AE BB DB Explorer


Action:
Author:
Search Terms (separate with commas, no spaces):


form_srcid: Turncoat

form_srcid: Turncoat

form_cmd: view_author

Your IP address is 67.202.56.112

View Author detected.

view author posts with search matches:

Retrieve source record and display it.

form_author:

form_srcid: Turncoat

q: SELECT AUTHOR, MEMBER_NAME, IP_ADDR, POST_DATE, TOPIC_ID, t1.FORUM_ID, POST, POST_ID, FORUM_VIEW_THREADS from ib_forum_posts AS t1 LEFT JOIN (ib_member_profiles AS t2, ib_forum_info AS t3) ON (t1.forum_id = t3.forum_id AND t1.author = t2.member_id) WHERE MEMBER_NAME like 'Turncoat%' and forum_view_threads LIKE '*' ORDER BY POST_DATE ASC

DB_err:

DB_result: Resource id #4

Date: 2007/12/04 21:52:14, Link
Author: Turncoat
Know your adversary.

It's not often you can hang a diagnosis on someone across the web. But reading Gil Dodgen's polymorphously pompous bit on his new role as evolutionary informaticist, and how he owed his scientific "vigilance" to dear old dad, "whom I admire more than any man I have ever known," I kicked myself for having gone so long without recognizing him as narcissistic.

Here's the list of diagnostic criteria in the DSM-IV. I say he hits regularly on five of them. You have to know somewhat more than the DSM says to understand what a red flag Gil's idealization of his father is. See Golomb, Trapped in the Mirror: Adult Children of Narcissists in Their Struggle for Self. Odds are good that the father is more narcissistic than Gil.

Date: 2007/12/04 21:57:40, Link
Author: Turncoat
Quote (Turncoat @ Dec. 04 2007,21:52)
Know your adversary.

Oops. Make that "Know your tard." I'm out of practice here.

Date: 2007/12/05 13:19:24, Link
Author: Turncoat
Gonzalez is an IDiot, but no idiot. He knew that he was destined either to a) gain tenure at a state-supported institution as a known ID advocate or b) sacrifice his career on the altar of intelligent design.

There has never been a losing outcome for the DI. One of its prime objectives is to rile the conservative-Christian rank-and-file. Given the low ebb of the ID tide, the DI gets more mileage out of a martyr than a tenured professor. The best of all scenarios for the DI is an extended sealing of the tomb, followed by a surprise ending in which the martyr emerges transfigured.

Date: 2008/01/09 12:46:32, Link
Author: Turncoat
Quote (Erasmus @ FCD,Jan. 09 2008,08:51)
What was semiotic007 hammering kf about?

I haven't the desire to go through his long yammering self-referential navel gazing posts.  And he is full of it too.  anyone copy that thread?

Short reply: The chattering ape drummed his chest, but ducked behind Mama Dave when Big Bad Semiotic rattled the cage.

At 198, kairosfocus personified the Wikipedia article on NFL, interrogated "Prof Wiki," and taunted when the encyclopedia did not reply. He signed the comment "GEM of TKI." But then he decided to give someone in the real world credit for his brilliance:
                 
Quote
199

kairosfocus 01/06/2008 4:51 am

PS: Kairos, we gotta get together sometime!

[Onlookers, Kairos and Kairosfocus are two very different persons. I am a J’can by birth and ancestry whose family has lived in several Caribbean territories, a physicist in my education base, an educator and strategic- management- in- the- context - of sustainability- of- development- in- a world- of- accelerating- and- transformational- change thinker in my work and service activities, and an Evangelical Christian [cf. this invited public lecture] living out here in Montserrat, where my wife hails from. I have come to use kairos as a key concept in light of its appearance in Paul’s Mars Hill address, circa 50 AD in Athens, as recorded in Acts 17:24 - 27. K, what about you?]

The post includes three links (I'm too lazy to copy them) to the web site for The Kairos Institute (TKI). Pages KF explicitly linked to identify Gordon Mullings and G. E. Mullings as authors. The site includes pages advocating intelligent design, e.g., An Apologetics Primer, by Gordon Mullings (see section 3.4). Further googling reveals that someone going by the name Gordon Mullings has openly advocated ID hither and yon. Today's entry in the KairosFocus blog, which includes entries by Gordon Mullings (including his email address), links to kairosfocus' "citizen's arrest, citizen's arrest!" post at UD.

Semiotic subsequently asked kairosfocus if Gordon Mullings would accept a wager. He never stated that kairosfocus was actually Mullings. KF wailed that he was going to get spam because Semiotic had posted his real name on the page, and that he might suffer persecution because Semiotic had tagged him an IDist.  :D

Date: 2008/01/09 13:38:24, Link
Author: Turncoat
dheddle,

"If you think you understand quantum mechanics, you don't understand quantum mechanics." --Richard Feynman

Date: 2008/01/09 14:14:38, Link
Author: Turncoat
Quote (Erasmus @ FCD,Jan. 09 2008,08:51)
What was semiotic007 hammering kf about? [...] anyone copy that thread?

It was a pun carried to lengths.  The "course" below is the Wiki article on NFL.
Quote
213

Semiotic 007 01/06/2008 10:45 pm

I wager US $25,000 that there is a “free lunch” in optimization. Details follow.
   
Quote

   I want my money back for this course!

   (Oops, it’s “free” — guess I shouldn’t expect to get the full straight dope easy for free . . . ?

   After all:

   there’s no free lunch!)

There’s no free mouth. I will offer even odds first to Gordon E. Mullings, a native of Jamaica. If he declines, a certain Dell millionaire will be welcome to take the bet. I will not insult Bill Dembski’s intelligence.

Definition: Let S = 10^90. [This is Seth Lloyd’s bound on the number of bits registered by the observable universe. I sincerely thank Dr. Dembski for getting me to read and ponder Lloyd’s article.]

Definition: Let X = {0, 1}^64. Also let F be the set of all functions from X^5 to X. [I’ve set this up so you can grasp F intuitively as the set of all functions y = f(x1, …, x5), where the arguments and value are 64-bit numbers. In practical optimization problems, F is often much larger. There is no genotype of a living organism that contains as few as 5 x 64 = 320 bits of information, by any reasonable measure.]

If algorithms to search functions in F are written as binary Turing machine descriptions (strings over {0, 1}) for a fixed universal Turing machine U, and the S-th cell of the tape from which U reads descriptions is set to 2 immediately prior to the operation of U, then there is no probability distribution on F for which all algorithms have identically-distributed sequences of observed values when presented to U. This holds even if algorithms (Turing machine descriptions) that are “redundant” in the sense of being longer than functionally-equivalent algorithms are eliminated from consideration.

I made the preceding proposition somewhat informal. I claim that a similar statement (to be included in a written agreement) will be proven a theorem in a journal in mathematics, science, or engineering within the coming three years, and that its negation will not. (Stipulation: The journal must have existed in printed form since 2003 or earlier.)

Both parties to the wager will deposit US $25,000 into an interest-bearing account. A third party will manage the account. If, at the end of three years, proofs of both the proposition and its negation appear, or no proof appears, in the specified journals, then each party will be returned his deposit plus interest. Otherwise, the party predicting correctly which proof would appear will receive all funds in the account.

I realize that Gordon E. Mullings, as a resident of an emerging country, may not have 25 grand on hand. I’ll go as low as US $5000 for him. So tell me, kairosfocus, if I contact the real Mr. Mullings, will he put his money where your mouth is?
 
Quote
214

Semiotic 007 01/06/2008 11:00 pm

P.S.–The journal article must be peer-reviewed, of course.

When kairosfocus responded that he could not, as a Christian, gamble, and complained that he was going to get spam because his name appeared on the page, Semiotic posted this:
 
Quote
232

Semiotic 007 01/08/2008 10:46 am

kairosfocus (227):

So would the real-world person you claim to be at least sign and return to me a hard-copy statement that the theorem below is false?

Note that every deterministic search algorithm satisfying the requirements of Wolpert and Macready (1997) implements a decision tree (English 2004).

Definitions: Let S = 10^90. Let F be the set of all functions from {0, 1}^300 to {0, 1}. Let D be the set of all decision trees for F.

Theorem (Practical Free Lunch): Let U be a universal Turing machine processing binary Turing machine descriptions. Then for almost all d in D the shortest U-description of a Turing machine implementing d is of length greater than S.

Loosely, the theorem says that almost all “basic approaches” to solving a relatively small search problem require more bits to program than the observable universe is believed to register. As a practical matter, an algorithm that can be physically encoded and executed by some physical process is superior in performance to all that cannot.

By the way, I want the signature to be notarized. When I scan the signed statement for display on the Web, I will obscure the signature, but not the notary’s stamp.

Among KF's many objections to this post were his concern that someone asking him to sign his name to a statement might be planning identity theft.

Date: 2008/01/09 14:29:50, Link
Author: Turncoat
dheddle,
 
Quote
If he means that in a real environment, there is some miniscule but nonzero probability that the air molecules will conspire to line up in front of the ball and stop it, and underneath it to hold it up, I suppose he is technically correct.

That's a specified event with probability much less than Dembski's universal probability bound, don't you think? ;)

BTW, fantastic quote of Sam Harris.

Date: 2008/02/28 03:34:10, Link
Author: Turncoat
Deletion without comment is the Czar's sincerest form of flattery. Formerly, and ever so briefly, at Wanted: More Greenhouse Gases:
                                                       
Quote
DaveScot and GilDodgen,

Like most guys who know a lot about computing, you know precious little about statistical inference. A downward spike in the latest sample of a time series that has shown a long-term upward trend is no basis for saying that you are "proven right." This is not to say that you are not right -- it's just to say that it's foolish to trumpet on the basis of so little information. It's obvious that neither of you has ever worked in prediction of nonlinear, nonstationary time series.

I genuinely do not know what to believe about anthropogenic global warming. But I do know that there is so much scientific uncertainty in the matter that any layperson who claims to have resolved it absolutely is pure bluster. And I know also that there are hugely different costs associated with different errors in inference. If CO2 emissions are in fact causing global warming, and we do nothing about them, then the cost of the error is astounding. If the emissions are not responsible for global warming, and we reduce them needlessly, the cost of the error is relatively low. Given the present scientific uncertainty, and the possibility that severe cost is associated with allowing CO2 emissions to rise, a prudent course would be to look for approaches to reducing emissions that are a) relatively high in efficacy and b) relatively low in impact on the economy.

Nothing forces us to do everything possible to reduce CO2 emissions or to ignore them totally. Only simpletons and blow-hards insist on giving all-or-nothing responses to ambiguous scientific information.


A funny aspect of my head-butting with Dave is that he doesn't know, and seems incapable of discerning, that my IQ is identical to his. Of course, I have four degrees that he does not, and a bunch of teaching, research, reviewing, etc., to boot. And unlike him, I can allow that some of you outclass me.

It happens that I once published results in prediction of annual sunspots numbers (a classical problem in statistics) that improved greatly on all in the literature. When I tell Dave he's an IDiot to claim "proof" that global temperatures are not rising when temperatures dip sharply for a year, he really should listen. But I think we all agree that Davie's wee-wee would have shriveled and fallen off if he had left my comment on the blog.

Date: 2008/02/28 03:43:43, Link
Author: Turncoat
I forgot to mention that I've been "Expelled" yet again.

I would love to see some of Dembski's old expulsion notices released to the media right after the film opens. Wes, are you reading this? Have you saved stuff going back that far?

Date: 2008/02/28 18:44:20, Link
Author: Turncoat
Quote (Reciprocating Bill @ Feb. 28 2008,06:35)
       
Quote (Turncoat @ Feb. 28 2008,04:43)
I forgot to mention that I've been "Expelled" yet again.

I would love to see some of Dembski's old expulsion notices released to the media right after the film opens. Wes, are you reading this? Have you saved stuff going back that far?

Turncoat -

You called out Dave's NAME and that's why you and your message were obliviated: "Only simpletons and blow-hards..."

Who were you at UD this time?

Cloud of Unknowing.

I've been expelled many times, but deleted only once before. That was after I challenged the Gish-galloping Gordon E. Mullings, by name, to a $25,000 wager regarding "free lunch" in optimization. (I'd have actually gone through with the bet, were he fool enough to take it, but I intended simply to jerk his chain.) He signed his comments "GEM of TKI" and linked to web pages giving his name, so I didn't think he had left himself privacy I could infringe upon. But DaveTard decided otherwise. I suppose homely cusses like Mr. Springer (barefoot and pregnant) and Msr. Mullings (3rd and 5th thumbnails) have to stick together.

I love it when the Czar gives me the boot and leaves his "rationale" on the blog. Perhaps he's finally caught on to the fact that I've been playing him against himself. He is, of course, his own worst enemy. But I suspect he's conscious that evidence of UD's hypocrisy is a bad thing when "Expelled" is about to be released.

Date: 2008/02/28 20:55:46, Link
Author: Turncoat
Quote (Erasmus @ FCD,Feb. 28 2008,18:48)
Why I have wondered is that seminal douche Kairosfocus not here anymore.  Perhaps you have answered the question.  I have suffered that fool, gladly, several occasions (so have Hermagoras and others).

I don't hang diagnoses on everyone at UD, but KF gives strong signs of bipolar disorder (and Gil Dodgen has essentially tagged himself as the narcissistic son of a narcissistic father). I predicted that he would swing into depression and disappear for a while, and I was right about that. We'll have to wait and see if he returns galloping.

I think there's some justification for getting the UDers to parade their... issues. But I don't want to be out-and-out cruel to Mullings. David Scott Springer is a different matter. The IDists think evolutionary theory has caused people to disavow moral responsibility for their actions. Well, I hold DaveScot morally responsible for his polymorphously unconscionable behavior.

Having worked on a psychiatric ward, I know just how pernicious mania is. If I'm right about Mullings, I sincerely hope he takes the psychotropics Denyse O'Leary says no ensouled body needs and gets better.

Date: 2008/02/29 07:39:50, Link
Author: Turncoat
Quote (Reciprocating Bill @ Feb. 28 2008,21:58)
       
Quote (Turncoat @ Feb. 28 2008,19:44)
     
Cloud! And your other incarnations? Rewiring minds want to know.

I really can't remember most of them. I used to be Semiotic 007. Some time before that, I went by my own name, and Our YEC Savior went to some length to make sure THE WORLD (he clearly believed that THE WORLD was reading UD) did not mistake me for the creationist by the same name. The funny thing is that the notion of creation is central to my belief system, but I have absolutely no use for the -ism. Gil Dodgen opened a thread addressing me directly when I pointed out that he had made absolutely moronic comments about simulation. And DaveScot also opened a thread addressing me directly -- if I recall correctly, he was responding to my observation that lay "science skeptics" are sorry substitutes for practicing researchers with PhD's. (I've actually insisted this in 4-5 incarnations now. It's a sure-fire way of getting the IDiots to demonstrate their circle jerk.)

You'll know me when I post again at UD. ;)

Date: 2008/02/29 09:27:15, Link
Author: Turncoat
Quote (Erasmus @ FCD,Feb. 28 2008,23:48)
Turncoat

I am surprised that you might think that Kairosfocus is somehow not all there.  I mean, not really, because ID requires some fundamental level of disconnect, but on the other hand this cat has all sorts of posted tard at the always linked discussion.  And that kind of shit is hard to argue about, clearly it is the product of a warped but focused mind.  

