Uncommon Descent Contest Question 10: Provide the Code for Dawkins’ WEASEL Program

_{Denyse O'Leary

August 26, 2009

Darwinism}

Share: Facebook; Twitter; LinkedIn; Flipboard; Print; Email

Special invitation for Richard Dawkins – but any civil person is entitled to enter.

There’s been some discussion here and elsewhere whether the the recent IEEE article by Dembski and Marks correctly characterizes Richard Dawkins’ famous METHINKS IT IS LIKE A WEASEL program.

Does the program ratchet correct letters or does it let them vary? One is a partitioned or stair-step search, the other a more realistic evolutionary search. From The Blind Watchmaker, where Dawkins describes the program, its performance suggests that it could be either of these options (though he doesn’t say).

On the other hand, from a (video-run of the program , go to 6:15), it seems to be the latter.

It’s easy enough to settle this question: Make the code for the program public. Perhaps Richard Dawkins himself or his friends at RichardDawkins.net can finally provide this code (apparently a program written in BASIC).

The prize is a copy of either Stephen Meyer’s new Signature in the Cell or Richard Dawkins’ soon-to-be-out The Greatest Show on Earth.

Should the winner choose the latter, I will ask Dawkins’s publicist to mail the copy. Given that at his site, he calls himself “the most formidable intellect in public discourse,” I would assume that if he signs the copy, it will be worth millions.

But wait. Let’s see that code first.

Comments

kf whines Worse, this is what I actually said about Mr Hoyle (and note the part that was conveniently omitted from the cite without even a ellipsis to warn the reader):
Remember you are in the presence of a Nobel Prize equivalent holder here. he may be wrong on points, but he is not going to be making simplistic blunders, and even his errors will be instructive. That is my experience from many years of running across his work, on a wide array of topics, starting with the Steady State universe Hypothesis. (He it is who gave the name “big bang” to the cosmogenetic theory of that name, but he did not intend it to be a positive term.)
In short, just the opposite to blind adherence to authority, I am saying that you need to treat competent workers with respect and highlighted how a major error of Sir Fred Hoyle – the Steady State theory – was highly instructive! Worse, this was suppressed and distorted by use of half-truthful selective citation in 271, which HAD to be deliberate. This is utterly inexcusable.
kf is correct that I omitted the text that he bolded. But since I did not quote anything AFTER the Nobel Prize line, ellipsis seems unneccessary. I might as well fault kf for omitting the preceding sentence “And please resist the red herrings and strawmen in the usual rebuttals of Hoyle a la Wikipedia.” when he quoted me quoting him. What, no ellipsis, kairos?? I omitted the sentence beginning “That is my experience…” because I was trying to avoid any implication that the “authority” kf was arguing from was himself, rather than Sir Fred. If I had been trying to quote-mine, I would have omitted the phrase “and even his errors will be instructive”, surely? In closing, I would like to thank kf for demonstrating that Weasel performs much better than a random search, and that mutation and selection can, cumulatively, produce results that appear directed/designed. That was, after all, CRD’s only point. I would also like to thank him for highlighting the sloppy writing seen in section E of the D&M paper. If someone as smart as kf did not notice the elision from Partitioned to Deterministic searches on pages 1055 – 1056, there is little hope that typical readers will be able to spot that eqn22 CANNOT describe a search that has generational champions, and therefore it CANNOT describe Weasel (irrespective of pseudo-quasi-implicit-latching).DNA_Jock_{September 2, 2009
September
09
Sep
2
02
2009
10:46 AM
10
10
46
AM
PDT}

kf waffles on
3] DNA-J, 269: for partitioned search without sample replacement, the target can be found in N-1 queries (p1056) irrespective of the target length.
The basic problem? Here is the M & D analysis of partitioned search in the IMMEDIATE context of Weasel, p. 1055: Assuming uniformity, the probability of successfully identi-fying a specified letter with sample replacement at least once in Q queries is 1 – (1 – 1/N)Q, and the probability of identifying all L characters in Q queries is q = (1 – (1 ? (1/N))^Q)^L . (22) Red herring led out to strawman again.
I understand where you may have gotten confused, kariosfocus: D&M’s article is poorly written. Page 1055 describes the mathematics of a partitioned search with sample replacement, but the authors then slip (without indicating to the reader that they have switched algorithms in midstream) into describing in words the behavior of a partitioned search without sample replacement on p 1056, the page I cited. The misleading paragraph begins on p1055 “The enormous amount of active information provided by partitioned search is transparently evident when the alphabet is binary. Then, independent of L, convergence can always be performed in two steps.” Err, only without sample replacement. They go on to spend a whole paragraph discussing a partitioned search without sample replacement. (For onlookers who are playing with EIL WeaselWare, this second search algorithm is mathematically the same as the so-called “Deterministic Search”, and it performs much, much better than the “Partitioned Search”, which reduces its “Active Information” by deliberately ignoring useful information provided by the oracle.DNA_Jock_{September 2, 2009
September
09
Sep
2
02
2009
10:25 AM
10
10
25
AM
PDT}

4,502 word salad. Yum. Is this a record, I wonder?
kf waffles
1] DNA-J, 265: …even if there were an explicit latching mechanism in Weasel[and all the evidence indicates that there is not one], that would not make it a partitioned search as described by D&M. The first reported Weasel run in TBW CANNOT be the result of a partitioned search. Only two letters change in generation 2.
The essential point in partitioning, according to M & D [p. 1055], of course – from their example — is that “Two of the letters {E, S} are in the correct position . . . In partitioned search, our search for these letters is finished . . . “ (That is, a partitioned search is one that identifies correct letters individually and by one mechanism or another preserves them from reversion, so that he progress of the run of generational champions is ratcheted, with the successful letters being latched. Latching of coruse means: “To close or lock with or as if with a latch.”)
Oh dear. We can agree that partitioned search, as accurately described by D&M in eqn22, latches. Unfortunately, you are using this fact to conclude that a (hypothetical) latching search is a partitioned search. All partitioned searches latch. Kf’s imagined Weasel latches. Therefore kf’s imagined Weasel is a partitioned search All ostriches have two legs Kf has two legs Therefore Kf is an ostrich. I don’t think so.DNA_Jock_{September 2, 2009
September
09
Sep
2
02
2009
10:22 AM
10
10
22
AM
PDT}

