Home » Comp. Sci. / Eng., Evolution, Religion » Bogus Computer Simulations

Bogus Computer Simulations

This one, published by New Scientist, really takes the cake. From the article:

God may work in mysterious ways, but a simple computer program may explain how religion evolved.

By distilling religious belief into a genetic predisposition to pass along unverifiable information, the program predicts that religion will flourish… The model assumes… that a small number of people have a genetic predisposition to communicate unverifiable information to others. They passed on that trait to their children…

The model looks at the reproductive success of the two sorts of people — those who pass on real information, and those who pass on unreal information.

It would be a colossal understatement to call this utter silliness, and it stuns me that anyone would take this seriously, much less allow it to be published as a “scientific” study by “The World’s No.1 Science and Technology News Service.”

I design real-world computer simulations in my work with a finite-element analysis (FEA) program called LS-DYNA, which is the world’s most powerful and thoroughly used and tested program of its kind. It originated at Lawrence Livermore National Laboratory, and has been designed and refined over decades by some of the greatest minds in the field.

LSD (as I affectionately call it) models the laws of physics and Newtonian mechanics with stunning fidelity, and material properties are well understood, tested and documented, and are modeled accurately as well. LS-DYNA is capable of analyzing and simulating extremely complex systems involving all kinds of non-linear dynamics. This program is so powerful that it is used heavily in the automotive industry to simulate car impacts, airbag deployments, occupant injuries, etc.

Yet, even with all this, and a programmer who is experienced and knows what he’s doing, the simulations must always be tested against reality to finally validate them. (This way you only have to crash one or two real cars instead of 50 or 60 to get things right.)

So, when I hear about a computer simulation that demonstrates how religion evolved (or how any living system evolved, for that matter), all I can do is roll my eyes in wonderment and disbelief that anyone takes this stuff seriously.

The dissertation above should also cast doubt on computer models that attempt to simulate climate change. There are simply too many inadequately understood non-linear interacting dynamics, including positive and negative feedback mechanisms, to have any confidence that these simulations reflect reality. And, they can’t be tested against the real world, except by waiting and seeing what happens decades later, at which time the faulty simulations will have been long forgotten.

As a final observation, I can make an LS-DYNA simulation do just about anything I want, by arbitrarily tweaking parameters and material properties. I have done this on occasion just for the entertainment value. Those who attempt to model biological evolution, religion, or climate change can do the same.

Just for fun, check out this LS-DYNA simulation of a car airbag deploying (a 2 MB AVI file). It’s pretty amazing.

If you watch the clock in the upper left-hand corner you’ll notice that the actual deployment takes only seven hundredths of a second. (The AVI runs in slow motion so you can see what’s going on.) Note how DYNA accurately simulates the dynamic behavior of the inflating gas and all the details of the unfolding and inflating fabric. The squares on the airbag are the finite elements.

Visit this page for more DYNA simulation AVIs. These are simple examples of many of the program’s simulation capabilities, including impact analysis, fluid dynamics, fluid-structure interaction (FSI) using Arbitrary Lagrangian Eulerian (ALE) methodology, heat transfer analysis, metal forming, explosives simulation, and more. In addition to producing the animations, DYNA generates mountains of data in many forms that can be studied and analyzed with other sophisticated software tools.

This is what real computer simulation is all about, and it actually works, if you know what you’re doing. It is because of the success in the real world of software programs like LS-DYNA that “computer simulation” has become a magic phrase, often used in an attempt to add credibility to a claim. Just wave the magic computer-simulation wand, and all doubts are expected to disappear.

But don’t be fooled. Bogus computer simulations abound, like snake oil in The Old West.

  • Delicious
  • Facebook
  • Reddit
  • StumbleUpon
  • Twitter
  • RSS Feed

22 Responses to Bogus Computer Simulations

  1. Starting with the assumption that ‘telling people unverifiable things’ is genetic? Goodness. So, what kind of genes did Socrates have? How about Plato?

    Though it’s amusing that the article (unwittingly?) implies that atheism was and may well still be a show-stopper for evolutionary development. Wouldn’t that make for an interesting article? ‘Computer simulation shows that atheism leads to lowered fitness and eventually extinction of populations.’

  2. It’s amazing to me that these so-called reputable science rags think that it is their job to try and convince people that religion is something without any basis in reality. The real purpose of that article and the entire evolution movement (yes, it has become a movement now in reponse to ID) is made crystal clear. They don’t seem shy in making it known that their “science” agenda is really a sham, it’s a bigoted exercise in religion bashing in a cynical attempt at preaching atheism.