Don't get me wrong, he is one of my favorite all time tards because of the swallowing of the baby-puke colored pill, but at the very least his highly systematic and seemingly ennumerated scheme for epistemology indicated that he was more than your average turettes syndrome basketcase pubic hair filer.

When someone with bipolar I disorder is on the upswing, but not yet in a psychotic state, he is "all there," and may be highly focused, sleeping little and working long hours on something that seems tremendously important at the time. I don't think the Kairosfocus who responded to Semiotic 007 at UD was in as good shape as the Mullings who wrote the documents on the web. Manics are big-time systematizers. They are often grandiose. They may speak rapidly and tangentially. Sound familiar?

ID requires some fundamental level of disconnect? The vast majority of IDists are "liars for God." They're creationists trying to circumvent federal law by obfuscating their theory that intelligence creates complex specified information. Beyond that, the notion of winning back science for Jesus, closely akin to the notion of winning back America for Jesus, is friggin' looney-tunes.

Date: 2008/02/29 09:56:39, Link
Author: Turncoat
Quote (Erasmus @ FCD,Feb. 29 2008,08:51)
ohhhhh, an expelled member of the lab, eh?

that's working from the inside out, huh.

how big was that broom closet, anyway?

I wasn't expelled. I was willing to stand up for Marks and his right to push evolutionary informatics and notions like active information. But when the names "Dodgen" and "Sewell" turned up next to mine, I was appalled. And I excused myself politely. Bob Marks is a decent guy, as best I can tell.

About thirty years ago, I was expelled from, and subsequently readmitted to, a Baptist college. My sin was to agitate for equality of women and men in education. My salvation was the many students and faculty members who spoke out on my behalf. Last year, I felt it was my turn to speak out for someone. I interacted behind the scenes, obviously to no avail, with regents and the executive vice president and provost of Baylor.

Fun facts to know and tell: I exchanged notes with a regent on the 51st anniversary of my birth in Waco. The site of my first home is now part of the Baylor campus.

Another important aspect of the story is that when I signed on with the EvoInfo Lab, I thought I had some solid "free lunch" results for optimization forthcoming. Bob Marks agreed up front not to censor my work, so my plan was to post at his site a research paper contradicting much of what he and Dembski have to say.

As things stand, I do have some "free lunch" results, but I consider them somewhat trivial. I'm giving that line of research a rest at the moment, and will return to it after revisiting some work in evolution of time series predictors.

All in all, the experience with the EvoInfo lab was embarrassing for me. But I feel I did what was right for me to do.

Date: 2008/02/29 09:59:49, Link
Author: Turncoat
Quote (Raevmo @ Feb. 29 2008,08:26)
Turncoat = Tom English

Only five other posts since October 2006. And identifying me rose to threshold. I am flattered.

Date: 2008/02/29 10:17:57, Link
Author: Turncoat
Quote (Richardthughes @ Feb. 29 2008,09:59)
Turncoat, were you on the top-secret ID list-serve?

edit? Yes.

No, I did not wind my way into the Heart of Darkness.

Dembski suggested that Semiotic 007 (what a big secret my real identity was!) contact Marks. I did. Marks reviewed my web site, complimented me on some of my poems, asked for comments on his and Dembski's papers, and linked me into the EvoInfo web site. I never commented on the papers. Eventually I asked Marks to remove me from the site. Marks agreed, sent his best wishes, and completed the job the same day. That's the full extent of our interaction.

I have more email text from regents and the EVPP of Baylor than I do Marks and Dembski.

Date: 2008/02/29 12:43:46, Link
Author: Turncoat
Quote (Raevmo @ Feb. 29 2008,10:59)
 
Quote
Turncoat: Only five other posts since October 2006. And identifying me rose to threshold. I am flattered.


It was your remark concerning sunspots that gave it away for me. I recalled you saying something about predicting sunspots with genetic algorithms on UD at the time, and I looked up your paper back then.

My writing style is distinctive, as is the combination of topics I comment on. It's never a secret who I am. The game is that as long as I don't say who I am and don't cross DaveScot, everyone pretends he doesn't know who I am.

Date: 2008/03/04 02:15:50, Link
Author: Turncoat
Quote (hooligans @ Feb. 29 2008,21:45)
Turncoat,

Did you actually step foot into the lab? Did you see photos of the equipment? What kind of staff does the lab employ?

I virtually set foot in the EvoInfo lab, just as I've virtually quaffed a few at PT. The inside dope is that Galapagos Finch handles development and testing of MATLAB programs.

Date: 2008/03/04 14:42:27, Link
Author: Turncoat
[Graffiti moved to Bathroom Wall. -Admin]

Quote (Alan Fox @ Mar. 04 2008,06:00)
Tom wrote:            
Quote
I genuinely do not know what to believe about anthropogenic global warming. But I do know that there is so much scientific uncertainty in the matter that any layperson who claims to have resolved it absolutely is pure bluster. And I know also that there are hugely different costs associated with different errors in inference. If CO2 emissions are in fact causing global warming, and we do nothing about them, then the cost of the error is astounding. If the emissions are not responsible for global warming, and we reduce them needlessly, the cost of the error is relatively low. Given the present scientific uncertainty, and the possibility that severe cost is associated with allowing CO2 emissions to rise, a prudent course would be to look for approaches to reducing emissions that are a) relatively high in efficacy and b) relatively low in impact on the economy.


Great minds think alike!

Well, Alan, weighing the costs is merely common sense. But common sense ain't all that common -- especially at Uncommon Denial.

I suspect that Mr. Springer's "thinking" is rooted in his investment portfolio. Most other IDiots believe that God gave humankind dominion over the earth, and that signs abound that Jesus really is coming soon. There's a large measure of hedonism in their Christianity.

I once went out with a hospital CEO who explained at length how Jesus had made her a success. Then she tried to get me to fuck her. (Sorry, but fuck is le mot juste.) Physically, she was an absolute beauty, but somehow I'd lost interest.

Date: 2008/03/04 17:27:48, Link
Author: Turncoat
Quote (Reciprocating Bill @ Feb. 29 2008,07:26)
 
Quote (Louis @ Feb. 29 2008,06:42)
Bipolar disorders (1, 2 or cyclothymia or whatever they are calling them nowadays) and clinical depression are indeed mental illnesses but they don't always render the patient a window licking unfunctioning ambulatory vegetable.

Kay Jamison's autobiographical An Unquiet Mind (which I read ~10 years ago), Touched with Fire: Manic-Depressive Illness and the Artistic Temperament and Exuberance: The Passion for Life go directly to these questions with respect to bipolar illness.

An Unquiet Mind is a good read (I haven't read the others). Jamison is one of the world's foremost authorities on the illness, has bipolar illness herself - and resisted pharmacological treatment for years because it was during early phases of her manic episodes (what she called her "white manias" - distinct from her subsequent inevitable "black manias") that she was most productive.

Diagnosis with the illness sometimes raises dilemmas for the individual, who may value elements of their high moods and must also confront the reality that their affect often can't be uncritically "trusted" - an enthusiasm, attraction, interest etc. may be the leading edge of a hypomanic mood and not "genuine." I knew a guy who would write plays, paint, compose and sing during hypomanic periods - all quite unproductively, because he couldn't finish projects. His diagnosis was a severe blow, because it undercut core elements of his identity and self-esteem that revolved around his creativity and sometimes seemingly boundless energy. He also struggled with the fact that, during the peak of his manic periods, there were psychotic elements to his thinking that he valued. He would, at those highest moods, walk to the beach and sit before the surf, which he believed "imparted information" to him that was unique and quite indescribable. He recognized the unreality of that phenomenon at other times, but his extremely high moods impaired his judgment about the quality of his thinking when active.

My undergrad degree is in psychology, and I worked on a psych ward for a time, and I was married to a psychotherapist. Now middle-aged and unmarried, I find that a hugely disproportionate fraction of available women about my age are trauma survivors -- usually child sexual abuse -- so I've read mostly about that awful stuff in recent years. But I have read An Unquiet Mind, and it's precisely the book I would recommend to anyone who wants to understand bipolar disorder. The story you tell is not uncommon.

Date: 2008/03/04 17:52:37, Link
Author: Turncoat
Quote (Louis @ Feb. 29 2008,05:42)
My boss at the time asked me a question once : "Why is it that people seem to think that the brain is the only organ which cannot get sick?"

Just saw this, Louis. You're on the money. In the States, it's conventional to draw an analogy between affective disorders and diabetes.

The worst thing going on in IDist circles, apart from HIV denial, is Denyse O'Leary's campaign against antidepressants. There's plenty of hyperbole in my anti-ID rhetoric, but none when I say that she's a menace to public health. I'm sure there are conservative Christians she has convinced to pray and "keep a positive mental attitude" instead of take pills.

Date: 2008/07/30 11:31:38, Link
Author: Turncoat
Quote (Assassinator @ July 30 2008,05:47)
The creo's are really feeling like stuck records, I'm getting bored of them. Let's hope they shed the skin called ID soon, and something new and fun emerges from underneath it.

Date: 2008/08/27 15:43:28, Link
Author: Turncoat
Quote (Wesley R. Elsberry @ Aug. 26 2008,18:14)
Antievolutionists want to confuse and conflate meaning and information. Spetner, Gitt, Truman, and Dembski... all of them want meaning to be folded within whatever sort of "information" they propose.

Shannon's discussion of information explicitly excluded meaning. Algorithmic information theory only cares about one aspect of meaning: what is the shortest program and input that can generate a string?

Critique of Dembski's "complex specified information"

Wes, the one exception to what you say is the Kolmogorov structure function of algorithmic information theory, which I have seen Paul Vitanyi relate to "meaning."

For criticism of the morph of CSI that came after the one you and Shallitt addressed, see this.

Date: 2008/08/27 16:55:23, Link
Author: Turncoat
The two leading notions of the quantity of information in an object are the self-information of Shannon and the algorithmic information of Solomonoff, Chaitin, and Kolmogorov (working more or less independently). Algorithmic information is widely known as Kolmogorov complexity, so distinctions between "complexity" and "information" are not clear-cut. What self-information and algorithmic information have in common is relation of information to description.

For Shannon, there is an objective probability distribution p on a set of outcomes of a random experiment, and the information in outcome x is -log p(x). (Let's say that the logarithm is base-2, and the unit of information is the bit.) A justification for regarding this as the intrinsic information in x is that if you want to transmit messages indicating the outcomes of repeated experiments to a receiver, and you want to minimize the average number of bits you transmit per outcome, then you and the receiver ideally agree on a code that associates a binary description of length -log p(x) with each possible outcome x.

A problem with Shannon's self-information is the issue of how we know the objective (actual) distribution p in terms of which ideal description length is defined. In algorithmic information theory, description is defined in terms of computation instead of probability. That is, the algorithmic information of a string (finite sequence) of symbols is the length of the shortest binary computer program that outputs the string and halts. To be more specific, the computer is a universal computer, equivalent in "computing power" to a universal Turing machine. If we restrict ourselves to "simple" universal computers, the program length varies little from one computer to the next. To make this more concrete, and to relate it to earlier comments, the program is like a self-extracting zip archive. The program is a compact description of the string x it outputs.

From the equality

description_length = -log p(x),

we may obtain

p(x) = 2^-description_length.

In algorithmic information theory, it is common to define probability in terms of description (shortest program) length. What is known as the universal distribution corresponds roughly to the latter of the equalities I just gave you. So, in a sense, information (ideal description length) "comes from" probability in Shannon's information theory, and probability comes from information (shortest-program length) in algorithmic information theory.

I'm sure I've just thrown way too much at some of you. But some of you have been in the ballpark with your remarks, and I hope this helps a bit.

Date: 2008/08/27 17:27:36, Link
Author: Turncoat
Almost all binary strings x have no program ("self-extracting zip archive") much shorter than themselves. A string that is in this sense incompressible is algorithmically random by definition. The definition makes sense, though, because an algorithmically random string x passes all computable tests of randomness. In essence, there is no effective procedure that allows you to say that the bits in x did not come from a random source.

The problem some of you have had in experimenting with compression of texts is that you haven't taken into account the fact that you're working with strings of (presumably) 8-bit characters. If I recall correctly, the bzip utility compresses to bit strings instead of character strings. This is what you need for your experiments.

For large N, a string of N digit-bytes (8 bits per byte) coming from a random source will compress from 8N bits (always) to about N log 10 bits (almost always). The high compression ratio comes only from the inefficiency of an 8-bit representation of textual digits.

There are similar considerations with Macbeth. There are not 2^8 = 256 letters, spaces, and punctuation marks in Shakespeare's text, so you get some compression simply because the initial 8-bit representation of characters is inefficient. At first blush, I would say that the number of distinct characters is no more than 64. Thus to get a better estimate of the compressibility of Macbeth, count 6 bits per byte of the source text, not 8. Count 8 bits per byte of bzip output. (Bzip may tell you the number of bits in the output -- I haven't used it in a long time.)

Date: 2008/08/27 18:08:12, Link
Author: Turncoat
Quote (Richardthughes @ Aug. 27 2008,17:00)
Thanks for your posts, turncoat. Now, if you can put your empirical hat on, why don't we see CSI calculations?

RIP those who were waiting with bated breath to see Dembski's computation of CSI for the bacterial flagellum.

To apply CSI, you have to come up with an upper bound on the probability that a configuration of matter arose by purely natural processes. In principle, that does not reduce to argument from ignorance. In practice, you run afoul of argument from ignorance whenever you go from something like coin tossing to the  flagellum. We know all we're going to know about outcomes of tossing a fair coin, but we don't know how much we might eventually know about natural processes giving rise to the flagellum. From a Bayesian perspective, scientific learning generally increases the probability of observed phenomena. How is Dembski going to place an upper bound on the probability a future model might assign justifiably to the emergence of the flagellum by natural processes?

I think that assignment of probabilities to explanations of historical events that occurred under largely unknown circumstances is a philosophical quagmire. Dembski complains that evolutionary models are not sufficiently detailed. I would contend that evolutionary models are correct in reflecting our ignorance of the details. We cannot turn the emergence of the flagellum into a repeatable, controlled experiment, and I have no idea how Dembski justifies his frequentism in the development of CSI.

Date: 2008/08/27 18:47:48, Link
Author: Turncoat
Quote (Henry J @ Aug. 27 2008,17:55)
Turncoat,

That's kind of what I was getting at in my post last night.

Henry

Yep. Hadn't read that far yet.

Date: 2008/08/27 18:53:58, Link
Author: Turncoat
Quote (Richardthughes @ Aug. 27 2008,18:35)
They have compressed it to "bac flag", though. Progress!
???

I saw "BF" yesterday. You may see that as further progress, but I say that when the information content of the message is zero, the waste of bandwidth is infinite.

Date: 2008/08/27 21:44:09, Link
Author: Turncoat
Quote (Richardthughes @ Aug. 27 2008,20:52)
All this experimenting is not very congruent with ID, folks. back to navel-gazing and hand-waving, please.

Have you forgotten how amazingly "creative" one can be with MatLab?