BillB: The game just changed. After the red herring, strawman and ad hominem fest overnight, you have proved that you were a harbinger. What is indicated now is apology and correction on your part; and that of several others. On basic civility. Absent that, discussion is over. G'day. GEM of TKI PS: Onlookers Point 5 above answers to anything on the serious merits BillB might otherwise have had. (And of course the just above from him concludes with yet another turnabout false accusation. I think astute onlookers can easily enough see that others, Joseph and I have taken a lot of time and effort to answer to genuine issues on the merits, ever since December last. To slander us as distracting and distortion to try to reduce us to immoral equivalency to those whose sleaziness is revealed above through documented misdirection, misrepresentation and mischaracterisation, is slander, willful and malicious, inexcusable slander. Period. [Save, that this last slander tells us a lot of what we need to know about the real balance of the case on its merits.])kairosfocus_{September 2, 2009
September
09
Sep
2
02
2009
06:36 AM
6
06
36
AM
PDT}

The issue here, at its core, is simple. Is the algorithm that Dembski and Marks describe the same as the one Dawkins describes. The answer, clearly: It is not. You have consistently failed to deal with this simple issue, resorting instead to verbal gymnastics to try and make it look like all the obvious differences between the two are not really differences, or are irrelevant to the point, and to making derogatory comments about the motivations of those who are questioning your dubious reasoning. Your profound inability to deal with these simple and straightforward issues is shocking, as is your contempt for academic standards. I see no point it continuing with this, you are beyond the grasp of reason, logic and evidence. Goodbye.BillB_{September 2, 2009
September
09
Sep
2
02
2009
06:15 AM
6
06
15
AM
PDT}

It seems pretty trivial to focus on getting the exact code that Dawkins wrote, it's a throw-away example - that any ape with a keyboard can code up in an hour. I wrote it in C++ and it works without locking. To focus on the exact code that Dawkins used seems trivial, it serves no purpose other than to attack the man himself. The important thing is whether the code works and any ape with a keyboard and compiler can demonstratively show it to work.Kel_{September 2, 2009
September
09
Sep
2
02
2009
04:12 AM
4
04
12
AM
PDT}

8] Gaz, 272: KF and Joseph, being creatures of Faith, clearly need there to be latching, even if it doesn’t exist, because certain prophets say it exists in the original Weasel. This commenter should do his homework first: it is Joseph and I who have been showing the credible presence of latching action in Weasel c 1986 as can be deduced from the o/ps and discussion at that time. We have never appealed to any authority beyond Dawkins to show that – and that only to show what was showcased and what was said about it. Gaz needs to first show that there is a better explanation of the evidence c 1986, other than the “don't believe yer lyin' eyez” type of argument. As to the sneer on “faith” the only faith we have used here is the faith that an inference to best explanation is a legitimate piece of empirical investigation. 9] NM, 275: I asked about the differences between the mechanism of natural selection and its first-order approximation in WEASEL. You answered by talking about Hoyle’s views on the origin of life. Misrepresentation based on selective summary. The key question I answered was yrs in 261, as I bolded in 264: What’s the conceptual difference between “proximity to target” and “fitness to environment”? Such a question NATURALLY raises the issue of the biological and pre-biotic realities and challenges of information generation, and I referred all the way back to Wistar 1966, where Schutzenberger et al challenged the Darwinist elites on the issue of plausibly getting to complex information-rich functionality by the mechanisms they champion. Hoyle picked up the same theme,a nd it continues unanswered to this day, misdirections such as Weasel notwithstanding. I then specifically addressed the difference between proximity and complex function:
3 --> . . . until one has spontaneously created novel complex information that functions through similarly sophisticated but spontaneously originated machinery, at a realistic level, one has no right to pretend that a targetted process that uses a hotter-colder distance-to target comparison of “mutant nonsense phrases” — i.e confessedly NON-FUNCTIONAL ones — is a good analogy of the power of random variation and natural selection, which is premised on the origin of information by chance variation, and the selection of FUNCTION on superiority of some sub-populations. 4 –> that was the heart of Sir Fred’s challenge and it is the heart of the design theory challenge to the claimed spontaneous origins of life and of novel body plans.
So, the distraction on long answers etc was a misdirection form your own failures of selective half-truthful citation and/or summary, and failure to address the matter cogently on the merits, other than by denial in the teeth of patent facts [cf 11 below]. Onlookers, the pattern is painfully clear. 10] Creationists love to harp on about Darwin’s “racism” as though [if it were true] it invalidates everything else he has to say. But they don’t mention much about the astonishing sexism in some of Hoyle’s novels. In context, this is an instance of the “Design theory = Creationism in a cheap tuxedo promoting a hidden theocratic tyrannical agenda” slander by false association and falser accusation. Worse, the issue with Darwin and his racist social darwinism from say Ch 6 of his Descent of Man, is that it documents a start-pint for a HISTORICAL chain of causes,t hat ended up with eugenics, euthansaia of “undesirables,” and genocide. That is a part of the Darwin legacy that needs to be honestly faced, but is often dodged as just shown. And, Darwin's scientific legacy stands or falls on its own merits. Darwinian mechanisms may account for some cases of micro-evolution, but – as for instance Behe's empirical edge of evolution illustrates and the sudden appearances of phyla in the Cambrian fossil strata reveal -- they have never adequately accounted for the macro-evolutionary changes required to turn a worm or primitive fish into us in about 600 mn years. As to Hoyle and novels (and claimed “sexism” therein), I have never hitherto heard of him as a novellist; I only have known of him as a major astrophysicist who keeps turning up in a lot of areas I have had interests in, and as a holder of a Nobel-equivalent prize. Prejudice based on sex is wrong, but respecting the differences of nature reflected in maleness and femaleness is right. (E.g. gentlemen use their generally superior physical strength to PROTECT women, and let ladies go first, save when the lion is coming.) 11] The distance of a given letter sequence from the target in WEASEL is a model of a gene’s fitness to its environment. There fore it is “a good analogy of the power of random variation and natural selection, which is premised on the origin of information by chance variation, and the selection of FUNCTION on superiority of some sub-populations,” because that distance is the way we happen to modeling functionality. A model is of course an analogy. But in this case, selection of generational champions on mere increments in proximity to target of “mutant NONSENSE phrases” is plainly dis-analogous to a competition on FUNCTION that leads to domination of an overall population by “favoured races.” [Onlookers, this last is from the subtitle of Origin, from its earlier edns.] Worse, Dawkins admitted as much, as I already cited from BW:
Although the monkey/Shakespeare model is useful for explaining the distinction between single-step selection and cumulative selection, it is misleading in important ways. One of these is that, in each generation of selective 'breeding', the mutant 'progeny' phrases were judged according to the criterion of resemblance to a distant ideal target, the phrase METHINKS IT IS LIKE A WEASEL. Life isn't like that. Evolution has no long-term goal. There is no long-distance target, no final perfection to serve as a criterion for selection . . . In real life, the criterion for selection is always short-term, either simple survival or, more generally, reproductive success.
Of course, survival and a living organism and reproduction embed vast levels of information rich functionally specific complex information, starting in the 100's of thousands of bits worth of DNA storage; for observed life. And, as I pointed out already, this is vastly beyond the 1,000 bit threshold where the entire resources of the observable universe could not credibly search more then 1 in 10^150 of the config space. That is, random search for required initial functionality-driving information is not a credible mechanism. And, until you achieve life function and reproduction – including the von Neumann replicator with blue print, code, reader and effectors – you cannot compete on relative function. Similarly, until a functional body plan exists, the level of evolution required to account for the 3 dozen or so phyla and sub-phyla in the Cambrian is not credible, esp. as we are now dealing with increments of 10's – 100's of m bits of required code based functional information. 12] the first letter of sequence corresponds to length of the neck of the virtual creature . . . Until you have a credible basis to account for the origin of the relevant information to build a neck, you cannot use this. 13] KF will probably regale you with his conspiracy theory about Lewontin and how evolution is all a materialist plot that depends on a material origin of life, which means OOL is relevant to everything, apparently. I suspect that if a step by step origin of life is ever established, KF will most likely say that it begs the question of where matter came from, and that that was the issue all along, and by concentrating on OOL we were simply distracting with clouds of oil of ad hominem burning straw men meant to poison and confuse … Onlookers, simply take a glance at my always linked. You will see that I start from the issue of information origination, then apply it in sequence to OOL, origin of body plan level biodiversity, and the issue of the fine-tuned functional complexity of the cosmos. All, on empirical evidence and reasoned argument linked thereto. I then point out -- with relevant evidence adn tellign citations including that from Mr Lewontin that would be so deftly brushed aside [and ignoring the actions of the US National Academy of Sciences that show that in recent years, the attitude in the 1997 quote by a NAS member has become official policy, with demonstrably unjust consequences] -- the impact of the neo-magisterium on origins science and science education, highlighting how the a priori imposition of evolutionary materialism censors science from being the unfettered (but ethically and intellectually responsible) pursuit of the empirically warranted truth about our world. The above, in short, is simply a jaundiced barbed dismissal that ducks addressing a cluster of evidence that seriously challenges the claims of the institutionally dominant materialism of our day. In a context where matters of basic justice are at stake. For shame! And given the pattern of sleazy rhetorical tactics I have had to expose overnight, that speaks volumes on the real balance of the matter on the merits. ______________ All of this is ever so sad. But, to be forewarned is to be forearmed -- if we are wise. GEM of TKIkairosfocus_{September 2, 2009
September
09
Sep
2
02
2009
04:10 AM
4
04
10
AM
PDT}

