# Take the Design Challenge!

Since posting my essay on Genetic Algorithms, I’ve since developed a brand-new C++ version of my Steiner Networks genetic algorithm, a vast improvement over the old Fortran number-clunker I developed five years ago.

And already, the new code is leading to some very interesting results.

In light of William Dembski’s remarks in No Free Lunch, basically arguing that in all Genetic Algorithms,

… the fitness function … is well-defined and readily supplies the complex specified information that an optimal crooked wire genetic antenna [or any other problem solved with Genetic Algorithms] seems to acquire for free,

I’m giving Intelligent Design proponents (and everyone else!) a chance to actually Design something!

As you recall, my algorithm involves finding Steiner Trees, the shortest networks of straight-line segments connecting a given collection of fixed points. These networks may include additional variable “Steiner Points” where segments may meet.

The Challenge Here is a collection of six fixed points. Designers, send your candidates for the Steiner Solution for this particular 6-point system to me at nmsrdaveATswcp.com (replace the AT with an @ if you please).

I will announce the winners (if any) in a week’s time, and also will present the answer that Evolution came up with. I’m interested in proposed solutions from any and all (you don’t have to be in the ID camp), but am especially interested in solutions by ID advocates, since y’all are saying that the solution is already implicitly defined in the statement of the problem (finding shortest connected networks). Here’s a Hint:

You can send solutions as images (GIF preferred), textual descriptions (e.g. “a point at 250,300 connected to a point at 350, 300”), plain text ( / | \ _ ) or web pages.

This invitation is cordially extended to the entire Software Engineering Team at Uncommon Descent, where Salvador Cordova recently wrote

This is not the first time Thomas put together a sham. A few weeks back he put together a sham on evolutionary algorithms. After I visited pandas thumb to called him on his little charade Comment 111151 he suggested I leave. In the meantime, the blogsphere from PZ Myers onward hailed Dave’s little gimmick as some kind of truth. Most of the software engineers here at UD saw right through it, but the eager consumers of Panda food gulped it down uncritically, believing the design argument was refuted by his software theatrics. This last little bit of misrepresentation by Thomas is par for the course for the Panda bloggers.

Salvador, I didn’t “suggest you leave” never to comment again, but simply to spin up on the concepts (like Irreducible Complexity, No Free Lunch etc.) that were being discussed in that post. Your Input is now formally solicited.

This will be the first installment of a 3 part series where I will explore the work on evolutionary algorithms by Chris Adami, Dave Thomas, Wesley Elsberry, and Jeff Shallit.

He followed this by stating

I will also post on Dave Thomas’s evolutionary algorithms, but in brief, his disproof can be illustrated by this fictional scenario:

PZ Myers finds DaveScot in the park one day, walks up with a paint ball gun and shoots DaveScot in the chest. Ouch! Shortly before DaveScot confers a little retribution for this act, PZ pleads, “Don’t be mad Dave, you were not the target of my gun. Honest, I was aiming at the shirt you happen to be wearing, not at you specifically, only your shirt.”

Mr. Cordova has additionally claimed that

Thomas was resorting to double speak: “oh, I didn’t specify the target.” Picture this, some kid with a paintball gun goes up to another and shoots his victim, and says to his vicitm, “don’t be mad, I wasn’t aiming at you, I was aiming at the shirt you were wearing.” Thomas’s double speak was hardly better.

What we have here is a case of Orwellian Double speak, where disingenous labels are affixed which are misleading. For example, let us say I created a novel and useful medicine. I could later call it “poison”. I could then argue “poison” is healthy. The arguments in favor of Avida are merely double speak.

My challenge to Salvador and the UD Software Engineering Team is simple and straightforward: if the Target’s “shirt” (a stated desire for the shortest connected straight-line networks) is indeed as “CLOSE” to the “Target” itself (the actual Steiner Solution for the given array of fixed points) as you say it is, then you and your Team should be easily able to deduce the proper answer, and send it along.