Date: 2008/08/27 22:26:05, Link
Author: Turncoat
Here's a little shell script for getting word counts (case insensitive) from a text:
Code Sample
#!/bin/sh

tr -d "'" |
tr -cs "[:alpha:]" "\n" |
tr "[:upper:]" "[:lower:]" |
sort |
uniq -c |
sort -rn |
awk '{ cum += $1; print  $1, cum, $2; }'


It says that Macbeth contains 18596 instances of 3379 distinct words. Here are the ten most frequent words:

740 740 the
579 1319 and
385 1704 to
368 2072 of
335 2407 i
284 2691 macbeth
253 2944 a
233 3177 that
207 3384 in
202 3586 you

The first column is word frequency, and the second is cumulative frequency. Applying this command
Code Sample
awk '{h += $1 / 18596 * -log($1/18596)/log(2)} END{print h}'

to the output, I get per-word entropy of about 9.357 bits. It would have been nice to treat punctuation marks as words, but I don't happen to have a script on hand that does that.

Date: 2008/08/28 01:05:51, Link
Author: Turncoat
Quote (Richardthughes @ Aug. 27 2008,22:32)
 
Quote
It would have been nice to treat punctuation marks as words, but I don't happen to have a script on hand that does that.


I'm guessing they have low compressibility given their single character nature? You could perhaps make capitalization at the start of a sentence a rule and then compress that way?

Periods and commas have low self-information (are given short encodings) because they are of high probability. Most capitalization can be recovered from punctuation. My transformation to lowercase of words that are always capitalized (e.g., "Macbeth") has no effect on the entropy calculation.

My brain is fried, so I'm not good for any real work. Perhaps I'll diddle a bit with punctuation.

Date: 2008/08/28 01:59:55, Link
Author: Turncoat
Here is a preprocessing script. Run Macbeth through this first, and then run the output through the filter (script) I gave above.
Code Sample
#!/bin/sh

# Spell out punctuation marks. Leave apostrophe as part of word.
# Treat end-of-line as punctuation for processing of verse.
# This works well for Shakespeare from Gutenberg Project.
# Manually remove copyright notice prior to each act.

sed '/^ *$/d'   |
sed "
s/'/papostrophe/g
s/,/ pcomma /g
s/\./ pperiod /g
s/\!/ pbang /g
s/\?/ pquestion /g
s/-/ phyphen /g
s/\[/ plbracket /g
s/\]/ prbrackt /g
s/;/ psemicolon /g
s/:/ pcolon /g
s/(/ plparenthesis /g
s/)/ prparenthesis /g
s/$/ pendofline/"

I decided that end-of-line should be treated as punctuation in verse. I also noticed that there were copyright notifications prior to Acts II-V that I had not removed in my previous analysis. Now there are 25,307 "words," and the per-word entropy is 8.228 bits. Yes, the per-word entropy went down because the empirical distribution concentrates  a great deal of probability mass on punctuation pseudo-words. Here are the twenty most frequent words and pseudo-words:

2625 2625 pendofline
1873 4498 pperiod
1646 6144 pcomma
736 6880 the
567 7447 and
385 7832 to
356 8188 of
318 8506 i
284 8790 macbeth
253 9043 a
238 9281 pquestion
229 9510 that
208 9718 psemicolon
208 9926 pbang
207 10133 in
202 10335 you
192 10527 my
183 10710 is
164 10874 not
155 11029 with

Date: 2008/08/28 15:55:23, Link
Author: Turncoat
Quote (Wesley R. Elsberry @ Aug. 28 2008,09:46)
That reminds me... has anyone seen the Marks/Dembski collaborations appear in print anywhere yet?

Though I suppose that if they do, notice is likely to be given the IDC equivalent of a ticker-tape parade, appearing on the DI blog, the ID-the-Future blog, UD, TT, and however many DO'L blogs there are at the time.

Nothing in print that I know of. The fact is that I'll be obsoleting anything they publish, anyway.

Date: 2008/08/29 00:59:42, Link
Author: Turncoat
Quote (Richardthughes @ Aug. 28 2008,15:58)
 
Quote (Turncoat @ Aug. 28 2008,15:55)
The fact is that I'll be obsoleting anything they publish, anyway.

*Proffers High-Five*  :p

My results are solid -- I'm preparing a presentation and a paper.

Date: 2008/08/29 14:22:40, Link
Author: Turncoat
Quote (Wesley R. Elsberry @ Aug. 29 2008,07:18)
     
Quote (Richardthughes @ Aug. 29 2008,01:31)
The math may be beyond me but I'm sure folks like Wes could offer you a good critique, if you'd like.

I can at least speak to Dembski's history of claims with respect to information theory and evolutionary computation. I was the guy who set him the task of explaining away EC back at the 1997 NTSE conference. My example there was a GA that produces short tours for the Traveling Salesman Problem. It's an example Dembski has never come to grips with. The class of problem, NP-hard, sets both intelligent agents and evolutionary computation on the same level, seeking approximate solutions rather than exact optimal solutions for any non-trivial number of cities in the tour. The evaluation function is too simple to even try to claim that the solution state is incorporated into it: total cost for each tour. Instead, Dembski has been making a career out of misunderstanding even Dawkins' pedagogical example, the weasel program. So far, he has not even managed to describe the weasel program correctly.

Personally, I would stay away from using solution of traveling salesperson problems with genetic algorithms as an example. TSP is the prototype of problems for which GA researchers have designed custom representations and reproduction operators. David Goldberg's Genetic Algorithms in Search, Optimization, and Machine Learning (1989) gives special attention to TSP.

The Dembski-Marks paper with the misrepresentation of the weasel program (and, trivially, the misspelling of "methinks" as "me thinks" in the target string) has been "forthcoming" in the Wistar Retrospective proceedings for quite a time. Dawkins' program is an instance of what folks in evolutionary computation know as a (1, ?)-ES [evolution strategy].

How does the weasel program know the space of strings on which fitness is defined? How does it know that fitness evaluation of some strings is not impractically slow? How does it know that lower "fitness" values are better than higher values? How does it know to terminate when it obtains a string of zero fitness? The weasel program, like most optimization programs, is highly informed of the optimization problem. (And I have said nothing about the fitness function guiding the program to the target string.)

Dembski is blinkered by his agenda, but he's not wrong in all points he raises. He got me to ponder what optimization procedures know about problems, and this led me to see that the NFL analytic framework predicates considerable prior knowledge. Take away some of that prior knowledge, and there are huge performance distinctions among optimization procedures. This is not a fine theoretical point. Wolpert and Macready's constraints on optimization problems are in no sense natural.

Wes, I told you last fall that I thought I had major results on the way. It turned out that I had to beat my sub-genius head against the wall quite a few more times to "break on through to the other side." At the moment, I'm preparing for a job interview, and I'm trying to work what I have into a crystal-clear presetation. When I'm through with that, I'll write a paper. I would appreciate it if you were to read a draft and give me feedback.

Date: 2008/09/01 00:10:35, Link
Author: Turncoat
Quote (Jkrebs @ Aug. 31 2008,09:50)
Damn science - always changing because there is new evidence!

When someone wrote, "I’ve been told by Darwinists they object to the term [Darwinism] because Darwin’s theory is 150 years old, there’s a lot that Darwin got wrong, and it implies there’s been no progress made since then. Neo-Darwinism or The Modern Synthesis is more acceptable",

StephenB replied,
   
Quote
Yes, when a theory is so fluid that it can add or delete certain features to accommodate new evidence, its earlier manifestations do have a way of becoming obsolete. So, finding the right label for it would seem to be an ongoing challenge. Can we settle on ECE (the ever-changing theory of evolution)?
Link

Yes, and ID theory includes a means of detecting supernatural causation with no false positives. No, wait....

Date: 2008/09/01 00:25:07, Link
Author: Turncoat
Quote (midwifetoad @ Aug. 31 2008,11:47)
What pisses the anti-evolutionists off the most is that the features they can't live with haven't changed since Darwin: the fact of common descent happening over hundreds of millions of years, and the lack of foresight in variation.

The YEC crowd can't accept the first, and the ID crowd can't accept the second.

What neither crowd can accept is the pan-Messiah. Here is the Second Coming on Planet Zarcon:

Date: 2008/09/02 03:34:20, Link
Author: Turncoat
Quote (Thought Provoker @ Sep. 01 2008,22:14)
To Liz Lizard (or alternate),

If you are here.  My compliments on your comment to UD.  I agree it was destined to get you banned but it was a very well presented point.

It shows that candid talk about religious motivations is an absolute "no no".

Saying something like "me thinks UD doth protest too much" is undoubtedly too much an understatement, especially in this thread.

Welcome to the very long list of those banned from uncommon descent.

P.S. It looks like I need to modify this.  The discussion that followed on UD did attempt to address it.  Having a code of ethics can be a pain at times.     :(

I suspect that Elizabeth "Lizard" Bailey knows a bit of Aikido, but lost her cool in the heat of the battle. She should have highlighted Fuller's contradictions of Dembski, reiterated her admiration for Fuller, and left it at that. But she might contend that one never really loses when provoking the Expelled to show how readily he expels.

I am considering traveling to Austin next weekend to visit with family, and then swinging through Waco on Tuesday for ASA executive director Randy Isaac's presentation. No matter that Fuller slams theistic evolution, the ASA should, in my opinion, push Fuller's "honest ID" as hugely preferable to the Discovery Institute's downgraded creationism.

Something many of you don't understand, I think, is just how highly Christian conservatives value honesty and fairness. If you do nothing but slam ID, which they see as obviously true, you make them believe that you are unfair, if not evil. If you instead find some good things to say about a brand of ID that emphasizes "trying to get into the mind of God," and that calls for open discussion (free expression) of the religious views of scientists, and that does not attempt to prove what most of them believe should be accepted on faith anyway, you are not likely to be taken for an atheistic dogmatist.

Liz Bailey thinks somewhat more highly of Fuller than I do. But I see considerable strategic value in lauding someone who insists that belief in ID is a heuristic guiding research, and not something to be proven correct by means of research. There's no creationism in Fuller's ID that I can make out. And the fact that Fuller served as an expert witness for the defense in the Dover case is something to be taken advantage of, not given as an excuse for an antigen response.

Date: 2009/01/07 16:33:30, Link
Author: Turncoat
Quote (J-Dog @ Jan. 07 2009,13:59)
     
Quote (Albatrossity2 @ Jan. 07 2009,13:30)
       
Quote (J-Dog @ Jan. 07 2009,13:22)
"Wayne J. Downs holds an M.Div. from George W. Truett Theological Seminary at Baylor University. His long-term research interest lies at the intersection of cognitive science and philosophy, and focuses on the dynamics of pain in its relation to establishing moral values. After completing his doctoral studies, he hopes to teach philosophy and religion at the college or university level. He is program director of the Trinity Institute, a nonprofit Christian retreat and study center located in Tehuacana, Texas. "

In other words, his "research" probably focuses on whether spanking before, after or during church services results a better establishment of "moral values."

I wonder if the Templeton Foundation would be interested in funding that stuff...

I think there has been quite a bit of research in this area already. :)

Kind of appropriate that Demsbki and his merry band of sycophants are right on the cutting edge of science that goes back to Paley...again.

Earlier Research - IDCist Scooped Again

But the research was conducted by atheistic materialists who wouldn't know a moral if it bit them on the ass. They never considered that parental savagery might have life-long spiritual benefits for children even when no short-term behavioral benefits relative to non-corporal punishment were evident. (HT: Denyse O'Leary)

Seriously, I wish I hadn't seen this. It makes polite exchanges with IDists awfullly hard. The oxymoron of Christian violence is always tough for me to abide, but when the violence is directed at children, I tend to go ballistic (yes, verbally violent - no excuse).

Does Dembski beat his "developmentally slow" son extra hard to make sure he turns out moral? I doubt it. He may take the child to the faith healer, but I'll bet that he also avails himself of all the non-mentalistic neuroscience explaining development and behavior in terms of the material brain he can come by. If he had an ethical bone in his body, he'd have taken O'Leary to task for her simplistic bullshit long ago.

Date: 2009/05/04 03:09:11, Link
Author: Turncoat
Quote (Quack @ May 04 2009,02:21)
   
Quote (sTARTERkIT @ May 04 2009,00:55)
Ah, but ice is water in a straight jacket.

SSSSoooooo, water by any other name.............

Try again.


         
Quote
yeah but how does the water KNOW to be ice.  if that ain't proof of the god of the bible i don't know what is!

Seems to this old totter that Dembski is on to something akin to the principles that makes
homeopathy work!

Makes me think of chiropractic.

   
Quote
Jailed: No Innate Intelligence Allowed

“Innate Intelligence is a chiropractic term to describe the organizing properties of living things. [...] This vitalistic concept states that all life contains Innate (inborn) Intelligence and that this force is responsible for the organization, maintenance and healing of the body.”

“It was presented by early chiropractic leaders as a part of chiropractic philosophy, that life is a triune of intelligence, force, and matter. [...] Because of this early metaphysical construct, the terminology of Innate Intelligence is considered potentially detrimental to the profession’s development and reputation as it seeks acceptance in the greater scientific community.”

[...]

“Universal Intelligence is in all matter and continually gives to it all the properties and actions. The expression of this intelligence through matter is the Chiropractic meaning of life; therefore life is necessarily the union of intelligence and matter. Force unites intelligence and matter. Universal Intelligence gives force to both inorganic and organic matter. That force which Universal gives to organic matter as a higher order of its manifestations, is called Innate Intelligence.”


There really was a time when chiropractors were jailed. More at UD.

Date: 2009/08/12 17:31:25, Link
Author: Turncoat
Yeah, Clive has blown a gasket in the "Barrett Clown" / "you can call me GEM, but you doesn't have to call me Mullings" thread. He deleted a fair comment of mine -- and I hadn't saved a copy. "You know, of course, this means war." Furthermore, "I'll be back."

Synopsis:

O'Leary not only participated in the outing of the Canadian Cynic, but chimed in with her approval when the Blog Czar demanded that all critics of Steve Fuller give their real names.

There's an odd asymmetry in views on anonymity at UD. The presumption is that ID critics want to talk trash and not have the smell stick to them, while ID advocates must avoid persecution for their views. We have a pretty good idea at this point of the identity of the "Rembrandt of flash animation" who rendered Judge Jones, and the only thing he had to fear was the embarrassment of having folks find out how someone of his professional stature spends his time.

Date: 2009/08/25 16:00:26, Link
Author: Turncoat
FWIW, a response to Dembski at NewScientist I don't mind owning up to.

Date: 2009/08/25 16:40:06, Link
Author: Turncoat
Quote (oldmanintheskydidntdoit @ Aug. 25 2009,14:07)
Gordon Mullings does not know when to shut up
   
Quote
I noted earlier that no sensible rape complainant in Barbados shows up in court dressed in a micromini.

BillB plainly thought that his had to do with my prejudices — but I was simply a spectator, not a participant in the trial. My opinion had zero weight on the outcome of the case.

(Let’s just say that I sat in a waiting room across from a late 20’s young miss dressed in a greyish, plaid micro mini, black sheer stockings and a low cut top, and chatted with a couple of friends in for a traffic violation. The topic came up as to why I was there, and I said that I was there because of a rape case accusation against a former student who 300 mi from home was on a rape accusation by another student. Imagine my shock when I got into the courtroom, to see the same young miss as the complainant! Her lawyer should have been fired from the bar for failing to advise his client on suitable court dress, especially with so weak a case as she actually had.)

