Intelligent Design

A More Realistic Computer Simulation of Biological Evolution

Spread the love

In another thread a fellow who goes by Legendary made some rather derisive comments about a suggestion I once made, concerning making computer programs that purport to model biological evolution more realistic. The suggestion was half serious and half tongue-in-cheek, since it would be impractical.

My argument was as follows: Computer programs that purport to model biological evolution invariably isolate the effects of “mutations” to only those aspects of the “organism” that have a chance of helping the organism approach the desired goal (EQU in the case of Avida, for example). But this ignores an extremely important aspect of modeling living systems.

Random mutations, if they are truly random, will affect, and potentially damage, any aspect of the organism, including its ability to survive and reproduce. The computer program, OS, and hardware represent the features of the simulation that keep the organism alive and allow it to reproduce, but this is artificially isolated from the effects of mutations.

Thus, a realistic simulation would allow the program, OS, and hardware to be affected in a random fashion, just as a real organism’s ability to survive and reproduce would be affected randomly by mutational interference. A mutation might cause an enzyme to malfunction and the organism would suffer an early demise, or it might be rendered sterile, and the beneficial mutations would never be passed on.

Of course, this would not be practical, and each “organism” would require its own computer, but the point should be clear: A simulation can’t just arbitrarily ignore aspects of the reality it purports to simulate, because taking them into account would be likely to result in an undesirable outcome.

As a footnote, I highly recommend reading Eric Anderson’s piece on Avida here.

55 Replies to “A More Realistic Computer Simulation of Biological Evolution

  1. 1
    Nakashima says:

    Mr Dodgen,

    How would you apply the same principles to simulations of airplanes flying in storms, or nuclear power plants?

  2. 2
    George L Farquhar says:

    Gil,
    Given that Dr Dembski and Marks have a website dedicated to “Computer Simulations of Biological Evolution” do you think their work is invalid in light of this?

    Thus, a realistic simulation would allow the program, OS, and hardware to be affected in a random fashion, just as a real organism’s ability to survive and reproduce would be affected randomly by mutational interference.

    Can the “hardware” be virtual? For instance, I’m typing this inside a virtual machine right now. If it happened to be running a “Computer Simulation of Biological Evolution” as well, inside the VM, and I started to “poke” random values into the memory space of the virtual machine would you consider that as being a “Random mutation” as defined here? I.E. the “machine” also was subject to random mutation?

    I don’t see how that would be functionally different to “mutating” an actual computer. Just a bit cheaper.

    As virtual machines can be created on demand would this remove the “impratical” objection to performing this more realistic simuation? Modern servers can create many thousands of VMs on demand.

    If you agree that this “virtual mutating machine running a mutating program” met your definition, and where such a thing to be done, if the results did not match existing observed data regarding evolving populations would you then discard the idea?

    Would you care to make a prediction in advance of the results of such a simulation, were it to be done?

  3. 3
    Tajimas D says:

    Sorry, but I don’t understand the point being made here.

    If you’re talking about things like genetic algorithms, they work by selecting the ‘most fit’ individuals from a population and allowing them to breed to produce the next generation (with a specific mutation rate). Those individuals that were not selected can be considered either ‘inviable’ or ‘infertile’ if you wish to carry over the analogy.

    I don’t understand why you’d need for the program to reach outside itself and affect the OS or the hardware when the simulation is self-contained. Deleterious mutations in real biological organisms, for instance, don’t make the universe explode.

  4. 4
    Nakashima says:

    I’m interested in the opinions of anyone in the LS-DYNA community on this subject. Does it pass the beverage through the nose test with them?

  5. 5
    Diffaxial says:

    Gil, these comments, made in September 2006 in response to your original post, are as on point as ever:

    Reciprocating Bill:

    This makes *exactly* as much sense as requiring that a supercomputer simulating a hurricane blow over tables and chairs, drench the operator, and cause widespread power outages.

    You have forgotten what Turing demonstrated: the power of computation lies in the independence of computational algorithms from the physical substrate upon which the computation is instantiated.

    Karl Pfluger:

    Absolutely right. Furthermore, allowing random errors anywhere in the simulation would be tantamount to allowing the laws of physics, geography, climate, etc., to change instantaneously, which of course does not happen in reality.

    The selective environment is not random, and so a realistic simulation of a selective environment cannot be random either.

    Tom English:

    Teaching computer science students from the undergraduate to the doctoral level, I encountered quite a few who were excellent programmers, but who could not begin to comprehend the notion of a model. The concept is simply too abstract for some people. They never catch on to it.

    A simulation model of evolution executes on a computer, but the computer, its operating system, and the runtime system of the programming language in which the simulation was written are not part of the model. Their function is to execute precisely the evolutionary model specified by the programmer. Any environmental cataclysms are simulated by the program itself, and are not a matter of failure of the computer hardware or the software operating system. That is, the environment is simulated by a properly functioning computer. The computer itself is not the simulated environment.