I’ll be waiting! See you next Monday, August 21st. - Dave

… the fitness function … is well-defined and readily supplies the complex specified information that an optimal crooked wire genetic antenna [or any other problem solved with Genetic Algorithms] seems to acquire for free

IDers are bending over backwards to deny that genetic algorithms can actually create something. It’s actually a really dumb argument that they’re using. Essentially what they’re saying is that if my boss comes to me and says, “design a flying car that gets 100 MPG” and I turn around and actually do it, that I haven’t done any actual work. According to them, all the design was done by my boss when he defined what he wanted: “a flying car that gets 100 MPG”. What a joke.

Dave: Did you try posting this challenge at UD? Or are you banned? In any case, I’m looking forward to their “designs”. Any plans to post the ID solutions (if any) somewhere? Love the challenge, by the way.

I sent my answer to Dave an hour ago.

Can’t they just figure it out with soap bubbles? Assuming they can remember the way to the lab, of course.

Seems that Salvador is obfuscating on Uncommon Descent as usual. What a hoot. I wonder what the people at Cornell’s IDEA club have to say about Sal’s response.

And people still wonder why scientists consider ID to be scientifically vacuous. One merely has to point to Sal to see an excellent example of why ID is doomed scientifically speaking.

Thanks Dave, you have done a marvelous job at exposing not only ID but also some of the more vocal ID evangelists

It’s strange: it’s almost as if critics of GA’s are being intentionaly deceitful. I went to the UD page linked under trackbacks below, and the criticism of the orginal fortran code was, get this, that it had a fitness function. See, since this fitness function didn’t “evolve” the whole thing is a charade.

*does a double take*

The whole purpose of this sort of an algorithm is to create an artifical microcosm within which a simulation of something natural can occur. There has been no claims made that the environment, the simulation itself, is naturally occurring. It is that which is simulated that is done using mechanisms that are analogous to the natural world. The simulation is OBVIOUSLY artificial. But the processes, the mechanisms coded as algorithms are analogous to natural mechanisms. That’s the whole point.

With this sort of attitude, scientists cannot perform ANY laboratory experiments where they set up initial conditions that are intended to mimick something in nature.

So, for instance, if I create an environment of self reproducing organisms, and put those organisms under a selection pressure, and I can see how the mechanisms of random variation and differential reproductive success cause that population to change its character, these people would respond that since I wrote all the code, (or in a laboratory experiment, set up the starting conditions) the simulation (or experiment) is a charade. Nevermind the fact that the resulting organisms within the simulation, which were arrived at using processes analogous to those found in nature, were never coded into the software. Because the environment is artificial it doesn’t show anything.

I’m flabbergasted.

Is it dishonesty, or is somebody just not getting it?

O dear sal. The front loading is the fitness function. Gotcha. Let me suggest the ‘universal fitness funtion’ - the fall.

http://news.com.com/Is+evolution+pr[…]6074543.html

??

Is it dishonesty, or is somebody just not getting it?

Both. Generally they don’t get it, but when they do they lie.

Ondoher wrote on August 15, 2006 at 12:58 AM:

…these people would respond that since I wrote all the code, (or in a laboratory experiment, set up the starting conditions) the simulation (or experiment) is a charade.

SteveS calls them information cooties. Put a human being anywhere on the same continent as the experiment and it is infected with them.

Paul

Is it dishonesty, or is somebody just not getting it?

You have to understand that the ID crowd pats themselves on the back for *not* getting it. They actually remind me of students I’ve been in class with that just don’t understand subject material and claim that the professor has it wrong (yes, I’ve actually seen this). They remind me of a bunch of poor students who pat each other on the back for *not* getting it and telling each other how smart they are for not buying into the “elitist” professor’s claims. And when they point out problems with GAs that aren’t particularly relevant, they seem to actually believe they’ve made a serious criticism, but the only thing that’s clear is that they really don’t understand what’s relevant about GAs and why their criticism isn’t really significant. But, hey, the more they can deny evolution, the better they feel about ID’s evangelical potential to bring people to Jesus Christ. They have an incentive to not understand the relevant material. How are you supposed to teach someone like that?

Most of the software engineers here at UD saw right through it

I call BS. I’m a software engineer and GAs relevance to evolution makes perfect sense to me. UD must have a few recalcitrant software developers on it’s site.

SteveS calls them information cooties. Put a human being anywhere on the same continent as the experiment and it is infected with them.

