Uncommon Descent Serving The Intelligent Design Community

# Don’t Trust Computer Simulations And Models That Can’t Be Tested Against Reality

Share
Flipboard
Print
Email

Computer simulations of global warming and Darwinian mechanisms in biology should not be trusted, because they can’t be subjected to empirical verification. In these two areas, computer simulations and models can degenerate into nothing more than digital just-so stories — in one category about the future, and in the other about the past. The programmer can produce whatever outcome he desires, by choosing initial assumptions and algorithms, and weighting various factors to produce a desired output.

Unfortunately, when those in the general public hear the words “scientific” and “computer model,” they often assume that unassailable truth has been established.

Check out this book review by Carl Wunsch in the American Scientist. The book is entitled, Useless Arithmetic: Why Environmental Scientists Can’t Predict the Future..

I’m a software engineer who works for an aerospace R&D company. My specialty is designing guidance, navigation and control software for precision-guided airdrop systems. This involves a lot of mathematics and computer modeling/simulation.

A lot of work — over many years, by many brilliant mathematicians, aeronautical engineers, and atmospheric scientists — has gone into developing the basic theory with which I work. Here is some of the basic mathematics:

But this is just the beginning. The real world is very complex and messy: the flow-field created by the size and shape of the payload below the parachute changes the effective drag coefficient of the parachute; wind gusts and thermal activity are thoroughly unpredictable; and there are other considerations like changes in air density caused by atmospheric lapse rate.

The mission-planner for the guided-airdrop system mentioned above is a computer model which attempts to simulate the theoretical flight path of a parachute and its payload. It is critically dependent upon up-to-the-minute wind data which we retrieve by deploying a windsonde from the aircraft just before the payload is released. Even with all of this, a guidance system is required to make real-time corrections to compensate for all the potential anomalies described above.

So, the next time a global-warming advocate or a Darwinist tries to convince you of the validity of a computer model that can tell you what will happen in the future or what happened in the past, don’t pay much attention.

Computer simulations can be tested. Computer simulations of how the cosmos evolved can be tested on what the universe looks like today via galaxy surveys. The large scale structure of the universe looks the same in both the simulations and real data once the correct parameters are put into the models. JIM Jim Rachetto
Come on guys. Believe the scientists I am a Christian and believe in an Intelligent Creator, how can one not, and yet do believe that the universe and life evolved. The Apostle Paul said that the universe was a revelation of God. So we should trust the scientists to interpet that part of HIS revelation. I see no conflict between evolution and the Bible. God was telling Moses basiclly who, not how. It takes a much bigger God to create a universe that can create itself than a snap creation. No I don't believe in the old standard Darwinian evolution and I doubt if many biologist do either anymore, JIM Jim Rachetto
DaveScot: Ã¢â‚¬Å“And engineers certainly do consider the possible choices that intelligent agents make that will interact with their designs.Ã¢â‚¬Â Freelurker (me): Ã¢â‚¬Å“Tell me, should engineers consider the possible choices of non-material intelligent agents?Ã¢â‚¬Â j: Ã¢â‚¬Å“Humans can do things Ã¢â‚¬â€ such as generate indefinite amounts of complex, specified information, at will Ã¢â‚¬â€ that material processes canÃ¢â‚¬â„¢t. It would seem that humans are non-material intelligent agents.Ã¢â‚¬Â Should engineers consider the possible choices of non-human, non-material intelligent agents? Freelurker
DaveScot wrote: Ã¢â‚¬Å“Engineering deals with non-material causes all the time. Such causes are called Ã¢â‚¬Ëœacts of God.Ã¢â‚¬â„¢Ã¢â‚¬Â I donÃ¢â‚¬â„¢t agree that engineers are (as a profession) referring to non-material causes when we use the term Ã¢â‚¬Å“Acts of God,Ã¢â‚¬Â so IÃ¢â‚¬â„¢ve started looking for engineering definitions of the term. IÃ¢â‚¬â„¢ve found only one so far, but that one is consistent with how I have seen engineers use the term:
ACTS OF GOD. (1) an extraordinary interruptions by a natural cause, as a flood or earthquake, or the usual course of events that experience, foresight or care cannot reasonable foresee or prevent; (2) an event in nature over which neither the owner nor the contractor has any control. [http://www.iienet2.org/Details.aspx?id=1740]
