When Improbabilities Become Exponentially Improbable

_{Gil Dodgen

September 15, 2006

Intelligent Design}

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

The first insight I had into the nonsensical nature of the random mutation (RM) part of the RM plus natural selection (NS) hypothesis came through my mathematical studies and experience in software engineering. See here for some probabilistic calculations about the most simple of all computer programs.

Of course, Darwinists always ask, How can you know that RM+NS can’t account for all of life? The answer is simple, and it’s called probabilistic combinatorics.

The underlying biochemical and information-driven functions of living systems are tightly integrated and controlled by an unimaginably complex, sophisticated, fault-tolerant, self-repairing, self-replicating computer program. Components of such a system cannot be altered to produce significant innovation without the simultaneous, coordinated alteration of the components with which they interact. This is what software engineers do, not copying errors.

This is a deafening cry of design.

A microbe did not mysteriously mutate into Mozart and his music, and most people, thankfully, are smart enough to figure out that this is a silly idea.

Comments

Tom English:
According to Arthur M. Lesk, the Shannon entropy of base pairs (nucleotides) in the human genome is 1.63 bits. Assuming 3 billion base pairs, the length of the maximally-compressed genome is no more than 4.89 gigabits. Assuming 3.5 billion years for that much information to enter the genome, the rate of information gain is just 1.4 bits / year.
Bad calculation. The first form of life is bacterial. It shows up about 3.5 billion years ago. The conditions for life on the early earth didn't begin until roughly that time. So I'm afraid you don't have 3.5 billion years to work with. If you look into it a little more closely you might find a different approach you can take though. In the end, the numbers might still suggest the same information increase per year. But that proves nothing. We live in a quantum world. Nucleotide A, doesn't just add itself to nucleotide B, and then A and B add themselves to nucleotide C. There are minimal requirments. Continual increase over time is not a plausible mechanism for the accumulation of information.PaV_{September 18, 2006
September
09
Sep
18
18
2006
08:29 AM
8
08
29
AM
PDT}

Tom English:
Could you give us some links to actual applications of probabilistic combinatorics by ID advocates?
Why dont' you look at Fred Hoyle's, "The Mathematics of Evolution", where he calculates the probability of cytochrome C coming about by chance. It's really quite easy to demonstrate that proteins have very little likelihood of forming by chance. Do you have some idea as to how they formed by chance, as in the first cell?
This is most certainly not the case. There are many applications in which algorithmic generation of pseudorandom numbers does not suffice.
What's the point here, Tom? What are you disputing? So there are now better generators than a Monte Carlo simulator. So what. How does that affect my argument?
You seem to have problems with the Second Law of Thermodynamics. That is why there must be occasional errors in reproduction.
And you seem to have problems with the repair systems that biological forms have in place. Next question please.PaV_{September 18, 2006
September
09
Sep
18
18
2006
08:19 AM
8
08
19
AM
PDT}

Well come on Tom English. Surely you aren't goimg to take that lying down or are you? We will soon see won't we? "A past evolution is undeniable, a present evolution undemonstrable." John A. DavisonJohn A. Davison_{September 18, 2006
September
09
Sep
18
18
2006
06:29 AM
6
06
29
AM
PDT}

Tom English: "I predict that as, say, quantum random number generators decrease in cost and become more widely available, evolutionary computation researchers will move away from pseudorandom number generators" Anyone with a sound card and some cheap software can get high quality random numbers due to the semiconductor-based noise generated in the electronics. For example: http://random-numbers.qarchive.org/mike1962_{September 18, 2006
September
09
Sep
18
18
2006
06:27 AM
6
06
27
AM
PDT}