This isn’t true just of creationists, sadly. There are many people who deny that a human-created robot (running some program) could be “really” (as opposed to “metaphorically” or by “anthropomorphic projection”) conscious/intelligent/creative/etc. because a programmer wrote the code – or the code that generated the code, or the code that generated the code that generated the code … I’ve seen various mathematicians, philosophers of mind, and people on this board take this view.

UD must have a few recalcitrant software developers on it’s site.

Um, everyone at UD is recalcitrant.

Shortly before DaveScot confers a little retribution for this act, PZ pleads, “Don’t be mad Dave, you were not the target of my gun. Honest, I was aiming at the shirt you happen to be wearing, not at you specifically, only your shirt.”

I’d note that this is directly comparable to real-life evolution. Evolutionary systems, as Dembski has pointed out ad nauseam, aren’t usually capable of directly finding useful proteins. However, they are perfectly able to maximise the underlying fitness function, of which these useful proteins just happen to be peaks.

Well done, Sal, for your excellent analogy.

Maybe it is a mistake to mention the Steiner solution. This might give the impression that the success of the programme is in how close it gets to the Steiner solution and that this is some kind of target. (I know Dave didn’t say that - but people can get the wrong idea). The real success is how the programme can generate novel solutions that give the appearance of being designed to have a short total network length even though the specific design was not in any way incorporated in the algorithm (and indeed was not known to the programme writer).

Salvador tries to draw an analogy with finding the sum of the first 1000 integers. This is the equivalent to finding the Steiner solution. We may not know the answer in the sense of being able to say what the number is, but it is unique and clearly defined and could be intepreted as a target. A better analogy might be a programme that generates numbers that are in some sense better than others without there being a best number. I can’t think of an example without giving the numbers some physical meaning as Dave has done - but maybe a mathematician can?

Looks like ID proponents are confused about whether the fitness function defines fitness De Re or De Dicto.

http://plato.stanford.edu/entries/p[…]ts/dere.html

ie, the fitness function looks for the shortest solution, the Steiner solution is the shortest solution, therefore the fitness function looks for the Steiner solution.

‘The shortest solution’ seems to be referentially opaque here. Compare to: I’m looking for the tallest man in the room (true), John is the tallest man (true), therefore I’m looking for John (false; never heard of the guy).

Wow, I finally used my philosophy degree for something. :)

Right, have sent in my solution. Everyone else: get cracking :)

The only real criticism I can imagine is that the program was written in C++ and not Java.

I do prefer C++ myself, but there’s still a case for Java.

The shortest solution’ seems to be referentially opaque here. Compare to: I’m looking for the tallest man in the room (true), John is the tallest man (true), therefore I’m looking for John (false; never heard of the guy).

But when you find the tallest man, he’s John, so you must have been looking for John!!!

(The three exclamation marks here are a special symbol that means that the sentence is in the IDiot-speak dialect.)

At the moment I’m the one doing some “seeing through,” and I’m seeing through Salvadore Cordova’s claim of “seeing through.” What he apparently doesn’t comprehend is that specifying parameters that dictate conditions for success is NOT the same thing as specifying a target. For example, specifying a “strength of gravity” in a computer simulation is not the specification of a target even while it constrains the solutions. Whether by lack of comprehension or by intention, the IDists misrepresent matters by engaging in equivocation. As usual.

Nice experiment, Dave.

To find a solution, one could try the software at:

http://www.diku.dk/geosteiner/

Salvador wrote To find a solution, one could try the software at:

(sigh) But that’s not the point.

The ID claim is that solutions to complicated problems are in the realm of design, not evolution. That evolution (random mutation and natural selection) is incapable of efficiently solving complex problems.

This un-intelligent genetic algorithm, which many believe to be a reasonable model of the driving mechanism of natural selection, stands ready to take the challenge.

You, in contrast, are an intelligent agent, and should be able to ‘whup it easily.

So go ahead, play God. Design. (but don’t cheat and look up the answer)

Comment #119718

Posted by Salvador T. Cordova on August 15, 2006 08:59 AM (e)

To find a solution, one could try the software at:

http://www.diku.dk/geosteiner/

I’m with Stevaroni. If the stated math problem is so hard that the UD Software Team has to go out-of-house to get the solution, then what does that say about evolutionary algorithms also being able to derive the correct answer?