5] On ratcheted search and implicit latching. Now, the observations to be accounted for are the Weasel runs c 1986 as excerpted and showcased. In these runs, it is strongly evident that with high probability on “good runs” Weasel latched already correct letters for its generational champions and so ratcheted to target. Such behaviour can be achieved explicitly or implicitly, as has been demonstrated. Now, too, when we interpret queries as being in effect mutant phrases, we see that the output to be accounted for is not raw queries, but generational champions. So, we see that Q = G*S, i.e. queries to date are the number of generations so far times size per generation, and we observe too that his generational clustering can be done for BOTH explicitly latched and implicitly latched versions of Weasel. That is, in an explicitly latched Weasel, we can set up a cluster of mutants on the generational seed, and then use the proximity filter to pick the closest to target, reporting it as generational champion. In the case where we have implicit latching, we similarly pick the champion on proximity to target. The difference between the two cases is that in the explicit case the successful letters to date are masked off by direct instructions in the code. By contrast, in an unmasked case, once population per generation, mutation rate per letter and filter are such that there will be a high probability of no-change cases in the generations, and the dominant changed case is the single letter change, the dynamics – BTW another word for “mechanism” [AmH D: 3. An instrument or a process, physical or mental, by which something is done or comes into being ] -- at work will tend to either preserve distance to target so far, or give a single step advance. (If the population size is large enough for a given mutation rate, double mutations may turn up in enough numbers that a reversion and substitution advance may happen: a correct letter reverts, and a previously incorrect one advances, preserving closest so far distance to target. This, too, has been demonstrated and may be important especially in “closing the deal” towards the end of the run. Such a run will tend to show what has been called quasi-latching: ratcheting with occasional slips as commonly happens when the dog on a baitcaster reel is wearing down.) Indeed, the showcased 1986 runs hit target in 40+ and 60+ generations, so no-change wins the generational contest about half the time, and single steps dominated the rest. (Recreations have bracketed this range, some being faster, some slower.) Then we see that the same mathematics of ratcheting applies to both. Just, we must observe that Q jumps in lumps of size G*S, and that ratcheting in the relevant sense attaches not to individual members of the population but the run of generational proximity to target champions. That mathematics of course gives mathematical substance to Mr Dawkins' observation c. 1986 that targetted, proximity-rewarding search massively speeds up run to target relative to unassisted random search. This can be seen in the EIL adjustable Weasel, where unassisted search does ot converge in any reasonable time, but both the explicitly latched and the proximity reward case do. For the latter, once population size, mutation rate and filter are working together “right” implicit latching happens often enough to be reasonably observable. 6] BillB, 270: Latching individual letters is a partitioned search, ratcheting towards a target is only a partitioned search if individual letters are locked out of the search when they reach their target. Agenda-serving question-begging definition imposed after the fact to be contentious. 7] DNA-J, 271: changing the subject to abiogenesis and arguing from the authority of dear departed Fred Hoyle . . . . Soooo, “tornado in a junkyard makes a 747” is an appropriate analogy because an eminent guy once said so and we should “resist the red herrings and strawmen” that might disagree. Of course a glance at 264 in response to 261 gives a very different picture from the above strawman distortion:
take as look at the always linked, starting with the App 7, especially where it discussed the back-story on Weasel:
10 –> Back story: Mr [Fred] Hoyle had raised the issue that the origin of a first life form — such as, roughly, a bacterium — is a matter of such complexity that the odds of that happening in a prebiotic soup by chance was negligible. In a more up to date form, this is the challenge that is still raised by the design theory movement: life shows a threshold of function at about 600,000 bits of DNA storage capacity, so before one may properly apply hill climbing algorithms to scale Mt Improbable, one needs first to have a credible BLIND watchmaker mechanism to land one on the shores of Isle Improbable; e.g. drifting by reasonable random configurations of molecules in empirically justified pre biotic soups and empirically credible pre-life selection forces. (But . . . this OOL challenge is still unmet. And similarly . . . the origin of body plan level biodiversity requiring 10’s – 100’s or millions of bits of functional genetic information, dozens of times over, is equally unmet.) 11 –> So, it looks uncommonly like Weasel distracts attention from and begs the question. That sums up the balance on the main issue . . .
2 –> In short, there is a major question-begging — indeed, dismissal: “single-step selection” — at stake in Weasel as presented [c. 1986]: the origin of complex, functionally specific information. 3 –> that is, until one has spontaneously created novel complex information that functions through similarly sophisticated but spontaneously originated machinery, at a realistic level, one has no right to pretend that a targetted process that uses a hotter-colder distance-to target comparison of “mutant nonsense phrases” — i.e confessedly NON-FUNCTIONAL ones — is a good analogy of the power of random variation and natural selection, which is premised on the origin of information by chance variation, and the selection of FUNCTION on superiority of some sub-populations. 4 –> that was the heart of Sir Fred’s challenge and it is the heart of the design theory challenge to the claimed spontaneous origins of life and of novel body plans . . .
Nor am I citing Hoyle as an authority to be bowed down and worshipped – not least, onlookers, one may look at App 1 the always linked where I have addressed the thermodynamics and information origination issues at what I believe is a useful introductory level, and in my own voice. (Notice how the Darwinists habitually dismiss or pointedly ignore easily accessible material evidence that does not suit their rhetorical agenda.) Worse, this is what I actually said about Mr Hoyle (and note the part that was conveniently omitted from the cite without even a ellipsis to warn the reader):
Remember you are in the presence of a Nobel Prize equivalent holder here. he may be wrong on points, but he is not going to be making simplistic blunders, and even his errors will be instructive. That is my experience from many years of running across his work, on a wide array of topics, starting with the Steady State universe Hypothesis. (He it is who gave the name "big bang" to the cosmogenetic theory of that name, but he did not intend it to be a positive term.)
In short, just the opposite to blind adherence to authority, I am saying that you need to treat competent workers with respect and highlighted how a major error of Sir Fred Hoyle – the Steady State theory – was highly instructive! Worse, this was suppressed and distorted by use of half-truthful selective citation in 271, which HAD to be deliberate. This is utterly inexcusable. [ . . . ]kairosfocus_{September 2, 2009
September
09
Sep
2
02
2009
04:00 AM
4
04
00
AM
PDT}

