evolutionary algorithms are not models of evolution. So all this bit about us conflating the model with the thing modeled is just so much cow dung. Red herring.

  267. 267
    Mung says:

    Meanwhile, in the real world:

    The primary reason for us to learn about algorithm design is that this discipline gives us the potential to reap huge savings, even to the point of making it possible to do tasks that would otherwise be impossible.

    – Robert Sedgewick, Algorithms in C

    and:

    Careful algorithm design is an extremely effective part of the process of solving a huge problem, whatever the application area.

    – Robert Sedgewick, Algorithms in C

    But evolutionary algorithms don’t require design.

  268. 268
    Zachriel says:

    Mung: Apparently this happens by magic.

    Evolutionary algorithms presuppose reproduction.

    Mung: And for the most part, evolutionary algorithms are not models of evolution.

    Many evolutionary algorithms are not models of *biological* evolution, but some are.

  269. 269
    Carpathian says:

    Mung:

    1. What if the length of the target string is not the same as the length of the population member string?

    2. How does the code determine that the end of the target string has been reached?

    3. Your while loop will terminate the first time a zero is encountered in either the target or the population member. Is that what you designed it to do or was that just an oversight?

    1. It terminates the compare and returns the count of chars that match up to that point. This is exactly what it should do.

    2. In C, a zero terminates a string while some languages prepend the array of chars with a length count. The end of the target and population string is therefore determined by encountering a zero. This is exactly what the code does.

    3. This exactly what it should do.

    Maybe you should get your code to compile and run before asking us to.

    You asked me to write this function because you didn’t understand how to do it yourself. If you had understood how to write the function, you never would have asked me to do it.

    I said anyone that feels like it could try and write the entire Weasel program. I had no intention of writing any code until you asked me to.

    Unlike you, I have an understanding of how to put together a Weasel program and don’t need to actually do it to understand that the Weasel code doesn’t know what the target is, which was the whole point of the exercise.

    My total investment in time writing the code was less than writing this response to your questions so while I expected some bugs, it turns out you haven’t shown me any yet.

  270. 270
    Carpathian says:

    Mung:

    I also know how C allows you to go off into memory locations that have nothing to do with your code. That’s apparently a lesson you haven’t learned yet.

    Show me where I did that.

  271. 271
    Carpathian says:

    mike1962:

    It is in the case of software replicators. It makes no sense to say a software object is a “replicator” if it has no processes that can replicate. Those processes are necessarily an essential property of the replicator object. Otherwise it’s not a replicator.

    Replicators by definition contain processes that reproduce themselves.

    Think about it.

    I intend to do what you suggest.

    The replicator code will be a part of the software being mutated.

  272. 272
    Mung says:

    Carpathian: You asked me to write this function because you didn’t understand how to do it yourself.

    No, I asked for it because your code referred to it but lacked an implementation of it. And lacking any implementation, I could not validate whether your function could do what you claimed.

    Carpathian: I had no intention of writing any code until you asked me to.

    And there’s the rub. All sort of claims are made, but lacking code, we can’t put them to the test.

    You were making claims about what could be implemented in code. I felt the best way to settle the dispute was to have you present the code that was capable of doing what you claimed. I honestly don’t think I was being unreasonable.

    Carpathian: If you had understood how to write the function, you never would have asked me to do it.

    And that’s just wrong.

    Carpathian: My total investment in time writing the code was less than writing this response to your questions so while I expected some bugs, it turns out you haven’t shown me any yet.

    And I haven’t seen your source code, so that’s a pretty empty boast.

Leave a Reply