DNA as Digital Technology

_{Barry Arrington

November 19, 2011

Intelligent Design}

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

Which Bible thumping ID nutbag wrote the following:

There is a sense, therefore, in which the three-dimensional coiled shape of a protein is determined by the one-dimensional sequence of code symbols in the DNA…. The whole translation, from strictly sequential DNA ROM [read-only memory] to precisely invariant three-dimensional protein shape, is a remarkable feat of digital information technology.

Comments

The intelligence involved in GA is simply a matter of trying to copy what occurs in nature and adapt the process for commercial use. You are asserting something equivalent to saying that because wine and beer and bread making systems are intelligently designed, fermentation is not a natural process. That's just silly. Or because roads are intelligently designed, tar and concrete cannot occur naturally. Evolution is the inspiration for GAs, but nature is still better at it. Microbes are still developing drug resistance faster than we can engineer drugs.Petrushka_{November 22, 2011
November
11
Nov
22
22
2011
09:43 AM
9
09
43
AM
PDT}

Petrushka, You are using something known to be intelligently designed as evidence of what can be accomplished without intelligent design. The very best examples of unintelligent design you can find required teams of programmers and fifty years of accumulated computer science plus the design and manufacture of the computers themselves. You're case would be a tiny bit more compelling if you found an example with just a little less intelligent design in it.ScottAndrews2_{November 22, 2011
November
11
Nov
22
22
2011
09:26 AM
9
09
26
AM
PDT}

Your fallacy is that you repeatedly limit the capacity of intelligence to the capacity of ‘computational resources’ when it is observed and known beyond a shadow of a doubt that they are vastly superior.
I would be glad to see the computational metaphor dead and buried. The 500 bits of dFSCI has never been anything but a canard. But your claim of the superiority of "intelligence" is silly. Any process that requires systematic application of physical principles has been handed over to computers. Engineering programs routinely handle material strength and balance utility and cost. Programs have long ago surpassed humans in the ability to balance multiple design dimensions. Shipping companies routinely use genetic algorithms to plan routes. the power grid is balance using genetic algorithms. There are scores of industries solving multi-dimensioned problems with genetic algorithms. The simple empirical question to be answered is not how many bits are in current genomes, but how they got there and whether they got there incrementally. More interesting is how they keep changing, even in the brief span of human history.Petrushka_{November 22, 2011
November
11
Nov
22
22
2011
07:59 AM
7
07
59
AM
PDT}

Petrushka,
My argument can be boiled down to the assertion that utility of coding strings cannot be computed with ordinary finite resources.
Name one worthwhile designed thing that can be computed with ordinary finite resources. Not a sentence I type could ever be "computed," because no unintelligent resource could ever possess a thought or the desire to express it with language. When you take design problems and recast them as computational problems, of course they become impossible. That's why we use intelligence, set goals, form intentions, and plan rather than attempting to compute everything. Your fallacy is that you repeatedly limit the capacity of intelligence to the capacity of 'computational resources' when it is observed and known beyond a shadow of a doubt that they are vastly superior. Even in cases where computational resources are effective, it is only by the design of and at the service of intelligence.ScottAndrews2_{November 22, 2011
November
11
Nov
22
22
2011
06:00 AM
6
06
00
AM
PDT}

But you should know that most of their properties can very well be computed. Folding can be computed. It’s just, as you yourself say, “due to limitations of computational speed”.
My argument can be boiled down to the assertion that utility of coding strings cannot be computed with ordinary finite resources. It's an empirical claim. It's a claim that could be addressed by ID research. It seems like ID proponents would want to confront head-on the problem of design rather than pretend it doesn't exist. In the real world, the problem of functionality and utility are addressed through directed evolution. Like most industrial processes, a natural process has been stripped down to essentials and put to work. In one sense, directed evolution is the oldest industrial process, since plant and animal breeding are among the oldest of human technologies.Petrushka_{November 22, 2011
November
11
Nov
22
22
2011
02:58 AM
2
02
58
AM
PDT}

Petrushka: Molecules have properties that cannot be computed from the properties of the constituent atoms. Maybe. Sometimes. I really don't want to be a reductionist. But you should know that most of their properties can very well be computed. Folding can be computed. It's just, as you yourself say, "due to limitations of computational speed". I am really tired of hearing the concept of "emergent properties" used to defend the view thta anything can happen. Emergent properties have become a new form of belief in magics. The properties of molecules are due to specific laws og chemistry. Those laws can be understood. It's not that we add hydrogen and oxygen and water comes out as an emergent property. There are specific reasons of physics and chemistry that explain why those two elements combine, and why the resulting molecule has the properties it has. All that can be understood and explained. And, as you know, if we have a good scientific explanation, we can usually make good predictions. Certainly, we don't understand everything. I will easily support that for any scientic knowledge. But that does not mean that things cannot be understood. Design is based on understanding. You are suggesting that the chemical properties of molecules cannot be understood. I don't know from which strange ideology you derive that concept, but it's simply not true.gpuccio_{November 22, 2011
November
11
Nov
22
22
2011
12:34 AM
12
12
34
AM
PDT}

