652 thoughts on “Evolving Wind Turbine Blades”

  7. Patrick:
    Frankie,

    I’ll just leave that to stand on its own.

    That is all you can do. Cantor obviously didn’t know that infinity is a journey.

  10. Let set A = {0,1,2,3,4,5,…}
    Let set B = {0,2,4,6,8,…}
    Let set C = {1,3,5,7,9,…}

    According to Cantor they all have the same cardinality, ie the same number of elements. However this leads to a logical contradiction as A-B=C and not 0. If A and B had the same number of elements subtracting B from A shouldn’t leave an infinite set behind.

    Just sayin’

  12. phoodoo,

    Allan Miller: But that isn’t the case. Sometimes the fittest organisms get hit by an asteroid, eaten anyway, crushed by a log …Fitness is not survival.

    phoodoo: You have no criteria by which to say the fittest didn’t survive, unless you change the definition of fitness.

    Yes I do. It’s the one with the higher mean offspring number. It’s dead simple phoodoo, and I’ve said it a million times. But note … [phoodoo! phoodoo! over here! Are you even fucking listening?] … note that the mean is a mean of a probability distribution. Therefore, on any given run – at any point in any real population – the one with the greater mean expected value must have a nonzero chance of NOT being the one currently winning. It might even go extinct. This is basic – couldn’t-be-much-more-basic – probability theory. Therefore you cannot use the result of one run to determine fitness. The survivor of one contest is not the fittest, it’s the survivor of one contest and may be more, less or the same, in terms of its intrinsic fitness.

    If you can claim that the fittest didn’t survive, because they were unlucky, then I can claim that the fittest ALWAYS didn’t survive, because they have always been unlucky. The fittest never survive. How can you refute that, if you can claim sometimes they don’t.

    If the fittest never survive, they almost certainly don’t have the higher mean offspring number. If you think the fittest NEVER survive, good luck to you. I would expect no more from someone who thinks that one M&M in a thousand will never be picked regardless how many random trials you perform.

    Unless we have a definition for fitness other than the one that survives best, you are painted into a corner.

    I have NEVER defined fitness as ‘the one that survives best’. I have ALWAYS defined fitness relative to mean offspring number. I called this ARGUMENT #1 several posts ago, to save me having to type it. Yet here I am typing it again, because apparently I never really typed it on any of the other occasions. I am happy to repeat it, so you don’t miss it AGAIN.

    You are simply being obtuse. Which is fine, I am happy to keep saying it for ever, if you are stupid enough to keep saying I’ve said something entirely different to what I have said, despite there being a written record.

  14. Allan Miller,

    You are the one not listening Allan, or at least not able to comprehend.

    You just made the claim that SOMETIMES the most fit one doesn’t survive best-because of luck perhaps. Yet overall, you are convinced the most fit will prevail. And how do you know this? Because the definition is the one which prevails most often! See the problem yet? Still not paying attention?

    There is no magical certificate that an organism gets which says it is most fit. You making an “expected value” for an organism does not deem it most fit. If sometimes the luckiest wins, rather than the most fit, then you have no way of saying that its not the luckiest that frequently wins. Does the luckiest win 10 percent of the time? 30%? 80 %? If its the luckiest genotype winning (which you admit happens) it could be that the luckiest won 80% of the time, in which case you are now forced to call that not the luckiest, but rather the fittest, because that is how you expected value makes predictions. Your circle CAN’T be broken, because the one who is winning most often HAS to be the fittest.

    Let me label this as Allan Being Obtuse Number 1, so I don’t have to type it again. When your definition of the one that survives best is the fittest (it changes nothing by saying the one you EXPECT according to the formula you chose), then the fittest is the one that survives best. It doesn’t matter if its because the “real world” fitter one got hit by lighting ten generations in a row. The gimp with the shriveled hand, and the rotting kidneys is the most fit, by virtue of the fact that they weren’t the ones struck by lightening. You getting to magically decide the role luck plays, the percentage you allow, the inherent assumption in your definition, that the fitter wins by percentages, is fabricated by your circular definition.

    Its a self-created assumption that your theory is correct, by ensuring that the definition says the theory is correct.

    Allan Being Obtuse Number 1

  16. Mung:
    LoL. I finally started watching the video. It’s design, you morons.

    Mung’s hilarious when he drunk posts. 😀

  17. If evolution can supposedly develop a bird’s wing or a whale’s tail, could it be used to develop a wind turbine blade? I’ve written software that essentially makes wind turbine blades live, mate, and die.

    Fabulous!

    Let’s start with wind turbine blades and evolve them to prove that wind turbine blades could have evolved.

  21. Mung:

    Fabulous!

    Let’s start with wind turbine blades and evolve them to prove that wind turbine blades could have evolved.

    Better yet – let’s start with human designed wind turbine blades and see if evolutionary processes can make them more efficient.

    Hey, whatta ya know! This evolution stuff works!