Mrs O'Leary (and onlookers): I have been wondering why the Darwinists have been so caught up in trying to prove Messrs Demsbski and Marks wrong on Weasel, especially in the context where it is pretty clear from the remarks made by Mr Dawkins, that:
1] Weasel's performance gain over unassisted random walk search is due to the presence of an embedded target, with a warmer-colder proximity based filter that rewards mere proximity of "nonsense" - non-functional -- phrases. 2] Thus, Weasel works by and demonstrates the power of active information, as is being introduced by the cluster of current Marks and Dembski articles from the EIL. 3] Weasel also shows cumulative selection, which was showcased by a the "famous" 1986 runs that showcased how the progress to target in the sampled runs is without reversion. (This is consistent with the basic, plain garden untwisted meaning of "cumulative": Increasing or enlarging by successive addition. A sense that we have every reason to see from his descriptions and examples as the one intended by Mr Dawkins c 1986 in BW; latterday “revisionising” notwithstanding.) 4] two mechanisms have been shown -- yes, SHOWN -- capable of producing runs of generational champions that produce the same effect: (i) ratcheting action based on explicit latching of successful letters to date in Weasel, and (ii) ratcheting action produced by implicit latching due to interaction between selected population sizer per generation, mutation rate per letter and selection filter characteristics. 5] Mr Dawkins has more or less has said that the actual code for Weasel c 1986 is not forthcoming, but the various algorithms and programs currently on the web are good enough to replicate the essential action, whch seems to have been implicit not explicit. 6] Once we see such generation champion ratcheting action, as marks and Dembski analyse in their IEEE paper, p. 1055, an analysis of its implications compared to the performance of a baseline random walk gives us a reasonable measure of the impact of the active information involved in the targetting and warmer-colder proximity mechanisms.
So, why is there so much digital ink being spilled in an attempt to discredit the above? ANS: I think we first need to recognise that over the past decade or so, the standard rhetorical tactic regarding the work of Dr Dembski has been to try to tag him as an ignoramus and bumbler, who does not know what he is talking about and is misleading ill-informed followers. To that end, it has been a routine to twist his terms and analyses, making up ad homiem laced strawmen that have been ignited to cloud, confuse and polarise the atmosphere. Such ruthless, manipulative and misleading rhetoric has been becoming increasingly threadbare in recent years, and the ongoing collaboration with a distinguished Electrical Engineering professor makes the holes in the elbows, the frayed cuffs and the frayed collar quite plain. So, the above plainly reflects an attempt to renew the old rhetorical line and give it new talking points. It is in that context that we need to pick up some "highlights" overnight: 1] DNA-J, 265: …even if there were an explicit latching mechanism in Weasel[and all the evidence indicates that there is not one], that would not make it a partitioned search as described by D&M. The first reported Weasel run in TBW CANNOT be the result of a partitioned search. Only two letters change in generation 2. The essential point in partitioning, according to M & D [p. 1055], of course – from their example -- is that “Two of the letters {E, S} are in the correct position . . . In partitioned search, our search for these letters is finished . . . “ (That is, a partitioned search is one that identifies correct letters individually and by one mechanism or another preserves them from reversion, so that he progress of the run of generational champions is ratcheted, with the successful letters being latched. Latching of coruse means: “To close or lock with or as if with a latch.”) In short, we see here a plain strawman distortion, led off by the direct denial of the evidence from the actual showcased runs of “cumulative selection” in action. For, in the actual case from 1986, in 200+ cases of letters that could revert, none do, AND where we see incorrect letters, they typically persist across at least one decadal sample. This sort of sleazy rhetoric is simply not good enough. 2] DL, 267: Explicit latching makes it a form of partitioned search. Dawkins algorithm does not explicitly latch. This red herring led out to a strawman conveniently omits the other possibility: IMPLICIT latching, which is a main point of the discussion above. And, since we have learned of Mr Dawkins' claims since about 2000, it is implicit latching that has been the inferred best explanation of the behaviour of the showcased runs of Weasel c. 1986. For instance, I again excerpt App 7 my always linked, dating to April:
13 --> Letterwise partitioned search is also a very natural way to understand the Weasel o/p in light of Mr Dawkins' cited remarks about cumulative selection and rewarding the slightest increment to target of mutant nonsense phrases. As such, it has long been and remains a legitimate interpretation of Weasel. However, on recently and indirectly received reports from Mr Dawkins, we are led to understand that he did not in fact explicitly latch the o/p of Weasel, but used a phrasewise search. 14 --> Q: Can that be consistent with an evidently latched o/p? ANS: yes, for IMPLICIT latching is possible as well. 15 --> Namely, (i) the mutation rate per letter acts with (ii) the size of population per generation and (iii) the proximity to target filter to (iv) strongly select for champions that will preserve currently correct letters and/or add new ones, with sufficient probability that we will see a latched o/p. (This effect has in fact been demonstrated through runs of the EIL's recreation of Weasel.)
At minimum, this is culpable, irresponsible negligence of duties of care to truth and fairness; at worst, it reflects blind parroting of outright willful deception by misdirection. (DL: It is all too easy to swallow the Darwinist partyline talking points and fail to check whether they are true and reflect the whole truth. So, a $64,000 Question: if the Darwinist leaders (remember men like Dawkins and Elsberry are involved in this since Dec last) are so unreliable and incompetent or on evidence that is directly accessible – indeed in front of them – why should we feel inclined to trust their reconstructions of a remote, unobserved world in the remote past?) 3] DNA-J, 269: for partitioned search without sample replacement, the target can be found in N-1 queries (p1056) irrespective of the target length. The basic problem? Here is the M & D analysis of partitioned search in the IMMEDIATE context of Weasel, p. 1055:
Assuming uniformity, the probability of successfully identi-fying a specified letter with sample replacement at least once in Q queries is 1 - (1 - 1/N)Q, and the probability of identifying all L characters in Q queries is q = (1 - (1 ? (1/N))^Q)^L . (22)
Red herring led out to strawman again. 4] BillB, 270: The issue of WEASEL’s biological relevance is separate and unrelated to the issue of whether the citation in Dembski and Marks paper is accurate. You have demonstrated that WEASEL does not require an explicit latching mechanism to produce any of the observed results, as we have been saying all along. Given this fact, and Dawkins description, and his statements about latching mechanisms, your claim that the existence of an explicit latching mechanism is a reasonable interpretation is not valid. First, the whole point of Weasel in BW was to create the impression via computer simulation, that complex biological information can be created de novo out of cumulative small random walks and cumulative natural selection. It is therefore highly relevant to observe that the simulation in question used targetted search with the target information preloaded, artificial selection of NON-FUNCITONAL “nonsense phrases” on mere proximity, and so begged the question in ways that Dawkins was forced to admit were “misleading.” (He doubtless relied on the rhetorical difference in impact between a spectacular simulation and the weasel words that qualified it.) Secondly, the reality of implicit latching has been hotly contested by Darwinists, and that has been so right up until I again reproduced actual implicitly latched runs of generational champions. So, the pretence that Darwinists have been saying that Weasel does not require an explicit latching mechanism to produce ratcheted progress to target -- notice, onlookers, the clever omission of just what those “observed results” were – is a brassy distortion of the truth. Worse, as can be seen from 2 above, from March-April on, the implicit mechanism was advanced and documented: this is the false presentation of my longstanding argument as thought it were a latterday concession. And, there is a further ad hominem-laced strawman: I have pointed out that on the evidence of Weasel runs and commentary on cumulative selection c. 1986, explicit latching is a letgitimate interpretation. I have also made the specific distinction that on the CLAIMS of Mr Dawkins we have been willing to accept that Weasel 1986 did not explicitly latch. (And, as already repeatedly noted, videotaped runs c. 1987 are not decisive evidence of the state of Weasel c. 1986.) Of course, absent CREDIBLE WEASEL CODE c. 1986 -- and this is the reason for the contest question 10 – we have no definitive answer beyond all doubt as to the state of Weasel c. 1986 as manifested in the runs published at that time. But, a subtler point lurks, as can be seen from the point 2 above: the Darwinist rhetorical intent is plainly to pretend that the D & M argument only applies to explicitly latched Weasels, so is a “misrepresentation” of Dawkins' only implicitly latched Weasel. (This has in fact been claimed by Darwinists, above.) [ . . . ]kairosfocus_{September 2, 2009
September
09
Sep
2
02
2009
03:58 AM
3
03
58
AM
PDT}

