Author Topic: Sharks and grass  (Read 26960 times)

Offline Botsareus

  • Society makes it all backwards - there is a good reason for that
  • Bot God
  • *****
  • Posts: 4483
    • View Profile
Sharks and grass
« Reply #15 on: March 10, 2005, 06:42:44 PM »
I am thinking the same way about it as PY.

Offline shvarz

  • Moderator
  • Bot God
  • *****
  • Posts: 1341
    • View Profile
Sharks and grass
« Reply #16 on: March 10, 2005, 06:42:49 PM »
No it will not.  Think about it.  If you have a system where gene=enzyme, then if you are not using the enzyme for survival, the gene is going to mutate into something useless and you won't even notice it.  Also, if you are spending energy to make enzymes that you don't need, then you are wasting energy - this will be selected against.  Evolution would sort it all out.
"Never underestimate the power of stupid things in big numbers" - Serious Sam

Offline Botsareus

  • Society makes it all backwards - there is a good reason for that
  • Bot God
  • *****
  • Posts: 4483
    • View Profile
Sharks and grass
« Reply #17 on: March 10, 2005, 06:46:35 PM »
"No it will not" and then he starts explaining what PY just proposed. Its the govt. or a skeptic man.

Quote: "then if you are not using the enzyme for survival, the gene is going to mutate into something useless and you won't even notice it."

Exactly what happens in Py's system


Quote: "Also, if you are spending energy to make enzymes that you don't need, then you are wasting energy - this will be selected against. Evolution would sort it all out.

Exactly what happens in Py's system
« Last Edit: March 10, 2005, 06:46:59 PM by Botsareus »

Offline shvarz

  • Moderator
  • Bot God
  • *****
  • Posts: 1341
    • View Profile
Sharks and grass
« Reply #18 on: March 10, 2005, 06:48:48 PM »
Bots, can't you read - PY said that all bots will be omnivorous.  I said NO, unused genes will be mutated and the bot will not be able to eat that stuff anymore.  THEREFORE, it will NOT be omnivorous.  Do I have to spell this out everytime?
"Never underestimate the power of stupid things in big numbers" - Serious Sam

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7742
    • View Profile
Sharks and grass
« Reply #19 on: March 10, 2005, 06:57:41 PM »
The problem (as Py sees it I imagine, and to a lesser degree myself) is that there is no incentive to be anything but an omnivore.

I agree with everything you've said schvarz, but we should add some kind of incentive to selectively use enzymes.  Maybe the more you produce an enzyme the cheaper it is to produce.

That would be a rather passive method.

A more active method is to make enzymes cheaper to produce or work better the fewer types you have in your stomach.

Either would (I think) work.
« Last Edit: March 10, 2005, 06:59:00 PM by Numsgil »

Offline Botsareus

  • Society makes it all backwards - there is a good reason for that
  • Bot God
  • *****
  • Posts: 4483
    • View Profile
Sharks and grass
« Reply #20 on: March 10, 2005, 07:23:41 PM »
Num it will work fine in the system by Shvartz , better , we can try bouth systems just to prove that it will work fine in any, (Just save a backup copy of the Project)

Then we can see if PY+NUM system works better then SHVARTZ system.

I dont know so I am not taking sides.

Ok lets do an example picture of how Shvartz method and Num+Py method are the same:

lets say enzymes or whatever are A,B,C, and D

In nums method there are  1 of A 3 of B 2 of C 4 of D (too lazy to rewrite in %)

In Shvartz method its

Start
D '1
C '1<---
A '1  
D '2
B '1
C '2 <---
D '3
B '2
B '3 <--
D '4 <--
Stop

Math Time

3 = 3
4 = 4
2 = 2
1 = 1

cya later alegater  :D
« Last Edit: March 10, 2005, 07:30:27 PM by Botsareus »

Offline shvarz

  • Moderator
  • Bot God
  • *****
  • Posts: 1341
    • View Profile
Sharks and grass
« Reply #21 on: March 15, 2005, 11:31:56 AM »
Quote
we should add some kind of incentive to selectively use enzymes


There would be:  using all enzymes requires upkeep, using only some requires less upkeep.

Quote
Maybe the more you produce an enzyme the cheaper it is to produce.
and
Quote
A more active method is to make enzymes cheaper to produce


seem the same to me.  But both are reasonable.

Quote
work better the fewer types you have in your stomach


This I don't like.  Does not make much sense.  It should be the other way around - the more different enzymes you have, the easier it it is to digest stuff made from several kinds of compounds.  Imagine dissasembling a car using only philips screwdriver (even if you have a hundred of them) - not an easy task :)
"Never underestimate the power of stupid things in big numbers" - Serious Sam

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7742
    • View Profile
Sharks and grass
« Reply #22 on: March 15, 2005, 12:03:47 PM »
Okay, this is really driving me nuts!  I keep seeing different posibilities that make different methods work better than others.

Can anyone find a link on why some animals are omnivores and some specialize.  I know generally why, but I don't understand the mechanics of it.  If an herbivore finds a dead animal, why doesn't it eat it?  Try to find resources for unicellular organisms, because on the macroscopic scale things like teeth make a huge difference.

