Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - viplex

Pages: [1] 2 3
Suggestions / gene pools
« on: March 31, 2007, 07:21:35 AM »
Quote from: Numsgil
Generally programs which demonstrate "awesome" results for evolution manage this because the evolution is directed.  Evolution is basically a search algorithm, and the broader the solution space, the harder any search algorithm is going to have it.

Toward your actual suggestion, I'm not sure what you mean by a "life pool" or, for that matter, a "gene pool".  A gene pool as I understand it is an implicit term for the combined genetic material in a population.  By this definition, any population with genes has a gene pool.  Second, what do you mean by "mutated in all ways"?  The bots' behavior is determined by their genetic code.  If you were to mutate every single base pair in the bot's genome, you would effectively have an entirely different bot.

Last, this whole section doesn't make sense to me as it's written:

Sorry, I tried to be short.
I forgot to mention that most importantly there is crossover mutation besides all the others (perhaps DB already implements it, but it would need much more usage of it.)
Gene pools: Contains genotypes that are to be compared and crossed over with one another. Bots from different gene pools can't cross over. There is one gene pool for carnivores, one for herbivores, one for veggies, for example. Or one for tieing scavengers, one for others etc.
From time to time, the program picks specific bots from one gene pool for contest. Puts them on the screen, and let them do as normally. The one with more of predefined fitness will be the winner, or simply the one that survives. The winner then will be reproduced in its own gene pool (not on the screen), also in mutated and crossed-over instances of itself. So it will not appear instantly on the screen, but only when the program picks it for another contest. There are loads of possibilities for organizing contests, like in case of chess competitions.
Resurrection control: rarely (like in 1 out of 1000 picks) the program resurrects an older, already disqualified bot into the gene pool. This increases variablitiy.
With separated gene pools, with crossovers, and disabled "repro", the experiments could be more stabile, variability could be assured, and cancerous mutations do not occur at all. Of course, it is just a guess, tell me if I am mistaken.
And lastly: evolution is heavily directed in DB too. You have to change enviromental factors all the time depending on what characteristics you prefer. Unfortunetaly, that stands for every program in this field.

Suggestions / gene pools
« on: March 30, 2007, 11:30:38 PM »
Maybe Darwinbots could use gene pools as to increase variety. While maintaining variety, this method could also prevent populations from dying out. Of course, in this way inner reproduction, "repro" would have to be disabled. What convinced me is a program called EvBrain (in which neural structures are evolving), and which shows awesome results in evolution, (partly) thanks to the following considerations and long-term experiences of the author:
How variety can be managed is more important than survival of the fittest. Accurate and ultimate selection will devastate the all species. Main evolutionary mechanism is not on natural selection or select fittest creature, but on killing unfittest creature. Implementation : Force resurrected creature from life pool to be mutated all ways in order to make same DNA hardly deposit on nature field. Resurrection control is most important to maintain variety.
So, what do you think?

Simulation Emporium / Guys, you have to see this!
« on: March 22, 2007, 09:04:19 PM »
Its laying eggs!

Suggestions / physics
« on: March 15, 2007, 06:06:29 PM »
Quote from: Zinc Avenger
I am running a similar idea at the moment, I use teleporters to link together different sims with different environment settings. The real trick is to size the outbound teleporters right so the bots spend enough time in each environment but actually have a chance to migrate.

You mean you are running parallel sims on one PC? How do you do that?  

Newbie / veggy help
« on: March 15, 2007, 06:01:16 PM »
stupid cancer      

Newbie / veggy help
« on: March 14, 2007, 01:46:49 PM »
I'm trying to set up a veggy-veggeater coevolution. I set my max veg population to 250. But the population is actually always above that number, and when reaches it from above, ALL the vegs reproduce because of a bad mutation by which the repro condition got erased (if energy>6000...). What can I do?
Thanks in advance

Newbie / How did you find DarwinBots?
« on: March 13, 2007, 11:40:28 AM »
Quote from: Testlund
I used google and I sometimes spend hours trying to find the ultimate artificial life program. I've tried a lot of stuff. I found DarwinBots to be most interesting right now, dispite all the bugs. Hehe. I was surprised to find this forum and that you are still working on the Darwin sim stuff. Allmost thought it was a dead project.  

I was searching a lot, too. There were a lot of artifical evo and alife link collections, but it seems they havent updated them for a lon g time now. Maybe this area is out of fashion.  
I found most of the alife progs having all their advantages and disadvantages. They just simulate different aspects and mechanisms. Darwinbots, framsticks and darwinpond for example, are very entertaining and picturesque, but they are slow, tierra and avida are quick and show inventive evolution, but very incomprehensibly.
After all I'm glad that darwinbots is now so stabile and getting more and more versatile. Thanks for it!