The actual case was stopped by the Judge as unsafe to further pursue when it turned out after a couple of hours of evidence, that the claimed crime occurred in a bed room in a rooming house full of students with no-one else in the house at the time reporting any untoward sounds or circumstances. A complainant who presented herself as she did, and whose case was as weak as that effectively asked for such a result.

Fuck off and die Gordon Mullings.

Just wait until a Homeland Security interrogator leaves you bleeding out the ass, Gordon E. Mullings of Montserrat, and a judge gives no credence to a foreigner who works 24x7 to document his manic delusions on the Web.

A woman who was sexually abused as a child may dress provocatively when it is inappropriate, as well as exhibit a propensity for getting herself into situations where she is at risk of rape.

Grandiloquent dumb-shit.

Date: 2009/09/05 22:54:46, Link
Author: Turncoat
Quote (Maya @ Sep. 04 2009,16:40)
Quote (someotherguy @ Sep. 04 2009,16:37)
All of us could vow to never make socks at UD again because, really, getting your laughs from watching a bunch of moderately crazy strangers act like idiots is kind of immature

I'll stop mocking them when they stop voting.

Inspiring words. I'm bored with UD, but digging handy trench latrines for the neo-creos to lay tard on the front is a civic duty. I'll be back.

Dog and Country!

Date: 2009/09/06 13:48:45, Link
Author: Turncoat
Quote (Bob O'H @ Sep. 06 2009,04:26)
Ha!    
Quote
20

BillB

09/06/2009

2:22 am

   
Quote
Give us your real name, as I have done. Otherwise, please do us the favor of blessing us with your absence.


GEM of TKI to that



I'll just comment that I do think this Gil baiting is a bit unkind.  Whilst some of his ideas are a bit, um, odd, he doesn't come across as being malicious or nasty.  And he has at least made something of his off-line life.

Now jerry, on the other hand, deserves the vitriol aimed at him.

Gil is highly narcissistic, in the clinical sense. I would not hazard to say from online interactions if a clinician would diagnosis him as a full-blown narcissistic personality, but a number of the signs are there. Narcissists commonly present fine fronts in public, and brutalize their families at home. They are in fact malicious and nasty.

What's missing from the Dodgenator 3000 is Gil's frequent self-description as expert in one of a rapidly growing list of fields. At best I can tell, he's an expert if he reads a book about something (e.g., computational number theory) and then writes a program related to it. I think this earns him some baiting.

As for making something of his off-line life, Gil has moved from tweaking a checkers-playing program to tweaking the trajectory of a parachute and its military payload. The checkers program entertained some people, and may have contributed indirectly to science by offering competition to programs developed by actual researchers in AI. The trajectory-control program serves primarily to hurt people. What would Jesus do with a computer?

Date: 2009/09/06 22:05:08, Link
Author: Turncoat
Quote (khan @ Sep. 06 2009,18:09)
 
Quote
Gil is highly narcissistic, in the clinical sense. I would not hazard to say from online interactions if a clinician would diagnosis him as a full-blown narcissistic personality, but a number of the signs are there. Narcissists commonly present fine fronts in public, and brutalize their families at home. They are in fact malicious and nasty.

What's missing from the Dodgenator 3000 is Gil's frequent self-description as expert in one of a rapidly growing list of fields. At best I can tell, he's an expert if he reads a book about something (e.g., computational number theory) and then writes a program related to it. I think this earns him some baiting.

As for making something of his off-line life, Gil has moved from tweaking a checkers-playing program to tweaking the trajectory of a parachute and its military payload. The checkers program entertained some people, and may have contributed indirectly to science by offering competition to programs developed by actual researchers in AI. The trajectory-control program serves primarily to hurt people. What would Jesus do with a computer?


I spent ~20 years tweaking code, but I can't play piano and don't have frilly shirts.

My first master's thesis was a collection of 40 poems along with a poetics. But I never wore a frilly shirt. I suppose that if I had so little shame as to publish my thesis collection at my website, I might have little enough shame to pose in a frilly shirt.

Date: 2009/09/21 16:07:11, Link
Author: Turncoat
Quote (Quack @ Sep. 20 2009,09:18)
If my understanding is right, another facet of the concept of fitness in evolution is that it works even with a moving target. Evolution has no target; it is always about immediate fitness. As long as the fitness landscape don't change too fast. Otherwise, extinction.

In a nutshell what Darwin said 150 years ago - and still going strong.

ETA:

Shouldn't it be possible to write a version of Weasel with moving target - and wouldn't that prove unworkable with latching?

I proposed this at UD some months back. There's a very simple approach. You use exactly the same routine to mutate the target sentence in each generation as you do to mutate each copy of the parent sentence. This models environmental drift. The parent sentence generally will track the environment well from generation to generation, provided the environmental mutation rate is sufficiently low. "Sufficiently low" depends on both the mutation rate for offspring and the number of offspring per generation. For instance, you can set the mutation rate for the environment equal to that for the offspring if you make the number of offspring large.

Date: 2009/09/22 00:02:46, Link
Author: Turncoat
Quote (Wesley R. Elsberry @ Sep. 20 2009,16:49)
               
Quote (Zachriel @ Sep. 20 2009,14:44)
                 
Quote (DiEb @ Sep. 20 2009,14:03)
2: For his interview, Dawkins needed the program to run for ~ 2000 generations. This could be achieved by the combination (10 children, 4% mutation rate) But I suppose that Dawkins just fooled around a little bit with his program to get an optimal number of runs, i.e., the program was running during the length of his interview…

Is that correct? Dawkins seems to be showing all the children, not just the parents. It's slower because it takes time to display on the antique system he's using. [...]

I think the number 2485 comes up at the end of the video as the number of individuals. If that is the case, Dawkins likely did have to find fairly particular parameters in order to terminate in the short time of the video sequence, and the slow display system likely did have an impact on that. I think I posted some numbers here before on the likely parameter space the video run's parameters were taken from.

In contrast, the mean number of trials for the simple hill-climber encoded in Weasel2 to reach the target is about 5800. The closer Weasel2 gets to the target, the greater the expected number of trials for an uphill step on the fitness landscape. The expected number of trials to reach the target is in the ballpark of 2485 only when 25 or 26 of the characters in the best-so-far sentence match the target.

1. Weasel2 never takes a downhill step.

2. Weasel2 mutates the character in just one position in each trial. If the character is already correct, the mutation cannot yield an uphill step. The probability of selecting an incorrect character for mutation is (28 - n) / 28, where n is the number of correct characters. The probability that mutation yields the correct character is 1 / 27. Thus the probability of an uphill step when n characters are correct is p = (28 - n) / 28 / 27.

3. The expected number of trials for an uphill step is 2 * (1 - p) / p. This is the mean of a random variable following a negative binomial distribution. See Wikipedia.

4. The number of correct characters in the randomly initialized sentence is binomially distributed with mean 28 * (1 / 27) = 1.037 and variance 28 * (1 / 27) * (1 - 1 / 27) = 0.999. That is, the mean and variance are both about 1, and 25 correct characters in the initial sentence (see above) would be about 24 standard deviations above the mean.

5. Here is code and example outputs for the Unix bc utility. The function f takes as its argument the number of correct characters in the initial parent, and returns the expected number of trials to reach the target. The "for" loop sums the expected numbers of trials for all the uphill steps required to maximize fitness. (The expectation of a sum of random quantities is the sum of the expectations.)

Code Sample
scale=25

define f(m) {
  auto p, n, t

  t = 0

  for (n = m; n < 28; n++) {
     p = (28 - n) / 28.0 / 27.0
     t += 2 * (1 - p) / p
  }

  return(t)
}

f(27)
1510.0000000000000000000586656

f(26)
2264.0000000000000000000594216

f(25)
2766.0000000000000000000662760

f(0)
5881.8826109171485399780675527

f(1)
5829.8826109171485399780674987

f(2)
5775.8826109171485399780673643

f(3)
5719.7287647633023861319134600

(f(0)+f(1)+f(2)+f(3))/4.0
5801.8441493786870015165289689


I could go further with the analysis, but I don't see why I should. It is highly improbable that Weasel2 was at work in the BBC show.

Date: 2009/09/22 00:15:59, Link
Author: Turncoat
For any IDiots lurking, what I just accomplished was not a design inference. I analyzed the performance of a program for minimizing the Hamming distance of a string of characters from a target string, and determined that it is highly improbable that the program performs as well as claimed.

It took me much longer to attempt to explain the analysis clearly than actually to do it. Dembski should be embarrassed to have mass-distributed tard mail without scrutinizing Weasel2 for himself.

Date: 2009/09/22 01:06:58, Link
Author: Turncoat
As for Weasel1, how does it happen that its initial parent sentence matches up with the initial sentence output by the single-step selection program? Let's call that other program, for which we don't have code, Monkey.

Weasel1 calls the Randomize procedure to automatically seed the random number generator. This means that you would have to run the program many times to stand much chance of seeing an initial parent sentence you've seen before.

How does Monkey come up with the same initial sentence as Weasel1? A call to Randomize is very unlikely to generate the same seed that was automatically generated in Weasel1. And Dawkins did not know what seed was automatically generated, because Weasel1 does not supply it as output. Did Dawkins explicitly enter the first sentence output by Weasel1 into Monkey? In other words, did Dawkins "smuggle information" into the program by supplying the first sentence he attributed to a "monkey at a typewriter"? I doubt it.

(Yes, I know about the one-character discrepancy in sentences. I think it's safe enough to attribute it to a typo.)

If we really must guess about programs, rather than ask Dawkins to clarify the algorithm he had in mind, then it makes sense to guess that Dawkins reused code in some way or another and seeded the random number generator explicitly. Personally, I would have written just one evolution strategy, and would have set both the per-site mutation rate and the number of offspring to 1 to model the monkey at a typewriter.

Date: 2009/09/22 16:13:29, Link
Author: Turncoat
Quote (Wesley R. Elsberry @ Sep. 20 2009,16:49)
Quote (Zachriel @ Sep. 20 2009,14:44)
 
Quote (DiEb @ Sep. 20 2009,14:03)
2: For his interview, Dawkins needed the program to run for ~ 2000 generations. This could be achieved by the combination (10 children, 4% mutation rate) But I suppose that Dawkins just fooled around a little bit with his program to get an optimal number of runs, i.e., the program was running during the length of his interview…

Is that correct? Dawkins seems to be showing all the children, not just the parents. It's slower because it takes time to display on the antique system he's using. [...]

I think the number 2485 comes up at the end of the video as the number of individuals. If that is the case, Dawkins likely did have to find fairly particular parameters in order to terminate in the short time of the video sequence, and the slow display system likely did have an impact on that. I think I posted some numbers here before on the likely parameter space the video run's parameters were taken from.

My previous comment regarding Weasel2 was wrong. I should have used Wolfram's MathWorld rather than Wikipedia to get the mean of the negative binomial distribution. Wikipedia explains the r parameter incorrectly. The mean number of trials required for an uphill step is (1 - p) / p, not 2 * (1 - p) / p.

The mean number of trials for Weasel2 to reach the target is about 2900. In simulation, 12% of runs require 2485 or fewer trials.

My apologies for not double-checking my work prior to posting here.

Date: 2009/09/23 02:48:49, Link
Author: Turncoat
Quote (didymos @ Sep. 22 2009,20:30)
Personally, I think it's high time Clivebaby started quoting Huey Lewis at every opportunity.  Or Emmanuel Lewis.

"I want a new drug. One that won't make me sick."

Date: 2009/09/24 13:50:49, Link
Author: Turncoat
Zachriel,

FYI, the standard ONEMAX problem is to maximize a function that counts the number of 1's in a length-N binary string over alphabet {0, 1}. Generalization to a function that counts the number of bits matching those in an arbitrary "target" string in {0, 1}^N is trivial and inconsequential. Generalizing further to non-binary alphabets gets you the problem addressed by the Weasel program.

Stefan Droste has analyzed the performance of a (1+1) evolutionary algorithm (one parent competes with its one offspring for survival) in the case that a binary target string "wanders" precisely as I suggested.

Quote
Analysis of the (1+1) EA for a Dynamically Bitwise Changing ONEMAX

Although evolutionary algorithms (EAs) are often successfully used for the optimization of dynamically changing objective function, there are only very few theoretical results for EAs in this scenario. In this paper we analyze the (1+1) EA for a dynamically changing OneMax, whose target bit string changes bitwise, i. e. possibly by more than one bit in a step. We compute the movement rate of the target bit string resulting in a polynomial expected first hitting time of the (1+1) EA asymptotically exactly. This strengthens a previous result, where the dynamically changing OneMax changed only at most one bit at a time.


This comes from a Dagstuhl workshop, "Theory of Evolutionary Computation." (I got an invitation, but was underemployed at the time, and had committed already to two overseas trips. Dammit.) I can't find the full text online.

Do results for the binary case generalize easily to the non-binary case? I don't know.

Date: 2009/09/24 14:53:09, Link
Author: Turncoat
Quote (Maya @ Sep. 24 2009,14:25)
Dembski, the weasel, puts code in Dawkins' mouth:

From
   
Quote
When I contacted Richard Dawkins to confirm their authenticity, he replied, in an email dated 9.21.09, “I cannot confirm that either of them is mine. They don’t look familiar to me, but it is a long time ago. I don’t see what more I can say.”

the loud-mouthed but gutless Dover no-show gets to
   
Quote
We therefore conclude, unless further evidence is presented, that the single-mutation algorithm implemented by WEASEL1 is the one used by Dawkins in TBW.

in an almost Gordon Mullings-like spew of illogic and baseless assertions.

Any chance of bothering Dr. Dawkins again about this, Mr. Elsberry, sir?

I had guessed that Dawkins would reply to everyone that Dembski addressed. I'd like to know at least he gist of what Dembski left out. If Dawkins did not emphasize that it's the principle, and not the program, that matters, I'd be surprised.

But something really surprises me in Dembski's quote of Dawkins. If Dawkins had ever programmed in Pascal with "caps lock" on, he would remember it. I've never known a Pascal programmer who would go from a mixed-case style in one year to an execrable all-caps style in the next. If you're in the habit of using mixed case, you can't stand to look at all-caps code. And it simply does not take that much time to use the shift key.

Date: 2009/09/24 15:43:56, Link
Author: Turncoat
Quote (Maya @ Sep. 24 2009,14:48)
Dembski claims "These programs were widely circulated at the time." but a search for any of the unique strings within them turns up just one hit:  Dembski's UD post.

Is he really stupid enough to a) think that he can just make stuff like this up and b) not realize that neither of these faux weasels correspond to his claims in his paper with Marks?

Okay, enough for today, back to real biology.

Sorry, Maya, but R0b is right in pointing out that neither Weasel implements partitioned search. The apocryphal Weasels do comport ever so wonderfully with other dumb shit Dembski and his acolytes have said over the years, however.

I'm not sure that the claim that the Weasel of the BBC Horizons program worked differently from the Weasel in TBW originated with Dembski. But Dembski never passes on an opportunity to make Dawkins out to be a cheat. IIRC, the key "evidence" is the mention of "tries" on the screen. This does not seem at all odd to someone who has compared the efficiency of evolutionary optimizers using populations of different sizes. The number of trials is the number of generations multiplied by the population size. Dawkins' random search essentially has a population size of 1, and Weasel program has a population size larger than that. To make a fair comparison of the two programs, Dawkins had to report numbers of trials for the Weasel program rather than number of generations.