Agreed, Bill. Actually, I've been thinking it would be an interesting interview question for a software developer candidate (implementing WEASEL, not necessarily doing a comparison with D/M). If they start talking about specified complexity, I'll know to lower my expectations :Dnephmon_{September 1, 2009
September
09
Sep
1
01
2009
03:55 PM
3
03
55
PM
PDT}

nephmon (274), "Did I really manage to mistype “LOL”?" Don't worry, the God of Latching will get that K to an L in no time - and keep it there! "Shurely shome mishtake!" Are you another "Private Eye" reader?Gaz_{September 1, 2009
September
09
Sep
1
01
2009
03:52 PM
3
03
52
PM
PDT}

nephmon: I'm afraid KF will probably regale you with his conspiracy theory about Lewontin and how evolution is all a materialist plot that depends on a material origin of life, which means OOL is relevant to everything, apparently. I suspect that if a step by step origin of life is ever established, KF will most likely say that it begs the question of where matter came from, and that that was the issue all along, and by concentrating on OOL we were simply distracting with clouds of oil of ad hominem burning straw men meant to poison and confuse ... I'm tempted to see what would happen if you gave one group of computer science students Dawkins WEASEL description, gave another group Dembski and marks' description, and asked them both to implement the algorithm as described without any other prompting or background. I can't imagine either group ever producing anything close to functionally equivalent programs. There is just no way you can read those two descriptions, implement them, and come up with the same piece of software.BillB_{September 1, 2009
September
09
Sep
1
01
2009
02:40 PM
2
02
40
PM
PDT}

kf: I'm very impressed by your ability to cite background material, use long words, and deflect the topic in order to suit your purposes. You ability to answer simple questions clearly and succinctly, not so much. I asked about the differences between the mechanism of natural selection and its first-order approximation in WEASEL. You answered by talking about Hoyle's views on the origin of life. (Aside: Creationists love to harp on about Darwin's "racism" as though [if it were true] it invalidates everything else he has to say. But they don't mention much about the astonishing sexism in some of Hoyle's novels. Oh wait,I guess Hoyle was just a product of his time.) Anyway, by definition, WEASEL is "picking up the story" some way down the line from its beginning. You can hark back to the start of it all if you like, but it's not really salient to this discussion. In any simulation and discussion of it, there have to be some "givens". Unfortunately, you don't seem to be prepared to "give" anything, so it makes it hard to have a reasoned debate with you. I totally disagree with your point (3) in 264. The distance of a given letter sequence from the target in WEASEL is a model of a gene's fitness to its environment. There fore it is "a good analogy of the power of random variation and natural selection, which is premised on the origin of information by chance variation, and the selection of FUNCTION on superiority of some sub-populations," because that distance is the way we happen to modeling functionality. You could easily imagine a much more sophisticated model where the letter-sequence is transformed into an object (in the computer science sense) whose attributes allow it to perform better or worse in a virtual world. Simple example to make it more concrete: the first letter of sequence corresponds to length of the neck of the virtual creature. One of the attributes of the environment is the average and SD of the height of foliage on the virtual world's trees. Part of the candidate sequence's "fitness" is how well its phenotype's neck length enables it to feed. This is just a more involved simulation than "the Hamming distance between the first letter of the sequence and the "m" of "methinks", but doesn't alter the fundamental nature of it.nephmon_{September 1, 2009
September
09
Sep
1
01
2009
12:33 PM
12
12
33
PM
PDT}