    How long before it becomes obvious to lawmakers that how evolution is taught makes it nothing more then a religious philosophy in it’s obvious attempt to preach atheism?

  3. I’m starting to get annoyed with the idea that all religious tenets are unverifiable. There are people and places identified in the Bible that have been verified by secular historians and archaeologists.

    The only people who would take this seriously are of the ‘village atheist’ mentality, who seem genuinely offended at the idea that some people might have something (namely, faith) that they do not.

  4. The model assumes… that a small number of people have a genetic predisposition to communicate unverifiable information to others.

    And they all post at Panda’s Thumb.

  5. 5

    Actually, this seems to be a variant of a statement I have seen on Free Republic that religious couples have more children on average than liberal couples, so it can be reasonably concluded that liberalism is a dying ideology. That idea is embraced by conservatives. Indeed, even here.

    Liberals are literally a dying breed. Conservatives on average have more children than liberals. San Francisco in particular has the lowest percentage of children of any large city in the nation. It’s no conincidence that it also has the highest gay population in the nation. As we used to say when I was young(er) – DUH!

    So you see, it’s all very Darwinian. Survival of the fittest and all that. Conservatives are being naturally selected over liberals.

  6. I can’t believe(well, I can – f.o.s.) that New Scientist is so uncredible and has stronger ties to religious philosophy than true science.

    How long has it been like this?

    As an avid reader, it was only around the time Expelled was released (and the same time I discovered this gem of a website), that I discovered my NewScientist magazine for that week had a glossy title page: Evolution, how it can never be disproved, etc etc. A story on the same day/week as the Expelled documentary? Anyways, my “eyes were opened”, i guess.

    But, can anyone tell me how long New Scientist had been like this?

    I mean, up until April, or whenever Expelled was released, I had ‘believed’ that New Scientist was completely trustworthy in reporting the latest in breaking technology and research. Mistakes could happen of course, but they would be honest ones that were later corrected when further discoveries brought enlightenment.

    What am I to read now??? :(

  7. 7

    The model assumes… that a small number of people have a genetic predisposition to communicate unverifiable information to others.

    You mean like UCD?

    The Irony. They do what they claim others do.

    The funny thing about this is that they use their model as evidence that evolution (and/or naturalism) is true in order to argue against those who question it and convert them to believing in UCD (& naturalism). However, if everyone converts to being an evolutionist (and a naturalist) then, according to them, their model would be false which would mean that, according to their model, evolution would be false which would mean that, according to their model, scientists would have to start disbelieving in evolution which would make the model correct (re – causing those same scientists to convert back to evolution since the model is right, which would then make it wrong, then they would have to convert back to being anti – evolution which would make it right, etc…).

  8. 8

    What am I to read now???

    I suggest reading the Bible.

  9. I quite my subscription to NS a long time ago. :D

    But Popular Science/Mechanics I still can’t get enough of.

  10. Gentlemen:

    1] GIGO — no computer simulation is equivalent to reality. None is any better than its input data, algorithms and underlying assumptions [not to mention error propagation etc etc etc]

    2] Evolutionism and associated evolutionary materialism were movements — indeed “crusades” — from the days of Huxley on.

    3] EM is a worldview and life agenda that often likes to dress up in a cheap lab coat — well, I know, I know, there are no designer-label, or bespoke tailored lab coats around. It needs to be sent to the nearest solid phil seminar room, for a proper education in basic, 101-level comparative difficulties analysis across competing worldviews.

    4] Whereupon, it will self destruct as inescapably self-referentially incoherent. [Cf APP 6 the always linked for some of why.]

    GEM of TKI

    PS: PS/PM rock!

  11. soplo caseosa (5)

    Erm, not quite. Religious conservatives are just as likely to have children who grow up to be gay as anyone else.

  12. I enjoyed this comment over on the NS site:

    “I have constructed a computer simulation that demonstrates how evolutionary anthropologists might have evolved.
    In most cases they go extinct but in the scenario than non-evolutionary-anthropologists give them material benefits in return for constructing vaguely plausible but superficial and unverifiable stories they are found to thrive – though they remain in small communities on the edge of society. “

  13. Religious conservatives are just as likely to have children who grow up to be gay as anyone else.

    I doubt that this is true, because the children of religious conservatives are taught (either explicitly or implicitly) to reinforce their “straight” impulses and resist their “gay” impulses. That was my experience. I CHOSE to be straight every time I resisted homosexual impulses in favor of a traditional Christian sexual orientation. As a direct result, I now have a wife and two kids. If I had gone with my feelings, I would likely not have produced any offspring at all. How can you say that religious training has no effect at all on sexual orientation?