Sal’s dilemma can be can be illustrated by this fictional scenario:

Teacher: So, what is the length of the Hypotenuse? Salvador: To find a solution, one could read the book at Amazon.com

And Sal wonders why he flunked.

Brit Wrote:

IDers are bending over backwards to deny that genetic algorithms can actually create something. It’s actually a really dumb argument that they’re using. Essentially what they’re saying is that if my boss comes to me and says, “design a flying car that gets 100 MPG” and I turn around and actually do it, that I haven’t done any actual work. According to them, all the design was done by my boss when he defined what he wanted: “a flying car that gets 100 MPG”. What a joke.

Here’s another example.

Imageine that I wish to arbitrarily assert that no work (in the human economic sense) can possibly be performed by anything other than human muscles.

As a couterargument, you show me a car. The car, you claim, performs the work of moving people and goods from point A to point B using power from an internal combustion engine. The ICE, you claim, is completely inorganic and is not deriving power from human muscle tissue.

If I were arguing in the style of the IDers, I would then counterargue that you are in fact pushing on the gas pedal and pumping gas into the car. Your pushing and pumping, I would argue, provides all of the work necessary for the car to get from point A to point B.

The utter absurdity of this argument is evident when you apply it in a different context than evolution. It is possible to give this argument some veneer of truth in the evolution context since genetic algorithms have so far only been capable of designing things within narrow and clearly defined spaces. As GAs and other evolutionary computation techniques advance, this argument will become more and more apparently absurd as such systems give rise to more and more impressive and elaborate results.

“Is it dishonesty, or is somebody just not getting it?”

They aren’t getting it. It’s called ideology-induced blindness. Religious fundamentalists are incapable of grasping evolutionary concepts for the same reason that doctrinaire Soviet Marxists were incapable of grasping certain concepts from economics and statistics. A suitably dogmatic mind will actually prevent itself from making connections that could threaten a cherished dogma.

If you discuss with such people, you can actually see this taking place. You will debate, and they will follow… up until you point out something that is a) obvious and b) contradicts their dogma. At this point, a “lock up” occurs. Their mind simply refuses to perform the required connection, induction, integration, etc. They simply will not allow A to be A. It’s quite astounding to witness. I have experienced it with all sorts of dogmatic people, not just fundamentalist Christians.

I’ve sent in my solutions.

SteveS calls them information cooties. Put a human being anywhere on the same continent as the experiment and it is infected with them.

Paul

I think I got that from someone else, but I don’t know who. But yeah, Salvador’s idiotic objections would make evolution uniquely impossible to model in silico, unlike every other area of science.

I think it is pretty pathetic that Salvador’s response is partially “It’s written in archaic Fortran, so its hard to solve.”

Sal boldly states:

But the most important thing in hitting such a target of opportunity is that by intelligently designing the right strategy, one can hit a target of opportunity without the target being explicitly described.

The target of opportunity described in the SLAM-ER example is not analogous to an EA search. SLAM-ERs may include “a man in the loop” and are not independent. With GPS updates they are also not independent. A SLAM-ER “(d)esignates the target through real-time digital pattern matching” a preprogrammed set of criteria of potential targets and not a search of leading to a solution of a problem.

Using Sal’s example, for ID to be effective it requires constant intervention by the designer a “man in the loop” or something analogous to a GPS system to monitor and redirect the system at appropriate times such as shifting environmental conditions, effectively recognizing new ecological niches and then replacing natural selection. If this is what is being proposed then ID becomes testable since it proposes monitoring device(s) keeping track of earths changing environments. All that is required is a search for the monitoring device(s).

With the rapid loss of species and shifting environments due to global warming new ecological niches are opening and the monitoring devices should be active. This would be a good time to begin the search for the mechanisms. If hidden somewhere on earth the DI needs to recruit geologists to examine designed geological features where these mechanisms may be hidden. Then by reverse engineering everyone would have a better understanding of the designer(s).

A disembodied designer keeping track of changing environments and intervening at the appropriate times has been discussed over and over again.