Did I really manage to mistype "LOL"? Shurely shome mishtake!nephmon_{September 1, 2009
September
09
Sep
1
01
2009
10:30 AM
10
10
30
AM
PDT}

LOK Gaz@272. So in other words, "Latchingdidit" :Dnephmon_{September 1, 2009
September
09
Sep
1
01
2009
10:29 AM
10
10
29
AM
PDT}

What a fascinating discussion! Most interesting, for me, is the parallels with the philosophies of the participants. KF and Joseph, being creatures of Faith, clearly need there to be latching, even if it doesn't exist, because certain prophets say it exists in the original Weasel. Now, they can't say that the prophets are wrong, despite the objective evidence, otherwise the rest of the philosophy as espoused by the prophets is questionable. So they go through all kinds of linguistic contortions to claim latching exists, even to the point of defining "latching" as something it isn't. As in this thread, so in their religious philosophies. For "latching" read "God".Gaz_{September 1, 2009
September
09
Sep
1
01
2009
06:27 AM
6
06
27
AM
PDT}

Kairosfocus- You gave Nephmon 1,259 words of tasty word salad, changing the subject to abiogenesis and arguing from the authority of dear departed Fred Hoyle.
4 –> that was the heart of Sir Fred’s challenge and it is the heart of the design theory challenge to the claimed spontaneous origins of life and of novel body plans. That is, we do observe FSCI of an order of complexity of 1,000 or morte bits forming all the time: by DESIGN. But, we have never observed this by spontaneous forces tracing to chance + necessity without intelligent intervention. And, on search resource exhaustion grounds — as CRD actually concedes in the relevant passage — it is utterly implausible that we will ever get to such complex functionality on the gamut of our observed cosmos. [And please resist the red herrings and strawmen in the usual rebuttals of Hoyle a la Wikipedia. Remember you are in the presence of a Nobel Prize equivalent holder here. he may be wrong on points, but he is not going to be making simplistic blunders, and even his errors will be instructive. [Emphasis added]
Soooo, “tornado in a junkyard makes a 747” is an appropriate analogy because an eminent guy once said so and we should “resist the red herrings and strawmen” that might disagree. Now, I might be persuaded that Sir Fred was a “Nobel Prize equivalent holder”, but I am certainly not in his presence, any more than I am in Sir Isaac’s presence (A man with some very interesting ideas about the Philosopher’s Stone and about the Holy Trinity, by the way). Much has been learnt since these eminent gents passed on, and if they were alive today, they might disagree with their previously held beliefs. And, anyway, it is an argument from authority. Again (see post 136). One small request though : could my word salad be on topic – whether eqn22 could possibly describe TBW Weasel? Thanks.DNA_Jock_{September 1, 2009
September
09
Sep
1
01
2009
05:38 AM
5
05
38
AM
PDT}

KF: The issue of WEASEL's biological relevance is separate and unrelated to the issue of whether the citation in Dembski and Marks paper is accurate. You have demonstrated that WEASEL does not require an explicit latching mechanism to produce any of the observed results, as we have been saying all along. Given this fact, and Dawkins description, and his statements about latching mechanisms, your claim that the existence of an explicit latching mechanism is a reasonable interpretation is not valid. Joseph: Latching individual letters is a partitioned search, ratcheting towards a target is only a partitioned search if individual letters are locked out of the search when they reach their target. Partitioning means to divide up the search into separate, independent units. WEASEL is not an algorithm that searches for individual letters, it is not partitioned.BillB_{September 1, 2009
September
09
Sep
1
01
2009
05:25 AM
5
05
25
AM
PDT}

Joseph, I understand that you are going for the second door: “A latching process is (necessarily) a partitioned search (as modeled in eqn22)” which is a much better choice than the third door: “Weasel is a partitioned search (as modeled in eqn22)”, which is blatantly wrong. The key aspect of the partitioned search is that it is a "divide and conquer" procedure(p1055), in which the search for each character is independent of the search for the other characters. Thus for partitioned search without sample replacement, the target can be found in N-1 queries (p1056) irrespective of the target length. There are whole categories of search algorithms that work by step-wise comparing a short search string (one or two letters) with the target (starting at one end and moving along until a match is found.) This information is then used to infer the FOO (p1054) for the target, and subsequent short search strings use the derived FOO. Useful if the FOO is unknown. It latches, but it ain't a partitioned search.DNA_Jock_{September 1, 2009
September
09
Sep
1
01
2009
05:21 AM
5
05
21
AM
PDT}

I would like to publicly thank Mrs. O'Leary for getting me removed from the banned list. She has been very gracious in private email, albeit strangely fascinated with hockey. ;-)DeLurker_{September 1, 2009
September
09
Sep
1
01
2009
05:14 AM
5
05
14
AM
PDT}

Joseph#265
Except that latching/ ratcheting makes it a partitioned search.
Explicit latching makes it a form of partitioned search. Dawkins algorithm does not explicitly latch.DeLurker_{September 1, 2009
September
09
Sep
1
01
2009
05:11 AM
5
05
11
AM
PDT}

Except that latching/ ratcheting makes it a partitioned search. That is if you understand English...Joseph_{September 1, 2009
September
09
Sep
1
01
2009
04:56 AM
4
04
56
AM
PDT}

As I said before:
...even if there were an explicit latching mechanism in Weasel[and all the evidence indicates that there is not one], that would not make it a partitioned search as described by D&M. The first reported Weasel run in TBW CANNOT be the result of a partitioned search. Only two letters change in generation 2. Latching is irrelevant to the accuracy of D&M’s citation. I am accusing D&M of mis-citation when they claim that eqn22 describes Weasel. I encourage everyone to go to EIL and play with Atom’s adjustable Weasel, specifically the Proximity Reward Search, and contrast its behavior with that of the Partitioned Search.
Now don't I deserve some of your delicious word salad, kairosfocus? Please may I have some more?DNA_Jock_{September 1, 2009
September
09
Sep
1
01
2009
04:21 AM
4
04
21
AM
PDT}

