Pages

Friday, 3 August 2012

Evolution For Creationists Who Can Recite Jack & Jill

Okay! If you're still having a problem understanding evolution, try this mind experiment. It should only take a few minutes.

Imagine you have thousands of robots which have only three instructions:
  1. If you can, step up.
  2. Never step down.
  3. Otherwise move in random directions.
Now, place those robots randomly in a landscape which includes hills and valleys, a bit like the one on the right, and leave them to wander around for a few months.

Now answer these two questions:
  1. Where will your robots be when you go and look for them?
  2. Will they still be randomly distributed in the landscape?
Correct answers win the Grand Old Duke Of York Award for understanding evolution.

For those few who still don't get it, I'll explain it in a few days.





submit to reddit




25 comments:

  1. 1) You would expect the robots to be at localized high spots closest to their starting point.

    There will be exceptions to this, mainly because the random walk the robots took may have taken them far from their starting point before they were eventually started their inevitable climb to the high spot where they get "stuck". It is also possible some of the robots got "stuck" next to sudden increases in height that they can not climb (like at the bottom of cliffs). One big factor is how long you take before you return and how relatively flat your area is. If you return quickly and the area is pretty uniform in height, the robots will likely still be in transit to where they will get "stuck". Also I assume that there is nothing in this landscape that could push the robots downhill. Like stray animals, winds, or water. If the later is true, then those robots could be anywhere. I would still search the high spots first, but I would likely not find them all there. Like for instant if one was knocked over by a wild animal into a river and swept down stream I might find him suffling around on a boulder at a bottom of a lake.

    2) That answer is utterly dependant on how randomly distributed the robot's starting points are, how randomly distributed the localized high spots are, and how much space there is for robots at the top of those localized high spots.

    ReplyDelete
    Replies
    1. A 'stuck' robot should eventually become unstuck by random movement, if only by backing out the way it went in or came up against the obstacle.

      Delete
    2. A robot can only step up. So once a robot is at a point that is higher than any other point around around it, it will no longer move. It is stuck at the high point. Rule 2 keeps it from going downward.

      Delete
    3. True, but I was referring to those > "stuck" next to sudden increases in height that they can not climb (like at the bottom of cliffs).<

      Delete
    4. A rise next a cliff might be hard for you to visualize. But think of it this way. If your goal was to climb a mountain, but you did not know how to climb and you were not sucidal, where would end up? At the bottom of a cliff on the mountain. Most mountains eventually have cliffs or features that most humans just can't get around. And if you followed that cliff all the way around, you would eventually find a bottom of the cliff that was higher than any other point, but it is still not the top of the mountain.

      Now if your robots can handle any change in elevation, then they would not get stuck at the bottom of cliffs. But I did not want to assume that your robots had infinite ability to change elevation.

      Delete
    5. In the scenario given, there was no goal.

      Delete
    6. I think a "stuck" robot fits perfectly with evolution. Think of these as victims of selection whose genes were not passed. The higher you climb the more succecful your line has been.

      Delete
  2. This comment has been removed by the author.

    ReplyDelete
  3. Firstly, it depends on where the robots are before these commands, so if should to step up, not step down and moving in random directions, and if all one thousand in the same place and move at the same time, under the command, all will go in the same direction and be in the same place in the landscape, BUT because can not be at the same place in the same time and must move one by one going in the same direction all be in the same landscape in a both case.Second, this example demonstrates, or suggests that there is no natural selection, but artificial in this case, because a thousand robots are moving as a single and more types of robots do not need, and need only one or not even this one, because the test shows that natural selection, with only three commands is unnatural, artificial selection, and for
    Third, Thirdly, in God's relation to the selection of this conclusion, if we consider that the relevant test with robots, God is pure absurdity, as evidenced by the creation of the first two were destroyed, that the third was satisfied, BUT in great mistake, because the how people behave, not respecting any of the three commands, they will to destroy itself, but in fact the god, so it was kind of unpredictable selection which leads again and again just to endlessly re-creation and what is the question here is what is artificial ( only one space ), (un)natural, so-called selection, people, or God himself.

    ReplyDelete
    Replies
    1. Do you understand what 'random' means?

      Delete
    2. >suggests that there is no natural selection<

      What function does natural selection have if it's not 'if possible, always step up'?