    I say categorically, as someone who has worked in evolutionary computation for 15 years, that Gil does not understand what he is talking about. This is not to say that he is trying to mislead anyone. It is simply clear that he has never grasped the nature of a simulation model. His comments reflect the sort of concrete thinking I have tried to help many students grow beyond, often without success.

    Reciprocating Bill:

    Consider this: One may run an evolutionary simulation on a virtual Apple II, emulated (simulated) by a PC, emulated by a Mac running Virtual PC – and so on ad infinitum, substituting at each level any computational device powerful enough to be a Turing machine (given enough time and tape).

    At what level are you going to introduce the random hardware failures? All of them? Into the Apple II emulator alone? Just at the bottom?

    Answer: NOWHERE, as such failures are not part of the computed evolutionary model.

    Which is not to say that we couldn’t simulate Gil’s suggested experiment!

    Links to the extensive discussions that followed:

    http://www.uncommondescent.com.....n-biology/

    http://www.uncommondescent.com.....ment-67208

  6. 6
    Dave W. says:

    It appears to me that Mr. Dodgen could answer his own objection…

    A simulation can’t just arbitrarily ignore aspects of the reality it purports to simulate…

    …if he were to ask himself one simple question: what do evolutionary simulations purport to simulate?

    What does Avida purport to simulate? What does WEASEL purport to simulate?

    Take a hypothetical simulator which purports to simulate nothing more than genetic mutations and selection leading to an increase in information. To fault it because it fails to simulate broken ribosomes, broken bones or shotgun blasts to its simulated creatures’ heads is to criticize it for something that it does not purport to simulate.

    In other words, to properly criticize evolutionary simulators, one must restrict one’s criticism to the claims made (what the simulators purport to simulate). One cannot, as Mr. Dodgen seems to want, make additional claims on behalf of the simulators for the purposes of knocking those claims down. That would be a classic strawman.

    So, Mr. Dodgen, what are the names of the people who are working on what they flatly claim to be a simulator of “biological evolution?” After all, the researchers that Eric Anderson freely insulted only purported to be working on simulating natural selection and genetic mutations, not on the entirety of “biological evolution.”

  7. 7
    GilDodgen says:

    The forest is not being seen for the trees. The point of the example is that critical aspects of a real-world system cannot simply be ignored, and victory declared on the basis of simplifying assumptions chosen for the purpose of producing a desired outcome. Clearly, the influence of mutations on the survivability and reproductive success of an organism, not related to the goal being sought, cannot be ignored without dramatically and artificially expanding the probabilistic resources.

    If you read Eric Anderson’s paper, you’ll find that the Avida folks made simplifying assumptions at every turn that were designed to produce results consistent with a conclusion that had already been reached.

    As far as LS-DYNA is concerned, I’m working with it on a daily basis. Some simplifying assumptions can be made, and others cannot. With experience, and comparing simulation results with empirical test results, it is possible to learn, over time, which simplifying assumptions can be trusted and which cannot.

    There is no such empirical verification methodology for bio-evo sims, which remain speculative, pie-in-the-sky, irrelevant, contrived flights of fancy of no significant relevance to the reality they attempt to represent.

  8. 8
    Granville Sewell says:

    Gil’s point is simple and brilliant. If you are trying to model how molecular errors could accumulate to produce animals and plants, why assume that the only molecules in the cell subjected to errors are the DNA? If you really want to see what unintelligent forces alone can accomplish, through accidents, you should assume accidents can occur in other parts of the organism.

  9. 9
    90DegreeAngel says:

    Perhaps a simple solution is to introduce into your simulation random accidents that kill off small populations within the program. A simulated “volcanic eruption” of sorts. However, this would still be in the program and thus part of the simulation. Perhaps one should instead throw a part of the computer that is doing the simulations into a volcano!

  10. 10
    Nakashima says:

    Prof Sewell,

    Unfortunately, that is not Mr Dodgen’s point. Mr Dodgen’s point is that there is no boundary between the simulation and the underlying reality. As a result, he recommends perturbing the hardware to introduce a perturbation into the data. There are more effective random number generators than leaving a lump of uranium on top of the server.

  11. 11
    Dave W. says:

    Mr. Dodgen,

    With your focus on Anderson’s insulting (not just to the Avida researchers, but to anyone who reads it) paper – with his clear lack of understanding of hypothesis testing – tell me again who is looking at the forest, and who the trees?

    I notice, for example, that you neglected to state what any simulation purports to do. Instead, you have a laser-like focus at one particularly bad “rebuttal” to one simulation.