Nephmon: You have asked an excellent question here, that goes to the heart of the issues at stake:
What’s the conceptual difference between “proximity to target” and “fitness to environment”? In both cases some offspring fail to reproduce (all but one in the harsh world of WEASEL). In the natural world, it’s the lack of reproductive success caused by poor fitness to the environment. In WEASEL this is simulated in a very simplistic way by representing the environment as a “perfect gene” that would be the fittest possible for the world it inhabits, and relative fitness is represented as “closeness” to that gene. What’s the fundamental flaw with this?
1 --> I first ask you to simply read the weak argument correctives above [and while you are at it you might want to take a look in a library at the Schutzenberger contributions on functional complexity dating all the way back to the Wistar high level summit of 1966], then take as look at the always linked, starting with the App 7, especially where it discussed the back-story on Weasel:
10 --> Back story: Mr [Fred] Hoyle had raised the issue that the origin of a first life form -- such as, roughly, a bacterium -- is a matter of such complexity that the odds of that happening in a prebiotic soup by chance was negligible. In a more up to date form, this is the challenge that is still raised by the design theory movement: life shows a threshold of function at about 600,000 bits of DNA storage capacity, so before one may properly apply hill climbing algorithms to scale Mt Improbable, one needs first to have a credible BLIND watchmaker mechanism to land one on the shores of Isle Improbable; e.g. drifting by reasonable random configurations of molecules in empirically justified pre biotic soups and empirically credible pre-life selection forces. (But . . . this OOL challenge is still unmet. And similarly . . . the origin of body plan level biodiversity requiring 10's - 100's or millions of bits of functional genetic information, dozens of times over, is equally unmet.) 11 --> So, it looks uncommonly like Weasel distracts attention from and begs the question. That sums up the balance on the main issue . . .
2 --> In short, there is a major question-begging -- indeed, dismissal: "single-step selection" -- at stake in Weasel as presented: the origin of complex, functionally specific information. 3 --> that is, until one has spontaneously created novel complex information that functions through similarly sophisticated but spontaneously originated machinery, at a realistic level, one has no right to pretend that a targetted process that uses a hotter-colder distance-to target comparison of "mutant nonsense phrases" -- i.e confessedly NON-FUNCTIONAL ones -- is a good analogy of the power of random variation and natural selection, which is premised on the origin of information by chance variation, and the selection of FUNCTION on superiority of some sub-populations. 4 --> that was the heart of Sir Fred's challenge and it is the heart of the design theory challenge to the claimed spontaneous origins of life and of novel body plans. That is, we do observe FSCI of an order of complexity of 1,000 or morte bits forming all the time: by DESIGN. But, we have never observed this by spontaneous forces tracing to chance + necessity without intelligent intervention. And, on search resource exhaustion grounds -- as CRD actually concedes in the relevant passage -- it is utterly implausible that we will ever get to such complex functionality on the gamut of our observed cosmos. [And please resist the red herrings and strawmen in the usual rebuttals of Hoyle a la Wikipedia. Remember you are in the presence of a Nobel Prize equivalent holder here. he may be wrong on points, but he is not going to be making simplistic blunders, and even his errors will be instructive. That is my experience from many years of running across his work, on a wide array of topics, starting with the Steady State universe Hypothesis. (He it is who gave the name "big bang" to the cosmogenetic theory of that name, but he did not intend it to be a positive term.)] 5 --> With these points in mind, let us look again at what CRD said in discussing Weasel: _________________ >> It . . . begins by choosing a random sequence of 28 letters ... it duplicates it repeatedly, but with a certain chance of random error – 'mutation' – in the copying. The computer examines the mutant nonsense phrases, the 'progeny' of the original phrase, and chooses the one which, however slightly, most resembles the target phrase, METHINKS IT IS LIKE A WEASEL . . . . What matters is the difference between the time taken by cumulative selection, and the time which the same computer, working flat out at the same rate [i.e. search resources come in here --KF], would take to reach the target phrase if it were forced to use the other procedure of single-step selection: about a million million million million million years. This is more than a million million million times as long as the universe has so far existed . . . . Although the monkey/Shakespeare model is useful for explaining the distinction between single-step selection and cumulative selection, it is misleading in important ways. One of these is that, in each generation of selective 'breeding', the mutant 'progeny' phrases were judged according to the criterion of resemblance to a distant ideal target, the phrase METHINKS IT IS LIKE A WEASEL. Life isn't like that. Evolution has no long-term goal. There is no long-distance target, no final perfection to serve as a criterion for selection, although human vanity cherishes the absurd notion that our species is the final goal of evolution. In real life, the criterion for selection is always short-term, either simple survival or, more generally, reproductive success. [TBW, Ch 3, as cited by Wikipedia] >> _________________ 6 --> In short, CRD acknowledges that the so-called fitness function he uses is nothing but a measure of distance to target which ignores and even dismisses the issue of first needing to be on at least the shoreline of an island of complex function before hill-climbing through warmer-colder signals can be reasonable. 7 --> Thus, he inadvertently testifies that it is the injection of artificially originated active information that makes the difference in achieving complex function within reasonable search resources. 8 --> We should note that he also realises at some level that the substitution he makes is "misleading." 9 --> Indeed. 10 --> Finally we should note that genes are in effect informaiton storage devices. Until the functional machines to interpret and carry out the information exist, they are jut polymer molecules. And, the machines in question have to be coded for and organised to function: ~ 600 - 1,000 k bits of info based on observed life. (Just 1,000 bits specifies a config space so large that the 10^80 or so atoms of our observed universe across its credible lifespan will not be able to scan through 1 in 10^150 of that, effectively the scope of feasible search is a fraction not materially different from zero.) This is the basic reason why OOL is such a challenge to evolutionary materialism. 11 --> And, when it comes to origin of body plans de novo, we are talking of not 600 - 1,000 or so k bits of information but 10's - 100's of mmega bits, all of which have to be in place before function is feasible. GEM of TKIkairosfocus_{September 1, 2009
September
09
Sep
1
01
2009
04:19 AM
4
04
19
AM
PDT}

I try it again: In their paper, Dembski and Mark they start with the phrase: SCITAMROFN*IYRANOITULOVE*SAM I calculated a next generation using Dawkins's algorithms with populations of 10,50 and 100 - and mutation rates of .04, .05 and .1. The tenth string in the list is the second generation given in the paper of Mark and Dembski. The differences with the first generation are in bold face: 1. SCITAMROFN*IYRANOIEULOVE*SAM 2. SCITAMROFN*IYRANOITULOGE*SAM 3. ECITAMRI*N*IYZANOITULOVE*SAM 4. SCITAMROFN*IYRANOITUL*VE*SAM 5. SCITAMROFN*IYRANOITULOVE*SEM 6. SCITAMOOLNOIYRAMOITULOVE*SEM 7. SCITANROFN*IYYANOITULOVE*SAM 8. SCITIMROFN*JYRANOITULOVE*SAM 9. SCITAMROFN*ICRHNOITSLOWE*SAV 10. OOT*DENGISEDESEHT*ERA*NETSIL kf - can anyone spot a difference in the design of the strings? Could you explain why it exists? Heck, I even calculated the probabilities for latching... (I hope I don't have to endure another twelve hours of moderation...)DiEb_{September 1, 2009
September
09
Sep
1
01
2009
01:11 AM
1
01
11
AM
PDT}