Petrushka, Intelligence doesn't "compute." That's what makes it so intelligent, and that's why it solves problems unattainable through searches. A search finds something. But truly functional solutions are never found. Intelligence processes the end goal and imagines components working in order, examining the big picture and all the small ones simultaneously. Intelligence can go from zero to 90 in a moment, from a blank slate to a light bulb over your head. Without begging the question of biological design, I have everything you've ever used or imagined or heard of in your whole life on my side. You've got some GA modifying an antenna after someone imagined and designed a radio and conceived of the need for an antenna and then imagined and designed a computer and then imagined and implemented a programming language and used to to implement a GA to look for a better antenna, entered the inputs, and tuned it to get the correct output and went on to design and implement a means for manufacturing the antenna. Everything is on my side. Your examples are on my side. I've got reality, all of it, and you've got abstract concepts that fill a thousand papers but never connect the dots from selection and variation and drift to any concrete examples or implementations more exciting than cichlid fish in different colors. All your bases are belong to us.ScottAndrews2_{November 21, 2011
November
11
Nov
21
21
2011
07:29 PM
7
07
29
PM
PDT}

When I was in high school I read that it was impossible for computer graphics to render a real-time landscape of three-dimensional shapes...
That was never true. It may have been impossible due to limitations of computational speed, but there's a difference between problems that are hard because they require a lot of computation, and problems that are hard because emergent properties cannot be computed at all. at least with the resources available in the physical universe. Molecules have properties that cannot be computed from the properties of the constituent atoms. It's amusing to watch ID advocates argue for reductionism.Petrushka_{November 21, 2011
November
11
Nov
21
21
2011
06:29 PM
6
06
29
PM
PDT}

If no one can even string a few of those together, then it’s not even hypothetical. It’s conceptual. Like when I imagine the flying car of the future without knowing how to build one.
A few have been strung together.Petrushka_{November 21, 2011
November
11
Nov
21
21
2011
06:01 PM
6
06
01
PM
PDT}

Petrushka:
the problem is that there is no shortcut to knowing what the effect of a coding string will be.
Well when looked at from a totally physio-chemical point of view, you would just have to consider the charges + / - and then the type of amino acids- hydrophobic or hydrophyllic- to see what shape would be formed. However if a chaperone is involved then the shape is pretty much already determined.Joseph_{November 21, 2011
November
11
Nov
21
21
2011
04:59 PM
4
04
59
PM
PDT}

"Why do you suppose problems like the traveling salesman problem are best solved with genetic algorithms?"
The TSP is not "solved" by genetic algorithms, in the sense that optimality is achieved; however a good (suboptimal) tour can be improved by them at the cost of computing time. So "best" here is subjective to purpose. Even GAs incorporate specific, designed target searches. A GA in this regard is an example of using a highly directed search, where the optimal solution cannot reasonably be known (in the case of TSP, because it takes n! iterations to discover an exact solution). Populations are not generated at random, but rather with fast heuristics known to ballpark optimization problems. For example, minimum spanning trees can be used to generate solutions, often within 20% of optimal, in as little as O(n logn) time. This means you could generate a tour of 512 cities in as little as 512*9=4608 computations. That is an estimated minimum -- some variations run in O(n^2) which would result in around 512^2=262144 computations. This is entirely non-random. A GA might start with a population of tours which have already been optimized by highly efficient heuristics, then combine or selectively mutate tour edges and test for improvements over the parent optimum. Iterative, selective, specific processes for generating variation might also be employed to improve effectiveness. It's likely that a properly designed GA, left to run for hours or days, might find a solution very close to the exact solution, depending on the number of points being considered. However it will undoubtedly rely on existing heuristic methods in order to be successful, by culling out bad solutions and starting with good ones. It is a tightly directed search, where populations begin near the target solution; and so the mechanism is not blind with respect to outcome, but rather has a target in mind -- and begins its search in the neighborhood of the target. For a sufficiently large set of edges, a successful GA will likely need to begin with reasonably "good" solutions, then perform selective optimizations (as mutations), all while steering toward a target goal of shortest route. With regard to fitness, initial population, and mutation or recombination, it is all very specific. The less specificity a GA has with respect to producing populations and variations, the less successful it will be in approaching an optimal solution. In any case, it must know what neighborhood it's target resides in. For most applications that would benefit from a good solution to the TSP, especially where efficiency is an issue, best to avoid GAs, IMO. To sum up, an effective GA for solving the TSP is not an alternative to other highly specific methods, but rather incorporates at least some of them.material.infantacy_{November 21, 2011
November
11
Nov
21
21
2011
04:51 PM
4
04
51
PM
PDT}