ThereÃ¢â‚¬â„¢s nothing in there about non-material causes. This raises an interesting issue. Notice that the term Ã¢â‚¬Å“Acts of GodÃ¢â‚¬Â is defined without ever mentioning God. You can bet that when great engineers of the past used the term Ã¢â‚¬Å“Acts of GodÃ¢â‚¬Â they were talking about things done by God. Engineers have now defined away consideration of God. WouldnÃ¢â‚¬â„¢t you say that materialist dogma has come to dominate the engineering profession? Do you guys have a problem with that or do engineers get a pass? (Someone out there may think that including God in the definition would make it circular, but that is not the case; itÃ¢â‚¬â„¢s Ã¢â‚¬Å“Acts of GodÃ¢â‚¬Â that is being defined. For example, one might define Ã¢â‚¬Å“Acts of GodÃ¢â‚¬Â as Ã¢â‚¬Å“manifestations of the will of GodÃ¢â‚¬Â without being circular.) Freelurker
Freelurker: "Why are there complaints about the materialistic bias in science but not about the materialistic bias in engineering? Why is a materialistic approach good enough for engineering but not good enough for science?" Engineers use whatever works well, including whatever can be shown to be valid scientific theories, materialistic or not. And, as DaveScot has already pointed out, engineers do often design to account for the intelligent (and unintelligent) agency of humans (ergonomics, etc.) But they can't use theories that are supported only by dogmatic assertions of their validity. The theories of physics and chemistry upon which engineering is based are experimentally verifiable. Such theories are also able to be accurately modeled using computer programs. But imagine trying to apply Neo-Darwinian hypothesis to engineering. It couldn't be done. It would be absolutely indefensible to rely on a "theory" that cannot be experimentally verified, nor even modelled using a computer! The neo-Darwinian hypothesis of evolution by random mutation and natural selection is worthless from the standpoint of engineering. (Genetic algorithms and other forms of evolutionary computation, which can find application in engineering, are not examples NDE, but of intelligent design. What they demonstrate is that random mutation and automated artificial selection, can be used for the development of solutions to pre-specified design problems. But of course the efficacy of artificial selection was recognized long before Darwin -- indeed, Darwin used it in one of his fallacious arguments in The Origin of Species.) Freelurker: "Tell me, should engineers consider the possible choices of non-material intelligent agents?" Humans can do things -- such as generate indefinite amounts of complex, specified information, at will -- that material processes can't. It would seem that humans are non-material intelligent agents. j
I am working today and will be very busy the next couple of days. Thank you for your responses; I am eager to get back to this. Freelurker
Hi Dave: I am adverting to a "strictly speaking" point in phil of science, that electrons and such like are in fact not directly observed but are inferred and/or imaged based on observations, classically spectroscopic ones especially. (Nor, strictly did you feel electrons but rather electrical effects held over the past 200 years or so to be traceable thereto, e.g. electrical shock or even burning. Do you have any interesting stores to match the IC that I accidentally spread out as a 6" fireball in front of my eyes, while trying to measure properties? I was tasting traces of the IC for weeks . . . ] Thus, e.g., my remark on bubble chamber tracks [i.e. for beta particles and of secondary netas which is how we sometimes infer gamma rays!] Molecules of Oxygen, similarly, are strictly speaking inferred and/or imaged, not observed. Indeed, that is a part of why there was a debate a bit over 100 years ago on the reality of atoms. One of Einstein's famous 1905 papers was a discussion of Brownian motion, which was deemed decisive: the tiny particles act as giant molecules and take part in the random motion of hte fluid in which they are suspended. [There was an interesting anticipation of his reasoning in Lucretius, think, though the scale of the dust particles is such that the mircocurrents in the air are a more plausible current explanation.] Milikan's experiment that measured the electron charge observed tiny light spheres floating in air, of oil if memory served, in an electric field. Then by adjusting voltages etc he got the drops suspended, thus inferred the charge to equate gravity and electric forces. As he reported, the charges were quantised relative to a basic unit, now known as e. [BTW, as I recall from one of my undergrad experiments, the manipulation to do this is very, very, very ticklish. I recall a MEU TV special in which his original notes were discussed. He cherry picked the data points that were reported in the literature . . . "Beaut! Publish!" in the margins are all too telltale.] Notice again, the electron is inferred adn its properties are calculated, not strictly observed. We could go on and on. Okay, have fun! (Don't "let the smoke and fire out" of any ICs . . . it makes them stop working.) GEM of TKI kairosfocus