      Delete
    3. I'm afraid the rest of your comment was incoherent gibberish. Was there something too complicated in the simple mind experiment for you to follow, or were the questions too difficult for you to answer?

      Delete
  4. What the hell do you mean "how randomly" random distribution means entirely randomly... there aren't degrees of randomness, it's either random or not...

    ReplyDelete
    Replies
    1. Actually there are "degrees" of randomness. :-) You can have "normal" distributions, "uniform" distributions, "discrete" distributions. All sorts of forms of random. Which one you are using is often dependent on context. For this context I assumed "uniform", or in otherwords, all possiblities were equally likely to occur.

      If was to write a program for this this robot it would be more specific:

      Step 1: Consult random number generator for a number between 0 and 360.
      Step 2: Observe spot that is one foot away that is at an angle equal to what the random number generator produced.
      Step 3: If that spot is equal to or higher that current location, move there and return to Step 1, else go to Step 4.
      Step 4: If that spot lower than current spot, do not move and return to Step 1.

      Delete
    2. Of course, I could have included instruction such as loading values into an accumulator, perform add with carry with a second value, etc. However, the result would still be the three instructions given, performed in that order of precedence.

      The point was to do the mind experiment and tell me the result, not to tell me how you could have written a different scenario.

      Delete
  5. robots are intelligently designed and in this scenario placed in a valley by an intelligence. That intelligence programmed the robots to perform its random step up and random movement. It might be a description of evolution but as I said knowledge was involved from the get go.That is also the problem in human experimentation whenever they attempt to use science to recreat building blocks of life.Their intelligent hand is all over it bringing the ingredients together.

    ReplyDelete
    Replies
    1. I see you dodged the questions.

      Unkind people might assume that that was because you didn't like the truth and couldn't bring yourself to write it down in case other people saw it too. They might then conclude that you have a 'faith' which makes you ashamed of, or afraid of, truth and honesty and which requires you to ignore them.

      What would you say to them?

      Delete
  6. Your questions are illogical for an atheist who plants Intel design & created robots in a valley of water to prove evolution on mount improbable. The analogy though would be consistent with TE.

    ReplyDelete
    Replies
    1. Sorry to have made you show that you are too afraid to answer them.

      Delete
  7. Intelligently designed robots placed in the valley by their Creator as you suggest. Would either continue to climb via a pathway or accessible route.others caught in snags and unable to move on but tread water.Some would get to the peak before falling off as they can't step up or again tread water.But given that before they set off they were already in existence and formed I admire your acceptance of special creation in this Robot Cambrian explosion.
    Morgan anonymous

    ReplyDelete
    Replies
    1. I'm mystified as to why you have introduced water where there was none.

      Was it in an attempt to obscure the issue again hoping other people would be confused and mislead by it, if so, I fear you may have underestimate the intellect of readers of this blog, or just your way of coping with something you can't refute?

      Delete
  8. Wtf am I missing here? All the robots would be at the highest level they could reach. Period. I don't understand why this is confusing people. I also (seriously) don't understand how it will help creationists understand evolution. And using an analogy with robots is begging 'them' to cry "yeah but you're the creator!!".

    I'm gonna show my wife this 'cause I currently have Lyme disease & can't think right (seriously, it's like having chemo-brain - f'ing awful) and maybe she can help me or maybe I'll stop by in a day or two & find out I'm not crazy and the point will be obvious.

    Keep up the good work!!

    ReplyDelete
    Replies
    1. The point you are missing is that the landscape inevitably forces order out of randomness, just as the environment inevitably produces order and direction in evolution without any complicated rules or external intervention.

      The environment forces organisms to reach the highest point available once they start ascending and they then become committed to that rising elevation with no way down.

      Delete
  9. It's a good analogy, inevitable some theist would attempt to use it to prove a designer, when it's just a thought experiment.

    ReplyDelete
  10. A perfectly fine analogy to demonstrate a point. Which went on to demonstrate that most religious people don't understand what 'thought experiment' means.



    ReplyDelete

Obscene, threatening or obnoxious messages, preaching, abuse and spam will be removed, as will anything by known Internet trolls and stalkers, by known sock-puppet accounts and anything not connected with the post,

A claim made without evidence can be dismissed without evidence. Remember: your opinion is not an established fact unless corroborated.