P: have you ever written a computer program?
http://itatsi.com I think you are agreeing with me that the resources of the universe are insufficient to store the possible coding strings and their attributes and levels of utility. And emergence means that there is no formulaic way of knowing or predicting the utility of a coding sequence.Petrushka_{November 21, 2011
November
11
Nov
21
21
2011
04:18 PM
4
04
18
PM
PDT}

P: have you ever written a computer program? This is linguistic and it is prescriptive information predicated on changing switches. FYI, the genetic codon table does allow prediction of results of changes in the bases. Indeed, there is a discussion of this just now on how experiments with this show an optimality in the code. The implication is you have to understand the language and the related controlled dynamics under the control of the language [the usual game for a control situation], i.e. a major challenge to intelligence. It is quite obvious that the complexity involved is vastly beyond the atomic resources of the cosmos, to arrive at the required proteins for first cell based metabolising life, which is again the same point. KFkairosfocus_{November 21, 2011
November
11
Nov
21
21
2011
02:55 PM
2
02
55
PM
PDT}

Petrushka, And you maintain, without a trace of evidence, that evolution can solve a problem that so far no one else has (while oddly asserting that because no one has, no on can.) The genetic algorithms that solve the traveling salesman problems are designed up and down, back and forth. They are one more of the countless examples that support my argument, not yours.
There are at least three scenarios that would make design possible. (Emphasis mine.)
You left out the one where someone else is just smarter than us and knows how to do something we can't, sort of like us compared to 100-years-ago us or 100-years-from-now us compared to us. To assert that an unsolved problem is unsolvable flies in the face of history and science. It's like an argument from ignorance with credit for future and eternal ignorance. Going "poof!" sounds more intellectual when you call it an 'emergent property,' but it's no less magical when you just tack it on as a post-hoc explanation without explaining a single detail of said emergence or the existence of the framework required for it to emerge. Poof! It emerged!ScottAndrews2_{November 21, 2011
November
11
Nov
21
21
2011
02:29 PM
2
02
29
PM
PDT}

Too tired is a rather lame excuse. The fact is you have no response. If you had a way of anticipating utility you would win a Nobel prize. And be wealthy.Petrushka_{November 21, 2011
November
11
Nov
21
21
2011
02:13 PM
2
02
13
PM
PDT}

But now we have the “it’s too complicated to be designed
Let's be clear. What makes design difficult or impossible is not simply "complication." the problem is that there is no shortcut to knowing what the effect of a coding string will be. You can imagine the is a shortcut, but lots of bright people have looked. And suppose you found a formula for coding proteins. You would still have the problem of sieving for utility. Simply having a protein doesn't make it useful. And having a useful protein doesn't tell you what its relative utility will be in an ecosystem. Why do you suppose it's so difficult to invent safe and effective drugs?Petrushka_{November 21, 2011
November
11
Nov
21
21
2011
02:05 PM
2
02
05
PM
PDT}

Petrushka: More probably, we are just tired to answer your repetitive non arguments.gpuccio_{November 21, 2011
November
11
Nov
21
21
2011
02:03 PM
2
02
03
PM
PDT}

Too complicated to have been designed without using evolution. Why do you suppose problems like the traveling salesman problem are best solved with genetic algorithms? There are at least three scenarios that would make design possible. One would be some way to anticipate emergent properties. That's not visible on the horizon. I do note the wishful thinking of some ID advocates. Too bad their own most capable supporters -- people like Douglas Axe -- don't buy it. Another scenario is that functional space is sufficiently connected to allow traversal by small changes. Another scenario is that the designer is omniscient and poofs things into existence.Petrushka_{November 21, 2011
November
11
Nov
21
21
2011
01:42 PM
1
01
42
PM
PDT}

I notice that no one has jumped in with an example of anticipating the utility of an arbitrary coding string. It would seem that not only is the ID movement without a designer, it is also lacking a conceptual framework that would make design possible. Is that the best you can do? Poof? The check's in the mail?Petrushka_{November 21, 2011
November
11
Nov
21
21
2011
01:31 PM
1
01
31
PM
PDT}

Petrushka,
utility is what cannot be anticipated.
I'm flabbergasted. For years folks have been criticized for the "it's too complicated to happen by itself, it must have been designed" argument. We've even been brainwashed into thinking it's a bad argument, when it's actually quite logical and rational. But now we have the "it's too complicated to be designed, it must have happened by itself" argument. Unlike the above statement, which is derived from the totality of human experience, this is derived circularly from itself. Its only evidence is that assumption that life was not designed, which in turn is defended with this assertion. It's too complicated to have been designed, it must have happened by itself. War is peace. Freedom is slavery.ScottAndrews2_{November 21, 2011
November
11
Nov
21
21
2011
01:04 PM
1
01
04
PM
PDT}