Suggestions / physics
« on: March 13, 2007, 11:10:00 AM »
Quote from: Carlo
It's exactly the original environment grid idea, that is, to divide the field into small cells each one with its own values for physical constants (things like friction, gravity, etc.., but also costs). I think it would be, in itself, a great step towards evolution of diversity.

Ah.. I thought enviroment grid was interconnecting screens via internet or somthing like that.

Suggestions / physics
« on: March 13, 2007, 09:18:57 AM »
It occured to me while thinking about inceasing enviromental diversity: what do you think about splitting the screen in terms of physical properties, like gravity? For example, the upper half of screen would be earth gravity, the lower no gravity, like a pond and its surface. Viscosity, light and such things could also be adjusted this way. I think that would force lineages to diverge, while allowing some slow transfer.
However, I would be suprised hadnt this been already suggested or considered.

Evolution and Internet Sharing Sims / diversity in enviroment
« on: March 11, 2007, 04:33:32 PM »
Quote from: Jez
How about finding several different veg, all nasty, fixing and removing .repro, then using starting positions for those veg to create different zones on the screen, each needing a different ability to cope.
I guess you'd need to adjust nrg input and maybe need to add a neutral veg but I don't see why this wouldn't force boundaries and the need for specialisation.

hmm thanx I'll try that

Biology / natural diversity VS simulated
« on: March 11, 2007, 04:29:42 PM »
Quote from: Endy
Your choice of plants can also have a huge impact. An evolving continually reproducing plant, will adapt to your animals' feeding methods making them adapt in turn.

I always use auto-repopulation for plants because genetic reproduction never works for me somehow: the population is wildly oscillating (when reaches max pop treshold, too many vegs reproduce at the same time) and many other issues I can't recall now. How do you set up a stabile veggy pop? Do the plants learn to move and shoot and things?

Biology / natural diversity VS simulated
« on: March 11, 2007, 04:21:38 PM »
Quote from: Jez
This reminds me of the internet sharing idea that has never been fully implemented, diverse environments joined by an internet pipeline.

That's an interesting idea, but
There is an evo software the name of which I dont remember, with rotating, stiff-body bots called "biots". Internet sharing works fot it. I evolved some cool biots and connected to other enviroments (to 4 directions, via borders of the screen). What do you think happened? There came some tens of thousands generations old terminators and quickly eliminated my whole population.
But if there was some way to standardize enviromental parameters, and then defining some kind of fitness for the creatures involved in the multiPC sim, then screens with close average fitness could be put next to each other, thus creating a great world of adjoining screens. That should elongate stability.

Biology / natural diversity VS simulated
« on: March 10, 2007, 11:52:58 AM »
Quote from: EricL
On the contrary, it's all about processor speed.  The more megaflops, the more physics that can be supportted, the more complex the interactions can be, the more bot-cycles per second per sim, etc.

you are right, but my point is...
if you have nothing interesting evolved in half an hour in your sim, processor time will not solve the problem, just like waiting for some hours more wont solve it. This late being an experience of mine.
So okay, this "completely irrelevant" is a bit exagerrated

Biology / natural diversity VS simulated
« on: March 10, 2007, 11:32:27 AM »
Quote from: Numsgil
If you look at it, most of the things that life takes for granted, such as protein folding, are not things that computers can do particularly well.  They can do them, but you'll always be pushing.  It won't come naturally.

One could argue that life can be abstracted away from the particulars of protein folding, but until it's demonstrated the possibility that life is protein folding, et al, can't be ignored.

Maybe its chemistry and all... and if we can make some abstraction and create some lifelike thing without chemistry, there is still a possibility that we will understand nothing of what is happening. If you know tierra or avida, with their "chemistry" being assembly, you can imagine what I'm speaking of: its hard to figure out how the simplest self-reproductor code works.
Though Darwinbots is very didactical, comprehensible to us, that exactly may be its weakness, since its language is too artificial (high level?)

Biology / natural diversity VS simulated
« on: March 09, 2007, 11:28:49 AM »
Quote from: EricL
Nature is many many many orders of magnitude more complex than DB w.r.t. environment, genetic degrees of freedom, physics, etc.  and yes, nature has more time, not just in terms of linear time but in term of interactions per unit time at everything from the molecular to the organism level.  Nature is parallel.  DB is not.  

The short answer is that DB is so amazingly simplistic in comparsiom that it would be incredible if anything evolves at all (IMHO, we have yet to really see anything actually evolve beyond the most basic seiving response to enviromental constraints).  I could make a strong case that DB has yet to evolve anything.

In that case, I suppose, processor speed is completely irrelevant regarding evolution efficiency. Yes, I have already observed something: if nothing interesting evolves in five minutes, you maight run it for five days with no interesting result at all.  
Maybe those quantum computers will help us.

Pages: [1] 2 3