BTW, I'm not really a fan of Dawkins, but he does impress me with his honesty and straightforwardness. Some years back, I sent Dembski email to explain how he'd gotten a fundamental aspect of the "no free lunch" theorems wrong. His response was something like "OK, but don't expect me to admit that." I had previously sent email to Dawkins to explain that he was wrong in some claim regarding the effect of mutation on the Kolmogorov complexity of a genotype. His response was something like "I think you may be right." He has since fixed the online article in which I spotted the problem.

Dembski constantly projects his own worst traits onto his adversaries.

Date: 2009/10/02 02:23:23, Link
Author: Turncoat
The inaugural post of Bounded Science: No Free Lunch for Intelligent Design, "Never Look a Gift Weasel in the Mouth," needs an infusion of wit. Come on over.

Date: 2009/10/02 12:39:49, Link
Author: Turncoat
Quote (CeilingCat @ Oct. 02 2009,06:58)
Quote (Turncoat @ Oct. 02 2009,02:23)
The inaugural post of Bounded Science: No Free Lunch for Intelligent Design, "Never Look a Gift Weasel in the Mouth," needs an infusion of wit. Come on over.

I'd like to try, but comments don't seem to be enabled.

I'm pretty sure Weasel 1 is not the program Dawkins used because IIRC he says in the book that he wrote it in Apple BASIC and then re-wrote it in Pascal.

I think that's been mentioned on UD once or twice too, but trust Dembski and company to miss and/or ignore that.

The "comments" link after the article does not stand out.

If you need an OpenID, look here.

Date: 2009/10/02 14:11:33, Link
Author: Turncoat
Quote (Richardthughes @ Oct. 02 2009,10:15)
Quote (Turncoat @ Oct. 02 2009,02:23)
The inaugural post of Bounded Science: No Free Lunch for Intelligent Design, "Never Look a Gift Weasel in the Mouth," needs an infusion of wit. Come on over.

Nice. Do you, Wes and DvK correspond?

Not on this. DiEb has a nice series of posts at DiEblog.

Date: 2009/10/05 17:57:37, Link
Author: Turncoat
The latest at Bounded Science:
 
Quote
Resolving a moral dilemma

I made a promise to Bob Marks that I would not divulge my correspondence with him regarding drafts of the paper that IEEE SMC-A published last month. But I did not know that he and Dembski would resort to trickery to get the paper published, and, after considerable agonizing, I've decided that the better course is to break my word.

False attribution of partitioned search to Dawkins is the not the full extent of the academic dishonesty in the article.

Date: 2009/10/05 18:56:29, Link
Author: Turncoat
Quote (sledgehammer @ Oct. 05 2009,18:06)
Quote (Turncoat @ Oct. 05 2009,15:57)
The latest at Bounded Science:
         
Quote
Resolving a moral dilemma

I made a promise to Bob Marks that I would not divulge my correspondence with him regarding drafts of the paper that IEEE SMC-A published last month. But I did not know that he and Dembski would resort to trickery to get the paper published, and, after considerable agonizing, I've decided that the better course is to break my word.

False attribution of partitioned search to Dawkins is the not the full extent of the academic dishonesty in the article.

This is the first I've heard of sneaky tactics or trickery in the path to publication.  Any details you can share? Give us dirty laundry!  We like dirty laundry.

Another blog entry is on the way. Omitting the names of the two evolutionary algorithms, as well as neglecting to cite the relevant literature, was a trick to keep the editors and reviewers from scrutinizing the redundant and/or worthless analysis Dembski and Marks provided. If there had been explicit mention of evolutionary algorithms, the editor-in-chief might have handed the paper off to a different associate editor, and the associate editor might have lined up better-informed reviewers.

Date: 2009/10/06 10:39:35, Link
Author: Turncoat
Quote (BillB @ Oct. 06 2009,00:31)
 
Quote (sledgehammer @ Oct. 05 2009,19:55)
That and "cleverly" hiding "Evolutionary Informatics" and "this is designed" in the starting strings for the partitioned search.
But I thought the most egregious error was characterizing evolutionary search as "monkey at a typewriter", the only entry in the section labeled "Critiquing Evolutionary Search Algorithms".
That should have been caught by the reviewers IMHO:

       
Quote

4 Critiquing Evolutionary Search Algorithms

Christensen and Oppacher note the sometimes-outrageous claims that had been made of specific optimiza-
tion algorithms" [7]. Their concern is well founded. In computer simulations of evolutionary search,
researchers often construct a complicated computational software environment and then evolve a group of
agents in that environment. When subjected to rounds of selection and variation, the agents can demon-
strate remarkable success at resolving the problem in question. Often the claim is made, or implied, that
the search algorithm deserves full credit for this remarkable success. Such claims, however, are often made
(1) without numerically or analytically assessing the endogenous information that gauges the difficulty of
the problem to be solved, and (2) without acknowledging, much less estimating, the active information that
is folded into the simulation for the search to reach a solution.

4.1 Monkey at a Typewriter
A "monkey at a typewriter" is often used to illustrate the viability of random evolutionary search. It also illustrates the need for active information in even modestly sized searches...

(my bold)
WTF is "random evolutionary search"?  Since "Active Info" is defined by them as the improvement over random search, so that last sentance is bullcrap.

Edited just because.

I'm not so sure the intention was to characterise evolutionary algorithms as a "monkey at a typewriter", although their wording is a bit cloudy.  They may be referring to Dawkins use of WEASEL where he compares it to the idea of moneys at typewriters in order to demonstate how cumulative selection gets round the proplem of a purely random sampling

Dawkins compared the number of trials the Weasel program required to obtain the target sentence to the expected number of trials for the "monkey at a typewriter" benchmark, uniform sampling with replacement. Dembski and Marks do much the same. They compare the probability of hitting the target with the alternative search to the probability of hitting the target with the benchmark search. That is, they take the ratio of the former to the latter.

For no good reason I have been able to discern, despite spending a lot of time looking for one, they take the logarithm of the ratio. (It looked reasonable at first.) The result is "active" information. Talk about a loaded and stupid term. There is nothing "natural" about uniformity in physical processes, and deviation from uniformity does not imply that something has acted to inform the process.

Date: 2009/10/06 10:50:36, Link
Author: Turncoat
Quote (sledgehammer @ Oct. 06 2009,01:36)
I see your point, Bill.  On a charitable re-read from that perspective, I concede that M&D might be trying to say something like "Monkeys at typewriters are sometimes used as a contrasting counterexample to demonstrate the effectiveness of evolutionary algorithms"
That still doesn't explain their use of the term "random evolutionary search", not to mention that it's immediately followed by that "mass of the earth" and "planet destroying" 147 bits hokum.  They then proceed to show how partitioned search can solve the problem "in milliseconds".  Nowhere do I see evolutionary algorithms critiqued, much less discussed, unless they somehow consider partitioned search as an example of an evolutionary algorithm.  I suppose that would be in line with their mistaken interpretation of Weasel.
 That plus the section title, "Critiquing Evolutionary Search Algorithms" followed by a single section titled "Monkey at a Typewriter" seems too deliberate from my admittedly cynical perspective.

They do analyze two evolutionary algorithms, but do not identify them as such, and do not cite the appropriate literature. I believe they wanted to avoid proper scrutiny of their results. See the email note to Marks explaining one of the algorithms and the comments on academic dishonesty at Bounded Science.

Date: 2009/10/06 10:59:03, Link
Author: Turncoat
Quote (Wesley R. Elsberry @ Oct. 06 2009,05:42)
Quote (Turncoat @ Oct. 05 2009,18:56)
 
Quote (sledgehammer @ Oct. 05 2009,18:06)
   
Quote (Turncoat @ Oct. 05 2009,15:57)
The latest at Bounded Science:
           
Quote
Resolving a moral dilemma

I made a promise to Bob Marks that I would not divulge my correspondence with him regarding drafts of the paper that IEEE SMC-A published last month. But I did not know that he and Dembski would resort to trickery to get the paper published, and, after considerable agonizing, I've decided that the better course is to break my word.

False attribution of partitioned search to Dawkins is the not the full extent of the academic dishonesty in the article.

This is the first I've heard of sneaky tactics or trickery in the path to publication.  Any details you can share? Give us dirty laundry!  We like dirty laundry.

Another blog entry is on the way. Omitting the names of the two evolutionary algorithms, as well as neglecting to cite the relevant literature, was a trick to keep the editors and reviewers from scrutinizing the redundant and/or worthless analysis Dembski and Marks provided. If there had been explicit mention of evolutionary algorithms, the editor-in-chief might have handed the paper off to a different associate editor, and the associate editor might have lined up better-informed reviewers.

I had already blocked out to have a section on the neologism issue in my response, but hadn't really taken it to the academic dishonesty conclusion. I was thinking more in terms of how it was another way that poor scholarship was expressed.

I'm a bit worried that lodging academic dishonesty complaints will fire up the old "expelled" propaganda mill. I really will listen to arguments against doing it.

Date: 2009/10/06 13:37:09, Link
Author: Turncoat
Quote (BillB @ Oct. 06 2009,11:56)
The alternative I suppose is to refute the paper from a technical standpoint through academic journals, and try and hint at the possible dishonesty through that means.

It often takes more words to rebut a smooth creationist argument than are in the argument itself. It's hard to explain what's wrong with some fundamental things that Dembski and Marks breeze right through, like the notion of search.

In the IEEE Computational Intelligence Society, which includes investigators in evolutionary computation, rebuttals go in a peer-reviewed "magazine," not the primary journals, and manuscripts are limited to 8 double-spaced pages. The notorious information-smuggling president of society, David Fogel, tells me that the rebuttals are of articles in journals published by our society, not those of other societies (e.g., Systems, Man, and Cybernetics). I need to check with the editor of the magazine to see if there might be an exception. I don't know him, and the matter is politically sensitive.

I have to decide what strong points I can get across in very few words, and forget the others, even if I think they're very important. I frankly am very poor at disciplining myself to do that.

Date: 2009/10/06 16:11:07, Link
Author: Turncoat
Quote (carlsonjok @ Oct. 06 2009,15:40)
Quote (Turncoat @ Oct. 05 2009,18:56)
     
Quote (sledgehammer @ Oct. 05 2009,18:06)
       
Quote (Turncoat @ Oct. 05 2009,15:57)
The latest at Bounded Science:
               
Quote
Resolving a moral dilemma

I made a promise to Bob Marks that I would not divulge my correspondence with him regarding drafts of the paper that IEEE SMC-A published last month. But I did not know that he and Dembski would resort to trickery to get the paper published, and, after considerable agonizing, I've decided that the better course is to break my word.

False attribution of partitioned search to Dawkins is the not the full extent of the academic dishonesty in the article.

This is the first I've heard of sneaky tactics or trickery in the path to publication.  Any details you can share? Give us dirty laundry!  We like dirty laundry.

Another blog entry is on the way. Omitting the names of the two evolutionary algorithms, as well as neglecting to cite the relevant literature, was a trick to keep the editors and reviewers from scrutinizing the redundant and/or worthless analysis Dembski and Marks provided. If there had been explicit mention of evolutionary algorithms, the editor-in-chief might have handed the paper off to a different associate editor, and the associate editor might have lined up better-informed reviewers.

Let me play the Devil's (heh!) Advocate here. You are assuming that D&M resorted to trickery to get their paper published.  I will grant that it is a safe assumption given those two's history (e.g. trying to sneak Dembski into the EIL ,and by extension back into Baylor, as a post-doc.)  However, you might ask yourself if the omissions you see as proof of academic dishonesty could, when divorced from the backstory, also be interpreted as academic sloppiness?  If there is latitude for someone at IEEE, who is not familiar with the evolution culture war like we are, to come to a different conclusion, then you might tread carefully about ascribing motive.

I am not saying that they weren't trying to pull a fast one and I am not saying you should not call them out for academic dishonesty.  But, just be aware that your reputation may be as much on the line as theirs.

Additionally, from my perspective*, the greatest indictment against Dembski and Marks is their willful misrepresentation of WEASEL as a partitioned search. There is a long, documented history of them being told otherwise. You specifically told Marks that it wasn't.  And Wes Elsberry had told the same to Dembski 9 years ago.

* Granted, I am not an expert in these technical subjects. Indeed, I don't even rise to the level of amateur.  But, I can understand the implications of their continued misrepresentation of WEASEL without having it stated specifically.

See what Shallit had to say about acknowledging priority at Recursivity. He seems to agree that it's a pretty big deal to leave out mention of prior work you know about.

Dembski and Marks certainly repeat published analysis in the case of the (1+1)-EA [the one with "elitism"]. Prior results make the approximation they derive for the (1,2)-EA essentially worthless. DiEb has done Markov modeling of the Weasel program as a (1,n)-EA [what most of us think it is, whether we know the nomenclature or not] to generate the wonderful plots at his blog. This is what Dembski and Marks evidently were unable to do for even the (1,n)-EA searching for a target sentence over a 2-character alphabet, with n restricted to 2. In other words, Dembski and Marks have little to show in comparison to what DiEb put out on his blog.

Hope that made sense. I've got to rush off to keep a date with my bicycle.

Date: 2009/10/08 01:29:53, Link
Author: Turncoat
Quote (REC @ Oct. 06 2009,22:05)
"Jeff Shallit charges Jonathan Wells with incompetence for claiming that duplicating a gene does not increase the available genetic information."

Dr. D decides to step it in by listing every type of information theory known to him. But logically, if a gene is duplicated, the duplicate serves as a rich founding source of genetic information. Note the word AVAILABLE. It can mutate and change in ways not constrained by the original function. The identical, duplicate gene may have no additional information at t=0 (depending on which rabbit you pull out of the hat to parse that), but if evolution has two templates to modify and select, thats TWICE as damn many as one.

In the original context-Hox gene evolution (see below) it is clear that this is the point being made.  The many  HOX duplications and diversification allow for the regulation of diverse body plans.

http://www.evolutionnews.org/2009....an.html

"The questioner became agitated and shouted out something to the effect that HOX gene duplication explained the increase in information needed for the diversification of animal body plans. I replied that duplicating a gene doesn’t increase information content any more than photocopying a paper increases its information content"

Now available at the UD page through Google Sidewiki:
Quote
Describe the population, not the offspring

Let's think of a simple biological type. Each individual has a single chromosome. Let's say also that there are only two distinct chromosomes in the population. Each individual has either chromosome X or chromosome Y.

You cannot describe the population of chromosomes just by writing out the genetic spellings of X and Y. You have to indicate how many X's there are, and how many Y's. Thus there is more information in the population than there is in X and Y. If an X-individual produces an offspring that is a perfect copy of itself, then the number of X's increases. There is generally more information in bigger numbers (ranging over the set of non-negative integers), whatever the measure of information.

The numbers matter immensely in biology. If we consider now random mutation of chromosomes in reproduction, there is a much wider range of possible outcomes for a population of one million X's and one million Y's than for one X and one Y. This fact is lost entirely if describe merely X and Y, and not the population of X's and Y's.

Date: 2009/10/08 01:39:02, Link
Author: Turncoat
Quote (oldmanintheskydidntdoit @ Oct. 07 2009,02:50)
O'Leary wishes that anything she writes will still be talked about decades later:
 
Quote
To the best of my knowledge, neither Newton nor Einstein wrote a long racist tract, as Darwin did (the Descent of Man).