    The real point is that if all a simulation purports to do is examine genetic mutation and natural selection, then the proper criticisms are that the mutations aren’t random with respect to fitness, or that selection is artificial (done outside the simulator) with respect to the fitness landscape. Anderson makes neither of those points (he doesn’t even try to), and so misses the point of Avida entirely (the point being that Behe claimed that IC structures are impossible for mutation and selection to generate, not just very unlikely or only under favorable assumptions – impossible).

    The forest: evolutionary simulators only test those aspects of biological processes that they purport to test, of which you have correctly described none.

    A tree: Eric Anderson’s insulting 2004 paper regarding Avida.

    (Yes, I am fully aware that Behe has since backtracked on whether it is possible for IC systems to evolve, but it’s hardly appropriate to fault the Avida researchers for not reading his mind.)

    Mr. Sewell: put in your terms, Mr. Dodgen’s point is nothing less than that without simulating the entire universe, down to every single quantum event, no conclusions can be reached about any sub-set of conditions within the universe. In other words, you are demanding that any simulation of any process take into account all possible relevant conditions in order to have any relevancy to the real world.

    If I create a simulation of a paint mixer, it is possible that where Mars is in its orbit will have some effect on actual paint mixing, but you would claim that since the simulation didn’t take Mars into account, the results should be discarded.

    Were he making such a “simple and brilliant” point, it would invalidate the conclusions Mr. Dodgen himself is reaching regarding the simulations he says he is running.

  12. 12
    IRQ Conflict says:

    Dave W, at what point would you suggest that it is enough to say this is as close to a realistic model as is reasonably plausible?

    In other words, where and when does/doesn’t the value of external stimuli begin and end with regards to influencing the outcome of the simulation.

    It is my understanding that in a simulation such as this anything that has physical contact with the material in question would have an effect on the outcome of the organism. i.e. the atmosphere.

    The gravitational effect that the moon has on the tides etc. Was the moon even a factor at the beginning of earths early history? If so, how much (if any) effect did it have?

    Your paint mixer analogy is good, however it is based on observed and observable science. Not on what we don’t know that is historical science.

    I hope that made sense?!

  13. 13
    PaV says:

    Fred Hoyle,using a path-integral formalism, came to the same negative conclusion about RM+NS : that is, when you take into account the greater likelihood of encountering harmful mutations versus beneficial ones, RM+NS can really do very little. His conclusion was the same as that of Behe: a couple of steps either way, and no more.

    What about the path-integral formulism? Well, it’s the formulism that Richard Feynmann won a Nobel Prize for in quantum mechanics. When you analyze the way photons interact with reality, there is an infinite number of ways in which light can so interact; or, put another way, there’s an infinite number of paths that light can take in getting from point A to point B. The formulism, which takes into account every possible path a photon can take and sums over all these possible ways, allows you to arrive at the correct answer without having to actually sum each and every possible path by generating equivalent integrals and then evaluating these integrals over the interval of concern. Thus, when Hoyle uses such a formulism, implicit in this method is an understanding that ANY possible route from let’s say, point A to point B, is included when evaluating the properly constructed integrals. (In Hoyle’s approach—and this is strictly from memory—he starts with the zeroeth generation, and then runs the integral through an infinite number of generations) The bottom line is that no matter what ‘simulation’ you run, all’s it does is to, roughly speaking, do the ‘summing’ of all the possibilities over the interval of interest. Thus, when Hoyle evaluates his integrals, he’s done everything that any possible simulation could come up with. And, his bottom line is that RM+NS can only move a genome a couple of nucleotides/amino acids one way or another.

    Now, if Hoyle had excluded harmful mutations—or taken unrealistic values for their frequency—he would have come up with a different answer, one that would have necessarily been more favorable to Darwinism. So, all you out there, here’s your choice: use realistic values for important parameters, or, come up with the complete improbability that RM+NS can do anything significant. Now, for those who are ‘true believing’, ‘card carrying members’ of the Darwinist Dream Squad, I know what they’ll prefer: unrealistic parameters.

    This is just another way of saying what Gil has already said.

  14. 14
    lamarck says:

    Gil is correct, computer simulations are all wrong. Futher, they are wrong because no one knows how evolution works. Recently e.g., it’s been found that many “silent mutations” stop and start protein making in new ways. The point is all of these silent mutations and junk dna might very well all have a function. There might not be neutral mutations at all. Until enough is known, you can’t run an accurate simulation.

    If biologists want to confirm that natselec/ranmut’s cause net info increase and large changes, speciation etc, they should start by realizing there’s no evidence for either of these points. So if they have a computer simulation confirming current neodarwinism, it’s automatically wrong.

    But in the meantime how about making it legal to say this in certain state’s biology classrooms in this country? If you persist in teaching kids accepted observed facts you will be hauled out at gunpoint thanks to Eugenie Scott and Ken Miller’s bizarre agenda.