Delta Pi Gamma (Scientia et Fermentum)

Dave Thomas wrote:

ID advocates, since y’all are saying that the solution is already implicitly defined in the statement of the problem

By the way Dave, that does not represent mine or any IDers position that I know of.

I can understand perhaps how you may have come to that conclusion. You might even be tempted to fault my infelicitous expression of ideas for your horrid misunderstandings and misrepresentations and mischaracterizations of what IDers believe. I presume the last thing you would blame this mischaracterization on would be the Panda’s propensity to uncharitably characterize what IDers say.…Fine!

The point is, what you said in your opening post does not accurately represent what I or other IDers believe.

I hope you’ll post an addendum somewhere on this site conveying the fact to the readers that what you said does not represent my position. If you have to sugar coat it by arguing that you were misled because you couldn’t decode what I was claiming, fine. But I request you withdraw your mischaracterization of what I believe.

To find a solution, one could try the software at:

http://www.diku.dk/geosteiner/

The question isn’t what the correct solution is; from the earlier post on this GA, it was made clear that evolution very rarely comes up with the optimal solution. The question is what evolution will produce. It doesn’t make any difference if either you or evolution comes up with the right answer. The ID claim is that the specification of the problem determines what solution the genetic algorithm will give you. Therefore, from reading the specification of the problem, you should be able to deduce what solution(s) will be produced.

If you can’t, it’s kind of damning.

Comment #119993

Posted by Salvador T. Cordova on August 16, 2006 10:33 AM (e)

Dave Thomas wrote:

ID advocates, since y’all are saying that the solution is already implicitly defined in the statement of the problem

By the way Dave, that does not represent mine or any IDers position that I know of.

I can understand perhaps how you may have come to that conclusion. You might even be tempted to fault my infelicitous expression of ideas for your horrid misunderstandings and misrepresentations and mischaracterizations of what IDers believe. I presume the last thing you would blame this mischaracterization on would be the Panda’s propensity to uncharitably characterize what IDers say.…Fine!

The point is, what you said in your opening post does not accurately represent what I or other IDers believe.

I hope you’ll post an addendum somewhere on this site conveying the fact to the readers that what you said does not represent my position. If you have to sugar coat it by arguing that you were misled because you couldn’t decode what I was claiming, fine. But I request you withdraw your mischaracterization of what I believe.

Oh really? Then what am I to make of this claim by Stephen Meyer:

these programs only succeed by the illicit expedient of providing the computer with a “target sequence” and then treating relatively greater proximity to future function (i.e., the target sequence), not actual present function, as a selection criterion. As Berlinski (2000) has argued, genetic algorithms need something akin to a “forward looking memory” in order to succeed. Yet such foresighted selection has no analogue in nature.

Or this one, by William Demsbki?

What’s more, evolutionary algorithms find their way through all the various zigzagging tangles—most of which don’t work—to one that actually does. This is remarkable. Even so, the fitness function that prescribes optimal antenna performance is well-defined and readily supplies the complex specified information that an optimal crooked wire genetic antenna seems to acquire for free.

I had an email exchange with Thomas (personal communication), and I pointed out that the program created no real novelty and no information besides the information originally contained within the fitness function itself. My logic was as follows: the desired solution has (1) all five points connected, and (2) the shortest path-length. The program selected for networks that (1) connect all five points, and (2) have shortest path-lengths. It is no wonder that the program converges regularly upon short, optimum networks; it has been told precisely what to do by explicit instruction in the fitness function.

Sal, you doth protest too much, methinks.

Dave

The point is, what you said in your opening post does not accurately represent what I or other IDers believe.

Salvador, maybe you should clearly explain your position to us. Specifically, does the fitness function contain at least as much CSI as the solution? If so, I have full confidence that you can explain what that means and how you arrive at that conclusion without begging the question.

secondclass: I think Sal is saying that the CSI is not exclusively “hidden” in the fitness function. He’s saying that CSI is “hidden” in the choice of the pair (problem, fitness function). A devious ID verifies which problems can be solved by using an evolutionary algorithm with an “innocent” looking fitness function, and then produces the results as evidence on how generic tools random mutation and cumulative selection are.

