Virtual Genetic Evolution

Sleeperservice writes "This story at New Scientist describes how, using cell simulation in computers, evolution can be simulated. How long until we can work out what the DNA sequence for a Dragon should be I wonder?"

Comments

How long before we get virtual humans?

[strredwolf]

They're speeding up evolution via a cell simulator. How long will it be before we get true AI by teaching a virtual human that was just freshly evolved? :)

Re:How long before we get virtual humans?

[khuber]

Can you speed up evolution? Can you simulate complex evolution with digital computers?

I don't know. This is very simplified "evolution".

Evolutionary systems tend to crap out and stop evolving quite quickly. I'd expect this one to do the same. To expect brains and intelligence to evolve is overly optimistic IMO, since nobody has accomplished that yet.

The Santa Fe Institute (I believe) had a much more interesting ALife system where the little guys would live in a distributed ecosystem on various internetworked computers.

It seems reasonable to me that you'd have to evolve an entire world/ecosystem, not just organisms.

-Kevin

Re:How long before we get virtual humans?

[powerlinekid]

They discuss the last part of your comment in the article. What drives the "evolution" is the complexity of the task at hand. Whether its pushing a box or something else, they say that it is very very important to keep coming up with new tasks to push the limits of the little cyber-critters. I would think that without an enviroment that demands adaptation, evolution would occur either very slowly or not at all. I mean how far do you think we would of evolved if say everyone had food and didn't have to worry about predators? We'd probably evolve to the point where we die off at an acceptable rate in order not to cause overcrowding and hurt the species in the long term. Of course some rouge critter would probably develop a gene to live longer, so maybe you can't stop evolution anyway.

Just in case someone says that evolution is "just gene mutation", I would disagree. I would say that evolution is "gene mutation" that benefits the entity that has mutated and made it better adapted to it's enviroment than say average-joe-critter.

I don't really see how this could work

[Anonymous Coward]

They're hoping that one day the artificial creatures will develop artificial intelligence. They have to tell this what to develop (like an lgiht sensor for sight or a limb for motion.) We don't know how intelligence came about so how could we tell this program how to develop some? It's just like all the other things where you tell a computer a bunch of stuff about the world and hope it becomes intelligent, sure it has all this information but without knowing how to use it on it's own, it becomes quite useless.

Re:I don't really see how this could work

[mofolotopo]

That's not actually how artificial life works. They don't tell them that they need eyes or bigger muscles or whatever, they just eliminate the ones that are unsuccessful. This process leads to the development of things like eyes and muscles because the ones that developed those by the process of mutation tend to be culled less than those that didn't. Tom Ray's Tierra simulation has absolutely NO global rules and no specifications about what sort of things are expected to develop, and he gets complex ecological interactions such as parasites and hyperparasites.

You have, however, hit on an important point. We know how to objectively judge box-pushing behavior, speed, and other physical measurements, so we can develop rule sets to encourage the development of traits to maximize those things. Quantifying intelligent behavior is much more difficult, making the evolution of intelligent behavior problematic. That doesn't mean it's impossible, it just means that we need to think more on the difference between simple problem solving and true intelligence, and most importantly how to judge what intelligent behavior is without trying to figure out how to create it.

Not the craziest thing I've ever heard of.

[Graelin]

I suppose when you get down to it we're all nothing more than electrical pulses running down organic "wires" anyway. That is, if you can ignore the support systems (cardio, and the like). It seems possible that this could be represented by the extremely logic nature of a computer.

The problem is they're not really simulating life since they pre-program the genome anyway. They are, in effect, telling it how a "complex web of neurons" (aka. brain) should look and operate.

The article claims they've created creatures with very primitive mobility and senses. Going from there to sentience, wants, needs, emotions, reproduction etc. will be very difficult to simulate in an environment where chaos does not exist.

OTOH, sounds like they've found a really clever compression scheme.

The sheer Arrogance of slashdot posters

[siewsk]

The sheer arrogance of the slashdot poster is unbelievable. Just listen to this:

This story at New Scientist describes how, using cell simulation in computers, evolution can be simulated. How long until we can work out what the DNA sequence for a Dragon should be I wonder?

Come on! The simulation is about as closer to reality as the economist assumptions of the rational consumers. You don't know how many corners they can cut in their simulation. It's just a simulation and it is as good as the assumptions made by the people who created the simulation. Real life is messy and difficult to predict. It will be hundreds of years before they get the simulation even close to the true working of a real cell.

Re:Until they can deal with ACGT...

[Daniel Dvorkin]

Um ... we know perfectly well what they do: make copies of themselves. Now, the circumstances under which they make copies of themselves, and close analogues of themselves in RNA, is in most cases still an open question -- but simulations like this are an important step on the road to figuring that out. Of course it's trial and error; that's how science works.

Re:Heathens

[EatYourGreens]

Before I reply to this article, I would like to say that I am an Evangelical Christian. I believe in God, the death and resurrection of Jesus Christ, and that the Bible is the revealed work of the Holy Spirit. ... and I believe that we are here as the result of many years of evolution. I am aware that this thread is not about the evolution/creation debate, so I am not going to try to counter the poster's comments point for point. I would just like /.ers to know that there are Christians who can still think straight (at least in Germany). I believe in God because I find the evidence overwhelming. I believe in evolution because I find the evidence overwhelming. It belittles God and makes Christians look stupid when Christians say that evolution runs counter to the second law of thermodynamics, or when they point to mistakes that scientists have made in the past while trying to formulate the theory of evolution. If evolution could be so easily disproved, then I think the thousands of scientists who have studied evolution over the years would have worked this out by now. I am sure I will be flamed for this, but I am prepared to take the risk. However, I have written this for all the Cristians in the world who have not had their brains removed, who can worship God and marvel at the amazing things evolution has made. For these Christian (I know of many who hold this view, several educated to Doctorate level) there is no tension.