Added: 3 years ago
From: kjlg74
Views: 11,845
Sort by time | Sort by thread (beta)

Link to this comment:

Share to:

All Comments (15)

Sign In or Sign Up now to post a comment!
  • I had no freaking idea there were artificial life programs like this!! So amazing to watch evolution in progress.

  • Its so majestic O:

  • ITS SO CUTE :D

  • i still dont get hw you make this

  • Hey cool I had one like this, but it was way better, it curled up in a ball and sprang out jumping, rolling then springing up again. But I think I deleted it

  • @Mrxb0x I'm sometimes surprised to see which creatures people on youtube enjoy vs those that I enjoy. I find creatures like this one fun to watch, but somewhat disappointing as evolution results. Nearly 4000 views though! I'd never have guessed it would be so popular when I first posted it.

  • Once ive done the full 600 generations, ill start on experiment 1... most of my experiments go terribly wrong so ill just try to do ones other people make but dont have much CPU power...

  • Ok, im on generation 140 atm... ive found out how to get the rendering alot faster... when asked for settings use OpenGL Rendering subsystem rather than the default one and then change Video mode to 640 x 480 and colour depth 16... it looks pretty ugly sometimes but its worth it in rendering time, but these settings sometimes dont work on some computers

  • Oh, thanks for letting me know. On my machine DirectX runs faster. I'd figured that would be the same for everybody, but now I know even that can vary from one machine to another. Interesting.

  • hhmmm, it turned yellow?

  • Oh yes. Colour changes occur from time to time. Colour is a completely neutral trait. It has no effect on fitness at all. So colour is subject to genetic drift as well as other effects like hitchhiking (a colour-change mutation occuring together with some advantageous mutation, and the two spread through the population together).

  • Well, at generation 56 its starting to figure out that the white bit at the top is unbalancing it more when its on its top meaning that it is jumping higher every so often...

  • So it's shrinking the white part? Or making more centered against the spiral? Just curious.

  • its getting rid of it completely... it helps to get it in agood position, but happens 4 maybe 5 times a generation

  • Should be interesting to see how much further it might evolve. It looks pretty crazy, but it does seem to eventually position it to make the final jump. Is there any selective pressure to be conservative with the amount of movement required to get to the result, or is the height of the jump the only factor in it's selection?

  • Only the height matters. To be more precise, the fitness is the largest net upward change in the vertical component of the creature's center of mass during allotted time. That value is also divided through by a measure of creature size to make sure big creatures have to jump higher than small ones.

    The suggestion about taking the amount of energy or movement into account is a good one! Someone else made a similar comment to me a while ago. I think it's something I should eventually implement.

  • Interesting. One thing I've noticed is that some of the creatures seem to be quite unconservative in their structure, and some equivalent of genetic drift seems to be creeping in and allowing extra bits to not be taken off because of the lack of a selective pressure to be efficient. If limiting the creatures' mass and energy requirements were selective pressures, I suspect that would generate more conservatively made creatures. I don't know difficult that might be to implement though.

  • Some of the fitness measures have a "scaling" option that divides values (like distance or height) by creature "size". That size is a function of creature volume, so that creates a pressure to remove unneeded body parts and to shrink those that are needed.

    I think there may be a plus to having that pressure turned off, though. The body keeps changing by the drift you mentioned, and I think that might be a good source of variation, alongside mutation. It would be an interesting thing to test.

  • Lol, i love that this program never has the same 2 results every time u run an evolution from scratch...

  • Yeah, they're pretty different every time. :)

Loading...
Alert icon
0 / 00Unsaved Playlist Return to active list
    1. Your queue is empty. Add videos to your queue using this button:
      or sign in to load a different list.
    Loading...Loading...Saving...
    • Clear all videos from this list
    • Learn more