I doubt anybody will even remember O'Leary after an equivalent amount of time has passed.
 
Quote
I’d like to know if, at any point, Darwinists have ever simply disowned the racism, instead of covering it up or misrepresenting it as not really Darwin’s views, to protect their prophet.

Have you stopped beating your Husband yet O'Leary?

"Has your husband stopped enjoying your beatings yet, O'Leary?"

[Fallacy of the complex and revolting question]

Date: 2009/10/08 02:41:14, Link
Author: Turncoat
Quote (Richardthughes @ Oct. 07 2009,10:23)
DR^2:

http://www.uncommondescent.com/intelli....-336511

 
Quote
9

William Dembski

10/07/2009

7:50 am
Nakashima & Frank: The ID movement does not use just one concept of information.


Because one definition does not lend itself to equivocation, bait and switch and goalpost-moving.

As in leaky conservation for complex specified information in No Free Lunch and strict conservation for active information, which has nothing to do with CSI.

Remember those little details he left out of the Jello book for readability -- stuff that anyone with some mathematical acumen should have been able to supply for himself?

Date: 2009/10/10 14:06:39, Link
Author: Turncoat
Quote (DiEb @ Oct. 10 2009,11:43)
I'd love to start a new topic - but it seems that I'm not allowed to do so. Could someone start a thread titled Conservation of Information in Search - factual errors in an article of R. Marks and W. Dembski (or something like that)? It would be nice to have a place to collect our critique of their weaselings...
One problem with the article is that it is so sloppily written: For instance, the references seem just to be copied together (like creationistic pamphlets of old) from earlier works. And every time, I touch one, I find an error. That's not the fault of the reviewers, as those aren't expected to check the references - it's the fault of Dembski and Marks. And it is a typical sign of sloppy research.

It happens that I woke this morning wanting to start an engineering-oriented discussion of the article somewhere.

I didn't even know that the "evolutionary computation" thread existed until now. The quality has been good, and there would be some advantage in just joining in. Then again, discussion of the D&M article deserves some prominence.

Hmm... I'm going to ask on that thread whether participants think there should be a thread devoted to the D&M article.

Date: 2009/10/10 14:12:15, Link
Author: Turncoat
DiEb and I want to get serious about identifying errors in the IEEE SMC-A article of Dembski and Marks. Do you folks feel that the discussion deserves its own thread?

Date: 2009/10/10 15:42:46, Link
Author: Turncoat
Quote (dvunkannon @ Sep. 09 2009,16:37)
Here's a little something related to genetic algorithms that I was thinking about during my recent vacation. I decided to write it down and share it with y'all in the hope getting some feedback on the idea. Now that I've got the idea sketched out, I'll implement it in the little GA I've been building.

Thanks in advance for any comments.

Something few people understand about the NFL theorems is that they level algorithms in terms of how well they do with n evaluations of the fitness function, and not how well they do with t units of running time. If you do not have prior knowledge of a problem instance that leads you to prefer some algorithm over others, or you choose to ignore that knowledge, then select the algorithm that runs the fastest (i.e., completes the greatest number of fitness evaluations in available running time).

You'll generally do better letting the computer architecture drive the choice of algorithm than by choosing the algorithm and trying to get its parameters right.

Back in 1991, before I'd hit on the NFL rationale (1994), I was working with a SIMD architecture with a toroidal mesh of processing elements. While giving a course on GA's, I designed a GA to run really, really fast on that architecture, abandoning bio-mimicry when necessary, and handed off the programming to two doctoral students. One of the students -- the one who subsequently presented our work at a conference -- liked to speak of my "big, messy genetic algorithm." The phrase was not in our paper, but it evidently made its way to Goldberg. There has been a fair number of BMGA papers to come from his lab.

The size of demes on individual processing elements was determined entirely by available memory. The parameter we played with most was the number of generations between migrations of individuals between neighboring processing elements. (I can't remember how we selected the migrants.) We thought, buying into Goldberg, that migration and recombination were essential. There is actually a theoretical argument for isolated parallel runs. So nowadays I would leave out migration unless I had reason to believe it would be beneficial, even though our program did not spend much time on it.

The program blazed, in terms of wall-clock time, and this was due to exploitation of the architecture. My students also implemented and studied many existing GA's, and we discovered just how much unreported parameter tweaking underlay reported results. There is no magic in any form of evolutionary computation, and exploiting the architecture should be Job One unless you have prior knowledge of the problem instance to exploit.

Date: 2009/10/10 21:00:33, Link
Author: Turncoat
Quote (Wesley R. Elsberry @ Oct. 10 2009,19:31)
There is the thread that started for critique of the "unacknowledged costs" essay by Dembski and Marks.

That is an earlier draft of the "Conservation of Information" article, isn't it? Much of the factual / technical content is in regard to the abortive analysis of ev. Also, the last comment is about 1-1/2 years old, and dvunkannon says on the "Evolutionary Computation" thread:      
Quote
I'd prefer for that discussion to happen here. Special purpose threads are hard to find (at least for me). I prefer general threads that pick up topics as necessary. Viz. the discussion of Mendel's Accountant and previous Weasel discussions on this thread.

Date: 2009/10/10 21:31:09, Link
Author: Turncoat
Quote (Turncoat @ Oct. 10 2009,15:42)
There is actually a theoretical argument for isolated parallel runs. So nowadays I would leave out migration unless I had reason to believe it would be beneficial, even though our program did not spend much time on it.

I should clarify that a bit. I would not run isolated GA's with small populations, high crossover rates, and low mutation rates in parallel. There would be a lot of wasteful reevaluation of individuals. I would run isolated algorithms with relatively high mutation rates in parallel. With high mutation rates, few individuals are evaluated more than once.

Date: 2009/10/11 23:36:53, Link
Author: Turncoat
Anyone else notice the attributed oxymoron here?
   
Quote
English’s Law of Conservation of Information (COI) [15] notes “the futility of attempting to design a generally superior optimizer” without problem-specific information about the search.

If you're attempting to design a generally superior optimizer, you are not looking at problem-specific information.

Mighty rude to complain about being made infamous, ain't it?

Date: 2009/10/11 23:50:01, Link
Author: Turncoat
Quote (dvunkannon @ Oct. 10 2009,22:17)
Quote (Turncoat @ Oct. 10 2009,16:42)
 
Quote (dvunkannon @ Sep. 09 2009,16:37)
Here's a little something related to genetic algorithms that I was thinking about during my recent vacation. I decided to write it down and share it with y'all in the hope getting some feedback on the idea. Now that I've got the idea sketched out, I'll implement it in the little GA I've been building.

Thanks in advance for any comments.

Something few people understand about the NFL theorems is that they level algorithms in terms of how well they do with n evaluations of the fitness function, and not how well they do with t units of running time. If you do not have prior knowledge of a problem instance that leads you to prefer some algorithm over others, or you choose to ignore that knowledge, then select the algorithm that runs the fastest (i.e., completes the greatest number of fitness evaluations in available running time).

Taken to the extreme, that says just run "Generate and Test" and give up on this fancy selecting, sorting, etc. NFL is NFL.

However, I think we agree that most human interesting problems are in a class where some level of GA can help if we don't already know a closed form solution. I realize there is a view that demes are just to allocate hardware appropriately, but I've also seen research that avoiding panmixis is a benefit irrespective of hardware. YMMV.

My main point was that hardware considerations must sometimes take precedence over algorithmics.

When we get into the Dembski and Marks article, you'll see that I believe we have learned a lot about problems, and that D&M ignore this source of information in their comments about evolutionary optimizers. I don't believe that their "search for a search" regress of probability measures models our learning through experience.

It seems to me that you have a metaheuristic, and that you would instantiate generic operations differently for different problems, presumably exploiting knowledge of the problem. It is not merely a fine theoretical point that different instantiations give different algorithms.

Date: 2009/10/12 15:51:23, Link
Author: Turncoat
I've added a Bad Science and Bad Theology comment to the Sidewiki for UD. (I can revise it, and suggestions are welcome.)

Google Sidewiki allows you to augment any webpage with comments in a sidebar. I believe that it will catch on. Installation is quick and easy.

The ordering of comments in the sidebar depends on votes on their usefulness. Please install Sidewiki and vote.

Date: 2009/10/13 14:15:32, Link
Author: Turncoat
Quote (Bob O'H @ Oct. 13 2009,13:55)
Quote
I don't believe that their "search for a search" regress of probability measures models our learning through experience.

I recall from a previous incarnation of this idea (when wMad assumed, for consistency's sake, that q<p and proved, 4 pages later, that log(q)<log(p)) I thought a nice metaphor would be that Dembski was saying it's easier to find a needle in a haystack than to find that huge electromagnet in the shed next to the haystack.

Neigh-h-h-h.

Date: 2009/10/13 14:46:05, Link
Author: Turncoat
Quote (midwifetoad @ Oct. 13 2009,14:08)
I can't evaluate the math, but I think people are being led down a garden bath with the concept of search.

I don't see any evidence that biology is modeled by a search algorithm. In biology a change might affect fitness for unknown reasons, and a subsequent change to the same position might further affect fitness. There is no single correct value for any given position in the string.

Behe and Dembski want you to believe that evolution must progress toward goals (consider Behe's obsession with the flagellum), but biology merely chugs along with whatever is adequate.

I agree entirely. I raised the issue of whether optimization was a good model of biological evolution in my first NFL paper, back in 1996. Now I am completely convinced that it is not. As Allen MacNeill rightly emphasizes, the consequence of variety, heredity, and fecundity is demography. A novel biological type can survive by virtue of its difference from the type that gave rise to it. There is not necessarily any basis for saying that the difference makes it better or worse.

Dembski and Marks have indicated that there are "implicit targets" for biological search. It's hilarious that they smuggle in teleology while accusing others of smuggling information into computational models of biological evolution. Creationists have long mistaken what did happen for what had to happen.

Date: 2009/10/13 15:50:30, Link
Author: Turncoat
In every introductory AI lecture I gave, I asked my students, "Which are more intelligent, cats or dogs?" The question is just as silly with "fit" in place of "intelligent."

I've read an interesting article by Sober on the role of reproductive fitness in biological modeling. As I recall, he indicates that its main use is in modeling at the level of a few alleles and traits (i.e., in population genetics). I am not sure that the concept is necessarily tautological at the level of whole organisms, but I definitely have seen many folks slip into tautology.

Creationist arguments that bacteria gain antibiotic resistance through a decrease in "real" fitness (i.e., the environment outside the hospital, in which the antibiotic-resistant strain fares poorly, is more real than the environment inside the hospital) are silly. And they go "poof" with emphasis on demography, as opposed to fitness.

Date: 2009/12/10 10:19:13, Link
Author: Turncoat
Well, I just made somebody or another at UD piss his pants. I posted, without the least meanness or gloating, an explanation of how Dembski and Marks got a crucial point wrong in the "search for a search" analysis of their latest publication. Someone deleted it, and I don't think Clive has the mathematical acumen to know just how devastating my comment was.

I lost what I wrote to a page reload - stupid me - but I've posted Blunder in the new Dembski-Marks paper on my blog.

Date: 2009/12/10 15:03:38, Link
Author: Turncoat
Quote (carlsonjok @ Dec. 10 2009,11:34)
Quote (Turncoat @ Dec. 10 2009,10:19)
Well, I just made somebody or another at UD piss his pants. I posted, without the least meanness or gloating, an explanation of how Dembski and Marks got a crucial point wrong in the "search for a search" analysis of their latest publication. Someone deleted it, and I don't think Clive has the mathematical acumen to know just how devastating my comment was.

I lost what I wrote to a page reload - stupid me - but I've posted Blunder in the new Dembski-Marks paper on my blog.

I am curious if you have submitted your critique of their first paper to IEEE?

Yes. And it's not purely a negative effort. The IEEE Systems, Man, and Cybernetics Society does not publish correspondence in response to articles in its three journals. But the IEEE Code of Ethics requires members "to see, accept, and offer honest criticism of technical work, to acknowledge and correct errors." The editor-in-chief of IEEE SMC-A has final accept-reject responsibility for articles. I hope to get the publications committee of the entire IEEE to require that all peer-reviewed publications be subject to peer-reviewed criticism and correction.

Date: 2009/12/10 15:08:56, Link
Author: Turncoat
Quote (Gunthernacus @ Dec. 10 2009,11:39)
Quote (Turncoat @ Dec. 10 2009,11:19)
Well, I just made somebody or another at UD piss his pants. I posted, without the least meanness or gloating, an explanation of how Dembski and Marks got a crucial point wrong in the "search for a search" analysis of their latest publication. Someone deleted it, and I don't think Clive has the mathematical acumen to know just how devastating my comment was.

Oh dear.  As one of the more Truly Articulate Representatives of Design once said:
   
Quote
Anyone who *truly* believes what they purpose[sic] would take on their top notch critiques[sic] at the drop of a hat.

But deleting valid criticism?  Well, as she might say, when the going gets tough - stay out of the kitchen.

Sorry, my error. I thought I'd been deleted and silently booted, but evidently my login expired.

Date: 2009/12/10 19:30:24, Link
Author: Turncoat
Quote (Kristine @ Dec. 10 2009,15:32)
Quote (Zachriel @ Dec. 10 2009,12:24)
   
Quote (Turncoat @ Dec. 10 2009,10:19)
Well, I just made somebody or another at UD piss his pants. I posted, without the least meanness or gloating, an explanation of how Dembski and Marks got a crucial point wrong in the "search for a search" analysis of their latest publication. Someone deleted it, and I don't think Clive has the mathematical acumen to know just how devastating my comment was.

I lost what I wrote to a page reload - stupid me - but I've posted Blunder in the new Dembski-Marks paper on my blog.

     
Quote
Bounded Science: Dembski and Marks believe that people search for search algorithms in a higher-order space Omega2. They write,

     
Quote
Let Omega2 be the finite space of all search algorithms on the search space Omega.

The word I've emphasized is wrong, wrong, and wrong. The set of all randomized search algorithms is infinite, not finite.

Quite an obvious error too—once you see it—, and something that many readers will be able to understand. Great catch!

-

Edited from Alpha to <span style='font-family:times'>Omega
!</span>

Just wondering - hopefully, this is not a stupid question, but do you think that they meant (or could come back and bluff that they really meant) a finite number of effective algorithms, and if so, were they hoping that this would make a difference? (Which of course it doesn't. As for the definition of "effective," this is Dembski we're talking about - Mr. Move-the-Goalposts himself.)

Man, I think you hit the cat out of the ballpark with this one! Elegant!

Christine, if we really want to talk meaningfully about the effectiveness of an algorithm for drawing a sample, we don't say, "Make Q queries. Use as much time and storage as you need." Instead we say, "Here's how much time and storage you have. Do the best you can to draw a sample that includes an acceptable solution."

A small, fast algorithm succeeds on many more problems, given limited time and storage, than does a big, slow algorithm. This gain in performance has nothing to do with problem-specific knowledge.

I don't think that Dembski and Marks want to tread this turf.