Freelurker re materialism in engineering Engineering deals with non-material causes all the time. Such causes are called "acts of God". http://scholar.google.com/scholar?q=%22acts+of+god%22+engineering&sourceid=navclient&hl=en Now that I've answered your question you can hopefully answer mine. DaveScot
kairosfocus have you ever SEEN an electron? No. But I felt them. I haven't seen a molecule of oxygen either. Have you ever seen your left ventricle? DaveScot
H'mm: A few notes: 1] FL, Nos 13 & 16: The model that you use and the way you use it are completely materialistic. This is consistent with all other engineering work I have ever seen . . . . Why are there complaints about the materialistic bias in science but not about the materialistic bias in engineering? Why is a materialistic approach good enough for engineering but not good enough for science? Here, there is a basic equivocation on the term, "materialism." For, materialism, primarily, is a worldview/philosophical position. So when UD's statement says "Materialistic ideology . . . materialistic worldview . . . materialistic theories of biological and cosmological evolution" they are plainly speaking in a philosophical context, then extending it inot the evolutionary materialist research programme "from hydrogen to humans." Now, of course, most science and engineering trained people as well as the general public today, have had little exposure to worldview thinking and analysis, so to some extent missing the force of the above is understandable. But, in a forum at this level, it is not to be shrugged off -- the conflation of materialism as worldview with inferences to chance and/or natural regulatriities in scientific work takes on the character of a pernicious, often overlooked fallacy. Lakatos gives a useful perspective. In identifying research programmes as a unit of analysis of sceince as a social and philosphical process, he points outt hat there is a worldviews core and a belt of associated theories. Consequently, empirical testing faces a much steeper challenge in addressing the pattern of interlocking theories than is at first apparent. Indeed, his conclusion was that it is only when there is a range of challenges that cutacross several of the theories, that the programme as a whole begins to be questioned. And, what ID has done is to raise the point that functionally specific, complex information is, on excellent empirical fgrunds, an artifact of design. Further, when we see multidimensional finetuning, that too is often a characteristic of design. So, the issue is to consider the full set of possible causal forces, chance, necessity and agency, when looking at cases where evidence of such CSI or finetuning is on the table. Worldview level question-begging and agenda serving redefinitions of "science" don't make the grade. 2] M7, 15: An intelligent mind produced them, based upon information from thinkers in the past who wholeheartedly believed in an Intelligent Mind that crreated the universe and left us clues all over with an intelligent mind of our own to interpret the information we see and hear. This adverts to the historical fact that modern science was by and large founded by Judaeo-Christian, theistic, design-orieneted thinkers. That particularly holds for the newtonian dynamics that is so much used in engineering, which is here applied science. And of course, the model being used to project the path of a parachuted object is a newtonian dynamical one. Further to this, we see a case of CSI, a complex string of digital data constituting a functional computer program; and it is lo and behold, produced by an intelligent mind. So, what about the similarly complex digital data in DNA and the associated nanotechnology of the cell? Especially, given that contingency eliminates natural regularities,a nd that chance has a maximally low chance of arriving att he codes and systems in the gamut of the observed cosmos? 3] DS, 18: Engineering doesnÃ¢â‚¬â„¢t deal with untestable abstracts. Theoretical sciences do In fact, though your intended point is valid,this claim is strictly not so. All sciences and applied sciences [i.e. engineering] do have to address unobservables, e.g. have you ever SEEN an electron? [We infer to the electron etc as unobserved explanatory concepts that make sense of what we do observe, bubble chamber tracks, meter readings, etc.] The other half is also problematic, as theoretical sciences do generate expected observables as predictions that are subject to empirical test. That is, we are looking here at abductive explanation as the framework of scientific thought. Pausing . . . GEM of TKI kairosfocus
