Author Topic: Cambrian explosion  (Read 2715 times)

Offline jknilinux

  • Bot Destroyer
  • ***
  • Posts: 468
    • View Profile
Cambrian explosion
« on: November 28, 2008, 10:13:31 PM »
Well, I've noticed that almost all evosims seem to be stuck in a sort of precambrian era- none of the rapid diversification that leads to co-evolution ever seems to occur. If you look at the precambrian, there's little life other than gloop- asymmetrical MBs at best, few carnivores, few species. Sound familiar?

Anyway, I was wondering if anyone had any ideas on how to solve this issue? Maybe we can make an official list of features that can encourage speciation for DB3?

Offline ikke

  • Bot Destroyer
  • ***
  • Posts: 300
    • View Profile
Cambrian explosion
« Reply #1 on: November 29, 2008, 07:49:22 AM »
Quote from: jknilinux
Well, I've noticed that almost all evosims seem to be stuck in a sort of precambrian era- none of the rapid diversification that leads to co-evolution ever seems to occur. If you look at the precambrian, there's little life other than gloop- asymmetrical MBs at best, few carnivores, few species. Sound familiar?

Anyway, I was wondering if anyone had any ideas on how to solve this issue? Maybe we can make an official list of features that can encourage speciation for DB3?
Some suggestions:
- Patience: life evolved about 3.5 billion years ago, the cambrian explosion was a mere .6 billion years ago
- multiple habitats to allow for diversity
- have algea and animals follow separate evolutionary steps for instance by different (contradictory) metabolism genes.

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7718
    • View Profile
Cambrian explosion
« Reply #2 on: November 29, 2008, 03:58:46 PM »
A couple things happened that IMO enabled the cambrian explosion.

1.  Snowball earth.  The cambrian explosion happened after a massive extinction event when the Earth supercooled and the planet was covered in over a mile of ice.

2.  Diploid DNA: at some point some life learned to merge during hard times to save resources, with both DNA strands being both in control at the same time.  This lead to...

3.  Sexual recombination, allowing safer exploration of the genomic landscape.  If one DNA strand can survive, and another one can too, combining them increases the probability of a successful novel combination while minimizing the risk of something that won't survive.

4.  Eukaryotes: eukaryotes have a far more organized structure than protists.  Centralized DNA, specialized organelles, etc.

5.  Multicellularity: once multicellular critters appeared, it sort of snowballed into massive diversification.  Apparently once you perfect multicellularity on a small scale, it scales to larger sizes pretty well.

Offline jknilinux

  • Bot Destroyer
  • ***
  • Posts: 468
    • View Profile
Cambrian explosion
« Reply #3 on: November 29, 2008, 04:51:21 PM »
Numsgil-

I'm afraid all those were around long before the cambrian explosion, except for number 1: that might have contributed.

But even bacteria can use sexrepro, while eukaryotes, which were around long before the cambrian explosion, all are diploid and use sexrepro. Multicellularity appeared in the precambrian, in the form of ediacaran biota. However, something turned one of those ediacaran things into the first bilaterian, and after that, life had a foundation to make all known complex organisms.

So, I don't know if that works out. One idea I heard was the emergence of advanced predators- in a world of a. minimalis, there emerged a quickdraw. This didn't kill off everything, but left an area where there were very few plants, similar to moonfisher's nnbot evosim. This is where intelligent/interesting behavior evolves, and as these dead zones grew, life inside the dead zones had to evolve in a way similar to bots in the IBBleague. So, basically we need a huge sim and need to wait a huge amount of time.

ikke-

what do you mean by "have algea and animals follow separate evolutionary steps for instance by different (contradictory) metabolism genes." Do you mean to make sure that sexrepro offspring of the algae and animals will not be viable?

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7718
    • View Profile
Cambrian explosion
« Reply #4 on: November 30, 2008, 03:03:30 AM »
"long" before is a relative term.  I'm saying those things had to occur before the explosion.  It wasn't a fated event, bound to happen.  It involved things that had to happen first, and then the proper sequence of events had to line up in the right order at the right time with a bit of luck.