  15. 15
    George L Farquhar says:

    Gil,

    Thus, a realistic simulation would allow the program, OS, and hardware to be affected in a random fashion, just as a real organism’s ability to survive and reproduce would be affected randomly by mutational interference.

    I’ve noted that you could achieve your goal of a “realistic simulation” in a pratical way via Virtual Machines.

    What do you say? Has the impratical become pratical?

    Could your work with LS-DYNA have it’s accuracy improved by running it in a virtual machine that was itself having a simulation of being dropped running against it? Have you suggested such in your day-job? If not, why not?

    Again, if we take “more realistic” to be “more accurate” this seems to me to be something that can be determined by experiment.

  16. 16
    niwrad says:

    In evolutionary computer simulation of biological systems there is sort of dichotomy: the computer here, the simulator program there. The latter should test the consequences of random changes in the events space of the simulator software, while the former is always granted perfectly working and without failures.

    In the reality of living biological systems such dichotomy doesn’t exist. Nowhere there are parts of the system granted to be a priori ok. Failures may happen everywhere and whenever. Unfortunately in the physical world there is no place free from entropy-disorder (both in its thermodynamics and information sense). Complex systems designers and administrators well know that these systems need a continue injection of intelligence (in a sense the inverse of entropy) to start and continue working ok. Hence designers and administrators usually don’t believe the naive evolutionary idea of complex systems arising and increasing complexity by means of randomness (that is another name of entropy itself). Whether Darwinian evolution were true we would have that increasing entropy decreases entropy! A clear logical antinomy.

    So Gil’s “don’t test the part, rather the whole” argument is reasonable, insofar a biological system is a computer (or even is a network of computers, being the biological cell sort of information processing and information exchanging system) and itself must be tested inside the simulation.

  17. 17
    Joseph says:

    How can a computer simulate biological evolution when we don’t understand biology enough to model it?

    For example we don’t even know what makes an organism what it is.

    We do not know what DNA sequence/ sequences are responsible for eyes/ vision systems.

    So the bottom line is before you can simulate something you have to have an understanding of it.

  18. 18
    EvilSnack says:

    The real weakness of computer simulation of evolution is that the effect of a given mutation at a given point in the simulated genome is often (if not always) defined in advance in the simulation. In real life, we really don’t know what effect changing this A, C, T, or G to something else will have on the resultant organism, until the change actually happens and the affected organism develops.

    Granted, this simplification is presently necessary because simulating the development of an actual organism from an actual sequence of DNA is a problem for which our current level of computer resources is hopelessly insufficient, and probably will be for the better part of a century.

    The upshot of all this is that the results of a simulation are not to be taken as conclusive except in extremely limited circumstances.

  19. 19
    Nakashima says:

    Messrs Lamarck, Joseph, EvilSnack,

    Here we have to distinguish between evolution as an abstract concept and the simulation of actual biology. In the spirit of some other comments on other recent threads, a GA is not simulating evolution, it _is_ evolution.

    I agree that we are still a long way from the day when every atom in a cell is simulated down to the quantum mechanical effects. That is why we have to work at several scales at once and with different (explicit, agreed upon) simplifying assumptions.

    This is why work like MESA and Mndel’s Accountant is used by ID friendly scientists. Neither of them comes with instructions to shake the hard drive while running the program.

  20. 20
    ScottAndrews says:

    If we were able to simulate actual macroevolution, the implications would be mind-bending. Living things, whether designed or evolved, demonstrate extraordinary innovation.
    With all of our technology and intellect, we only strive to imitate them.
    But if we could, for example, throw a bunch of cars into a simulation, enable them to reproduce and mutate, program them to select for increased mileage, lower maintenance, and better safety, perhaps they could outperform our engineers.

  21. 21
    Dave W. says:

    IRQ Conflict, your question makes perfect sense, but it’s a question for Mr. Dodgen, who claims to be running simulations and drawing relevant conclusions from them. The primary complaint here seems to be that we don’t know what we don’t know, so we don’t know what’s relevant in a simulation and what’s not, but that’s true of all disciplines, and not just evolutionary biology.

    So without simulating the whole universe, how does Mr. Dodgen get any good data at all regarding whatever it is he’s simulating? Simple: he makes simplifying assumptions, and leaves out whatever he thinks won’t affect the real-life system he’s simulating. He claims to be testing some of those assumptions, but the fact is that there are a lot more things he hasn’t even thought to test that he’s simply left out (for practical purposes, of course).