Freelurker: "Is a materialistic perspective corrupting to science but perfectly fine for engineering?" You don't understand. A materialistic perspective is perfectly fine for "science" as long as it doesn't become a dogma that precludes following the evidence where it leads. When design becomes apparent -- indeed, when design SCREAMS WITH A LOUD AND SHRILL VOICE -- it should not be ignored in favor of a precommitment to a philosophy. DaveScot, I get frustrated with the distinction between engineering and science. Engineering is science, that deals with the messy and complex world I mentioned above. Theoretical science has contributed a lot, and I use all kinds of theoretical science in my work, but engineering is held to a much higher standard: The products of the engineer's science must work. Perhaps this is why some engineers tend to be skeptical of Darwinian claims. There is no accountability for the Darwinist to demonstrate that his stuff actually works. GilDodgen
DaveScot:
Engineering doesnÃ¢â‚¬â„¢t deal with untestable abstracts. Theoretical sciences do. Historical biology is a theoretical science. Therein lies the difference and the reason why engineering (and experimental sciences) are inextricably anchored in measurement and observation.
But what about materialism? Is a materialistic perspective corrupting to science [or certain sciences] but perfectly fine for engineering?
And engineers certainly do consider the possible choices that intelligent agents make that will interact with their designs. In fact thatÃ¢â‚¬â„¢s just about a given in most engineering from bridges to computers to spacecraft to weapons to drugs and just about everything designed to be used and possibly abused by intelligent agents in all kinds of anticipated and unanticipated ways.
And IÃ¢â‚¬â„¢ve never said otherwise. Tell me, should engineers consider the possible choices of non-material intelligent agents?
A year ago you posted a comment here Ã¢â‚¬Â¦ I replied that Ã¢â‚¬Â¦ Can I get my reply now?
I gave a reply, but it never showed up. That explains why I havenÃ¢â‚¬â„¢t been back here much. IÃ¢â‚¬â„¢ll be happy to talk again about design recognition in engineering, but first IÃ¢â‚¬â„¢d like to find out what IDists think about materialism in engineering. Freelurker
Freelurker Engineering doesn't deal with untestable abstracts. Theoretical sciences do. Historical biology is a theoretical science. Therein lies the difference and the reason why engineering (and experimental sciences) are inextricably anchored in measurement and observation. Organic evolution is a non-repeatable phenomenon that happened in the past and if it's even still happening today it's not demonstrable because it ostensibly happens too slow for its major claims to be observed in any realistic timeframe. By major claims I refer to the creation of novel cell types, tissue types, organs, and body plans, not the kind of evolution that makes a tiny change to a digestive enzyme in a bacteria or alters the nominal size of finch beaks but real anatomical novelty never seen before that would place the organism into major new taxonomic category. And engineers certainly do consider the possible choices that intelligent agents make that will interact with their designs. In fact that's just about a given in most engineering from bridges to computers to spacecraft to weapons to drugs and just about everything designed to be used and possibly abused by intelligent agents in all kinds of anticipated and unanticipated ways. A year ago you posted a comment here asking for any examples of engineers doing design recognition claiming you couldn't find any. I replied that if you were really an engineer you should have answered that question for yourself by knowing what phrase to google. Then you disappeared for almost a year without replying. Can I get my reply now? DaveScot
Computer simulations, unlike exact mathematical solutions in the form of formulas and equations, do not give immediate insights into the relative influences of different parameters. For example, the formula for kinetic energy, one-half the product of the mass and the square of the velocity, immediately shows that doubling the velocity has twice the effect of doubling the mass. Parameter studies of computer simulations can be performed by making many runs to see the effect of varying different parameters, but that is time-consuming. Therefore, an exact or even an approximate formulary or equational solution for an approximate model of a system is often better than an exact computer simulation (which is usually not even