652 thoughts on “Evolving Wind Turbine Blades”

  1. Frankie:
    dazz,

    LoL! Elizabeth thinks that NS optimizes. You guys need to get on the same page

    Optimizing is not the same as searching. It optimizes design because natural selection is better at finding creative solutions to certain problems than we are at designing

  2. dazz: You and Mung keep asserting stuff but provide no substance.

    Perhaps that’s the correct way for us to respond to people who keep asserting stuff but provide no substance. 🙂

  3. Mung: Perhaps that’s the correct way for us to respond to people who keep asserting stuff but provide no substance.

    like FIASCO CSI?

  4. Mung: Perhaps that’s the correct way for us to respond to people who keep asserting stuff but provide no substance.

    Do you have any specific examples? I’d be interested in such.

  5. Frankie: Unfortunately all you have is luck to find the proper path if one even exists

    How does Intelligent Design help find the “proper path”?

    Given that you are an intelligent designer, Frankie, can you design a better turbine?

  6. Allan Miller: I always get a kick out of watching people try to deny that GAs have any relevance to evolution. The early designers of the same took as faithful a representation of the natural process as practical, generating variation, copying and an environmental reason for variable copy numbers, and discovered that it made a powerful tool – for both search and optimisation applications.

    I always get a kick out of people denying it’s a search because it has two solutions and not just one.

  7. Mung: I always get a kick out of people denying it’s a search because it has two solutions and not just one.

    i don’t care either way. But let’s say that it is, and it comes up with a solution that is bigger than its own size. Has it created information?

  8. petrushka,

    I don’t mean you can eliminate all possible paths. I mean that given X differences and Y time, one could infer the probability that X occurred via small incremental change.

    Yes, but small incremental change isn’t the only mode the critic needs to close down. If we have no precursors or relatives from which we can infer the sequence of events, the critic is completely in the dark in trying to eliminate all paths. I think it an impossible task. We can do ancestral protein reconstruction and show that something is possible, but the critic has the harder task of proving something impossible.

    Darwin’s challenge is to prove that it could not have arisen. He was obviously not familiar with genetics, and was more committed to small change than some of his supporters. But gene duplication, for example, can make a big change to the landscape. It does not happen incrementally, but still happens with reasonable frequency. Or endosymbiosis and other large-scale LGT events, rare but not impossible.

    It strikes me as a non-starter to eliminate all possibilities.

  9. Mung: I always get a kick out of people denying it’s a search because it has two solutions and not just one.

    Nobody said that. OTOH you can’t answer to the simple question of how do you know you’ve found a solution for your search, because you can’t

  10. Richardthughes: I don’t think you understand the problem.

    I think we do understand the problem.

    Either the size of the search space has been reduced, or the number of solutions has been increased, or the search is guided. ID 101.

  11. Mung,

    I always get a kick out of people denying it’s a search because it has two solutions and not just one.

    I always get a kick out of critics thinking there is recourse in semantics.

  12. Mung,

    What utter fucking gibberish, Mung. Why don’t you quote the full post, so folks can see what was actually being talked about?

  13. Mung: Either the size of the search space has been reduced, or the number of solutions has been increased, or the search is guided. ID 101.

    You are right, it’s ID 101 as you’ve drawn a box around all the possibilities and said it’s one of them.

  14. dazz: Nobody said that. OTOH you can’t answer to the simple question of how do you know you’ve found a solution for your search, because you can’t

    In this case the solution was found when no further changes were made to the blades. And no further changes were made because the final blade had the highest output.

  15. dazz: Optimizing is not the same as searching. It optimizes design because natural selection is better at finding creative solutions to certain problems than we are at designing

    Natural selection is not a search and natural selection is impotent

  16. Mung: I always get a kick out of people denying it’s a search because it has two solutions and not just one.

    If evolution is a search, what’s it searching for? Will it ever find it?

  17. Do “guided” evolution proponents who think evolution is a search, contend that evolution stopped when it found humans?