    Really, the solution to Mr. Dodgen’s specific complaints is to have a simulation which just randomly kills off some percentage of the population. That would simulate not only broken cellular machinery, but falls from cliffs, cancer, axe murderers, lightning strikes, etc. I’m sure that (for example) Wesley Elsberry could work that into his WEASEL sims as an extra variable in no time flat, but I would also guess that he’d say that the same effect could be had by simply lowering the mutation rate.

  22. 22
    Dave W. says:

    ScottAndrews wrote:

    But if we could, for example, throw a bunch of cars into a simulation, enable them to reproduce and mutate, program them to select for increased mileage, lower maintenance, and better safety, perhaps they could outperform our engineers.

    That’s already been done for radio antenna design and electronic/logic circuits. The results have often included odd bits and pieces that seemed like they should be non-functional, but when snipped out caused performance to suffer.

  23. 23
    ScottAndrews says:

    That’s already been done for radio antenna design and electronic/logic circuits. The results have often included odd bits and pieces that seemed like they should be non-functional, but when snipped out caused performance to suffer.

    That’s consistent with observed microevolution. Tweak something randomly, and you might get a result that increases survivability under specific circumstances.
    I’m looking for something more. Evolution goes outside the box. It invents lungs, wings, eyes, and even brains that in turn invent more things.
    Therefore it’s entirely reasonable that a simulation that mutates cars and then selects for safety, manufacturing cost, maintenance, and fuel efficiency should produce the vehicle of the future. I don’t mean just a varied shape – I’m talking Knight Rider. Flying, bulletproof, solar-powered cars. Eventually they should overtake and kill us.
    Okay, that’s a bit much. I’ll settle for the solar power and the flying.
    The known accomplishments of mutation and design surpass every man-made design. Why wouldn’t we put it to work?

  24. 24
    ScottAndrews says:

    That’s “The known accomplishments of mutation and selection. Always typing when I’m in a hurry.

  25. 25
    Cabal says:

    One of the better examples of the use of evolutionary algorithms must be in the design of microwave antennas. I guess somebody may provide a link, or maybe a little googling will do. Another interesting use of computer programs is in the design of multi-band antennas for cell phones.

  26. 26
    ScottAndrews says:

    Here’s a NASA link.
    Isn’t it funny how this brings us right back to micro- vs. macro-evolution. By applying brute force to execute a random search and then refining its results the process improves upon an existing design.
    But the process has limits. An outside intelligence is required to coerce antenna “reproduction,” or else nothing will happen. But more importantly, it produces nothing but variations on an existing design.
    Through a simple process of mutation, replication, and selection, evolution creates self-replicating, self-repairing machines. Then it gives them eyes, ears, wings, tentacles, and brains.
    Given that it works so well and outperforms human designers*, shouldn’t we seek to replace intelligence with evolution as the preferred means of innovation? If not, why not?
    *Yes, it takes forever, but computers can reduce reproductive cycles from months to milliseconds. And the elegance of the so-called designs is unparalleled.

  27. 27
    R0b says:

    Gil:

    The point of the example is that critical aspects of a real-world system cannot simply be ignored, and victory declared on the basis of simplifying assumptions chosen for the purpose of producing a desired outcome.

    Certainly, conclusions drawn from a simulation shouldn’t exceed what is warranted by the fidelity of the model. You continue to make that point, which nobody disputes, and in the meantime you haven’t disputed the point made by your critics. So it seems we’re all on the same page.

    “Simulator” is a misnomer for programs like Avida and ev. They are instances of random variation and selection, not simulations of it. (I think the creators of Avida agree with this. I once had a professor who threw fits when we called our neural nets “simulators”.) They simulate biological evolution only in a ridiculously weak sense.

    The idea behind programs like Avida is to provide insights into evolutionary principles. You could dispute the applicability of such insights to biology, and I would probably agree with you. But unwarranted conclusions from lo-fi models is a problem on both sides of the ID divide. Does modeling the evolution of the bacterial flagellum as random combination tell us anything about reality?

    The rational solution is to tone down our conclusions and/or incorporate more detail into our simulated environments. But your solution is to tweak the environment in which the simulation runs. Do you understand why this doesn’t make sense?

  28. 28
    Nakashima says:

    Mr ScottAndrews,

    As with other methods of design optimization, evolutionary methods can produce results that exceed humans, the question might be – is this cost effective for a particular application? EC methods usually require testing a large number of examples. This may be cost or time prohibitive.

    In other cases, it can shift the work of the engineering staff to a higher level of abstraction. Instead of creating a particular design, the human now is concerned with the quality of the fitness function that scores any design.

    In this sense, EC is an intellectual amplifier for human directed processes, not a replacement. To use a physical analogy, humans are no longer the galley slaves pulling the oars, they are freed to be the captains, sailing the ship.