  25. Mung: If evolution can supposedly develop a bird’s wing or a whale’s tail, could it be used to develop a wind turbine blade? I’ve written software that essentially makes wind turbine blades live, mate, and die.

    Fabulous!

    Let’s start with wind turbine blades and evolve them to prove that wind turbine blades could have evolved.

    This is so dumb it’s supernatural.

    It’s evolution, not the origin of life. It’s not supposed to simulate a process whereby something that can reproduce comes into existence, it’s supposed to simulate how something highly adapted came to be highly adapted.

    I’m however surprised to see that you think that once something is subjected to cumulative selection, it’s evolution is pretty much guaranteed. Heck, even I wouldn’t go that far. Mung the Evolutionist champion.

  26. Mung: “I’ve written software that essentially makes wind turbine blades live, mate, and die.”

    Tell me how you would make a simulation of evolution only, without first having to program into it a population of reproducing organisms.

  27. phoodoo,

    You just made the claim that SOMETIMES the most fit one doesn’t survive best-because of luck perhaps. Yet overall, you are convinced the most fit will prevail. And how do you know this? Because the definition is the one which prevails most often! See the problem yet? Still not paying attention?

    Wrong. The definition is not the one which prevails most often. It is the one with the higher mean fitness (mean offspring of carriers). Having higher mean fitness will cause you to prevail more often.

    Let me label this as Allan Being Obtuse Number 1, so I don’t have to type it again. When your definition of the one that survives best is the fittest

    Except that it isn’t, it is the one with the higher expected value for mean offspring numbers. For the million and first time.

    (it changes nothing by saying the one you EXPECT according to the formula you chose),

    The word ‘expect’ in ‘expected value’ is misleading you to think that someone is doing some expecting. It isn’t. The expected value could equally be called the ‘convergent value’ – the value that will be converged upon over many trials. It happens that a genotype increasing with a higher exponent than another will tend to outcompete it. You don’t define that genotype’s propensity as ‘more likely to outcompete’, it is a consequence of its propensity.

    then the fittest is the one that survives best.

    Ultimately, fitter genotypes – those with higher mean offspring numbers – will tend to fix more often than those with lower. But the definition of fitter genotype is not “will tend to fix more often”. Having more offspring has that effect, over the long term. You seem to think this needs proof So what other possibilities are there? What needs to be demonstrated to satisfy you that having more offspring (on the average) will tend to make a genotype more common (on the average)?

    Its a self-created assumption that your theory is correct, by ensuring that the definition says the theory is correct.

    Nah. You could conceivably make an observation that genotypes producing more net offspring over the long term typically go extinct more often than those producing fewer, instead of fix more often. This would be somewhat counterintuitive, but it is a possible observation. That observation is not made, but unfalsified does not mean unfalsifiable. Though it may pain you to admit it, the fittest genotypes are not typically defined as ‘more likely to survive’. But they certainly are more likely to survive, as a matter of fact. Do you have a reason to doubt that reproductive differentials exist and have effects on genotype frequencies, beyond assumed circularity?

  28. A couple of folks have expressed amazement that I could carry on for dozens of posts without ever having watched the video. There’s no magic secret here people. Nothing miraculous about it. Here’s how it’s done.

    Don’t make statement about the content of the video.

    So, for example, my first post in the thread. Video not required.

    My first comment from page 3. Video not required.

    My first comment from page 4. Video not required.

    My first comment from page 5. Video not required.

    My first comment from page 6. Video not required.

    My first comment from page 7. Video not required.

    See how it’s done?

    It’s amazing to me that someone would bother to count the number of my posts in the thread without bothering to look at their content [or lack thereof] before accusing me of arguing without having watched the video.

    Richardthughes: So you’ve been arguing before watching?

    Yeah, so? Relevance?

  29. Allan Miller: I would expect no more from someone who thinks that one M&M in a thousand will never be picked regardless how many random trials you perform.

    ok, so one M&M out of 6,227,020,800. 🙂

  30. Rumraket: Tell me how you would make a simulation of evolution only, without first having to program into it a population of reproducing organisms.

    GA’s are not simulations of evolution.

  31. Allan Miller: You don’t define that genotype’s propensity as ‘more likely to outcompete’, it is a consequence of its propensity.

    Expected propensity?

  33. Mung,

    Lovely. But your first comment on page 2 included

    A very powerful method for exploring searching large phase spaces.
    Fixed that for you.

    which is about the video.
    Yes, I do “see how it’s done”.
    LMAO

  34. DNA_Jock: which is about the video.

    But a very powerful method for exploring large phase spaces, such as … the travelling salesman problem (which is a real world problem).

    Right. It’s about the video, it’s not about the method. petrushka’s comment should not have been taken to refer to anything other then the video. Certainly not anything like the TSP. One could not possibly offer a legitimate response to petrushka’s comment without have watched the video.

    You’ve certainly got the spin down. Now work on the shape.

  35. Mung: But a very powerful method for exploring large phase spaces, such as … the travelling salesman problem (which is a real world problem).

