My Final Post at UD
| April 25, 2009 | Posted by GilDodgen under Intelligent Design |
Last evening I posted the following, and within a short period of time the Darwinbots descended upon it, challenging my expertise in two highly sophisticated areas of computational science, AI and FEA, fields in which I have the goods to demonstrate that I know what I am talking about. One commenter even asserted that the physics involved in an LS-DYNA simulation cannot be represented with mathematical precision. Yes they can. And it works.
At this point I decided that I have nothing further to offer. If some people cannot recognize that the information-processing systems encoded in biological systems defy naturalistic explanation and suggest a design inference, I cannot help them, and they are free to continue to pursue a phantom.
Farewell, and best wishes to all.
Gil
A number of years ago I developed an interest in AI (artificial intelligence) games-playing programming, and pursued a research project in that arena with so much success that I eventually lost interest, because there were no remaining human opponents to challenge. You can read about the project at my website. Real-world experience demonstrated the success of the project.
I now earn my living as a software engineer in aerospace R&D with a specialty in computer simulations, and as a result have pursued another interest: transient, dynamic, nonlinear, finite-element analysis (FEA) using a simulation program called LS-DYNA, originally developed in the 1970s at Lawrence Livermore National Laboratory to simulate underground nuclear tests.
My company sent me away to LS-DYNA school after I volunteered, but I was warned that it would be really, really difficult, and that I had better bone up on the relevant math and engineering concepts. I took this advice to heart, and spent at least 200 hours preparing for the five-day course. Even with vast experience in software engineering and this preparation, it took everything I had to keep up with the instruction. The LS-DYNA course was a huge eye-opener concerning computer simulations and reality.
On the first day of the course our instructor, Dr. John D. Reid, who was absolutely fantastic, commented that it is really easy to make “cartoons” with LS-DYNA. (Dyna not only produces vast quantities of data, but generates AVI animations of the simulation.) By that he meant that without a thorough understanding, it is easy to make a Dyna simulation produce whatever results you like, that might look cool, but have no correlation with reality.
LS-DYNA has been under development for more than 30 years by the most brilliant scientists in the field, and its simulations have been compared repeatedly against real-world results. Material physical properties are well known, tested, and quantified (Young’s modulus, Poisson’s ratio, mass density, area moment of inertia, etc.), and the physics involved can be simulated and represented with absolute mathematical precision.
Yet with all of this, the results of a simulation must be scrutinized and evaluated against reality, because a single erroneous assumption or programming error can render the simulation completely invalid.
Which brings me to the point of this essay: The notion that any computer simulation of biological evolution has anything to do with reality is a complete fantasy. And the notion that any computer simulation of the earth’s climate into the distant future can be relied upon is an equivalent fantasy.
These computer simulations are cartoons.
77 Responses to My Final Post at UD
Leave a Reply
You must be logged in to post a comment.
I juxtapose names, and leave it for the reader with half a wit to wonder how much Tom Cech sounds like Gil Dodgen.
and leave it for the reader with half a wit to wonder how much Tom Cech sounds like Gil Dodgen.
Assuming he accepts the appeal to your authority as to how Tom Cech sounds.
Hamlet is no longer with us. After repeated warnings to stop beating Gil personally, he just could not seem to help himself.
There is a fine line between attacking people and attacking their arguments. This is particularly the case if part of their argument consists of arguments from authority, and the authority cited is themselves. Under such conditions, it becomes difficult to attack such an argument without seeming to attack the person.
Consequently, I have attempted to frame my own arguments in such a way as to cite only published evidence, and to omit references to my own expertise in a field. I am the first to admit that this is not always possible, and even that I have been tempted (and sometimes have given in to the temptation) to cite my own knowledge or training in a particular field when discussing some bit of evidence from that field.
Ergo, mea culpa, and I will attempt to stick to arguments of substance, rather than resort to arguments from authority (including my own) henceforth, and would hope that the rest of us will do the same.
As for this being Mr. Dodgen’s last post, I do find that unfortunate. I am not conversant in the field of computer simulation, and therefore have found the debates about the various merits and limitations of this technique to be generally enlightening, especially when they have focused on substance, rather than personalities or questions of authority.
That said, I must also state that I am generally skeptical of any claims for the usefulness of mathematical or computer simulations on either side of the EB/ID divide. I have written elsewhere of my own belief that the reduction of evolutionary biology to mathematical models (i.e. the epistemological core of the “modern evolutionary synthesis”), while it served a useful purpose at the time, has now outlived that usefulness and may in some cases now represent a stumbling block to further progress in evolutionary theory.
All of that is beside the point when considering Mr. Dodgen’s withdrawal from the fray here at UD. I have often strenuously disagreed with him, especially when he has posted unsupported condemnations of evolutionary theory, but due to my lack of expertise in the field of computer simulation, I have generally withheld commenting on his posts.