Tom English I love your admission that I am a "dangerous man." What a tribute! Is that why you refuse to answer my challenges, why you pretend that I do not exist, why my name is not to be mentiond at After The Bar Closes and why you go right on blindly supporting the most failed hypothesis in the history of science, oblivious to the fact that it is a proven hoax? Being "dangerous" is what science is all about and when an adversary refuses to respond to another's challenges he has forfeited his right to be a scientist. Furthermore, he has displayed with that stance that he is an intellectual coward. The "establishment," to which you so obviously belong, is a highly organized bunch of "groupthinking" ethical degenerates who have survived for one reason only. They are hamstrung by a congenital view of the world that cannot realize that which is so obvious to those that do not suffer from that malaise. They are "born that way" and there is nothing that can be done for them. Their fate, like that of every other aspect of the world, past and present, was "prescribed." That establishment survives now exactly as it has in the past by pretending that the Darwinian fairy tale has never had any critics. You are doing nothing more than continuing that tradition when you dispense with me as you just did by describing me as a "dangerous man". Neither you nor any other Darwinian has ever answered my several challenges to your mindless dogma either here at Uncommon Descent or anywhere else. I have been summarily banned at every Darwinian sronghold on the internet. I managed just one message at P.Z. Meyers' "Pharyngula," to be greeted with "your stench has preceeded you" and instant bannishment. I love it so. You are a hero to the "After the Bar Closes" crowd and so apparently is Alan Fox. What more need be said? I don't exist over there either, just as I don't at ARN and EvC and God only knows how many other ""closed union shops" too numerous to mention. I love it so. Apparently I am not nearly "dangerous" enough. I will try to do better in future. "When all think alike, no one thinks very much" Walter Lippmann "You ain't heard nothin' yet." Al Jolson "A past evolution is undeniable, a present evolution undemonstrable." John A. DavisonJohn A. Davison_{September 18, 2006
September
09
Sep
18
18
2006
03:32 AM
3
03
32
AM
PDT}

"According to Arthur M. Lesk, the Shannon entropy of base pairs (nucleotides) in the human genome is 1.63 bits. Assuming 3 billion base pairs, the length of the maximally-compressed genome is no more than 4.89 gigabits. Assuming 3.5 billion years for that much information to enter the genome, the rate of information gain is just 1.4 bits / year" Pretty non sense. First, most of the significant advancments in the genome complexity did occur in a very narrow time range; second, the more complex genomes have very limited number of organisms to decently maintain RM+NS operation; third, and very important, arguing about the yearly rate in #bits is completely meaningless without any information about the actual regularity of the search space.kairos_{September 17, 2006
September
09
Sep
17
17
2006
11:48 PM
11
11
48
PM
PDT}

According to Arthur M. Lesk, the Shannon entropy of base pairs (nucleotides) in the human genome is 1.63 bits. Assuming 3 billion base pairs, the length of the maximally-compressed genome is no more than 4.89 gigabits. Assuming 3.5 billion years for that much information to enter the genome, the rate of information gain is just 1.4 bits / year. I assume that with the long repeating segments in the genome, someone has managed to compress the genome to considerably less than 4.89 gigabits.Tom English_{September 17, 2006
September
09
Sep
17
17
2006
11:07 PM
11
11
07
PM
PDT}

John, You warned me some time back you were a dangerous man, and I took you at your word. ;)Tom English_{September 17, 2006
September
09
Sep
17
17
2006
09:23 PM
9
09
23
PM
PDT}

Gil: "To be fair youÃ¢â‚¬â„¢d need to expand on which the source code is dependent and include the linked libraries, compiler, OS, etc." To be fair, you would code the program in the lambda calculus and submit it to a universal combinator. The universal combinator is a universal computer that has been encoded in just 425 bits.Tom English_{September 17, 2006
September
09
Sep
17
17
2006
08:52 PM
8
08
52
PM
PDT}