  14. tribune7 Great one! And how true.
    —-
    Computer simulations of the evolution of religion? What utter nonsense.

    Unverifiable information? Like frog to prince evolution?

    Like ‘the brain is an outgrowth of the eye’ stories supposedly explaining the evolution of eyes? I kid you not, right HERE.
    - from a paper by Dr. Ma’ayan Semo, University College London
    Strange how normally very intelligent people can say such dumb things whenever Darwinism is involved.

  15. I think it’s only fair to note that there’s a lot of words and phrases in there suggesting that this isn’t and wasn’t meant to be definitive or in any way related to reality. The author spends a lot of time talking about “believers in the unreal” as opposed to the religious, and there’s lots of speculating and talking about “baby steps”. I think it’s probably more a case of an academic trying a new tool out on an old problem to see if anything interesting comes up. Given the high profile of the whole science vs. religion debate lately thanks to expelled and other things, the angle of the story is probably more dependent on the journalist looking for something likely to draw more traffic. I imagine the actual article is a rather lot more conservative in its statements.
    But still, I’m curious. If it’s utter silliness to perform an admittedly highly simplified, but real, computer simulation on human behavior to suggest an explanation for belief in religion… Then what is positing an utterly impossible computer simulation to suggest a reason to disbelieve evolution?

    Also, about those FEA animations. I’m hoping most of them are of somewhat better quality than this. Though the trailing end warping due to absolutely nothing at all is certainly interesting.

  16. Thanks, Gil. It is amazing, sad really, what gets passed off as “evidence.”

    Just this morning I happened to stumble across a post we had done quite some time ago on Avida, so I enjoyed the irony of checking UD today and finding your post. Seems like you have enough material for a whole category of Bogus Computer Simulations.

  17. Everyone should read Eric Anderson’s great piece on the Avida program:

    http://www.evolutiondebate.info/BitByte.pdf

    The following link might be of interest as well:

    http://www.uncommondescent.com.....mulations/

  18. Nullasalus said “Starting with the assumption that ‘telling people unverifiable things’ is genetic? Goodness. So, what kind of genes did Socrates have? How about Plato?”

    To answer your question: Socrates had socratic genes while Plato had platonic genes. All quite simple don’t you know. ;)

    On a more serious note, doesn’t it amaze you at the pervasiveness of reductionary thinking within modern science. Most do not even realize that it is a philosophical assumption. If rationality, or real knowledge is reducible to genes, then how do we distinguish which genes lead to real knowledge? How can you possibly trust your own genes, less it be out of blind faith? In other words, I concur!

  19. I blogged about this here, asking the question as to why these scientists never seem to get around to doing investigations into the origins of atheism:

    http://mothwo.blogspot.com

    David

  20. davidanderson

    You make an exellent point.

  21. I think that NS has some nerve throwing the smoke-bomb term “pseudoscience” at ID if this is what it is churns out on its pages these days. It is utterly embarrassing.

    “Evolutionary simulations” are 10-a-penny these days. If I were an atheist eager to employ my skills at propaganda-programming, perhaps I’d employ the following recipe:

    1. Try and imagine how a false belief could possibly benefit a member of my simulated population. (Of course, it is a dangerous question to ask exactly what it means for a simulated phenotype represented in the memory of a souped-up Turing Machine to hold a “false belief”. Who cares? Scientists trump philosophers in today’s market. I’ll accuse them of sophistry if they even dare to challenge me.)

    2. Initiate and breed population so that there will be a tendency to preserve those members which hold false beliefs. Wait a minute! My palms are starting to sweat. What happens if the results aren’t as I wish? Step three…

    3. Failing the desired outcome, run the program again! It doesn’t cost taxpayer money to repeat this kind of experiment with the parameters tweaked, and I don’t have to inform anyone of the outcome. How wonderfully unaccountable!

    [I have worked with EAs before. I know how awkward it is to get them to evolve in the "right" direction.]

    4. If (3) doesn’t produce the goods, I can show that the unstoppable march of evolution weeds out “believers in the unreal”. NS will surely publish that too. It’s win-win: I can either demonstrate that religious people shouldn’t exist, or else show why they do! (Oh oh. Philosophy alarm again. Does evolution say what does happen, what did happen, what could happen, or what should happen? I can’t remember. Who cares? As a scientist, my only job is to humbly tremble before the evidence. As long as it suits me.)

    5. One of steps (3) or (4) is guaranteed to work. But what if I am taken to task by somebody who has ever read a book? I will couch the whole thing in tentative language so that I have full deniability—but everyone will know what I am getting at. Because many today are eager to hear a Scientist debunk “religion” [again], it will sell like a hot cake.