possible) of the exact model. Where possible, formulas or equations should be used to verify the results of computer simulations and also to guide the direction of studies using computer simulations. Another problem with computer simulations -- even exact computer simulations -- is the "butterfly effect" of chaos theory, which says that small changes in initial conditions can sometimes produce huge changes down the line. "Butterfly effect" is the idea that the flapping of a butterfly's wings can determine whether or not, say, there is going to be a hurricane. Indeed, one result of the butterfly effect is that sometimes even direct tests of the system itself cannot predict the system's future behavior. Larry Fafarman
But why are scientific models of origins any more materialistic than engineering models of parachutes? [My intention here is to use the word Ã¢â‚¬Å“materialismÃ¢â‚¬Â in the same way it is used in the UD mission statement. If there is an official UD definition of the term, IÃ¢â‚¬â„¢ll work with that.] Why are there complaints about the materialistic bias in science but not about the materialistic bias in engineering? Why is a materialistic approach good enough for engineering but not good enough for science? Freelurker
Freelurker, Why are they "materialistic" models? An intelligent mind produced them, based upon information from thinkers in the past who wholeheartedly believed in an Intelligent Mind that crreated the universe and left us clues all over with an intelligent mind of our own to interpret the information we see and hear. "Materialistic ideology has subverted the study of biological..." Evolutionist devoted to materialism produced "vestigial organs" and "JunkDNA" ID is not afraid of iterative processes or genetic models. It incorporates it as part of an overall thinking approach. Whereas accidental, materialistic philosophy cannot begin to conceive of how information is created in the form of functional codes like DNA. Check out Trevors and Abel discussing FSC - functional sequence complexity. They acknowledge Dembski and Behe's thoughtful objections to a purely random model have opened a place for serious debate in this area. These models mean nothing without an intelligent mind to observe and then predict outcomes. They are only materialistic if not observed. But once an observer, indeed a creator of such models is aware and utilizing such information to forcefully manipulate natural causes then the material is transposed into the non-material. Michaels7
PaV, "In the case of evolution, this is near impossible since we canÃ¢â‚¬â„¢t reconstruct how evolution unfolded in any precise manner. So it ends up one guess against anotherÃ¢â‚¬â€œwhich means you can end up with anything you want; so you end up, likely, with the result you had in mind when you began the whole thing. Sounds like GW computer models, doesnÃ¢â‚¬â„¢t it?" Uh huh and good point. Part of this is circular funding for a specific outcome. Most research projects are good for funding. For example, more efficient technology for biofuels, or solar power, or any number of medical breakthrus that we know are practical in terms of reach. Other research projects are so far off the map in terms of reasonable models to even quantify the variables of input. Much less a robust and powerful enough system to then manage all the variables input into a manegeable output prediction model. Schwartz nailed it when he stated that "evolution is undemonstrable" and Doolittle nailed Darwin's Tree of Life issue with the "quixotic pursuit" label. Which in my mind, notes it is largely a great waste of time and money. And it goes back to what Dave is fond of noting that it is largely stamp collecting and then guessing the value of such stamps. Who knows? No one really does, but National Geographic, SA, and other media, plus our Museums, our Universities and so many other areas are tied up in this for literally billions of dollars. So I understand now, how such a large group of people get fearful when their pet money theory gets knocked down. They all have so much invested from years of work, books, to paid benefits and a gravy train to defend and relish. They're not about to give in without a fight. But they are getting more desperate I believe as the Origins of Life Prize indicates. They're out of answers for how FSC and information formed. To the point of putting up Prize money for anyone to put forward a reasonable scientific paper as rebuttal to Dembski and Behe. The naysayers can yell all they want. Dembski and Behe are now recognized for the challenge they have thrown down by many leading scientist. Michaels7
I read your post and I read the UD position statement that appears next to it. The model that you use and the way you use it are completely materialistic. This is consistent with all other engineering work I have ever seen. Is a materialistic perspective corrupting to science but perfectly fine for engineering? Freelurker