Also, can you explain how what you term as "implicit latching", i.e. lack of reversion of correct letters to incorrect ones, which is a result of the strategy of selecting the fittest of a large population of offspring, is fundamentally different from what's observed at in real organisms at the gene level? If the human genome mutates at, say, 100 bases pairs per sexual reproduction (which is towards the upper bound of the estimated range), what's the chance of any of those 100 out of the 3.4 billion or so base pairs in the genome are going to mutate back again in the near (or even distant) future? Rather small, I'd imagine. Thus once a beneficial mutation has made it into the genome, it will very likely be passed down for all intents and purposes "forever", in other words, it's latched. In what way is this different from happens in WEASEL?nephmon_{September 1, 2009
September
09
Sep
1
01
2009
12:28 AM
12
12
28
AM
PDT}

kf:
PPS: Nephmon similarly, cannot agree that targetted search — with a built-in target phrase! — using a proximity to target filter to “cumulatively select” increments of proximity of “nonsense phrases” is a case where “the program contains information that allows it to solve the problem.”
What's the conceptual difference between "proximity to target" and "fitness to environment"? In both cases some offspring fail to reproduce (all but one in the harsh world of WEASEL). In the natural world, it's the lack of reproductive success caused by poor fitness to the environment. In WEASEL this is simulated in a very simplistic way by representing the environment as a "perfect gene" that would be the fittest possible for the world it inhabits, and relative fitness is represented as "closeness" to that gene. What's the fundamental flaw with this? Actually, I did mention what RD had to say about the drawback of the program containing the target phrase, and I said that I think it's an implementation detail. How "external" to the running code of the program would you need the target/environment to be before you accepted that the program itself didn't contain the "information" that it's working on? I already posited an example where the string could be read from an input stream, the other end of which could be a program running halfway across the world. Is that external enough? If one states the problem as "show how the successive choice of the fittest of a group of randomly mutated offspring to be the sole parent of the next generation can quickly converge on the optimally 'fit' entity," you need at some point to provide a quantity to measure fitness against. For convenience, it can be hardcoded into the program itself, but as I've mentioned a few times now, that isn't necessary for the program to function, neither is it necessary for the target to be static. Out of interest, in what way does WEASEL fail most egregiously as a simulation of natural selection as you think it's accepted by evolutionists? As I've freely admitted before, there are many simplifications implicit in it, but of the major components of "the natural selection of fittest offspring that are subject to random mutation", where to you feel WEASEL most comes up short? (Please don't concentrate on the "natural" aspect; by definition any simulation won't be "natural" - there isn't really mutating DNA in the computer.)nephmon_{September 1, 2009
September
09
Sep
1
01
2009
12:13 AM
12
12
13
AM
PDT}

Joseph and CJY: Thanks. Appreciated. It seems the Darwinists -- oh, the irony! -- are in denial. Maybe, they need to read this classical philosophical story, and then "wheel and tun and come again." I think a spiritual issue is at the root. I suggest meditation and prayer on the lines of this classical text from the Sermon on the Mount. G'day. GEM of TKI PS: Onlookers, it seems BillB thinks I am now "insane" and/or "stupid" -- sounds familiar? -- to have (months ago) analysed, identified and defended then DEMONSTRATED the mechanism that triggers implicit latching as an explanation of Weasel 1986 [probably the only reason why it is suddenly "obvious" to all and why it is held that it is my intellectual incapacity that prevents me from seeing it in his estimation!], and to point out that the Marks and Dembski analysis starts from credibly OBSERVED ratcheting (oops, he apparently denies that the cumulatively selected o/p of Weasel 1986 evidently latches!), and deduce its implications for active information. This, from one who evidently has a difficulty with explanation of program o/p's by credible causal mechanism tracing to i/p's and processing behaviour. There is something plainly deeply and sadly wrong in the state of Darwinland -- and it isn't the name of the river of Egypt. PPS: Nephmon similarly, cannot agree that targetted search -- with a built-in target phrase! -- using a proximity to target filter to "cumulatively select" increments of proximity of "nonsense phrases" is a case where “the program contains information that allows it to solve the problem.” [He needs to read what CRD said in his own words on this, as is repeatedly cited above, e.g. last at 221, and as is explained in my online note app 7].Something is indeed deeply, sadly wrong in the state of Darwinland.kairosfocus_{August 31, 2009
August
08
Aug
31
31
2009
09:55 PM
9
09
55
PM
PDT}

At least I can put the onus that no one answers to my edits on the fact that these stay in moderation for a half a day and not on a lack of quality of the edits themselves :-)DiEb_{August 31, 2009
August
08
Aug
31
31
2009
03:16 PM
3
03
16
PM
PDT}

I like Joseph's demand for proof from the text of TBW that it explicitly precludes ratcheting. Somewhat akin to asking for proof that God (sorry, an intelligent designer) doesn't exist. And in both cases the response is the same: instead of demanding proof for the non-existence of something, why not provide proof for its existence? In the case of WEASEL, please do what BillB asked for in #245. I suspect that TBW doesn't explicitly say "positions aren't latched once they have their correct values" because such an outrageous hack would never have occurred to RD. It just doesn't represent what he was trying to model. Joseph, please just confirm that you understand this: letters don't revert to incorrect values because the candidates that do are never the fittest amongst the generated offspring. This is true even for small values of n (about 50 from my experiments, depending on p(mutation)). Make n smaller, and yes, you WILL see reversions, and in the limit, as n -> 1, you're basically making a random mutation to a single offspring, and off course you'll never get closer to matching the environment string (unless you employ ratcheting, a la D/M). You do understand all this, right? On this issue of "the program contains information that allows it to solve the problem" as raised by kf and others, I disagree: this is just an implementation detail. Instead of the target string being harcoded into the program (or input at the start), let's read it from an input stream instead. Moreover, let's read it at the start of every generation, allowing it to change over time. The program will still work just as well, even though its only knowledge of the target string is garnered at the beginning of each generation, making it similar to an "environment" in which the offspring are generated that I mentioned in a previous post; the progeny that goes forward to the next generation is the one that fits best the current environment, not some predetermined "distant target".nephmon_{August 31, 2009
August
08
Aug
31
31
2009
11:02 AM
11
11
02
AM
PDT}

Prev 1 2 3 4 5 6 … 13 Next

You must be logged in to post a comment.

Leave a Reply