Of course, even the above position would admit that there exists CSI which can be produced by non-intelligence. Let me guess. The argument continues: this CSI, however, can only be procuded in the world of math and computation, instead of the natural world free of human manipulation, and it can only be detected by ID? However, even this strange postition would require IDers to retreat from the claim: “CSI can only be produced by ID”.)

Evolution, I’m given to understand, has no goal.

Your understanding is wrong…survival is the very explicit “goal” of evolution/natural selection.

Not a “deliberate” goal in that a bacterium wakes up in the morning and says, “I am going to do everything I can to survive today,” but a de-facto goal in that those organisms that survive produce progeny.

Comment #119985

Posted by Michael Suttkus, II on August 16, 2006 09:29 AM (e)

No, no no! Intelligent Design is not a theory of intelligence or design! It’s a theory of design DETECTION! See, if you make a solution, they will scan it and see if it is intelligent with their classic and undeniable, “Yes/Maybe Yes” test of power! They don’t design anything. Only intelligences can design things. If they could design something, don’t you think they would have designed a scientific theory by now?

Tell you the truth, the whole “detection thing” reminds me of the Doctrine of Signatures:

The “Doctrine of Signatures” has been an idea of herbalists for centuries, but it did not become part of the medical thinking until the middle of the seventeenth century. In simple terms, the “Doctrine of Signatures” is the idea that God has marked everything He created with a sign (signature). The sign was an indication of the purpose for the creation of the item. The “Doctrine of Signatures” was popularized in the early 1600s by the writings of Jakob B�hme (1575-1624), a master shoemaker in the small town of G�rlitz, Germany. At the age of 25, B�hme had a profound mystical vision in which he saw the relationship between God and man. As a result of the vision, he wrote “Signatura Rerum; The Signature of all Things”. His book espoused a spiritual philosophy; however it soon was adopted for its medical application. The Doctrine states that, by observation, one can determine from the color of the flowers or roots, the shape of the leaves, the place of growing, or other signatures, what the plant’s purpose was in God’s plan. A good example would be Hepatica acutiloba, a perennial wildflower that is common in the eastern U.S. Hepatica has a three-lobed leaf that supposedly bears a resemblance to the liver. Because of this, herbalist believed the plant to be effective in treating liver ailments. Even though the “Doctrine of Signatures” was a product of medieval alchemists and herbalists, a similar belief was held by Native Americans and Oriental cultures. Today you can still find a similar belief among old-timers in the Appalachian region.

Something though-up by a shoemaker having hallucinations.

Moses Wrote:

Tell you the truth, the whole “detection thing” reminds me of the Doctrine of Signatures:

The doctrine in it’s Chinese form is a major reason rhinoceros continue to be poached. The big phallic horn on their nose is “proof” that they are a cure for impotence, so Chinese herbalists pay big for rhino horn and another animal dies from human ignorance.

A modified doctrine of signatures forms the basis of the homeopathic “Law of Similars”, the rather surreal idea that if a substance CAUSES a symptom in a healthy person, a dillution of that substance, in a person suffering those symptoms, will be a cure. Taken to it’s logical conclusion, since arsenic causes all the symptoms of arsenic poisoning, a dillution of arsenic can be used to cure arsenic poisoning. The method of identifying the signatures is different, but it’s the same concept, and just as silly.

(I must be a total idiot. I’m already embroiled in fights on vaccine denial and “organic” food on other threads, and here I am ticking off the homeopaths.)

caligula Wrote:

I think Sal is saying that the CSI is not exclusively “hidden” in the fitness function. He’s saying that CSI is “hidden” in the choice of the pair (problem, fitness function).

If that’s Sal’s position, then he has some ‘splaining to do. According to Dembski, Thomas’s program must contain at least as much CSI as the solution. But his program doesn’t contain a problem statement, so it would seem that the CSI must reside exclusively in the fitness function.

caligula:

“But I personally think that evolution does *additionally* create brand new information, actually new conceptual layers where new kind of complexity can be measured. The most obvious example is the evolution of multi-cellular life.”

I’m not so confident discussing information without a tangible definition. KC IT defines how much there is, Shannon IT how it is transmitted, and Maxwell’s demons how it relates to entropy and memory. That doesn’t help here.