Interesting interview with Michael Crichton on Charlie Rose. Here is the link to the transcript. http://www.crichton-official.com/charlierose-021907.pdf He makes an excellent point of scientific research taking on a different tone after Lyndon Johnson declared war on cancer. To get funding the research needed to be linked to fighting cancer. He also notes an interesting stat that about only 9% of scientists are absolutely certain humans are the primary cause of global warming. The scientists were anonymous on their answers. I wonder what survey in the biological sciences would look like asking if scientists believed (A) Evolution is true (B) Are you absolutely certain NDE is the primary cause of evolution. late_model
johnnyb : great link - Crichton should take a closer look at ID vs Evo. Don't know where he stands on that but, if he can't see the same symptoms of diseases of the mind at work on the Darweenie's side, he has himself been duped by what he describes in his own lecture. Borne
Many years ago I represented a multi-national accounting firm in a securities fraud case. It seems my clientÃ¢â‚¬â„¢s client had developed a complicated computer model of its business projects and then used those projections to sell several hundred million dollars of stock to investors. My client, the accounting firm, had Ã¢â‚¬Å“certifiedÃ¢â‚¬Â the projections. Basically, this meant they had looked at the companyÃ¢â‚¬â„¢s assumptions and determined those assumptions to be Ã¢â‚¬Å“not unreasonableÃ¢â‚¬Â and verified the mathematical integrity of the model (i.e., the numbers added up). Well of course the business went bust and the investors sued my client for securities fraud. One of our defenses in the case was Ã¢â‚¬Å“the projections were just a computer model, not a performance promise.Ã¢â‚¬Â IÃ¢â‚¬â„¢m not sure the plaintiffs and their lawyers ever understood the difference. BarryA
A point of interest: As a backup (in the event technical problems preclude the use of windsonde data), we program the guided-airdrop systems in advance with trajectories computed from forecast winds. These proprietary forecast winds are provided by the military and are generated with the best information available and with the best computer models. By the time we deploy on a mission, these wind forecasts are a few hours old. Sometimes they work, and sometimes they don't. In only a few hours winds can completely reverse direction or change in velocity significantly, and this is impossible to predict reliably. If wind predictions a few hours from now are hard to make, what chance do we have of accurately predicting the climate a few decades from now? GilDodgen
Way back when, I took a course in numerical methods, Runge-Kutta, etc. I quickly lost interest in the class because these methods, which I'm sure underlie much of computer simulation/models to this day, were so darn simplistic. There's nothing like an analytical solution to a problem. What I saw amounted to more or less an educated guess. The simulator was just guessing that he had the right equations. Once you set up the gridwork for the simulation, the computer just cranks out numbers. There's no way of knowing whether it is correct or not, unless you can compare it to what reality tells you. In the case of evolution, this is near impossible since we can't reconstruct how evolution unfolded in any precise manner. So it ends up one guess against another--which means you can end up with anything you want; so you end up, likely, with the result you had in mind when you began the whole thing. Sounds like GW computer models, doesn't it? As to GW models, it is possible for them to "predict" the past. You set up a model, input the data averaged over the last thirty years, but beginning 40 years ago, and then you simulate the following ten years--for which, of course, we have ACTUAL data. I believe they've done this, and can't get the models to "predict" actual weather. But let's spend trillions of dollars on it anyways! PaV
ajl:
Naturalists can never show this level of detail and testability for their Evol. models yets continue to say its as established as gravity. t
And when was the last time you heard a physicist who specializes in gravitatiion theory say that their theories are as well established as evolution? Just a thought! DonaldM