  29. 29
    ScottAndrews says:

    Mr. Nakashima (Formal, but okay)

    As with other methods of design optimization, evolutionary methods can produce results that exceed humans…

    The example of the antenna appears to demonstrate that. A simulated evolutionary process improved upon the existing design in ways that a human designer might never have thought to test.
    But it neither demonstrates nor attempts to demonstrate that such a process could not only improve upon a design, but create one.
    I can’t fault any simulation for not evolving a new virtual life form or a better car, because the technology doesn’t exist yet.
    But I find it amusing that NASA’s page hints that the experiments mirror how life evolved (references to crocodiles and dragonflies) when all they can demonstrate is improvement upon an existing design.

  30. 30
    Nakashima says:

    Mr ScottAndrews,

    You can look up the “Humies” (Human Competitive Awards) given out at the recent GECCO confernces for more examples. The rules are pretty specific that designs have to exceed human performance is certain ways, for example be patentable.

    There are other examples of antenna design that just start with “wire”, not any working or complex antenna design. It is not necessary to start with an existing human design in the first generation.

  31. 31
    ScottAndrews says:

    Mr. Nakashima

    You can look up the “Humies” (Human Competitive Awards) given out at the recent GECCO confernces for more examples. The rules are pretty specific that designs have to exceed human performance is certain ways, for example be patentable.

    An improvement need not be innovative to qualify for a patent. I looked over several of the award winners and saw just that – improvements, not inventions. It confirms what you had said earlier,

    EC is an intellectual amplifier for human directed processes, not a replacement.

    These are designed constructs that employ technology to do some heavy lifting and search out a number of possible solutions for the ideal one. They love referring to “evolution,” but their simulations never do what evolution claims to – invent something new.

  32. 32
    T M English says:

    Nakashima-san,

    Outstanding comments at 26. There are now many engineering problems that humans know how to make amenable to solution by evolutionary computation, but do not know how to solve analytically. The two types of knowledge are fundamentally different in kind.

    Humans indeed bias and constrain the evolutionary search for good solutions to engineering problems. The prior knowledge we use to do so is often generic and, in a sense that is intuitively clear but difficult to formalize, inexpensive to obtain. Good solutions are typically high in value relative to the cost of running evolutionary algorithms on computers for long periods of time.

    For example, I know very little about the dynamics of solar weather, yet I applied an evolutionary algorithm to obtain what was, at the time, by far the best predictor of annual sunspots counts. My algorithm selection reflected my knowledge of certain modeling principles, my belief that the sequence of counts was chaotic, my belief that a class of models was good for chaotic time series prediction, and my knowledge of what works and what does not in evolution of models in that class. After consuming several weeks of computer time in execution of the algorithm, I had an outstanding predictor of sunspots counts. And no person knows, or is likely to figure out, what the model, a combination of more than 20 thousand smaller models, “knows” about sunspots. The information I added to the evolutionary search is of a type that I can use, with slight and rapid adaptation, for many problems. The information that came out of the search is of a type I would never have obtained with my unaided intellect.

    The distinction in kind and value of the information used in algorithm selection and the information obtained through algorithm application is missing from all information-theoretic analyses of search. Not all bits are created equal. There is a sort of “free lunch” when valuable information is purchased by supplying cheap information to a cheap computational process.

  33. 33
    T M English says:

    That was Nakashima at 28, not 26.

  34. 34
    Nakashima says:

    Mr ScottAndrews,

    They love referring to “evolution,” but their simulations never do what evolution claims to – invent something new.

    Evolution claims to change allele frquencies over time. If you don’t think what is happening in EC is invention, please explain.

  35. 35
    Joseph says:

    Nakashima:

    Evolution claims to change allele frquencies over time.

    ASnd that is what I meant by equivocation and evolution

  36. 36
    Nakashima says:

    Mr Joseph,

    I am being the opposite of equivocal. I am being explicit about which of several possible meanings I am using. Not that it matters in this case, as none of the definitions you refer to in your blog entry match the claim that “evolution claims to invent”.

  37. 37
    ScottAndrews says:

    Mr. Nakashima:

    Evolution claims to change allele frquencies over time. If you don’t think what is happening in EC is invention, please explain.

    I’m sure there are hairs to split over the meaning of the word “invention.”
    A simulation based on mutation and selection might improve upon an antenna, but it won’t figure out how to manipulate and receive radio waves to send information.
    Perhaps that’s a way to falsify ID. Generate irreducible complexity or CSI in a simulation of random changes and selection. I’d be surprised and disappointed if no one is already trying.

  38. 38
    Nakashima says:

    Mr ScottAndrews,

    Perhaps that’s a way to falsify ID. Generate irreducible complexity or CSI in a simulation of random changes and selection.

    That is not how it is going to happen, I think. Generating CSI is just attributed to the active information content of the EC system. While active information has been defined, we don’t know how to account for it, how to distinguish the relative importance of the fitness function, the clock, the history, and the random number generator.