Nevertheless, I think you should also add two earlier sources here. “other DNA carriers in the ecosystem” (coevolution between species) and “the other genes of the same genome” (coevolution between genes) are probably part of your conceptual layers too. At least they are qualitatively different in the sense that coevolution apparently break NFL theorems.

“Does this stuff cure cancer? is it gone?”

At some point, mankind will cure cancer. Please note that I’ve done the front-loaded, CSI-laden hard part above. So when they’re handing out those nobel prizes, I *FULLY EXPECT* to be top of the list.

secondclass Wrote:

If that’s Sal’s position, then he has some ‘splaining to do. According to Dembski, Thomas’s program must contain at least as much CSI as the solution. But his program doesn’t contain a problem statement, so it would seem that the CSI must reside exclusively in the fitness function.

I think Salvador has this CSI on two different conceptual layers”, so to speak. (How CSI maps from two different conceptual layers to just one layer, the solution, is beyond me though.)

Yes, some of the CSI might be “hidden” in the fitness function. Maybe some of it is also “hidden” in the problem specification.

But, most importantly. If I understand correctly, Salvador believes that there is only a tiny (and alien-to-nature) set of problems which are solvable with GAs by using an “innocent” looking, minimal fitness function. As opposed to the vast majority of problems which, if solvable with GAs at all, require a very lenghty and CSI-rich problem specification. So the very detection of problems belonging to the first group requires CSI and an ID! And this detection is only possible in the world of math, which supposedly is larger than life; problems such as the Steiner problem don’t exist in wild nature and thus have nothing to do with biological evolution.

If the above is Sal’s position, I agree that some serious explaining is in order.

The doctrine in it’s Chinese form is a major reason rhinoceros continue to be poached. The big phallic horn on their nose is “proof” that they are a cure for impotence, so Chinese herbalists pay big for rhino horn and another animal dies from human ignorance.

Slight correction.

I think Chinese herbalists are well aware that things like rhino horns and tiger testicles etc are absolutely useless for Chinese medicine.

The thing is the rich would pay insane amounts of money for anything that sounds like it can prolong life. And the remedies must sound exotic.

Strangely, there are some Chinese medicinal mixtures which requires the ash of human hair. Human hair, fingernails and rhino horns are made out of the same thing (keratin, right?). But who wants to pay for human hair they can get from the barber’s when they can get a few grams rhino horn for 100 000 000 yuan and boast about it?

It’s all about economics. The rarer a commodity, the more you charge for it.

So it is due to human ignorance, but mostly the ignorance of the rich and those who the scams are designed for.

I wondered:

The IDers still call back to Steiner solution, which can be found even without genetic algorithms (although hardly by their own power). How about trying the experiment in some way that would DISQUALIFY Steiner solution? For example, imagine putting a line segment somewhere within the picture, and specifying that no line of the solution can cross that segment. What is known about this modified problem?

Then, you can put the segment in such a way that it would completely eliminate Steiner solution from the list of possibilities, and THEN let them try to come up with a good solution :)

I wondered:

The IDers still call back to Steiner solution, which can be found even without genetic algorithms (although hardly by their own power). How about trying the experiment in some way that would DISQUALIFY Steiner solution? For example, imagine putting a line segment somewhere within the picture, and specifying that no line of the solution can cross that segment.

If you imagine the problem as finding, say, shortest set of straight roads connecting cities, this is an obvious generalization - what if there is a fence you can’t cross or some part of terrain where you just can’t build the road over? What if there is a lake?

Then, you can put the segment in such a way that it would completely eliminate Steiner solution from the list of possibilities, and THEN let them try to come up with a good solution :)

Syntax Error: not well-formed (invalid token) at line 1, column 72, byte 72 at /usr/local/lib/perl5/site_perl/mach/5.18/XML/Parser.pm line 187.

demallien sneaked in this:

That’s not a bad idea actually [DISQUALIFYing the Steiner solution]. If a GA can come up with a better solution than the IDist, who presumably feel that they are more intelligent than a GA, then it completely blows away many of ID’s criticisms of evolution as a theory, in a way that they can understand, ie a blind GA created a clever solution that an intelligent agent couldn’t find.