That said, I do believe that having him contribute here has merit, and hope that he changes his mind and decides to re-enter the lists, fully armed and ready for combat.
To paraphrase a well-known quote,
This is why I take out a significant portion of my days to comment here and to post on my blog, and hope that the rest agree that knowledge is always a good thing, no matter how painfully attained.
Mr MacNeil,
I think most people who have worked with simulations are painfully aware of their limitations. I would take a comment that simulation has outlived its usefulness in evolutionary theory much more seriously from someone who claimed some familiarity with them.
However, I am intrigued – what are the areas in which you feel evolutionary theory needs to progress, and why is mathematical and computer simulation not likely to help in these areas?
Forgive me if I seemed to imply that mathematical models or simulations of evolution are not likely to help us understand how evolution works. What I meant to say is that they are of limited utility in understanding how particular cases of descent with modification have occurred. Furthermore, I find that there is a tendency among some theoreticians (i.e. folks who do mathematical modeling and simulations, but do not actually test them against reality in the nature) to assume that, if something observed in the field does not fit the model, it is the observation that is the problem, rather than the model.
IOW, I always remember that there is a world of difference between the moon and the finger pointing at the moon…
If I may, could I add my voice to those calling for a separate “meta” thread on the subject of moderation, and also suggest (based on our experience with the thread on “Bleak Conclusions” that a “meta” thread on the subject of “natural law” might also be quite interesting and productive? I believe that evolutionary biologists (and indeed, most other empirical scientists) have a very different idea of what “natural laws” might be, and it would be instructive for all of us to see what such differences might be. These threads could live in the sidebar with the Moderation Policy and the “What Arguments Not To Use” policy. It would certainly be helpful to people just joining the conversation, and might help to reduce the level of rancor here.
Just a suggestion…
It’s unfortunate that Gil’s competence was challenged, but those things happen, even on this board. For instance, recently a well-published computer scientist/mathematician, whose work has been cited in support of ID, was told, “trying to explain computational number theory to you would be as futile as trying to explain calculus to someone who can’t add fractions.” We all need to work on being more respectful.
Gil, Please don’t let a handful of bad mannered Darwin-bots (to quote Denyse) run you off. Barrett in post #28 is right. There are many visitors here who never post and your expertise adds much to the discussion.
I’m formally asking you to re-consider!
Donald M
Allen suggests “that a “meta” thread on the subject of “natural law” might also be quite interesting and productive?”
I agree. There does seem to be a langauge gap between those who appeal to that term in a design vs. non-design context. Is it possible that confused definitions are responsible for some of the less-than-civil (sometimes coming from my corner) interchanges and accusations of “stonewalling?” I, for one, would like to find out.
Allen (68),
There are big differences between analytic modeling (think in terms of classical population genetics) and simulation modeling. As you are well aware, theoretical biologists often introduce rather extreme simplifying assumptions in order to make mathematical analysis tractable. It is commonly possible to simulate conditions much more realistic than those that permit analysis.
Simulation results do not prove anything, in the ordinary sense of proof, but they sometimes establish that analytic results are incorrect under realistic circumstances. For instance, so-called evolutionarily stable strategies are sometimes not stable under evolutionary dynamics (Fogel GB, Andrews PC, and Fogel DB, 1998, “On the Instability of Evolutionary Stable Strategies in Small Populations,” Ecological Modelling, Vol. 109, pp. 283-294):
The assumptions I’ve emphasized often underlie theorems in genetics beloved by “Mendel good, Darwin bad” IDers. Replacing a random variable with it’s expected value is often mathematically convenient, but is usually dubious. And reproductive isolation of small populations has been a key component of theories of speciation since the 1950′s, IIRC.
The question marks in my last comment are ligated “fi” in the text I copied and pasted. The messed up words are crucial ones: finite and infinite. They looked fine in the preview.
sal gal:
You are quite correct, and if I erred in overstating the similarities between mathematical modeling and simulations (as it appears to me on reading your comment), the error was entirely my own.
Allen,
That was the point of my essay, and you have summarized it far more eloquently and succinctly than I did. In any event, I’ve decided to move on and invest my time in stuff that really matters, like family and ministry that edifies individual human lives.
Gil,
Clearly there are good reasons for spending your time elsewhere, and there are good reasons for staying. For my part, I have been edified, encouraged, and strengthened by your posts. Your presence would be missed. If you leave here shaking the dust off your feet, you would be following good precedents. I definitely agree that family is more important than writing to a web audience, and other more personal ministries may be much more in line with where the Captain is directing you to spend your effort in the ongoing race.
Whatever you decide, thanks for your past contributions and for investing your gifts in this area.
Lars
Dear lars,
Thanks much. My calling is to follow the Captain to the best of my ability. Life is ephemeral and I have only one wish when it is over, to hear the words:
“Well done, good and faithful servant.”