Gil, You had a lot more in mind than to give folks a lesson in "probabilistic combinatorics." Given that your "66-character combinations" are actually permutations, I don't peg you as someone who does much work in the field. Your point was to say that if discovering such a simple program by chance is so unlikely, then how could neo-Darwinian processes obtain "biochemical and information-driven functions of living systems [that] are tightly integrated and controlled by an unimaginably complex, sophisticated, fault-tolerant, self-repairing, self-replicating computer program"? But I truly am interested in combinatorics and probability and information, and what I want to know is how many bits of (algorithmic) information there are in the compressed human genome. I would divide that number by, say, 3.5 billion years to estimate the mean number of bits of information added to the genome each year. For now, what I know is that there are about 3 billion base pairs. This indicates a gain of less than one base pair per year, and considerably less than 2 bits of information per year (because the genome is compressible). To claim that evolution has picked up less than 2 bits of information per year is hardly to worship chance. My argument is sound, and the only way you have to counter it is to oppose gradualism. That is why the notion of irreducible complexity is vital to ID.Tom English_{September 17, 2006
September
09
Sep
17
17
2006
08:32 PM
8
08
32
PM
PDT}

This is for Tom English and all other Darwinian zealots. Since no one pays any attention to me anyway, here as elsewhere, I will let another with more illustrious credentials speak for me, the greatest Russian biologist of his day and, in my opinion, the greatest evolutionist of all time. "The struggle for existence and natural selection ARE NOT progressive agencies, but being, on the contrary, conservative, maintain the standard." Leo Berg, Nomogenesis, page 406, my emphasis. "Hereditary variations are restricted in number and they develop in DETERMINED direction." ibid, my emphasis. "Evolution is in a great measure an unfolding of PRE-EXISTING rudiments." ibid, my emphasis. Ergo - The Prescribed Evolutionary Hypothesis. So much for chance and accordingly so much for Darwinism in any form, the biggest, most infantile and persistent hoax in the history of science. How do you like them apples Tom? "A past evolution is undeniable, a present evolution undemonstrable," John A. DavisonJohn A. Davison_{September 17, 2006
September
09
Sep
17
17
2006
04:35 PM
4
04
35
PM
PDT}

My Hello World program serves to illustrate how quickly combinatorics create huge search spaces. To be fair you'd need to expand <stdio.h> on which the source code is dependent and include the linked libraries, compiler, OS, etc. Natural selection does indeed not have a goal, but it must produce islands of function and steadily increasing functional complexity and sophistication. As Stu Harris pointed out:
Your "Hello World" program's existence is even more wildly improbable to arrive at by chance and selection than you have stated. It is written in C. This presumes some BNR definition of the C language to begin with. How did that come about by chance and selection? It assumes a compiler written in some other language that can compile and interpret the C code. How did that come about by chance and selection? The compiler runs on an complex specified operating system, which runs on complex hardware which is composed of metallurgical, mineral, and plastic complexities that... well you get the picture. Multiply the probability of a "Hello World" program by the probability of the pyramid below it and you can come close to the un-stateable improbability of the world your program is trying to greet.
The bottom line is that there is no credible evidence that RM+NS has the creative power attributed to it in biology. It is an article of faith based on presupposition. This should be admitted, but it never is, because the entire house of cards would be seen to rest on a very shaky foundation.GilDodgen_{September 17, 2006
September
09
Sep
17
17
2006
04:08 PM
4
04
08
PM
PDT}

P.S. -- The "#include" of stdio.h was truncated by WordPress above, but I did count all the characters in the line.Tom English_{September 17, 2006
September
09
Sep
17
17
2006
02:54 PM
2
02
54
PM
PDT}