LOL .…or one the ‘intelligent agent’ never looked for, or worse couldn’t be bothered, or even, while preening in front of the cameras facinated by his own reflection was thinking of marrying Jocasta.

demallien Wrote:

If a GA can come up with a better solution than the IDist, who presumably feel that they are more intelligent than a GA, then it completely blows away many of ID’s criticisms of evolution as a theory, in a way that they can understand, ie a blind GA created a clever solution that an intelligent agent couldn’t find.

Sounds like they should start worshipping the GA were that to happen. ;)

Here are the 4 proposed networks

In Cartesian coordinates, letting Vertex 1 = be the origin, or equivalently Vertex 1 = (0,0), and Vertex 6 be at (x,y) = (800,300), I propose the following 4 solutions of equivalent length:

Solution 1:

steiner points:

A = (86.6025, 150) B = (313.3975, 150) C = Fermat Point joining vertex 5,6,3 = (730.42, 224.88)

A connects to Vertex 1, Vertex 4, Steiner Point B B connects Steiner Point A, Vertex 5, Vertex 2 C connects to Vertex 5, Vertex 6, Vertex 3

Solution 2:

steiner points: A = (86.6025, 150) B = (313.3975, 150) C = Fermat Point joining vertex 2, vertex 6, vertex 3 = (730.42, 75.13)

A connects to Vertex 1, Vertex 4, Steiner Point B B connects Steiner Point A, Vertex 5, Vertex 2 C connects to Vertex 2, vertex 6, vertex 3

Solution 3:

steiner points:

E = Fermat Point joining vertex 1,vertex 4, vertex 5 = (69.58, 224.88) F = (486.6025, 150) G = (713.3975, 150)

E connects to Vertex 1, Vertex 4, Vertex 5 F connects to Vertex 5, Vertex 2, Steiner Point G G connects Steiner Point F, Vertex 6, Vertex 3

Solution 4:

steiner points:

E = Fermat Point joining vertex vertex 1, vertex 4, vertex 2 = (69.58, 75.13) F = (486.6025, 150) G = (713.3975, 150)

E connects to Vertex 1, Vertex 4, Vertex 2 F connects to Vertex 5, Vertex 2, Steiner Point G G connects Steiner Point F, Vertex 6, Vertex 3

Here is the sum of all the lengths where each solution has the same length, but I’ll us solution one to calculate length, and then one can assume the other 3 solutions have the same length:

1 to A 173.2050808 4 to A 173.2050808 A to B 226.7949192 B to 5 173.2050808 B to 2 173.2050808 sub total left 919.6152423

5 to C 338.8521647 3 to C 235.4003099 6 to C 102.3907822 sub total right 676.6432568

total 1596.258499

Each Steiner point has degree 3 with each connected edge opening 120 degrees from the adjacent one radiating from the same steiner point. These are necessary but not yet sufficient conditions for a steiner solution. I await Dave’s correct solution, but this one I think these are at least MacGeyver’s.

Here are the 4 proposed networks

Show your work, Sal. There is not a hint here of how you derived these proposed solutions from Dave’s code. If you don’t do that, then this doesn’t in any way help the bald assertion that the solution’s information is already somehow in the program.

I gotta question with which to bamboo-oozle you, oh Sal-of-the-structural-grasses:

When I-wish Dementski exposes his stupidity so publicly, his not-clever-beyond-measure cover story is that he’s “really” engaging in *koff koff* street theater.

So when you pull stunts this ridiculous, hanging your intellectual cheeks out, forcing the entire internet to cover its metaphorical peepers in an effort to avoid seeing the unmentionable, whadda you call it?

Street cleaning? Squid porn? Science friction?

On second thought, I don’t really wanna know.

THE CHALLENGE HAS BEEN CLOSED

Now that the ID community has officially weighed in, and since I’m frantically working on tomorrow’s posting of the results, I declare this Challenge to be officially Closed.

I’ll see you Monday with the results.

Dave

P.S. It’s funny – of the over 30 other entries received, only Salvador Cordova’s was posted publicly as a comment. Everyone else was able to follow the instructions, though. Curious, that.