10 Replies to “Further evidence that biological innovation belongs to the science of engineering and not to the Victorian myth of unintelligent evolution

  1. 1
    timcol says:

    I don’t understand why this is considered ‘evidence’ of anything. Sure, the maple seed provided a rough concept for building the nano air vehicle (although the NAV has far more capabilities many of which did not derive directly from the natural world). And of course process and materials to actually construct the NAV were entirely different from that used to create the maple seed. But why should it be surprising that we wouldn’t take something tried and tested from nature and learn from it — after all the maple seed has to operate in the same environment as the NAV and follow the same laws of physics in regards to aerodynamics and resistance. The only difference of course is that the process to create the maple seed was a far longer one and required much trial and error to get to its current stage. Sorry, but this is another example of ‘pubjacking’ that seems to be grasping yet more straws.

  2. 2
    Strangelove says:

    Notice as well, that the biological sciences learned nothing from the engineers in this case. Rather, the engineers learned from biology, despite what the headline of this post implies.

  3. 3

    The biological sciences aren’t trying to build something so why would they gain insight from the engineers?

    But…the engineers gained insight from biology but not the biologists. In other words, they gained insight by treating the maple seed as a design, not as a random accident. That’s the point.

  4. 4
    Strangelove says:

    “In other words, they gained insight by treating the maple seed as a design, not as a random accident.”

    How can you tell how they treated it? How would an engineer treat something that has been designed differently than something that had been tweaked through the tried and true process of trial and error for millions of years? But seriously, how can you tell?

  5. 5
    BC says:

    Naturalistic evolution (by which I mean random mutation + selection + reproduction) is a type of designer. Further, it’s quite capable of creating designs that are beyond human capabilities. Describing evolution as “random accident” is to ignore 2/3rds of the evolutionary equation.

    Genetic algorithms (which are modelled on random mutation + selection + reproduction) are quite capable, and in some cases are being employed to design products instead of using living, breathing engineers:
    http://www.popsci.com/popsci/p.....drcrd.html

  6. 6
    Mats says:

    BC,
    The site you alluded starts with the words “John Koza Has Built an Invention Machine”. A machine that is built to reconize certain patterns, and that is PROGRAMED to select what might work, is hardly evidence for naturalistic unguided evolution.

  7. 7
    BC says:

    The site you alluded starts with the words “John Koza Has Built an Invention Machine”. A machine that is built to reconize certain patterns, and that is PROGRAMED to select what might work, is hardly evidence for naturalistic unguided evolution.

    It’s not “built to reconize certain patterns, and that is PROGRAMED to select what might work”. It’s figuring out which “organisms” of the current generation work best (run the fastest, pickup radio signals best, etc) and then using them in the next generation. It’s not picking the ones that “might work” or “recognizing certain patterns”. There is no foresight or predetermined genetic codes; it’s just taking what seems to work in the present (fitness) and then using it (with some additional random modifications) in the next generation. In this case, it is using artificial selection (not natural selection) to determine who breeds and who dies. An artificial selection that grades on the current organism’s fitness (without foresight) isn’t terribly different than a natural selection mechanism that grades on the current organism’s fitness. If, for example, I wanted an organism that runs fast, I can hand-select the fastest individuals for breeding. On the other hand, if I put a bunch of gazelle in a forest and the lions chase down the slowest ones, you get the same result. There is essentially no difference in the selection mechanisms (artificial versus natural) in which individuals get to pass on their genes to the next generation and there is no difference in how powerful the mechanism is.

    Further, if you are drawing attention to the word built as evidence of intelligent design, then yes, the machine is intelligently designed. But, a human building a machine which uses random mutation, selection, and reproduction is analogous to a God designing a universe where random mutations, selection, and reproduction (NDE) can produce organisms. If it were analogous to biological intelligent design, then it would require the human inventor to occasionally step in and create whole new sets of irreducably complex systems or whole new sets of “genes” in the “organisms”. That’s not what’s happening. In fact, I would argue that Koza’s machine would be less impressive if he had to occasionally step in and alter things by hand, although this is what biological ID says is going on in our universe.

