Quote: Original post by ketek
next time before killing a bug or a small animal, or you when you feel the impulse to do some experimenting on something alive, slam an hammer on your hand and see if it hurts

Quote: Original post by owl
Does the fact pursuing a zebra first to the left, then to the right, then sourounding it, jumping over it and biting it's neck (which we will take as an optimal hunting strategy) gets recorded, submited to analysis/comparision with other experiences, encoded and stored into DNA after being qualified as a good way of getting food?

My beliefs on this are that what is encoded and stored are the fundamental actions and instincts. Complex behaviours like hunting are learned and for most species, this learning is initiated through "play" as a youngster. This play time serves multiple purposes, but it enables the young to build basic sets of complex behaviours, such as 'chase' and 'evade, from those that are hard-wired. Through social education (usually from parent to young, or herd to young) the animal learns even more complex behaviours, but more importantly, learns "appropriate" behaviours (i.e., when to use certain actions). This can be seen in different groups of the same species using different group hunting strategies and tactics. You can even remove several group members and the strategies survive when those members are replaced. The remaining members are able to teach the new members their roles... and through observation and interplay, the newer members learn the roles of others in the group, so that if one of them is lost, they too can teach a replacement.

My earlier point was that no one expects this "social education" to be encoded. The building blocks are encoded and the environment decodes that information (which includes the social mechanisms for building more complex actions from the basic set) to produce the 'intelligent' behaviour.

Of course, that's just my view point... and it could be wrong! ;)

Cheers,

Timkin

Quote: And maybe it's descendency will too.

Of course the point is moot here, since we're talking about ants, and workers are
infertile.

Quote: Original post by Timkin
Complex behaviours like hunting are learned and for most species, this learning is initiated through "play" as a youngster.

Yeap. There is a very nice article about playing here on gamedev.

Anyway, I'm not sure if lions, for example, are taught by the adults on how to hunt.

Cats, as other example, know how to hunt a bird without having to be taught how to do it. The chasing play they do, comes up instinctically, nobody teaches that to them. I think that's also true for lions.

Quote: Original post by Timkin
Of course, that's just my view point... and it could be wrong! ;)

You know, I'm posting this exactly with the hope of being proven wrong. I wouldn't like to waste my time thinking about things that aren't right. :)

I think we agree quite a lot in our points of view anyway.

[Edited by - owl on November 24, 2005 8:44:55 PM]

Quote: Original post by LeGreg

Quote: And maybe it's descendency will too.

Of course the point is moot here, since we're talking about ants, and workers are
infertile.

That seems to be true! So, HTF do they evolve? Do they? :)

Quote: Original post by owl

Quote: Original post by LeGreg

Quote: And maybe it's descendency will too.

Of course the point is moot here, since we're talking about ants, and workers are
infertile.

That seems to be true. So, HTF do they evolve? Do they?

I dont believe too much in natural selection, but assuming it is true:
An ant queen that gets a mutation that make her produce better workers will have more chance of seeing her nest survives long enough for her to pass that gene into daughters that will make other nests with superior workers and spread that good gene.

Thus sterile ant workers can evolve from mutation of their queen. I would go as far as thinking that since genetic material comes from a very small percentage of the population, a good mutation can spread faster than with your average "everyone can spread their juice" mammal.

That would be an interesting experience with GA... Queens vs Couple evolution deathmatch.

Quote: Original post by Steadtler

Quote: Original post by owl

Quote: Original post by LeGreg

Quote: And maybe it's descendency will too.

Of course the point is moot here, since we're talking about ants, and workers are
infertile.

That seems to be true. So, HTF do they evolve? Do they?

I dont believe too much in natural selection, but assuming it is true:
An ant queen that gets a mutation that make her produce better workers will have more chance of seeing her nest survives long enough for her to pass that gene into daughters that will make other nests with superior workers and spread that good gene.

Thus sterile ant workers can evolve from mutation of their queen. I would go as far as thinking that since genetic material comes from a very small percentage of the population, a good mutation can spread faster than with your average "everyone can spread their juice" mammal.

But workers have special behaviors and characteristics. Are their behaviors product of "random" mutations on a creature (the queen) that do not behave like them? I find that hard to sustain logically and mathematically. Given an infinite number of posibilities for a mutation to be beneficial or detrimental, on a finite number of ants, there are big changes of having a long period of detrimental mutations that would doom the species.

Evolution does not work by recording the experiences and choices of individuals. Rather the biology of the individual affects what choices it will make. A large part of the structural biology is determined by the genetics. If the genes affect the individual's behavior in a way that increases survivability, that individual is more likely to pass its genes along to future generations. The genetics can also be affected by mutations when the genetic code is not copied exactly from one generation to the next. Mutations that cause biological advantages, including improving instinctual behaviors, are more likely to survive to future generations.

In the case of ants and similiar animals, all of the ants will possess the same genetic code as the queen. If a mutation in a queen ant results in the workers performing better (for example making them behave in a smarter way that enhances their survivability) then that will also help the colony to survive. If the colony survives, then that queen ant with the beneficial mutation will survive and produce more offspring, some of which will be queens which will form more colonies.

Wild animals that instinctively know how to hunt will survive better than animals that don't and thus will produce more offspring. An animal that learns how to hunt cannot pass this knowledge through genetics. Intelligence is a useful trait because it allows the parents to teach the offspring, so intelligence can enhance the survivability of a species and cause it to produce more offspring.

Quote: Original post by Anonymous Poster
Evolution does not work by recording the experiences and choices of individuals. Rather the biology of the individual affects what choices it will make....

Fair enough. Now, why does the biology of the individual changes in the first place?

Quote: Original post by Anonymous PosterThe genetics can also be affected by mutations...

Read my post above. If mutation were the only factor of evolution then life should have ended long time ago. What else causes species to evolve?

Quote: Original post by Anonymous PosterThe genetics can also be
Wild animals that instinctively know how to hunt will survive better...

We agree. But how do they know how to hunt instinctively? Or how to build a nest, a trap, or to bring a pray to their mate to gain it's sexual favors? Through "random" mutations?

Quote: Original post by owl

But workers have special behaviors and characteristics. Are their behaviors product of "random" mutations on a creature (the queen) that do not behave like them? I find that hard to sustain logically and mathematically. Given an infinite number of posibilities for a mutation to be beneficial or detrimental, on a finite number of ants, there are big changes of having a long period of detrimental mutations that would doom the species.

I dont see what is the problem, logically or mathematically. First, that the queen (and princess, etc) does not behave like the workers is inconsequential: Only the queen (and the few males) can pass genetic information. Thus anything that doesnt come from each worker's experience or reflexion comes from the queen. The queen is not a worker, but she holds every gene needed to create a worker. Physical or behavioral, it does not matter.

Second, according to the evolution theory, a bad mutation would decrease the chances of survival of that particular nest of ants, thus would probably be eliminated through competition with the other nests and species.

What I am thinking, is that the "queen" model would accelerate that evolution. On the normal "couple" model, a beneficial mutation will take many, many generations to spread. On a "queen" model, it could spread to almost the entire colony in a few generation. Likely, a detrimental one will spread as fast and doom the nest fairly quickly. On an evolution point of view, it is like each nest is a single individual. Which is not that crazy, since they almost all share the same genetic code...