  39. 39
    GilDodgen says:

    Dave W.: …the researchers that Eric Anderson freely insulted only purported to be working on simulating natural selection and genetic mutations, not on the entirety of “biological evolution.”

    Oh really? Check again. The “researchers” begin their paper with obligatory genuflecting to Darwin (“one of the greatest scientific achievements of all time”), and the entire purpose of the exercise was to demonstrate that the Darwinian mechanism in biology can produce the irreducibly complex systems identified and elucidated by Behe.

    Furthermore, Anderson did not insult anyone; he simply pointed out that the entire project was contrived from the beginning to produce a desired result, and assumed in advance what it attempted to demonstrate.

  40. 40
    PaV says:

    [32] T M English:

    In using your EC to find a predictor for sunspot numbers, did you, or did you not, use the actual observations of sunspots in the past in evaluating the efficiency of your EC? I suspect your answer is that you did, indeed, use such observations. That then became your ‘target’, and you simply waited until the program reached the ‘target’. But evolutionists insist that evolution is ‘undirected’, i.e., that evolution is NOT looking for any particular ‘target’ beforehand. Thus, while you might call your algortithm ‘evolutionary’, it does not, in fact, imitate evolutionary processes. It is no more than a blind search through a search space that is amenable to such a search. Your wrote, after all, that you “consumed” several weeks of computer time in finding the desired predictor.

  41. 41
    Nakashima says:

    Mr PaV,

    Evolution happens whether the selection is guided or unguided. I think Darwin in OoS spends some time talking about breeding animals for this reason.

    In this particular case, using the data as part of the fitness function does ot make the data set a target, in the direct sense that most people have from discusing WEASEL. The organisms in the population are models, sets of parameters or actual programs and parameters. In some paradigms, some of the data might be available to a model, while the rest is held out to test fitness. In other paradigms the models never see any of the data.

    At no point is the search blind. It always has some of the history of previous samples available.

  42. 42
    Dave W. says:

    Mr Dodgen:

    Oh really? Check again. The “researchers” begin their paper with obligatory genuflecting to Darwin (“one of the greatest scientific achievements of all time”)…

    For one thing, the researchers made a true statement: Darwin’s theory “is widely regarded as one of the greatest scientific achievements of all time.” If you can present evidence that it is not “widely” regarded as such, regardless of your own feelings about it, please do so.

    Second, your mimicking of Anderson’s ad hominem attack on the researchers is a waste of my time and yours.

    …and the entire purpose of the exercise was to demonstrate that the Darwinian mechanism in biology can produce the irreducibly complex systems identified and elucidated by Behe.

    Exactly correct, which is why I stated (also correctly) that they did not purport to be simulating all of “biological evolution.” If you think that “biological evolution” is synonymous with “random mutation and natural selection,” then you need to learn more biology.

    Furthermore, Anderson did not insult anyone; he simply pointed out that the entire project was contrived from the beginning to produce a desired result, and assumed in advance what it attempted to demonstrate.

    You invoked the same ad hominem attack as Anderson, but Anderson went further, suggesting that the reason the paper was published in Nature was because of the authors’ “genuflecting” to Darwin. This sort of argumentation is massively insulting, because it is clear that Anderson must have thought his readers to be morons for him to consider that a good point to make. Anderson further compounds the insults by asserting motives for the authors for which there is no evidence, and by mocking the real hypothesis testing that was being accomplished.

    But really, the greatest insult provided by Anderson to all his readers (you included, Mr. Dodgen) is that he never criticizes the aspects of evolution which were actually being tested, but then in the conclusion he flatly claims that Avida failed those tests. He must think us all very stupid, indeed, to think that anyone would be fooled by such an empty argument.

  43. 43
    lamarck says:

    Mr. Nakashima,
    “Here we have to distinguish between evolution as an abstract concept and the simulation of actual biology. In the spirit of some other comments on other recent threads, a GA is not simulating evolution, it _is_ evolution.

    I’ll assuming GA means genome analysis? Why would evo biologists be developing a type of evolution that has nothing to do with biology? If they were interested in mirroring what is known of biology to gain understanding, wouldn’t they be striving for a program that allows for no macro evolution, and only small changes to existing structures, as well as larger micro evolution but only in brackets, while at the same time their computer genome would start out ordered and move towards chaos? Also wouldn’t they need to input for reducing phyla and an inverted tree of life?

    IE. aren’t Neodarwinian evo biologists designing their simulations backwards?

  44. 44
    GilDodgen says:

    The bottom line is that the Darwinian mechanism — which is based on complete and utter 19th-century ignorance of the underlying information-based nature of biological systems — is perfectly inadequate to explain what is observed, beyond the trivial and obvious.