  19. genetic algorithm:

    A genetic algorithm is a class of adaptive stochastic optimization algorithms involving search and optimization. Genetic algorithms were first used by Holland (1975).

    Wikipedia concurs and Wikipedia is sacred for evolutionists

  20. Allan Miller,

    I always get a kick out of watching people try to deny that GAs have any relevance to directed evolution. Even grade school kids can see that they are

  21. dazz: Nobody said that.

    That’s exactly what people are saying. I’ve just turned it into a reductio to show how silly the objection is. How many potential solutions before it ceases to be a search? Three? Five?

    dazz: OTOH you can’t answer to the simple question of how do you know you’ve found a solution for your search, because you can’t

    I can’t list all the possible solutions, so it’s not a search? That’s your argument? The logic is even worse than I thought.

  22. Frankie: In this case the solution was found when no further changes were made to the blades

    No. The simulation was stopped when the user felt like it.

  23. dazz: Do “guided” evolution proponents who think evolution is a search, contend that evolution stopped when it found humans?

    Isn’t it obvious? But then, I’ve only been sampling here at TSZ.

  24. Richardthughes: i don’t care either way. But let’s say that it is, and it comes up with a solution that is bigger than its own size. Has it created information?

    Let’s go with this Mung. What say you?

  25. Frankie: Wikipedia concurs and Wikipedia is sacred for evolutionists

    Is what evolution is searching for set by the Intelligent Designer?

  26. Richardthughes: The simulation was stopped when the user felt like it.

    And I could create a program that generates random strings of characters that only stops upon user intervention.What would it tell us?

  27. According to the video the guy stopped the program after noticing no further changes were being don e to the jellyfish blade. That no further changes were being done means the program found its solution.

  28. Frankie: No

    How do you know the guidance has not stopped? What specifically?

    I.E. if it stopped one day, would you know? How?

  29. Mung: I can’t list all the possible solutions, so it’s not a search? That’s your argument? The logic is even worse than I thought.

    Try reading for comprehension. I didn’t ask to identify all possible solutions, or the solution if there’s one. I asked how do you know you’ve found ONE solution in an algorithm that can go on forever.

    We’re not talking about applications where some solution is known to exist, but the exact solution is unknown. In that case, that would be a search.

    We’re talking about GA’s that “closely” model evolution. Let me quote myself:

    “A simulation is always going to have a finite set of possible outcomes, but for all intents and purposes it can be considered infinite in the sense that one could never evaluate every possible specification in a true search for the optimal one, and that’s why a GA is used instead.

    Seems to me that in a GA with a huge set of possible outcomes, there’s never a point where one can be sure it’s reached the optimal solution. If the GA is kept running a better solution may always emerge, particularly if the random variation allows to overcome getting stuck in local maximums (recombination?)”

  30. Mung: And I could create a program that generates random strings of characters that only stops upon user intervention.What would it tell us?

    Better ask KF about that, I feel.

    Tell me, is there any difference between generating a string of random numbers and evolving turbines from a programmer’s point of view?

  31. Is the concession on whether GAs are a valid model of biological processes contingent on whether one agrees that both are searches? If so, I vote yes, they are both searches. I never knew it could be so easy.

  32. Mung: And I could create a program that generates random strings of characters that only stops upon user intervention.What would it tell us?

    That it is probably not a search, or not a very good one. You should not check the same box twice.

    Amazing how effective these things modelled after evolution are, isn’t it?

  33. Richardthughes: Amazing how effective these things modelled after evolution are, isn’t it?

    Intelligent Designers studying evolution? Whatever will they think of next?

    You’d think that IDers studying evolution knowing it was the product of ID would have an advantage over others doing the same without that insight. So those IDers would be the biotech leaders, the big patent holders.

    But they are not. Are they? How very strange. I guess ID brings nothing to the table of any use except the knowledge that something is designed.

  34. Personally I am quite happy to say that the turbine development program is directed evolution, because in my view all adaptive evolution is directed. If there was no direction there would be no adaptation. Of course in real life there is always direction, although this can come from different sources.