Gil, If the only tool you have is a hammer, then every problem is going to look like a nail. I am a computer scientist. I would love to make the genome into a program. But the "genetic program" metaphor of the 1950's simply doesn't make a good scientific model. Genetic regulatory networks are modeled in various ways. Stuart Kauffman published work on one approach, Boolean networks, in 1969, so this is nothing new. But if you did want to pursue a programming model, something like a C program would be inappropriate. Genes are weakly analogous to rules in a production system (e.g., OPS5). The rules can be arranged in a network or collection of networks. Now what is interesting about this is that one of the advertised virtues of production systems is that can add, remove, and modify rules without changing other rules. In other words, there's a high degree of modularity. And you can have rules that, in their present form, never get used, and that have little impact on the performance of the system. What I'm driving at here is that all you have to do is to change the programming model from imperative programming to rule-based programming and evolution becomes much more feasible. (In fact, our mutual friend in La Jolla and I were once going to work on evolution of rules, but I decided on another approach.) Of course, the genetic programming community likes to work with LISP, a language with almost no syntax rules. I have to say that in the example you link to you repeat a fallacious argument that creationists advanced a number of years before ID came on the scene. You cannot argue against neo-Darwinism by making it into something it is not. Neo-Darwinian evolution has no goal. Yet you set up a goal and argue that it would be impossibly hard to achieve. If you want actually to challenge neo-Darwinian theory with a counting argument, you have to take it as given by scientists. Another point about your example is that, even if I count 7 bits for each and every character, I get only 525 bits. Thus if an evolutionary process could pick up only a bit of information per year in interaction with the "environment," it would have the C program in a mere 525 years. You get a "deafening cry of design" only if you slip in an assumption that gradual gains in information are impossible. I should mention that a fair amount of your code is not necessary. Here is a 280-bit program that outputs something: #include main(){printf("H");} This is an approximation to your full C program, and 2 ^ -280 is well below Bill Dembski's universal probability bound. Furthermore, in a 64-character LISP implementation, (print "hello, world!") is a 144-bit program. A question I have about your C program is, if I came up to you and said I generated it all at once by chance, how would you compute the complex specified information to show that I had not?Tom English_{September 17, 2006
September
09
Sep
17
17
2006
02:46 PM
2
02
46
PM
PDT}

Just as a reflection, it seems to me that it would be a great thing if biology students were required to take a class on computer programming.... Anyone having gone through that experience would instinctively Ã¢â‚¬Å“knowÃ¢â‚¬Â that RM+NS just isnÃ¢â‚¬â„¢t going to get it done.
Speaking as someone with a bachelors in Computer Science (software engineering concentration), with two decades of programming experience, and as someone who completed the undergraduate work in pre-med, the evolutionary mechanism of RM+NS makes very good sense to me. You have to understand that genes are not like sections of computer code. I think you are going to underestimate the usefulness and flexibility of RM+NS if that's your analogy. I think that's why engineers and programmers get hung up on RM+NS: they are thinking about evolution through their experience of programming languages and engineering tasks. It's a natural human trait to think about new information through the lense of past experience, but in this case, I think the analogy doesn't hold very well and it leads to some misleading conclusions.BC_{September 17, 2006
September
09
Sep
17
17
2006
01:06 PM
1
01
06
PM
PDT}

Sorry Tom English, but it won't wash. There was never a role for randomness in any aspect of phylogeny just as there is now no role for it in ontogeny. To claim that which cannot be demonstrated is intellectally dishonest and pure fantasy. "Neither in the one nor in the other is there room for chance," Leo Berg, Nomogenesis or Evolution Determined by Law, page 134 Please note the complete title of Berg's opus magnus, in my opinion the most significant single work in all of the evolutionary literature. We will soon be discussing Bergian evolution, as the Darwinian myth descends into oblivion to join the Ether of Physics and the Phlogiston of Chemistry, nothing more than a footnote. How it has survived as long as it has is a mystery. It should have died twelve years after its inception with the publication of St George Jackson Mivart's book with the tongue-in-cheek title - Genesis of Species. That is roughly when the Ether of Physics bit the dust and a century after Phlogiston had also gone the way of all inventions of the human imagination. It is hard to believe isn't it? I love it so! Who is next? "A past evolution is undeniable, a present evolution undemonstrable." John A. DavisonJohn A. Davison_{September 17, 2006
September
09
Sep
17
17
2006
12:53 PM
12
12
53
PM
PDT}