    6. Finally, dealing with criticism.

    (i) If a religious person criticises me, I will say they are “infected” with a meme. Fantastic! Nobody can prove me wrong, and I can make a fool out of them in front of my friends.

    (ii) If a philosopher criticises me, I will point out that my elite has done more for the consumer than they ever did. People want cures, robotic limbs, speech recognition like Star Trek, tackling world hunger? They NEED me.

    (iii) If a scientist criticises me, I will at first attempt to discredit them somehow. If I am trumped by a “bigger scientist”, I will be crestfallen but employ clever rhetoric to demonstrate how this study is of limited, but nevertheless important value. “Baby steps.” I never /claimed/ to have the complete answer. Anyway, hopefully other big scientists will come to my rescue. They can endorse me [I mean, my work], I’ll endorse them [oops---theirs].

    (iv) If a layman or somebody online criticises me, I’ll just launch into them no-holds-barred. Don’t they know who I am?? I’m smarter than you: I’ve got a PhD.

    The wonderful thing is, you can decide the outcome of all this before you even declare a variable.

    I am sorry if this offends anybody, and I honestly don’t go out of my way to be cynical. I love science and many of my mentors have been scientists. I have passionately followed theoretical computer science since I was young. (A bit odd, I know.) I am annoyed to see the field abused by hired goons like this. But, sadly, this is how I perceive many establishment scientists these days. I’m sure the public must do too, which is why NS don’t deserve such a wide audience. It is only by the recipe above that I can possibly see how this tripe could end up in print.

    Rant over.

  22. Quite simply, the journalist, Callaway, betrays that he does not understand the scope of what the researcher, Sosis, is doing:

    Richard Sosis … says the model adds a new dimension to the debate over how religion could have evolved

    So Sosis puts this within a defined context, an existing debate.

    The headline is what is wrong: “Religion is a product of evolution, software suggests.”

    In it, Callaway betrays that he is ignorant of what can be concluded from what: He writes with clear premise language: “distilling religious belief into a genetic predisposition to pass along unverifiable information…” as the model of religion considered in the simulation (I respect that that he didn’t say “false” or “imaginary”). But it is modeled as a genetic predisposition.

    Sosis’ wording suggests no more than that he is adding to an already existing debate, Callaway’s summary tries to broaden the implication to span to an area larger than what was tested.

    It is quite possible that in investigating “how religion may have evolved” the end goal is to document that it did evolve. Once they get a reasonable enough model, someone can indeed evoke the typical ad Ignorantium that tends to be used here and say that they have given a reasonable enough case for evolved origin and the challenge is to prove “how else” religion came to be.

    What’s interesting is Callaway’s faith in the process that requires a fit model even while the model is being disputed, despite the model being the only suggestion of verifiability in the endgame argument. The ultimate argument says that provided that we have given you a solid model for religion as evolved, you have to prove to us “how else”. Here is a case that the model has omitted an angle (or this research is not significant) or even has previously run counter to this model, and yet despite the model being in process, it remains a sufficient enough “verification” that I would accept that Callaway’s own misunderstanding does not fall under the exact pattern of Sosis’ model. Truly that Sosis’ study “suggests” that religion was evolved outside of a context that doesn’t suggest religion was evolved is not verified–at least within Callaway’s text. So how much, given that Callaway’s take is suspect, does Callaway’s interpretation of Sosis’ study fall–especially given that publication in the NS more than constitutes “passing along”–within the parameters specified as religion.

    Ultimately, I don’t think Sosis has a case. He doesn’t address how avoidable “unverifiable” propositions are. I love the idea that gets propagated about the “verifiable in principle” which suggests that we can impose an unverified suppositions on what is “verifiable” but not as yet verified. Based on what we believe is likely to be verified–yet unverified. The funny think is is that this often gets put into the mix as just as important as things that we have already verified. Imagine for example that the value of a computer circuit were on the basis that it was theoretically verifiable that electricity can be conducted through a medium in a constant state, so as to hold states and create signals. Would we have “technology” that was only “theoretically” verifiable?

    “Theoretically verifiability” is only an ossification of an answer to the problem of any definite answer to what is worth pursuing from what is already known. The harder problems of skepticism are not to everyone’s liking so they invent these little schemes of “unverified” knowledge.

    Taken at it’s base “verifiable” means “able” to be verified. How able? Well, theoretically able. How soon until we hear that “scientific consensus” verifiability is the only verifiability that could make sense? So the consensus is verified by its being consensus. You see, the ad Populum argument doesn’t work in the world of Science, where we need consensus, and that it was theorized as a fallacy by ancient Greeks not currently up to snuff on 21st century science.

Leave a Reply