Date: 2009/12/11 20:11:57, Link
Author: Turncoat
Quote (Kristine @ Dec. 10 2009,15:32)
   
Quote (Zachriel @ Dec. 10 2009,12:24)
       
Quote (Turncoat @ Dec. 10 2009,10:19)
Well, I just made somebody or another at UD piss his pants. I posted, without the least meanness or gloating, an explanation of how Dembski and Marks got a crucial point wrong in the "search for a search" analysis of their latest publication. Someone deleted it, and I don't think Clive has the mathematical acumen to know just how devastating my comment was.

I lost what I wrote to a page reload - stupid me - but I've posted Blunder in the new Dembski-Marks paper on my blog.

       
Quote
Bounded Science: Dembski and Marks believe that people search for search algorithms in a higher-order space Omega2. They write,

         
Quote
Let Omega2 be the finite space of all search algorithms on the search space Omega.

The word I've emphasized is wrong, wrong, and wrong. The set of all randomized search algorithms is infinite, not finite.

Quite an obvious error too—once you see it—, and something that many readers will be able to understand. Great catch!

-

Edited from Alpha to <span style='font-family:times'>Omega
!</span>

Just wondering - hopefully, this is not a stupid question, but do you think that they meant (or could come back and bluff that they really meant) a finite number of effective algorithms, and if so, were they hoping that this would make a difference? (Which of course it doesn't. As for the definition of "effective," this is Dembski we're talking about - Mr. Move-the-Goalposts himself.)

Man, I think you hit the cat out of the ballpark with this one! Elegant!

Christine, thanks for the good question. I was on the road 12 hours today, and I spent a lot of time thinking about what Dembski and Marks might do to patch things up. I came up with several approaches, but none that suits their ends. And, of course, they're not out to investigate computational search, but to make a case for what they "know" must be true.

They might define a uniform distribution on search algorithms of bounded length. But this is entirely mathe-magical. There is no algorithm to decide if a sufficiently short algorithm searches Omega. Rice's Theorem, which Dembski and Marks mention often, implies this. Without a way to calculate whether or not a sufficiently short algorithm is in Omega_2, there's no algorithm for drawing uniformly at random for Omega_2. That is, a randomized "search for a search" algorithm cannot select any search algorithm it cannot recognize as a search algorithm.

This is probably incoherent. Perhaps I'll get around to saying more on my blog.

Date: 2009/12/11 21:06:16, Link
Author: Turncoat
Quote (Kristine @ Dec. 11 2009,18:45)
       
Quote (Bob O'H @ Dec. 11 2009,16:20)
           
Quote
It's a contributed paper in conference proceedings.  Review standards are fairly lax for publications of this sort.

I think conference proceedings are more important in computer science than elsewhere, so it might have undergone "proper" peer review.  But I'd be interested in comments from a computer scientist - I'm not one myself.

As far as I can see, the paper just shows why avida does better than a random search.  But framed to suggest that this is teh evil.

That's enough technical comment from me for tonight.

Ask and ye shall receive:
Quick Critique of Dembski and Marks

         
Quote
Personally, I come from a background where I look at these things in K-C terms. And one of the major results in K-C theory is that you can't really quantify information very well. How much information is in a particular search? Damned hard to say in a meaningful way. But D&M don't bother to address that. They just use a very naive quantification - the same logorithmic one that Dembski always uses. I dislike it intensely, because it essentially asserts that any two strings with equal numbers of bits have the same amount of information - which is just wrong. ("00000000" and "01001101" have different amounts of information, according to information theory.) The reason that Dembski can quantify things specifically is really because he's handwaved that important distinction away - and so he can assert that a particular search encodes a certain number of bits of information. What he actually means is that a given search algorithm can be said to contain something equivalent to a string whose length is N bits - but he doesn't say anything about what the actual true information content of that string is.

Second, he rehashes some old, sloppy examples. As has been discussed numerous times, Dembski likes to harp on the "Methinks it is like a weasel" example used by Dawkins. But he misrepresents the example. In Dawkins' presentation of that example, he used a program that searched for a string based on its closeness to the target string. It did not lock positions in the string. But Demsbki has always presented it as if it locks.


Dembski Stoops Even Lower

MarkCC writes posts that make me want to propose marriage.

Wow, I've never been so motivated to work at my blog. Do I stand a chance? Might you go for a Connery wannabe with an excellent letter of reference from an excellent belly dancer? Be sure to consider my application of algorithmic information theory to analysis of search-based optimization in Optimization Is Easy and Learning Is Hard in the Typical Function. That's from ten years ago. Just imagine the sweet somethings I could whisper in your ear now!

Date: 2009/12/11 21:54:45, Link
Author: Turncoat
Quote (Benny H @ Dec. 11 2009,21:36)
Check this out: http://www.asa3.org/ASA/meetings/baylor2009/papers/ASA2009Marks.pdf

I'd like to know what Marks said about the second slide. The IEEE Computational Intelligence Society publishes the IEEE Transactions on Evolutionary Computation.

Date: 2009/12/12 17:34:32, Link
Author: Turncoat
I'm sorry, folks, but I have trouble attending to the minutiae when ED&M are fundamentally wrong in their orientation. From Wikipedia: "Avida is an artificial life software platform to study the evolutionary biology of self-replicating and evolving computer programs (digital organisms)." As I recall, the ability of a virtual organism to reproduce itself depends on how much work it can do, which is determined by its fitness. Different organisms must do different amounts of work to reproduce. Right?

Avida, with the fitness function addressed by ED&M, does not engage in optimization, let alone search. The virtual organisms are operating in a virtual environment offering different payoffs for different behaviors. The  population of virtual critters in fact changes to obtain greater payoffs, but this does not mean that the purpose of Avida is optimization. An outcome is not a purpose. More importantly, when a run of Avida does not yield a critter of maximum fitness, the run has not failed. There are still a bunch of "living" virtual organisms, and they always exhibit greater complexity than did the organisms in the initial population.

The objective of the Avida research under consideration was to address emergence of complex features in artificial life. ED&M analyze runs of Avida as though they were runs of an optimizer, and then argue that the ALife simulator is not a good optimizer. This is the kind of abysmal stupidity you get from pretend-scholars who are on a "mission from God" to deliver payback.

Date: 2009/12/12 17:54:51, Link
Author: Turncoat
Oops. Avida is an ALife program, not simulator.

Date: 2009/12/15 15:14:53, Link
Author: Turncoat
Quote (sledgehammer @ Dec. 13 2009,18:59)
Quote (Maya @ Dec. 11 2009,13:48)
Dembski is crowing about getting another article published.  Has the IEEE always had such low standards?
 
I note that these last two "papers" are conference proceedings.
The IEEE standards for accepting a conference paper and those for publication in one of their journals are vastly different.
 Conference papers are accepted purely on the basis of an abstract, with no peer review except by the session chairs, who often have a quota of slots to fill.
Later, after the conference, authors are often invited to submit a "proper" paper for peer review and publication in the appropriate journal.

ETA:  I see olegt has already made this point three pages ago! I'm late and lacking, as usual.

Sorry, but this is simply wrong.

Date: 2009/12/16 07:22:33, Link
Author: Turncoat
Quote (sledgehammer @ Dec. 15 2009,20:27)
   
Quote (Turncoat @ Dec. 15 2009,13:14)
       
Quote (sledgehammer @ Dec. 13 2009,18:59)
       
Quote (Maya @ Dec. 11 2009,13:48)
Dembski is crowing about getting another article published.  Has the IEEE always had such low standards?
 
I note that these last two "papers" are conference proceedings.
The IEEE standards for accepting a conference paper and those for publication in one of their journals are vastly different.
 Conference papers are accepted purely on the basis of an abstract, with no peer review except by the session chairs, who often have a quota of slots to fill.
Later, after the conference, authors are often invited to submit a "proper" paper for peer review and publication in the appropriate journal.

ETA:  I see olegt has already made this point three pages ago! I'm late and lacking, as usual.

Sorry, but this is simply wrong.

Yes, all generalizations are false. I should have said some, or many, IEEE conferences accept papers on the basis of abstracts or summaries alone.  I know LEOS, CLEO, IMS and other physics conferences do, because I've had  papers  accepted on the basis of 500-word summaries.
 It could well be that the computer science societies are more selective because they have more submissions?  I've never had to pay to have a paper included in a conference, like this one seems to.

Sorry I didn't read the next page before responding. The IEEE is an excellent organization. Although I'm a computer scientist, I dropped my membership in the Association for Computing Machinery and stuck with the IEEE.

The norm for IEEE conferences is three reviews of full-paper submissions. Two reviews is a minimum. Reviewers typically get three or four papers apiece, and have a month or so to complete reviews. Authors are on their honor to revise as instructed by the reviewers. Organizers of special sessions commonly present in their own sessions. But they do not oversee review of their own papers. If there are two organizers, each can handle reviews for the paper of the other.

The IEEE does not publish many journal pages per year per member. There is some high quality stuff in IEEE conference proceedings that does not make its way into IEEE journals. There has also been, in my field anyway, an increasing number of conference papers that should not be read and that never get cited. As best I can tell, they serve no purpose but to boost attendance. I'm not sure if conferences are money-making propositions, or if it is just that bigger is better for organizers' reputations.

The thing to understand about Bob Marks is that he actually has done outstanding work, and deserves the fine reputation he has within the IEEE. I strongly suspect that his Handbook of Fourier Analysis, published by Oxford University Press, is a rigorous treatment.

I find myself wondering how Dembski and Marks coordinate their efforts. I think that many of their errors are attributable to Dembski's gelatinous treatment. Show me a definition of "search algorithm" in their papers. In fact, show me an algorithm that is written out as an algorithm, rather than explained casually. If they actually wrote out their claims in formal terms, they could not help but catch many of their own errors.

Date: 2009/12/16 07:27:35, Link
Author: Turncoat
A humdinger of an error the reviewers of the IEEE SMC-A article did not catch is that Dembski and Marks measure work and call it active information.

As the old blues song goes, "If Kristine don't love me, I know her sister will."

Date: 2010/03/16 14:59:22, Link
Author: Turncoat
Write a little note to Dembski and Marks, e.g.,

Dear Profs. Dembski and Marks:

This is to request that you comply with the IEEE Code of Ethics, and note in the online version of "Conservation of Information in Search: Measuring the Cost of Success" the errors that have been called to your attention.

The mathematical errors evidently were first identified in a RationalWiki article, "Conservation of Information in Search - Measuring the Cost of Success."

Sincerely,

xxxxxxxx


The email addresses are  wdembski@designinference.com and Robert_Marks@baylor.edu

Date: 2010/05/08 01:07:32, Link
Author: Turncoat
Quote
1 Corinthians 1:21-25 (New International Version)

For since in the wisdom of God the world through its wisdom did not know him, God was pleased through the foolishness of what was preached to save those who believe. Jews demand miraculous signs and Greeks look for wisdom, but we preach Christ crucified: a stumbling block to Jews and foolishness to Gentiles, but to those whom God has called, both Jews and Greeks, Christ the power of God and the wisdom of God. For the foolishness of God is wiser than man's wisdom, and the weakness of God is stronger than man's strength.


I'd forgotten this gladly, but saw some of it in a Powerpoint show by Bob Marks, Science, God, and Christ: Genesis and Science: Compatibility Extraordinaire.

I regard Jesus as an incredible teacher -- particularly the Jesus of The Gospel of Thomas. As for Paul, the author of the passage above, I'd say that the depiction of him as a huckster in The Last Temptation of Christ is about right. The "Christ crucified" is of course also the Christ resurrected, so the notion that "what was preached" had no "miraculous signs" is garbage. Paul got carried away by his own rhetoric, which he of course meant to be taken as very wise.

Date: 2010/05/09 08:42:48, Link
Author: Turncoat
Oops, KF, you've departed twice from that point in a single thread.
     
Quote
Just for a point of departure, here is Plato on the issues, in The Laws, Bk X: [BIG snip]

(I really don't read his comments. Just caught my eye.)

Date: 2010/05/09 18:37:06, Link
Author: Turncoat
Near-death guffaw. KF:
Quote
Above: yup. Sadly. DaveScot played a very important role in setting up a forum that was reasonable; loo. . . oong story. G

Date: 2010/05/11 08:17:39, Link
Author: Turncoat
Quote (sparc @ May 10 2010,23:23)
Off topic: Haven't the previuous casey luskin graduate award winners been announced in May? Some of them have preferred to stay anonymous, though.

Winston Ewert and George Montanez, I'll bet.

Date: 2010/05/11 08:31:42, Link
Author: Turncoat
If Casey's stopped handing them out, what stops me from making "Casey Luskin Graduate Awards" in his honor? I'll send checks for 98 cents to kids descending into IDiocy. Anyone have a certificate?

Date: 2010/05/11 09:43:37, Link
Author: Turncoat
Quote (Richardthughes @ May 11 2010,09:19)
Quote (utidjian @ May 11 2010,08:36)
Does Gordon E. Mullings have a MSc. in Physics? Linky
Apparently his office isn't just in the Administration Building, it IS the Administration Building. I guess it depends on how you interpret it. Seems he is teaching an evening class in Intro to Philosophy. Gosh that must be a fun class. Anyone know what his day job is?

-DU-

Circumlocutious Internet blowhard.

Logorrheaic, narcissistic, hypocritical, sanctimonious, supercilious, whiny, ...

Date: 2010/05/16 17:20:57, Link
Author: Turncoat
Quote (keiths @ May 16 2010,16:35)
'above' can't understand why atheists don't like him:
 
Quote
I actually had an atheist recently try to play that trick on me. After exposing him for lying and pointing out his ignorance and circularity of logic – although I must admit I took no great measures to sugar-coat it for him, but neither was I rude – he turned around and called me hateful.

"Above" is a liar for God, claiming to be neutral when it is anything but. It smells like one of Dembski's cultural warriors in training.

Date: 2010/05/20 12:07:06, Link
Author: Turncoat
Quote (didymos @ May 19 2010,21:17)
Batshit:    
Quote

as to the rest of your post Lenoxus: Genetic Entropy!


Yeah, Lenoxus.  Screw that: batshit's got crap Sanford made up and can't substantiate.  Also: Quantum Mechanics!  Jesus! Here's a random Youtube video!  Take that, Evolution!

Excuse a random rant. If the genetic-entropy proponents were just kooks, I'd laugh at them. But they are bald-faced liars for God. They know that they're trying to dress up The Fall as science. And they know they're involved in a power play to insinuate their religion into public institutions. Hypocrites!

ID does not predict that an evolving system should only deteriorate. Hell, practical design of a system to evolve goes back at least to G.E.P. Box's EVOP (evolutionary operations, 1950's).

Genetic entropy is predicated on the theological notion that an omniscient God made things perfect in the first place, and that "random" variation can only make them worse. It's friggin' stupid logic. If the biosphere changes dynamically, then "perfect" life has to change also. Who's to say that God should have made life static? Static sounds mighty lifeless to me.

Date: 2010/05/20 13:28:28, Link
Author: Turncoat
Quote (midwifetoad @ May 20 2010,13:15)
Quote
I have a friend who is using genetic algorithms to control tethered inflated aerofoils - the kites people use for kite surfing.  He is using GA's because applying conventional control theory to the task is rather difficult - the control laws are hopelessly complicated - yet a genetic algorithm is able to solve the problems fairly easily!


Of course the UD people haven't, in the course of ten years, been able to follow the 15 lines of code needed to implement Dawkins' Weasel.

Obviously your friend worked out the parameters of his kite control system manually, and smuggled the information into the GA program.