PaV: "And as to 'randomness', anyone doing any kind of programming at all knows that the only kind of 'randomness' you want in a program is the kind the 'program in', for example, a Monte Carlo simulation program." This is most certainly not the case. There are many applications in which algorithmic generation of pseudorandom numbers does not suffice. Pseudorandom number generators implement small algorithms, and this means that they can supply only a very limited amount of algorithmic information. There are web sites at which you can get random numbers obtained through measurement of natural phenomena, e.g., http://www.fourmilab.ch/hotbits/ http://www.random.org/nform.html I have shown in published work that, in practice, only a small fraction of all possible searches can be implemented with pseudorandom number generation. I predict that as, say, quantum random number generators decrease in cost and become more widely available, evolutionary computation researchers will move away from pseudorandom number generators (though they will still use pseudorandom generators to achieve repeatable results while developing software). PaV: "Thus, it is not surprising that biological forms use 'randomness' here and there; but that 'randomness' is very likely 'programmed in'. So, paraphrasing the world-renowned Richard Dawkins, 'Biology is the study of life forms that give the appearance of randomness.' I believe that this strikes much more closely to the truth of life." You seem to have problems with the Second Law of Thermodynamics. That is why there must be occasional errors in reproduction. Nothing in nature is "trying" to introduce random errors. Nothing in nature is faking them. Physics necessitates random errors, and they are a source of information in evolution. I think the biggest problem IDists have with this is that they don't know enough information theory to understand that an entity is most informative when it is "most random."Tom English_{September 17, 2006
September
09
Sep
17
17
2006
10:43 AM
10
10
43
AM
PDT}

DLH: "Here are some interesting searches on Probalistic Combinatorics. Look forward to combinatorial specialists adding serious reviews." I know enough to know what is and what is not probabilistic combinatorics. (I have published a bit on the properties of decision trees corresponding to randomized search algorithms.) Gil has done only elementary counting -- there's no "probabilistic" to it. I would like to see some links to actual applications of probabilistic combinatorics by ID advocates.Tom English_{September 17, 2006
September
09
Sep
17
17
2006
09:56 AM
9
09
56
AM
PDT}

DLH: "Here are some interesting searches on Probalistic Combinatorics. Look forward to combinatorial specialists adding serious reviews." Well, probabilistic combinatorics is a fancy term. And it's also an interesting field, for those of us who have actually looked into it. (I have published a bit on the properties of decision trees corresponding to randomized search algorithms.) The problem is that Gil does only elementary counting -- there's no "probabilistic" to it. Could you give us some links to actual applications of probabilistic combinatorics by ID advocates?Tom English_{September 17, 2006
September
09
Sep
17
17
2006
09:48 AM
9
09
48
AM
PDT}

But any self evidence is almost never a primary thought; it is the result of a reasoned judgement on which among several explications for a given event is really correct. Take a similar example of design recognition: a person sees an alien articraft and is able to recognize it as a non-natural but designed thing, and this without any error (you would see "self evident"). But how could have he produced this judgement without fast and unconscious rejecting the natural option and then accepting the design one?kairos_{September 17, 2006
September
09
Sep
17
17
2006
07:26 AM
7
07
26
AM
PDT}

kairos I never looked at it that way I guess. You aren't the first to be thorougly confused by my statements. Life is like that. "A past evolution is undeniable, a present evolution undemonstrable." John A. DavisonJohn A. Davison_{September 17, 2006
September
09
Sep
17
17
2006
07:09 AM
7
07
09
AM
PDT}

#7,11 "Design is not an inference and should never have been presented as such. "I have always regarded Intelligent Design as self-evident which is all that I meant or intended". John, I'm a bit confused by your statements. If one finds that ID is self evident this is just because he has actually done an inference to the best explanation of the biological world.kairos_{September 17, 2006
September
09
Sep
17
17
2006
06:47 AM
6
06
47
AM
PDT}

Carlos I have always regarded Intelligent Design as self-evident which is all that I meant or intended. Each of us has his own interpretation of the world he peruses. That is hardly a "bone of contention." I believe it is known as freedom of speech, and is covered by the United States Constitution. Correct me if I am wrong. "A past evolution is undeniable, a present evolution undemonstrable." John A. DavisonJohn A. Davison_{September 17, 2006
September
09
Sep
17
17
2006
04:18 AM
4
04
18
AM
PDT}