IMO, what happened was the invention of a whole new way to structure and organize and reproduce.  The eukaryote.  But ecosystems are built like a house of cards.  The base is built from what comes before.  The old protists were filling niches and couldn't be ousted, so the eukaryotes couldn't live up to their full potential.  It took a dramatic event.  The almost obliteration of life.  Then the old ways were cleaned out and new ways could flourish.
« Last Edit: November 30, 2008, 03:03:55 AM by Numsgil »

Offline ikke

  • Bot Destroyer
  • ***
  • Posts: 300
    • View Profile
Cambrian explosion
« Reply #5 on: November 30, 2008, 03:08:34 AM »
Quote from: jknilinux
ikke-
what do you mean by "have algea and animals follow separate evolutionary steps for instance by different (contradictory) metabolism genes." Do you mean to make sure that sexrepro offspring of the algae and animals will not be viable?
Not directly. What I meant to say is that animals and algae should have diverging evolutionary pathways each with their own increasing fitness. Currently veggies are a tick in the appropriate control box, not a set of genes making a bot fit to photosynthesis  (Spelling?? Not my native language...). Furthermore these (absent) genes should exclude efficient animal genes. While there are efficient hunters that are green, this is not due to the fact that they are autotrophe.

Offline jknilinux

  • Bot Destroyer
  • ***
  • Posts: 468
    • View Profile
Cambrian explosion
« Reply #6 on: November 30, 2008, 01:34:04 PM »
Numsgil-

You mean the snowball earth? Well, we don't have any limitation like that in our evosims, so it must be something else, right? Otherwise, we should have passed the precambrian stage by now.

ikke-

Oh- you mean we need a red queen's race (see above link)? But that should've been around long before the cambrian, since whenever there are predators, there's predprey co-evolution. And, since predators were around long before the cambrian, that doesn't seem to explain it.

However, I agree there should be a gene like ".makechlorophyll", instead of just checking a box. Perhaps the makechlorophyll gene should automatically root the bot in place?
« Last Edit: November 30, 2008, 01:35:54 PM by jknilinux »

Offline shvarz

  • Moderator
  • Bot God
  • *****
  • Posts: 1341
    • View Profile
Cambrian explosion
« Reply #7 on: November 30, 2008, 03:07:20 PM »
One thing that DB definitely lacks is the ability to simulate millions of organisms for billions of cycles. So there is a hard-wired limit for you. That said, I think I heard somewhere that cambrian explosion was mainly due to abundance of oxygen in the atmosphere. Suddenly energy became much cheaper.
You can simulate this by turning down all the costs in your sim and I've actually tried that and it works - you get tons of weird bots this way, because selection is not so hard anymore. Your bot counts shoot up as well, so can't really run a sim this way for long.
"Never underestimate the power of stupid things in big numbers" - Serious Sam

Offline jknilinux

  • Bot Destroyer
  • ***
  • Posts: 468
    • View Profile
Cambrian explosion
« Reply #8 on: November 30, 2008, 04:49:32 PM »
shvarz-

Hmm. Extra energy, instead of slowing evolution, speeds it up... That makes sense, in a weird way, but how does it do that? What prevents the bots from devolving, now that they are given so much freedom?

Or, perhaps, that was only a brief period- like lowering costs for only a million cycles, then raising it back up to where it was before to allow evolution to start again with all the weird new life forms it just made? That may be one answer, though unfortunately it requires having really low costs for a while- meaning about 50,000 bots on average, meaning 1 cycle per month on average...
« Last Edit: November 30, 2008, 04:49:46 PM by jknilinux »

Offline Ta-183

  • Bot Destroyer
  • ***
  • Posts: 105
    • View Profile
Cambrian explosion
« Reply #9 on: November 30, 2008, 04:54:50 PM »
Quote from: jknilinux
shvarz-

Hmm. Extra energy, instead of slowing evolution, speeds it up... That makes sense, in a weird way, but how does it do that? What prevents the bots from devolving, now that they are given so much freedom?

Or, perhaps, that was only a brief period- like lowering costs for only a million cycles, then raising it back up to where it was before to allow evolution to start again with all the weird new life forms it just made? That may be one answer, though unfortunately it requires having really low costs for a while- meaning about 50,000 bots on average, meaning 1 cycle per month on average...

It doesn't prevent anything from devolving, it just makes it harder for mutated species to die off. Makes selection more forgiving of slight alterations.

Offline jknilinux

  • Bot Destroyer
  • ***
  • Posts: 468
    • View Profile
Cambrian explosion
« Reply #10 on: December 01, 2008, 01:32:14 PM »
Well, there's not much we can do about that for DB3, except maybe to offer a minimalist hyper-speed mode, where the program just disables certain features, like E-grid, tie physics, population/mutation tracking, and potentially any other parts of DB that are not absolutely necessary. That way, we could run the sim in the minimalist mode under lowered costs to induce speciation and not have to wait a year for it to finish the million cycles required.

Any other ideas?