    It’s equivalent to trying to turn lead into gold through chemistry. It doesn’t work that way. It is a categorically erroneous explanatory approach.

    Neo-Darwinists are latter-day alchemists who have been mysteriously immunized against following the evidence where it leads. What an embarrassment, when the science is screaming design from every corner.

  45. 45
    Dave W. says:

    The bottom line, Mr. Dodgen, is that neither you nor Mr. Andserson have critiqued the mechanisms under test in the simulations. You instead fault the simulations for failing to address aspects of evolution and/or biology that they were never intended to address. And when that point is made, you dismiss what you fail to critique as inadequate, trivial or obvious.

    Mr. Behe stated that IC was impossible for mutation and selection to accomplish. The EQU test was set up to test just those aspects, and when it succeeded, Mr. Anderson argued not one bit against either the mutation or selection processes, but instead invented a raft of transparent objections which all missed the point, and then had the gall to claim that Avida failed. Or that it accomplished something trivial which Mr. Behe claimed to be impossible. Yes, according to Anderson, what Avida did was trivial and successful and impossible, all at the same time.

  46. 46
    Nakashima says:

    Hi Mr Lamarck,

    Sorry I was unclear, GA means Genetic Algorithms. The field is based on taking the basic insights of evolution and applying them to optimization problems and other uses in computers.

    Dr John Holland is one of the pioneers of the field, and in his book Adaptation in Natural and Artificial Systems he proves what is called the Schema Theorem. The Schema Theorem explains why, in the simplified population genetics of a computer simulation, the fitness of the population will increase over time.

  47. 47
    Nakashima says:

    Mr Dodgen,

    On the contrary, it is a fairly sophisticated hiding of implementation details. Evolution works, no matter if you implement selection with natural or artificial selection. Evolution works, no matter if you implement variation with epigenetics or DNA mutations.

    In your last sentence, I think you might mean “when the evidence is screaming design” not “the science”.

  48. 48
    ScottAndrews says:

    Nakashima:

    Evolution works, no matter if you implement selection with natural or artificial selection. Evolution works, no matter if you implement variation with epigenetics or DNA mutations.

    You’ve compared the success of natural selection to that of artificial selection. What would you cite as the single most impressive product of evolution via artificial selection?

  49. 49
    RDK says:

    You’ve compared the success of natural selection to that of artificial selection. What would you cite as the single most impressive product of evolution via artificial selection?

    You’re obviously reading into things that aren’t there. Go back and read what he wrote:

    “EVOLUTION WORKS, no matter if you implement selection with natural or artificial […]”

    All he said was that it works; he said nothing about the size or amount of the results of success.

  50. 50
    ScottAndrews says:

    RDK @48:
    I never thought or said that Nakashima overstated his claim.
    If I were to split hairs, I might point out this statement: statement

    Evolution works, no matter if you implement selection with natural or artificial selection. Evolution works, no matter if you implement variation with epigenetics or DNA mutations.

    Evolution is change, and we’ve known that such change occurs for thousands of years. To say that it works, regardless of the mechanism, does not require any science. (The sun rises, whether it’s the earth rotating or it gets pulled by a chariot, or something else.)
    Do we know which mechanism caused generations of pawed mammals to slowly develop hoofs, or do we name several possibilities and say that it must have been one of them?

  51. 51
    T M English says:

    PaV (40):

    The key Darwinian principle is that variety, heredity, and fecundity yield demography. Many evolutionary algorithms more or less reduce demography to fitness, but the one I applied does not. See Nonlinear Combination of Neural Networks from a Diverse and Evolving Population [warning: PDF renders slowly] for a concise description of the actual approach.

    What is relevant to your comment is the success in prediction of a chaotic time series generated by a process that (pseudo-)randomly changes from one unobservable state to another. The dynamics in the two states are radically different, and it is quite unlikely that any one model would capture the dynamics of both states. I would argue that there is, at least as a practical matter, no clear-cut target for the evolutionary algorithm. Pay special attention to the concept of niche implicit in the ranking of models. A model with relatively high error can be selected as a parent because its errors are dissimilar from those of other models. The evolutionary algorithm is not “trying” to make all of the models in the population fit the data as well as possible.

  52. 52
    T M English says:

    The missing link.

  53. 53
    Nakashima says:

    Mr ScottAndrews,

    You’ve compared the success of natural selection to that of artificial selection. What would you cite as the single most impressive product of evolution via artificial selection?

    Corn.

  54. 54
    ScottAndrews says:

    It seemed like a great question at the time. But it really wasn’t.

  55. 55
    Nakashima says:

    Mr ScottAndrews,

    Don’t feel bad. Some of the best proofs of God involve picnics, sweet corn, fresh butter, salt and pepper!

Leave a Reply