    Right. It’s about the video, it’s not about the method. petrushka’s comment should not have been taken to refer to anything other then the video. Certainly not anything like the TSP. One could not possibly offer a legitimate response to petrushka’s comment without have watched the video.

    You’ve certainly got the spin down. Now work on the shape.

    OMG, you really cannot help yourself, can you? Only in the video is the blade-optimization method described.
    Why on earth did you feel the need to use ellipses to shorten petrushka’s comment today, when your original post quoted petrushka thus:

    But a very powerful method for exploring large phase spaces, such as the space of all possible blade shapes, or the travelling salesman problem (which is a real world problem)

    You’ve certainly got the spin down.
    ROFLMAO

  37. keiths:
    That’s pitiful, OldMung.

    What is pitiful is saying that GAs are a demonstration of Darwinian evolution. And yet you and yours do that very thing

  38. DNA_Jock: Why on earth did you feel the need to use ellipses to shorten petrushka’s comment today, when your original post quoted petrushka thus

    Because I was hoping to draw your attention to the fact that one could respond to that comment without regard to whether or not it mentioned the video, which it simply offered as an example. An example of what? Oh, well, let’s not worry about that. Let’s pretend that because he mentioned the video my response had to be contingent on my having watched the video, or not.

    It’s really hilarious that this is the absolute best you and can keiths can come up with. But it’s not surprising.

  39. Mung,

    ok, so one M&M out of 6,227,020,800.

    Same. Don’t know why you would think the actual finite number N chosen would make any significant difference to the probability of picking 1 of N objects in N trials. I can see why phoodoo would, because his threshold seems to kick in below N=1000.

  40. Mung,

    Expected propensity?

    If you like, though it somewhat equivocates on the word ‘expected’, compared to the mathematical usage in ‘expected value’. Propensity is not a mathematical term.

  41. Mung: Because I was hoping to draw your attention to the “fact” that one could respond to that comment without regard to whether or not it mentioned the video, which it simply offered as an example.
    [airquotes added]

    Exactly. Your ellipsis was an attempt to draw our attention away from the fact that you were disputing petrushka’s characterization of a method (as described only in the video) when you were wholly ignorant of that method, the sole topic of discussion. It appears you believe that your [explore->search] edit would apply to all GAs, and that you are willing to make such blanket statements without actually being faffed to educate yourself about the particular method under discussion. An argument from ignorance. But that’s not surprising.

  42. Mung: GA’s are not simulations of evolution.

    Then write a simulation of evolution. How would you do that?

  43. Rumraket: Then write a simulation of evolution. How would you do that?

    He’d probably write a GA that fails, like running out of memory… oh wait

  45. Rumraket: Then write a simulation of evolution. How would you do that?

    Why don’t you start an OP on the difference between a model and a simulation and what it would mean to simulate evolution. Are you talking about simulating changes in gene frequencies in a population?

  46. dazz: He’d probably write a GA that fails, like running out of memory… oh wait

    Do you have any evidence whatsoever that I wrote a GA that failed because it ran out of memory? Any at all? I’m guessing that the answer is no, you don’t.

  47. OMG! Three posts in a row in this thread that don’t depend on knowledge, or lack thereof, of the contents of the video in the OP! DNA_Jock is probably contemplating suicide.

  48. Mung whines: Do you have any evidence whatsoever that I wrote a GA that failed because it ran out of memory? Any at all? I’m guessing that the answer is no, you don’t.

    Well, it wasn’t a GA, but you did claim possession of a program that failed because it ran out of memory, writing

    Well, my program eventually stopped due to running out of memory after more than 1.685B permutations. Guess maybe the built-in permutation function isn’t the way to go.

    But now that I think about it, I was actually able to count the number of permutations in a previous run so maybe it was something else that hogged the memory. I’ll have to take another look.

    So it isn’t unreasonable to suggest that “He’d probably write a GA that fails, like running out of memory…”.
    Later, after complying with a request to post your code, you then conveniently claimed that you had someone else “write that program for you”.
    You appear rather inconsistent. Confused, even…

    Mung: OMG! Three posts in a row in this thread that don’t depend on knowledge, or lack thereof, of the contents of the video in the OP! DNA_Jock is probably contemplating suicide.

    Nah. Much to my amusement, the vast majority of your comments here are entirely free of content, so they could be equally well posted on any thread. It is only when you claim subject-matter knowledge (e.g. that the method under discussion is a search, not an exploration) that there is any need to dust off the BS-detector. For the rest of your contributions, pointing and laughing silently suffices.

  49. dazz: He’d probably write a GA that fails, like running out of memory… oh wait

    Mung: Do you have any evidence whatsoever that I wrote a GA that failed because it ran out of memory? Any at all? I’m guessing that the answer is no, you don’t.

    DNA_Jock.: Well, it wasn’t a GA.

    I agree.

  50. DNA_Jock: Later, after complying with a request to post your code, you then conveniently claimed that you had someone else “write that program for you”.

    And yet the code you’ve produced is zero.

Leave a Reply