    Direction can take many forms. In breeding it is the human breeders, if you so wish you could say it is ‘intelligence’ that guides the process towards a desired outcome. In Nature it is the environment (including other organisms) that directs the changes towards better adaptation. In software it is the fitness function that guides the program towards a solution that satisfies the pre-set specifications.

    All evolution is directed. But that is not the real issue in this debate, isn’t it?

    fG

  35. Good thing he turned the wind off when speaking, but those plosive pops make it quite hard to listen to.

  36. faded_Glory: Personally I am quite happy to say that the turbine development program is directed evolution, because in my view all adaptive evolution is directed. If there was no direction there would be no adaptation

    All adaptive evolution is directed, OK could be interpreted that way, but is all evolution adaptive?

  37. Frankie:
    According to the video the guy stopped the program after noticing no further changes were being don e to the jellyfish blade. That no further changes were being done means the program found its solution.

    Frankie can read my posts. Its a miracle!

    “the guy stopped the program” – we agree.

    “That no further changes were being done means the program found its solution” – So why did two runs get two different answers?

  38. faded_Glory,

    Actually that is the real debate. Get your idea published and someone may listen to you. Until then evolution is blind and mindless, ie directed only by the surviving reproducers. Eliminating the less fit does not direct anything to be more fit

  39. It’s false that “no further changes were being done”. New “mutants” were being produced, but the results didn’t top the provisional best because of the limitations of the mutational model that would’t allow to get out of a local maximum, but how does Frankie know if at some point, maybe in two hours, maybe 20 days, one mutation may get it out of that local maximum and allow it to explore other areas of the fitness landscape? He can’t. He just set an arbitrary “solution” in a local maximum.

  40. A genetic algorithm is simply the algorithm used to simulate evolution. It takes candidate solutions, selects some of the best using user-defined evaluation functions, applies user-defined transformations (often called mutation and crossover, but implementations of these depend on the problem), and makes new candidate solutions.

    – http://www.c2.com/cgi/wiki?GeneticAlgorithm

    No worries though. It’s all just semantics.

  41. This popped up in a Youtube sidebar. It’s a pretty good overview from a non-biologist’s perspective. Amusing part where he describes GAs as, like some movies, ‘inspired by real events’.

  42. Mung,

    No worries though. It’s all just semantics.

    The way you play it it is. Got any substantive observations?

  43. Mung: No worries though. It’s all just semantics.

    3x + 1 = 10.

    x = 4 is a candidate solution, It’s not *the* solution, though.

  45. Mung: No worries though. It’s all just semantics.

    “Candidate solutions” produce Candidate solutions” that produce Candidate solutions” that produce Candidate solutions” that produce Candidate solutions” that produce Candidate solutions” that produce….

    So all the elements in the process are solutions! Exactly like evolution, everything that passes selection is a solution and it’s not a search for any particular one, and never stops producing “solutions”

    Glad we finally reached an agreement!

  47. It should be obvious that a static fitness landscape will converge on an optimum of some kind. Whether you want to continue the program is really up to you. But offspring keep on being produced. The process does not stop, it just runs out of things to find, and hence interest to the observer, unless the landscape changes. Real adaptive landscapes aren’t static, because there is a wider linked genome, a population, an ecology, a shifting planet, and change can come from anywhere stimulating a new track. The ‘halting issue’ is a non-issue.

  49. Mung: I’m ignoring it. Dodging it and ignoring it are not the same thing.

    Ignoring why? If it’s because the answer is inconvenient to you, then that’s dodging.

  50. Allan Miller:
    This popped up in a Youtube sidebar. It’s a pretty good overview from a non-biologist’s perspective. Amusing part where he describes GAs as, like some movies, ‘inspired by real events’.

    Thanks for posting that, the first video is that Karl Sims project Prof. Felsenstein linked a while back. I’ve been trying to find it to run it here but it seems it’s old and not available. It’s so cool

Leave a Reply