Design is not an inference and should never have been presented as such. I'll leave this matter between you and Dembski. If I understand your respective views correctly, this is a bone of contention.Carlos_{September 16, 2006
September
09
Sep
16
16
2006
09:45 PM
9
09
45
PM
PDT}

ID exam follow-up question: Had I written, "A microbe did not mysteriously mutate into Mozart and his music, and most people, mercifully, are much too smart to swallow this a silly idea," would you have detected alliterative design? How many alliterative M's would it take to make design an obvious, slam-dunk conclusion? Explain.GilDodgen_{September 16, 2006
September
09
Sep
16
16
2006
08:44 PM
8
08
44
PM
PDT}

ID exam question: Is the "m" alliteration in the last sentence of my post by design or by chance? Did you detect it? Explain.GilDodgen_{September 16, 2006
September
09
Sep
16
16
2006
07:05 PM
7
07
05
PM
PDT}

Design is not an inference and should never have been presented as such. It is a mandatory and self-evident feature of every living thing, its development and its evolution. A role for chance has been completetly eliminated from consideration as being involved in either ontogeny or phylogeny. If not chance then what? I say the Prescribed Evolutionary Hypothesis is the only conceivable alternative to the most failed hypothesis in the history of science. I regard it as firmly established and will until it has been discredited by controlled laboratory experiment. So far everything we are learning pleads for the great antiquity of biological information, much of it until recently thought to be of recent origin. The source of that information has always been the chromosome which is the seat of all meaningful information. How much of that original information still remains is problematical but that it once existed cannot be denied except by those that worship the Great God Chance. Variations in the particulate gene had nothing to do with either ontogeny or phlogeny which is why all living thimgs are so remarkably similar at that genetic level. Evolution, a process no longer in progress, was largely, if not entirely, a "position effect" phenomenon utilizing a common mechanism for which a controlled modification of the intrachromosmal environment served to alter the rate and timing of key developmental events. All that allelic (Mendelian) mutations ever did was introduce pathological modifications of an otherwise predetermined sequence. "The theory of the genes and of the accumulation of micromutants by selection has to be ruled out of this picture." Richard B. Goldschmidt, The Material Basis of Evolution, page 396 "A past evolution is undeniable, a present evolution undemonstrable." John A. DavisonJohn A. Davison_{September 16, 2006
September
09
Sep
16
16
2006
01:12 PM
1
01
12
PM
PDT}

Here are some interesting searches on Probalistic Combinatorics. Look forward to combinatorial specialists adding serious reviews.DLH_{September 16, 2006
September
09
Sep
16
16
2006
12:02 PM
12
12
02
PM
PDT}

Re: (4) Strict empiricism may assist the criticism of NDE, but it won't assist the promotion of any alternative. On the contrary, strict empiricism turns science into stamp collecting. ID, after all, is not strictly empirical. One does not observe design; one infers it. (That's why one of Dembski's books is The Design Inference, right?) On the other hand, if the demand for empiricism is relaxed, then one can consider different possible inferences, evaluate the probability of each one, the relationship between different lines of evidence, and the explanatory potential of different theories, etc.Carlos_{September 16, 2006
September
09
Sep
16
16
2006
11:24 AM
11
11
24
AM
PDT}

Natural selection is the cornerstone of Darwinism. It has been since its inception. It was a logical argument devoid of empirical support at the outset and still is. Advocates of NS point to sequence homologies as evidence of common descent and then argue that NS would preserve advantagous genetic changes. But what observations actually document multiple mutations leading to novel complexity? The out is that such occurences take millions of years so it is not realistic to expect to see such changes in a lab envirnment even for rapidly reproducing unicellular organisms. It it a good means of isolating the theory from testing. What we are left with is arguments based on homologous genes.pk4_paul_{September 16, 2006
September
09
Sep
16
16
2006
06:54 AM
6
06
54
AM
PDT}

Prev 1 2 3 Next

You must be logged in to post a comment.

Leave a Reply