I'll follow whatever resources we can find.  Post the links.

Offline PurpleYouko

  • Bot God
  • *****
  • Posts: 2556
    • View Profile
Sharks and grass
« Reply #23 on: March 15, 2005, 12:13:23 PM »
I still prefer all this stuff to be completely invisible to the robot and to the DNA.

I don't like the idea of having to actively produce specific enzymes as this implies that all robots are capable of making all of the available enzymes.

I like the system that Num suggested way back.

It uses a series of "bits" in a particular pattern to designate enzyme production systems that are on or off in a particular robot. If a bit is on then the robot can digest that kind of food, if not then it can't.
The bits can be set in the start of the DNA file just like custom variables for engineered bots but will mutate to different forms with successive generations.

Multiple bits set for a particular food type will double or triple specialize for greater efficiency but at the cost of losing other bits for different food types.

Using this system, existing robots (withouyt this bit pattern) will be automatically adjusted by the program. A fighting bot will become a specialized carnivore while any robot with the autotroph box checked will become a specialist veggie, able to take energy from waste (nitrates from the e-grid) and sunlight.

No DNA commands to store enzymes.
No genes to keep track of.

ALL behind the scenes and only effected by engineering design or mutation.

Simple, Easy to use, uncomplicated, Doesn't mess up old bots!

 :D  PY  :D
There are 10 kinds of people in the world
Those who understand binary.
and those who don't

:D PY :D

Offline shvarz

  • Moderator
  • Bot God
  • *****
  • Posts: 1341
    • View Profile
Sharks and grass
« Reply #24 on: March 15, 2005, 01:18:40 PM »
Hmm, I did a bit of research in this question.  Here are a couple of links to papers that deal with this stuff, but they are pretty tough for non-biologists:
http://www.pubmedcentral.nih.gov/articlere...bmedid=15289609
http://www.pubmedcentral.nih.gov/articlere...bmedid=15150419

Basically we were both correct.  There seems to be two major hypothesis on how specialization develops, and there are cases supporting each of them.

Hypothesis one (Shvarz): Specialization through mutation accumulation.  Organism can have as many genes as they want and improve them as much as they want, but those genes that are rarely used will accumulate bad mutations and disappear.

Hypothesis two (Nums and PY): Specialization through pleiotropic fitness costs.  If organism gets better at doing something, this automatically leads to it being unable to do something else.

So, there is no clear answer.  Both seem to be happening.
"Never underestimate the power of stupid things in big numbers" - Serious Sam

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7742
    • View Profile
Sharks and grass
« Reply #25 on: March 15, 2005, 01:40:16 PM »
Thanks Schvarz, I'll plow through them later when I have some time.  It's good to know we're both right.  Everyone gets a pat on the back!

My hypothesis is really a subset of yours, so I think we can do both at the same time, and see what develops.

Maybe we can even write up a paper in a few months on our experiments in DB with:

Pleiotropic and Mutational Accumulational Effects on Artificial Life Forms in Darwinbots - A Case Study

:P

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7742
    • View Profile
Sharks and grass
« Reply #26 on: March 15, 2005, 02:15:22 PM »
Okay, I've had a brainstorm.  Something you said, Schvarz.  Can't remember exactly what it was.

Okay, before I had an idea of having a fixed limit on the number of bits an enzyme could have.  32 was what I had.

Imagine that there is no limit.  The enzymes can get as big as they want.  Larger enzymes are more likely to get multiple functionality (if you don't understand why, go back to the other posts I had explaining my bit enzyme system.  I'm afraid I can't find the post right off the top of my head).

However! (and this is the second new part) Larger enzymes take more time and/or energy and resources to transcribe.  As before enzymes are continually wearing out and needing to be replaced.

How does that sound?  I think that works marvelously.

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7742
    • View Profile
Sharks and grass
« Reply #27 on: March 15, 2005, 03:16:56 PM »
Okay, found this article on the computational power of cells.

Which has me thinking...  What if we allow more enzymes or mechanics as the surface are of the bot increases?

Any mechanic that needs to, in some way, access the outside of the bot has to take up physical space on the outside of the cell.  We could define the amount of space a mechanic takes up.

Read the article to see where I"m coming from.  I think this could be an interesting area.

Offline shvarz

  • Moderator
  • Bot God
  • *****
  • Posts: 1341
    • View Profile
Sharks and grass
« Reply #28 on: March 15, 2005, 04:44:32 PM »
No, that guy is a looney.  He ignores one huge difference between biological computers and hardware computers - cells don't take information in bits.  In biological computers it is almost never 0 or 1.  Instead, they are more like quantum computers, but even more vague and parallel-processing.  They give values that range between 0 and 1 and depend upon other values.  Size rarely matters.
"Never underestimate the power of stupid things in big numbers" - Serious Sam

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7742
    • View Profile
Sharks and grass
« Reply #29 on: March 15, 2005, 04:59:50 PM »
:pray:

If you say so.

:P