Bodies and motor control have coevolved.

Date: 2010/05/20 17:12:56, Link
Author: Turncoat
Quote (Quack @ May 20 2010,16:24)
   
Quote
Genetic entropy is predicated on the theological notion that an omniscient God made things perfect in the first place, and that "random" variation can only make them worse. It's friggin' stupid logic. If the biosphere changes dynamically, then "perfect" life has to change also. Who's to say that God should have made life static? Static sounds mighty lifeless to me.

For 'genetic entropy' to be true, ought not life have reached the 100% entropy level by now? Since it has not, isn't that evidence that genetic entropy is as fake as the three dollar bill?

This by John Hawks is my favorite response. If I understand correctly, Sanford believes that culture (esp. medical technology) is reducing the selection pressure that has slowed genetic entropy.

Date: 2010/05/21 09:50:09, Link
Author: Turncoat
Quote (dvunkannon @ May 20 2010,23:10)
Quote (Jkrebs @ May 20 2010,23:23)
Congrats to Nakashima, if this story about a ring on June 4th is true, and to any and all alter egos he may have.

Oh, its true all right. I'm invited. A "ring side" seat, you might say...

Ceremony to be held at the former site of the archives of the Czechoslovakian secret police.


OK, it was monastery before that, and is again.

As long as the volcano cooperates.

Brno? The women there are gorgeous.

Date: 2010/05/21 16:40:12, Link
Author: Turncoat
Congrats and best wishes to the ruggedly handsome Nakashima, who's no slouch intellectually, from what I know of his resume.

Date: 2010/05/21 18:19:37, Link
Author: Turncoat
Quote (midwifetoad @ May 21 2010,12:04)
What became of Gil's Checkers speech? I thought he played a perfect end game, but he seems to have resigned.

http://www.uncommondescent.com/intelli....-355135

My old pal Sooner tells me that this has been stuck in moderation for over 24 hours:
         
Quote
38
Sooner Emeritus
05/20/2010
3:31 pm

Your comment is awaiting moderation.

Something Dembski and Marks do not begin to address is that there are different kinds of knowledge, and that some kinds are harder to come by, and thus are more valuable, than others. David Fogel moved from checkers to chess, and started out with a public-domain program that humans had been tweaking for years. After a couple days of computational evolution (again, evolution of the static board evaluator), the program had gone from master to grandmaster rating (370+ gain in rating, IIRC).

Now, no one would say that Fogel didn’t use any sort of knowledge in structuring the static board evaluator. But the knowledge he used was generic, and easily obtained. It was relatively “dumb” knowledge. The evolutionary process yielded, in a sense that has a precise operational definition, “smart” knowledge. Knowledge that makes a chess player a grandmaster rather than a master is something very special, at least in the estimation of humans.

“Dumb input, smart output” is not a phenomenon that will be explained away by counting bits of “active” information.

I'd add that Dembski and Marks have made utter asses of themselves by suggesting that the virtual environment's determination of wins, losses, and draws is some sort of information smuggling. What constitutes a win, loss, or draw is determined by the rules of the game, just as other aspects of the virtual environment are. Reversing the senses of "win" and "lose" in a game gives a different game. There is not an external entity providing information by "announcing" the outcome each time two individuals compete.

The pathetic blunders of intelligent men laboring to "prove" their religious beliefs never cease to amaze me.

Date: 2010/05/21 18:29:58, Link
Author: Turncoat
Oops. Clive let it through while I was writing. I also see some more comments on the "impossibility" of building ornithopters, and the importance of materials in robotics.

Date: 2010/05/24 01:09:17, Link
Author: Turncoat
Quote (Richardthughes @ May 23 2010,15:13)
*bump*

I had thought of contacting the editors of the Rocky Mountain News with an "angle" on an op-ed piece, but Reciprocating Bill has the goods on Barry already.

Some ideas:

The point in going to the mainstream media is to persuade the "Conservative Republican = Good Guy Like Us" crowd that Barry is actually not a good guy like them. You have to steer clear of suggesting that it is anything but fine to be a conservative Republican. Any reference to Barry's advocacy for gun owners has to be handled very carefully.

Nobody likes a liar. Play up the fact that the "intelligent design" brand of creationism originated with a (botched) global search-and-replace revision of a creation science textbook, and that it was subsequently dandified by a fast-talking lawyer. Point out that ID creationists have developed elaborate ways of denying that they are creationists. That is part of the subterfuge. I would suggest that it's OK for religious people to seek support for their beliefs in science, provided they are honest.

Candidate Arrington is also a lawyer, and is not the sweetest smelling of them, when you take a good sniff.

In the school-board debacles regarding science standards, no board member has been as devoted to creationism as Arrington. The claim of a slick lawyer-fox that he won't dine if the voters let him into the hen house seems a bit thin. Dembski recently posted on Arrington's blog, "Perhaps when Dover II rolls around...." Would the candidate promise to recuse himself from all board activities related to biology education?

Date: 2010/05/24 01:37:47, Link
Author: Turncoat
Quote (Richardthughes @ May 23 2010,15:12)
Here's the barry arrington thread.

http://www.antievolution.org/cgi-bin....;t=5332

I commented in the other thread. It appears that Reciprocating Bill could do a great job with an op-ed piece. I'd be glad to edit.

Date: 2010/06/28 11:59:23, Link
Author: Turncoat
My slow-motion reply to Dembski's post on Yarus' reuse of the Weasel program:

Willie can't stop whipping the Weasel.

You'll find a few zingers, and also some simple analysis that's worth knowing. For instance, in Yarus' "nothing in biology makes sense except in the light of evolution" example, only 91 of the 200 offspring in a generation are mutants, on average. I also show that the programs Dembski is pinning on Dawkins are REALLY screwy.

I got sidetracked by various things, e.g., reading some of the literature that Dembski, Marks, and cubs should have, and proving that a certain non-magical setting of the number of offspring works well. More to come, dog willing, and the Creek don't rise.

Date: 2010/06/28 12:58:01, Link
Author: Turncoat
Dembski haplessly admits to plagiarism?

Date: 2010/07/30 06:03:30, Link
Author: Turncoat
Some of you may be interested to know that the pretty plots in the last paper the Baylor cubs, Dembski, and Marks published were cattywampus. Mean guy that I am, I've kept this to myself for a couple months. Link

I emailed the link Ewert and Atom, and copied Marks and Dembski.  :p

Date: 2010/08/18 22:23:01, Link
Author: Turncoat
All of my UD identities I can remember:

  1. Tom English
  2. Thom English
  3. Thomas English
  4. T M English
  5. austin english
  6. Turner Coates
  7. Cloud of Unknowing
  8. Semiotic 007
  9. Liz Lizard
 10. Sal Gal
 11. Mystic
 12. Oatmeal Stout
 13. Atticus Finch
 14. CEC09
 15. Hamlet
 16. Sooner Emeritus

Rumination here.

Date: 2011/02/28 10:48:37, Link
Author: Turncoat
Quote (Erasmus @ FCD,Feb. 28 2011,09:55)
so youns got me all farred up to go see what kinda apoplectic fits the tards were having over papaw granfools latest breakfast-on-chin mishap and i was disappointed.  but the "recent comments" did not fail to please.  

Gordon Mullings is too stupid to know that he is a liar.

 
Quote
The digital code in D/RNA is there for all to see [it is not an analogy, it is instantiation, of a 4-state digital code that is also an algorithm, implemented through cellular, molecular scale nanomachines . . .which is what ribosomes etc plainly are], and is so integral to how cells work that it was there from their origin.


so rong you can't do nuttin with it except point and laugh.  i would like to see gordon mullings fight the inquisition.

What would the Jamaican Inquisition be? A gang of Rastafarians?

Date: 2011/03/14 08:56:28, Link
Author: Turncoat
From Wikipedia:
Quote
The Diagnostic and Statistical Manual of Mental Disorders fourth edition, DSM IV-TR, a widely used manual for diagnosing mental disorders, defines narcissistic personality disorder (in Axis II Cluster B) as:[1]

   A pervasive pattern of grandiosity (in fantasy or behavior), need for admiration, and lack of empathy, beginning by early adulthood and present in a variety of contexts, as indicated by five (or more) of the following:

      1. Has a grandiose sense of self-importance (e.g., exaggerates achievements and talents, expects to be recognized as superior without commensurate achievements)
      2. Is preoccupied with fantasies of unlimited success, power, brilliance, beauty, or ideal love
      3. Believes that he or she is "special" and unique and can only be understood by, or should associate with, other special or high-status people (or institutions)
      4. Requires excessive admiration
      5. Has a sense of entitlement, i.e., unreasonable expectations of especially favorable treatment or automatic compliance with his or her expectations
      6. Is interpersonally exploitative, i.e., takes advantage of others to achieve his or her own ends
      7. Lacks empathy: is unwilling to recognize or identify with the feelings and needs of others
      8. Is often envious of others or believes others are envious of him or her
      9. Shows arrogant, haughty behaviors or attitudes

Date: 2011/03/22 05:13:56, Link
Author: Turncoat
I have just submitted this tip to some news outlets.
[quote]I have retrieved from the document archive at Baylor University a thesis that plagiarizes extensively two publications coauthored by William A. Dembski, a leader of the intelligent design movement, and Robert J. Marks II, the "Charles Darwin" of intelligent design. The signature page indicates that Marks, who holds the rank of Distinguished Professor of Engineering at Baylor, was a member of the thesis committee. Thus Marks apparently condoned plagiarism of his own work. He has been listed as one of the twenty most influential Christian scholars.

I have posted further details, as well as contact information for the involved parties, on my blog,

http://boundedtheoretics.blogspot.com/
The thesis is apparently Winston Ewert's. Perhaps only the Baylor Lariat and the Waco Tribune will follow up. But that should be enough to create a stir at Baylor. Tell your friends and neighbors.



Date: 2011/03/22 05:16:02, Link
Author: Turncoat
Oops.

Date: 2011/04/18 16:11:49, Link
Author: Turncoat
Quote (paragwinn @ April 18 2011,13:23)
For Noesis' presumption of analyzing the Dr Dr's hallowed thoughts re: CSI, Clive prepares the Mod-Hammer:
             
Quote
You seem to think that a guy could only have one point of view or observation, which is ridiculous, and don’t presume to know his mind and psychologize him, this irritates me to the point of moderating you.


I'm way past irritation at this point. Half of the IDiot comments indulge in 'presuming' and 'psychologizing' of the 'Darwinist, atheist, materialist mind' of their critics. Is it me? It's Clive, right?

He had effectively expelled me, and wasn't honest enough to say it outright.

I also doubt that Clive was honest about what irritated him. I commented primarily on the psychology of "true believers" like him who would be crushed if their authority figures admitted to fallibility.

It's hilarious that UD constantly pushes the fallibility of science, but cannot admit that an ID theorist ran into a dead end.

Date: 2011/06/08 14:50:35, Link
Author: Turncoat
Quote (Sealawr @ June 07 2011,18:16)
Cross posted from EN&V biut appears to be on topic here:

http://www.evolutionnews.org/2011....21.html

The journal paid Sewell $10,000 and issued an apology?
What were they thinking?

The Disco 'tute road show (West, Behe, Richards, Luskin) made a stop here in OKC several months ago. You can't grasp how despicable John West is until you've sat in the sanctuary of a Christian church and watched him shrewdly manipulate the audience.

It annoys me to see West reduce David vun Kannon to a mere Web entity, exploiting the humorous self-deprecation on his blog, and then engage in the umpteenth instance of inflation of the credentials of an ID creationist. Anyone who has read the writings of DvK and Granville knows that the former is the intellectual superior of the latter. And the fact is that David has done pretty damned well in his career.

But the issue in the lawsuit was not the word of DvK versus that of Comb-Over. The editor violated his journal's editorial policy by rejecting the paper after it was accepted.

I suppose an expansion of this belongs on my blog. The fact is that dealing with medical problems is taking up much of my time. I did, however, check on Winston Ewert's thesis a couple days ago. Amazingly, it now includes a preamble that admits to plagiarism in the originally accepted version. See this at Bounded Science.

Date: 2011/06/08 15:12:53, Link
Author: Turncoat
Quote (oldmanintheskydidntdoit @ June 08 2011,10:59)
To me one of the defining characteristics about your average ID supporter is that they can never ever be wrong about anything, no matter how trivial.

True. And what distinguishes the ID leader from the average ID supporter is the ability to recognize when his arguments have become untenable, and to move on to new ones — without, of course, admitting to error in the old ones.

Behe abandoned his original notion of irreducible complexity. Dembski abandoned the explanatory filter, as well as complex specified information. But Darwin's Black Box and No Free Lunch continue to sell.

Date: 2011/06/09 00:26:42, Link
Author: Turncoat
Quote (Richardthughes @ June 08 2011,14:56)
Hey Tom - get well if you're not and stay well if you are!

"Better living through pharmaceuticals" has quite a different meaning when you're 54 than when you're 24.

Date: 2012/04/27 16:27:10, Link
Author: Turncoat
Forgive an ad for what I suspect is the most devastating post I'll ever make to my blog. Other folks have pointed out that "active information" is nothing but a performance measure. But I show that Marks has stated outright his misunderstanding of the NFL theorems, and explain how his misrepresentation of performance as information follows. All of you will understand the most important parts of the post, and many of you will understand all of it.

Sorry that I haven't been coming around to play. Keep giving 'em hell.

Date: 2012/04/27 16:50:09, Link
Author: Turncoat
Quote (Febble @ April 27 2012,16:34)
Quote (Turncoat @ April 27 2012,16:27)
Forgive an ad for what I suspect is the most devastating post I'll ever make to my blog. Other folks have pointed out that "active information" is nothing but a performance measure. But I show that Marks has stated outright his misunderstanding of the NFL theorems, and explain how his misrepresentation of performance as information follows. All of you will understand the most important parts of the post, and many of you will understand all of it.

Sorry that I haven't been coming around to play. Keep giving 'em hell.

Ooh, can I crosspost it at The Skeptical Zone - or would you like to?

I'd like very much for you to do that. Thanks so much.

Date: 2012/04/27 22:03:50, Link
Author: Turncoat
Quote (Febble @ April 27 2012,17:02)
                 
Quote (Turncoat @ April 27 2012,16:50)
                 
Quote (Febble @ April 27 2012,16:34)
                   
Quote (Turncoat @ April 27 2012,16:27)
Forgive an ad for what I suspect is the most devastating post I'll ever make to my blog. Other folks have pointed out that "active information" is nothing but a performance measure. But I show that Marks has stated outright his misunderstanding of the NFL theorems, and explain how his misrepresentation of performance as information follows. All of you will understand the most important parts of the post, and many of you will understand all of it.

Sorry that I haven't been coming around to play. Keep giving 'em hell.

Ooh, can I crosspost it at The Skeptical Zone - or would you like to?

I'd like very much for you to do that. Thanks so much.

Thanks!

Done:

http://theskepticalzone.com/wp....4....wp....4

And I've added your blog to my blogroll!  Hope that's OK.

I'm genuinely flattered. I'll soon unload with the other barrel. But that post has to be more technical. You've evidently got a fine stat background, and it will be perfectly clear to you, when I eliminate the obfuscation of "search," that we're looking at biased sampling. The model I have now is so simple that I feel embarrassed to have missed it for so long.

 

 

 

=====