Author Topic: My simulation: Evolving social bots  (Read 5488 times)

Offline meph

  • Bot Neophyte
  • *
  • Posts: 4
    • View Profile
My simulation: Evolving social bots
« on: April 25, 2010, 05:42:19 PM »

Hello, I just thought some people here would be interested in seeing a simulation that I put together a while ago on this subject:
http://www.youtube.com/watch?v=GvEywP8t12I



I'm not too familiar with darwin bots but there are some differences that I picked up on right away.
I guess the biggest one is that I do not use a coding-based DNA. Instead, my DNA specifies the structure, connectivity, and parameters of the perceptron network. My brains are therefore automatically robust to small mutations because of the invariance in the sigmoidal activation function.
However, compared to Darwinbots I think I am working on a much lower level and can probably accomplish less elaborate tasks per parameter.

The sensory modalities are also a little different: my bots can "smell density of bots" and "hear speed of bots", and also sense the colors, but on a little lower resolution.

Oh and I don't have pretty graphs

I should check out darwin bots sometime
Cheers,
meph

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7742
    • View Profile
My simulation: Evolving social bots
« Reply #1 on: April 26, 2010, 12:45:46 PM »
Looks pretty cool   Have you gotten anything like a stable ecosystem going?  (eg: predators and prey following a sinusoidal graph).  That's always been the hardest part in Darwinbots (it's been done, it just requires a lot of tweaking to get the initial parameters set up right).

Offline meph

  • Bot Neophyte
  • *
  • Posts: 4
    • View Profile
My simulation: Evolving social bots
« Reply #2 on: April 29, 2010, 01:10:24 PM »
Quote from: Numsgil
Looks pretty cool   Have you gotten anything like a stable ecosystem going?  (eg: predators and prey following a sinusoidal graph).  That's always been the hardest part in Darwinbots (it's been done, it just requires a lot of tweaking to get the initial parameters set up right).

Hi: well that's just the thing, I can't make those plots because I don't have this distinction. These bots just do their thing (whatever that is), and you can only qualitatively decide, sometime. Many times it is clear, but many times they are in this grey area in between where they are gatherers but then flip randomly and attack something.

If Darwinbots does have this clear distinction where every bot must be one or the other, then I would consider that as an instance of where you are putting in too much structure. You are baking the solution in, not seeing it emerge... which takes some fun out of it for me


Offline ashton15

  • Bot Builder
  • **
  • Posts: 99
    • View Profile
My simulation: Evolving social bots
« Reply #3 on: April 29, 2010, 03:03:47 PM »
Nay, darwinbots definitely isn't black&white at all, I personally find though that genrally most bots have no need for specializing, hopefully in the upcoming version of darwinbots more emphasis will be placed on speciality, the idea is to encourage multi-cellular creatures though hopefully it will also help provide abit more structure to the ecosystem. Not to much though

Offline peterb

  • Bot Destroyer
  • ***
  • Posts: 148
    • View Profile
My simulation: Evolving social bots
« Reply #4 on: April 29, 2010, 05:01:21 PM »
WOW THATS AWESOME !!!  


that video is amazing, and you solved the problem of how to train a neural net in such environments i think thats.. pretty amazing. un-trained NN
Oh i love to see this combined with darwinbots, where bots can evolve in shapes (multibots)
I've been thinking on how to create something like it, the darwin bot code is different currently small 'genes' with functions (often very complex functions)
But they are of human design. (with exception of zero bot leagus)  (both options would be nice.. human bots and NN bots)

Would be nice if DNA would rather describe structure (cell types, propulsion, eye, shell,venom..) and a neural AI would be the brain.
Even the structure of the brain, could be maybe described by DNA  (so you get darwin rules combined with neural brains).
The same DNA would connect structure and the input out to this structure (connect between eye and propulsion)
In an environment much like the current DB
« Last Edit: April 29, 2010, 06:36:07 PM by peterb »

Offline Botsareus

  • Society makes it all backwards - there is a good reason for that
  • Bot God
  • *****
  • Posts: 4483
    • View Profile
My simulation: Evolving social bots
« Reply #5 on: April 30, 2010, 11:17:47 AM »
This stuff is CRAZY!
I want the source code!  

Offline ashton15

  • Bot Builder
  • **
  • Posts: 99
    • View Profile
My simulation: Evolving social bots
« Reply #6 on: April 30, 2010, 11:57:49 AM »
Quote from: Botsareus
This stuff is CRAZY!
I want the source code!  

Nice decided to click the link as I don't normally bother with them to what it was seems like a pretty alright thing you've made there... could probably help darwinbots a bit...

Offline happyhamsterchan

  • Bot Destroyer
  • ***
  • Posts: 118
    • View Profile
My simulation: Evolving social bots
« Reply #7 on: May 02, 2010, 08:09:32 AM »
I don't really see the benefit of using an artificial neural network, especially if were talking about single-celled entities... It seems to be a bit constrained, like were forcing the bots to conform to our definition of intelligent, instead of evolving their own novel abilities. Sure, The same behaviours can be accomplished whether your using an ANN or evolved spaghetti code, but somehow the point of alife seems to be lost. Not to mention, you get a lot more efficiency from using evolved spaghetti code, instead of evolved spaghetti ANNs.
I AM THE...

AWESOME...

FUN...

TURTLE!!!!!

Offline Moonfisher

  • Bot Overlord
  • ****
  • Posts: 592
    • View Profile
My simulation: Evolving social bots
« Reply #8 on: May 03, 2010, 12:54:23 PM »
I think the point is to accelerate the evolution of behavior by narrowing down the options.
Either way it looks realy cool, I like how a NN is visualized in the background too.

Offline meph

  • Bot Neophyte
  • *
  • Posts: 4
    • View Profile
My simulation: Evolving social bots
« Reply #9 on: May 03, 2010, 08:48:36 PM »
I agree with what has been said.

The use of Neural Network is loosely inspired by the brain. (But way too loosely )

Its interesting to experiment with brain modules. Interestingly I also tried Radial Basis function networks and they perform MUCH worse.
I also tried implementing more higher level neuron types... so some neurons would not compute sigmoid, but do something like a MAX() of inputs, or StandardDeviation() of inputs, or they would be a little if_then_else module working with inputs. Those brains are therefore more quickly higher in the hierarchy of capabilities. Sigmoids are universal approximators so you could do all of this with simple perceptrons but may just need too many of them.
It is interesting by the way that the above idea did not really work too well. I think maybe these brains are harder to "train", I'm not sure. Just plain perceptron brains worked best so far. (weird)


Anyway, my current goals are to change the simulation to something more along the lines of DarwinBots, where the brain is a "program" and see what i get. I will also make the code nicer and maybe release it sometime in the future. My schoolwork has just finished, so I can start working on this now

Thank you for comments!
cheers
meph