I don't have it backwards at all. The "information" determines the production of proteins, but chemistry determines the utility of a given protein, and utility is what cannot be anticipated.Petrushka_{November 21, 2011
November
11
Nov
21
21
2011
11:39 AM
11
11
39
AM
PDT}

Petrushka: "Nor can one derive the properties and phenomena of chemistry from the abstraction commonly referred to as information theory." What?! This doesn't even make sense in the context of DNA. You've got it exactly backwards. What you should have said is that the chemical properties of the molecules in DNA do not produce the information life is based on. Rather, it is the information that is imposed on the medium that results in the particular chemical configuration.Eric Anderson_{November 21, 2011
November
11
Nov
21
21
2011
10:57 AM
10
10
57
AM
PDT}

You are an example to people like me who say too little with too many words.ScottAndrews2_{November 21, 2011
November
11
Nov
21
21
2011
08:19 AM
8
08
19
AM
PDT}

Joseph, Yes, I suspected that ) Petrushka, I remember you asking if I meant something like vitalism when I said that life was not reducible to chemistry/physics. I will answer by quoting. Here is what Chaitin says about irreducibility (in his "The limits of reason", SciAm, 2006). "In a way, saying something is irreducible is giving up, saying that it cannot ever be proved." There are definite signs of intelligence out there. Further than that we can't come (due to limitations of science which Chaitin talks about in that article about the Omega, the probability of halting of a random program). We can't rigorously prove that with respect to life (although we can well prove that e.g. in forensics or medicine). However what we can say is more than nothing. We can say design explains life better than chance and/or necessity on their own. And that is scientific enough to me.Eugene S_{November 21, 2011
November
11
Nov
21
21
2011
07:35 AM
7
07
35
AM
PDT}

Petrushka's entire diatribe simply assumes the existence of recorded information. He hasn't even gotten to first base. Hell will freeze over first.Upright BiPed_{November 21, 2011
November
11
Nov
21
21
2011
07:08 AM
7
07
08
AM
PDT}

Petrushka:
My point remains that design requires anticipation of utility, and the “language” of DNA does not allow anticipation of utility.
Maybe not to us, right now. But that does not mean the designer(s) of living organisms could not anticipate the utility of DNA.Joseph_{November 21, 2011
November
11
Nov
21
21
2011
06:57 AM
6
06
57
AM
PDT}

Petrushka, When I was in high school I read that it was impossible for computer graphics to render a real-time landscape of three-dimensional shapes composed of polygons in such a way that the areas facing the viewer were visible while the back sides were hidden. (They were usually rendered as wireframes with all the lines visible.) By 1995 every frame of every video game was full of thousands of them with varying textures applied to different facets. Since then it's replaced hand-drawn animation in cartoons. To rule out design because it's just too hard doesn't fly at all. Is there even a name for that fallacy? Argument from we-don't-know-how-to-do-that-yet? Argument from if-Douglas-Axe-says-it-can't-be-done-so-be-it? But the mind-blower is to take an apparent product of forethought and planning, declare it too difficult for intelligence, and attribute it to some vague concept of emergence from random variations instead.ScottAndrews2_{November 21, 2011
November
11
Nov
21
21
2011
06:48 AM
6
06
48
AM
PDT}

My point remains that design requires anticipation of utility, and the "language" of DNA does not allow anticipation of utility. If you don't believe me, ask Douglas Axe.Petrushka_{November 21, 2011
November
11
Nov
21
21
2011
06:38 AM
6
06
38
AM
PDT}

Eugene, I believe Petrushka was just playing. It's like when someone asks if you heard them and even though you did you say "what?"- and they repeat what they said.Joseph_{November 21, 2011
November
11
Nov
21
21
2011
06:31 AM
6
06
31
AM
PDT}

Petrushka, I guess what Joseph means is all new information must be explainable in the existing language to maintain communication. That is perfectly true. And in fact it is what routinely happens e.g. when a child is learning. Sensible parents/teachers explain things so that children can understand. So first meaning -> then a common alphabet and a common language -> then information transfer & processing. As was pointed out a number of times on this blog, the remarkable thing about information transfer and processing in a living cell is that between the sender and the receiver there may be no a priori physical/chemical contact. IMHO, this fact alone is telling of life's irreducibility to chemistry/physics.Eugene S_{November 21, 2011
November
11
Nov
21
21
2011
06:15 AM
6
06
15
AM
PDT}

Prev 1 2 3 4 5 Next

You must be logged in to post a comment.

Leave a Reply