Well I would hope that a good portion of the population is well enough educated to realize what a computer simulation is and what its limitations are. They do however represent our best and only tool for predicting complex events. Those simulations will always be limited in accuracy by our knowledge of the forces and mechanisms at play in the system we are modeling. When you are talking about the planet Earth and climate modelling the potential for error approaches 100%. So several models with different assumptions are done to give us our best range of possibilities. It is unfortunately the more outlandish scenarios that get hyped because people love sensationalism and are generally not interested in complexity. There are those that manipulate this for political ends. I have a serious dislike for those people. That said we should not write off climate models. As our knowledge improves so should the accuracy of those models. Our best models today are worth something, but probably not worth making policy over. As far as evolutionary models go, I thought they were meant solely as a refutation to an IC argument. To show that it could be done, however seemingly unlikely, would logically defeat an IC argument as far as I can see. jmcd
The truth about global warming: The aliens caused it. I've posted this before, but it's worth repeating. johnnyb
Gil: my guess is the FAA (or whoever monitors your work) would never, ever let you do your drops unless you showed mathematically that it would work. And not in a subtle way, but a totally convincing mathematical proof. Naturalists can never show this level of detail and testability for their Evol. models yets continue to say its as established as gravity. ajl
"Argumentation cannot suffice for the discovery of new work, since the subtlety of Nature is greater many times than the subtlety of argument." -- Francis Bacon In the case of evolutionary computation, one of the values of computer models is, of course, the demonstration, in principle, of what algorithmic processes are capable of. But the problem for blind watchmaker evolution is that no such models exist showing that it is capable of generating CSI. This post also reminds me of something I read once, but can't remember exactly where it was. (If anyone knows what I'm talking about, please let me know -- I'd like to have an actual reference for it.) As I remember it: Someone was commenting on the use of computer simulations in education, and asked whether, in the future, all education might not be accomplished using such (accurate) simulations only. When this possibility was run past physics professors, the overwhelming response was that, no, they would never go for that, because there's no substitute for actual interaction with the real world in one's physics education. Simulations can be helpful, but labs are an absolute necessity. One has to get their hands dirty to get a true appreciation of nature. j
Following up: This excerpt is also useful food for thought, on how computer modelling has diverged from insightful mathematical modelling, with what implications:
Before cheap, large, fast computers existed, "mathematical modeling" was indistinguishable from the construction of mathematical "theories" describing particular phenomena. Calculations were commonly done by scientists who had a grounding in differential and partial differential equationsÃ¢â‚¬â€a grounding that was often based on fluid dynamics, electromagnetic theory, SchrÃƒÂ¶dinger's equation and the like. Those scientists (like their counterparts today) were familiar with a wide range of approximate and asymptotic methods . . . . With modern computers, it is now possible for a graduate student or a practicing engineer to acquire a very complex computer code, hundreds of thousands of lines long, worked over by several preceding generations of scientists, with a complexity so great that no single individual actually understands either the underlying physical principles or the behavior of the computer codeÃ¢â‚¬â€or the degree to which it actually represents the phenomenon of interest. These codes are accompanied by manuals explaining how to set them up and how to run them, often with a very long list of "default" parameters. Sometimes they represent the coupling of two or more submodels, each of which appears well understood, but whose interaction can lead to completely unexpected behavior (as when a simple pendulum is hung on the end of another simple pendulum). One hundred years in the future, who will be able to reconstruct the assumptions and details of these calculations?
How many people really understand the latest climate models in depth, line by line, module by module, interaction by interaction,a nd globally? Evolutionary models? Econometric models, etc etc? How well are these calibrated against material, well warranted data sets? What track record of successful prediction obtains, and what are the environamental constraints that may invalidate such predictive power? [Known unknowns? Unknown unknowns? . . .] And more . . . Maybe we need to do some serious rethinking? "Mirror, mirror on the wall . . .?" GEM of TKI kairosfocus