Hi Salva,

I'm not sure if this answers your question, but in our institute a modelling study about self-organisation (of mussels in a mussel bed) at different spatial/organisation levels revealed that these nested levels help to make the total system (mussel bed) more resilient. See: http://www.nature.com/ncomms/2014/141022/ncomms6234/full/ncomms6234.html

I hope this gives some food for thought?

Regards, Jim

Hello Jim,

Definitely a very interesting work, looks like a new unit of selection on the making! This has provided me with some very fine papers I new nothing about and sparked a some thoughts on the subject, I highly appreciate you took the time to answer.

However, in my question I was aiming at more abstract models in artificial life (like Tierra & Avida) precisely because of their contrast to systems designed to model actual processes.

After posting this question I found about EVOLVE 4.0 (which seems like an updated version of this very same models I mentioned before). But there seems to be not as much information as I would like on the subject. What do you think about it?

Hi

To me this is a basis of thermodynamics, the reciprocal flow of matter energy, and ways of evolving strains of organisms in the random process of chaos by the expression of a multidimensional perfect niche for survival, expressed in their balance population of their genes. A complex network in narture demand a large transfer of energy flow.

In the artificiality that would operate in the same way, from simple to complex. As we want to advance in complexity must use more variables and dimensions. Obviously more energy demand.

I like your idea work.

Hello Fernando,

I do agree on the conceptual framework that you are exposing, since Ilya Prigogine the termodynamical vision of complexification is a common theme to which we resort to explain certain features of dissipative systems, like living ones. Yet complex phenomena (Chaotic dynamics & Self Organized Criticality to name some that you also mention) can be observed in "very simple" systems with not-so-high dimensionality.

I wonder if it will be possible for us to recreate the emergence of complex living systems without including arbitrarily large and complicated structures. Nonetheless, if you know an artificial life program following the strategy you suggest that deals with the issues we are discussing I'd very much like to hear about it.

Thank you for posting an anwer,

Salva

Hi Salva,

I was thinking a bit more about your question. Maybe what you are looking for is already happening? Your assumption is that organisms need to become more and more complex in the artificial world. I don't think there is a good reason for this to happen in models like Avida and Tierra. However, this does not mean that complexity is decreasing (simplification). I think the opposite is happening. You get more different species (incl. parasites and hyperparasites) which are dependent or even interdependent on each other for their survival in this artificial world. Thus, complexity is increasing. You cannot simply characterise the system by number of species or so. You need to include their connectivity, which emerged from the model. So in my opinion this is an increase in complexity of the community.

I think to get higher levels of organisation within the organisms, you have to embed them in a more complex environments. For example by adding space, the organisms may need to develop locomotion or other adaptation to get food or become very efficient with very little food. I particularly like models that include L-systems to mimic e.g. plant morphology that become fit to adapt to a mix of environmental pay-offs. Think about harvesting sunlight and water/nutrients at the same time.

Maybe an interesting paper is: http://www.sciencedirect.com/science/article/pii/S0304380010004850

I hope this spurs thinking/discussing this issue further. I'm curious about your thoughts about this.

Regards, Jim 

Hello Jim,

First of all, thanks again for taking the time to answer. I find this discussion really interesting and I absolutelly think you hit crux of the matter.

As you say, in Tierra and Avida we do observe an increase of the total complexity of the system by the creation of new species that, although "simpler" internally, create a complex network of ecological interactions that cannot be reduced to the properties of the individual agents.

That said, in the real evolution of our biosphere we do observe not only this process, but also the emergence of new levels of selection and more and more complex individuals. So the thing is, are we missing some key ingredient when we program artificial life models that forbids creativity and complexification?

You mentioned embedding the artificial organisms in a more complex environment. Do you think that organismal complexity is, in a sense, inherited from the physico-chemical environment? Putting it in other words, does the physical substrate act as a template for complexity of biological entities? If so, do we have to include a complete / fairly complete description of our world in order to get the same kind of phenomena?

Onto the last part, the paper on plant morphology you provided is superb. I came across some similar work by Karl Niklas (and also Richard Dawkins' biomorphs) some time ago and I was just amazed by the generative potential of grammars. I wonder if there is some lesson here waiting to be found...

Salva

Hi Salva,

In short. Yes and No. ;)

Yes I think that if you embed artificial organisms in more complex environments that this will lead to more complex organisms if in this complexer environment multiple ways of surviving are possible. Like the virtual plants, in the paper I suggested, develop different morphologies and life-history traits to cope with stress and disturbance in the landscape. Thus, the physical template will (to a certain extant) explain organismal complexity.

But, no... I don't think this is the whole story. And no I don't think it is a fruitful endeavour to model more and more complex environments to see if this will improve organismal complexity. Maybe it is good to do it as a sort of 'proof of concept' to show that this is one mechanism to get organismal complexity to increase. But I general like models for what they are not: copies of reality. To me modelling is a way of 'extended thought experimentation'. If you follow a certain set of rules consistently, what would be possible? Artificial life is a nice example, that might help us understand how organisms in the real-world came to be.

One last thing I think might make these artificial life models more interesting is, if organisms feedback on the complexity of the environment. I think this could lead to real interesting (emergent) behaviour. If the evolution of organisms contributes to the complexity of the environment, and the environment complexity contributes to the evolution of the organism, their is the potential for a feedback that propels increasing organismal complexity. I'm not sure if this has been done before, but I think that would be a real interesting endeavour.

What do you think? Would that be an interesting model to develop and explore?

Regards, Jim

Nice!

So we should include specific ways for organisms to manipulate environmental complexity at all scales. I'm not sure how to formalize this... niche construction should be the key aspect of the model. Anyway, thank you so much for your replies, I will check arround if there is something similar to what you suggested and I will leave it here.

Salva

Hi Salva.

You let me think about it and I have no knowledge of an algorithm that can give us an idea of using it and get artificially modeling complex or something simple.

I Dream in parsimonious nature in its random trends and limitations of excess to the limits of its ability to be allowed. For natural populations carrying capacity becomes the mold as possible and the way of their ability so that each individual can develop their potential multidimensionality for their life choices.

Think of some organisms to maintain their populations must have both sexes and complex behaviors to ensure replicability. There are also other similar organism in biological levels, where they only have females and have ability to replicate through a cloning without further energy expenditure of defending territories, have an expensive courtship or compete for effective partners. Here I see no complexity and simplicity. The issue here is sexual reproduction and parthenogenesis as a possible example of an issue that has been talking with Jim.

Sorry if I spilled the subject but I think even nature is chaotic for me and their strategies are carried parsimony to be a limit.

Fernando

May want to consider Richard Dawkins efforts at simulating evolution- and the premise that evolution involves successive and cumulative complexity. In particular the links deal with his WEASEL simulation (this deals with increasingly complex intelligent behavior): #1 link

#2 link: discusses a simulation that Dawkins created called "arthromorphs"

#3 link: discusses a simulation called biomorphs inspired by Dawkins

http://tccsa.tc/articles/weasel.html

http://arthromorphs.sourceforge.net/

http://www.staff.science.uu.nl/~beuke106/dawkins/dawkins.html