  8. 8
    j says:

    From the Popular Science article that BC linked to:

    Now 62 and an adjunct professor at Stanford University, Koza is the inventor of genetic programming, a revolutionary approach to artificial intelligence (AI) capable of solving complex engineering problems with virtually no human guidance.

    Come back when the word in boldface isn’t needed.

    …the machine’s method: Darwinian evolution, the process of natural selection.

    Don’t the editors of that magazine read the articles they publish? What’s described is not Darwinian evolution. Darwinian evolution has no goal. Try again.

    From there, it’s Darwinism 101. The invention machine mates some systems together, redistributing characteristics from two parent lens systems into their offspring. Others it mutates, randomly altering a single detail. Other lenses pass on to the next generation unchanged. And then there are the cruel necessities of natural selection: The machine expels most lenses with low fitness ratings from the population…so their genetic material won’t contaminate the others.

    Replace “Darwinism 101” with “Artificial Selection 101.” Replace “natural selection” with “artificial selection.” Darwinian evolution has no goal.

    …the real challenge will lie in deciding what to create. This is no trivial matter—to invent a car, even with an invention machine, you must be able to conceive of wanting a horseless carriage in the first place.

    This is true. Darwinian evolution has no goal.

  9. 9
    BC says:

    with virtually no human guidance.

    Yes, a human has to input the scoring system. If you want a system that solves complex engineering problem X, you need to tell the system what the device should actually accomplish. The machine is not a mind reader. For example, if you want an antenna that receives radio signals, you create a scoring system that rates ability of devices to receive radio signals, kill off the low scoring versions, allow the high-scoring versions to reproduce. If a someone does not input a scoring system based on what they want the machine to create, then all outputs are equally valuable and you come out with random noise.

    But, nature has a built-in goal. The “goal” (if you want to call it that) of evolution is survival. The organisms that survive the best move on to the next generation. What exactly helps an organism survive? It’s a wide variety of things – camouflage to hide from predators, being able to run fast, being poisonous so predators don’t want to eat you, being capable of spoting predators with eyes and ears. These are all traits you see in animals in the natural world. Nature creates it’s own grading system – it allows the survivors to reproduce.

    The most important fact about genetic algorithms isn’t whether it mimics naturalistic evolution in every way, but rather that certain major features of it – features which are the major cornerstones of evolution, features that creationists say can’t possibly work – actually work. Pointing out that there are differences, and acting like those differences allow you to ignore the entire system and all its parts is just a way to ignore the major features that validate portions of naturalistic evolution.

  10. 10
    j says:

    But, nature has a built-in goal. The “goal” (if you want to call it that) of evolution is survival. The organisms that survive the best move on to the next generation.

    The survival of the fittest. Who are the fittest? The ones who survive. The survival of those who survive. A tautology. Tautologies don’t explain anything.

    The most important fact about genetic algorithms isn’t whether it mimics naturalistic evolution in every way, but rather that certain major features of it – features which are the major cornerstones of evolution… – actually work.

    Genetic algorithms do not “mimic naturalistic evolution” at all. Naturalistic evolution — i.e., Darwinian evolution — is devoid of intelligence. Without intelligence nothing interesting happens.

    Professor Colin Reeves, of Coventry University, in a letter to the Telegraph back in January:

    In my own research area of evolutionary algorithms, intelligent design works together with evolutionary principles to produce better solutions to real problems. Sometimes the results are novel and surprising, but, on reflection, they were always inherent in the initial formulation. Without the initial activity of an intelligent agent, the evolutionary mill has no grist to work on.

    [Reeves is the Professor of Operational Research in the School of Mathematical and Information Sciences. He has a BSc in Mathematics, an MPhil in Operational Research, and a PhD in recognition of the contributions made by his publications to the development of the field of evolutionary computation. It is in the latter area that his research now specialises. He has written over 80 journal and refereed conference papers, as well as authoring and editing several books and conference proceedings.]

    You think that a purpose-driven, constrained trial-and-error system for locating optimized solutions to problems, says something about blind/dumb/purposeless Darwinian evolution? I agree. It says it’s missing the most important thing — intelligence.

Leave a Reply