Radically speed up evolution

Purely for scientific curiousity, what is the plausibility of speeding up evolution so to speak, to see what certain animals wuld become in this experiment.

perhaps by choosing one randomly selected sperm and another egg from a mouse, and the moment you are able to extract another sperm and egg do the same, and again and again, for perhaps 4000 - 10000 times, accounting for a substantial amount of time, and see what happens?

is there a quicker way than waiting until one can extract the sex cells?

Would not work.

Evolution doesn't occur "just because". You need mutations. Then, you need those mutations to give the organism an advantage in it's environment so they are more successful, breed, and spread the mutated gene around. Almsot all mutations are bad anyway and cause the organism to die, it's the lucky select few that give an advantage and get apssed one.

You could artificially speed up the breeding cycle of a species, but it wouldn't be 'evolution' proper, it would just be random mutations. It's not 'natural' selection, sicne it's a scientist choosing what genes get passed on regardless of whether they give an organism an advantage or not.

The only way to see what an organism will truly evolve into is to watch it occur naturally.

Its easier just to do it with virtual worlds. Saw a show a while back about experiments using supercomputers programmed with a few environmental variables and some initial life starting points of various as well as routines to generate new characteristics and eliminate others as advantagous. The goal for every simulation was to obtain a piece of food. The computer ran through everything from strange snake-like creatures to 8 legged monsters that could not be stopped, virtual species rose and fell and in the end of every run the end result was essentially us.. bipeds. I dont have a source but Im sure the exact info is easily searchable if necessary. This was many years ago so today I would think you could do the same with far more variables and simulated details.

.
You could do it by artificially mutatating gene by doing genetic insertions.

But to create a completely new gene you would have to start closer to scratch. To do this with any high degree of success would entail an understanding of protien folding and sculpting that we simply don't have yet.

Maybe the most profound developments of evolution will always depend on nature, because we are unable imagine some of the possibilites.
.

Originally posted by Esoterica
The only way to see what an organism will truly evolve into is to watch it occur naturally.

Very well said, although I like apc's idea. That really sounds fantastic. pm if you find the link.

Evolution is always as a result of the environments or a random mutation. Either way, it's incredibly slow and not doable by hand. But if you use a computer simulation, then you can do it much faster. I've done the same thing with astronomy. You wanna see how the universe started? You take what you know, plug it into a simulator and run it. If the result isn't (and it won't be) how it is know, you change some of the variables and run it again.

When it comes to huge things YOU can't change and that move SLOW, use a computer. Why? Because they're obviously smarter and more powerful than we are.

apc: They should've put in cyborgs, see how they would've done. Or magic!

I think to do the sim accurately would require exponentially more computing power than we currently have.

Sure, you could run a limited simulation, with a couple of interacting organisms, and a few environmental parameters, with a handful of variables. But that's nowhere near representative of the actual process by which organisms evolve.

But imagine trying to compute the earth...

I'd love to do it one day, but it's just not feasible right now.

Maybe a realistic goal would be to compute one isolated ecosystem, like..Madagascar (my personal pick). Then at least you would only have a few hundred thousand species, clustered in one environment, and a million or so variables, in terms of weather patterns, sunlight, breeding cycles, predation, disease, mutation, etc., etc..

In any case, it's a daunting task, but one well worth the effort.

If anyone wants to begin the process, start a company, get the ball rolling, look me up. I'd be more than happy to play data entry monkey or researcher for such a noble, fascinating endeavor. Evolution is my passion, and I'd absolutely love to see it represented accurately in a computer model. That would be like a dream come true.

Maybe a realistic goal would be to compute one isolated ecosystem, like..Madagascar (my personal pick). Then at least you would only have a few hundred thousand species, clustered in one environment, and a million or so variables, in terms of weather patterns, sunlight, breeding cycles, predation, disease, mutation, etc., etc..

Blue Gene could probably do that, not sure of the specs of the comp though. It would take quite a while to compute. What about a SETI@HOME type of project?

I know that it would not show how the animal would evolve naturally, obviously. it would be purely a scientific curiosity project. the mutations would still occur.

it is similar to breeding, which is like artificial evolution. a person chooses the male dog with longest fur and female with longest fur, does the same with the offspring, and so on, and so on, until a dog is born with sooooooooo long fur.

this would be one step further. perhaps without knowing it the scientists choose a cat with the longest teeth, and again and again until a sabertooth species of cat is produced.

anything could happen....

Sardion
Very good point. It might be the only way, without using warehouses of dedicated slave computers. Theres no way the financial gain from this sort of program could offset that sort of cost. Unless of course we have a wealthy philanthropist in our midst... No? Damn.

I too thought about a time share, network type approach, like SETI, it could probably work for doing the actual computations.

Obviously it would require a good amount of participation, and the data bases would probably still have to be compiled and stored somewhere, there's no way individual computer users would allow gigs upon gigs of disk space to be hijacked for Golden Lemur behavioral algorithms.

Then, the participating computers could work on a given number of interactions, either between two species, or between climate and species, disease and species, or breeding combinations, between wind and spores, etc., then take the result of that interaction and feed it back into a separate database, which would begin to organize the results into coded databases, accessible by the simulation software. Once you had all the possible results accounted for, stored and verified, you could begin the process of setting up the world, virtual spatial relations, and probabilities could be established for the various interactions.

I agree it could work, but it would still require an enormous amount of computer power, and god forbid you should ever lose the data... Years of work down the drain... Ack!

Anyway, it would be a technical challenge, a HUGE technical challenge, but I really believe it would be worthwhile.

Totally. There's no way a conventional computer could pull something off like that, the idea was to just test a few things out.

I don't know how well an @home thing would work. You wouldn't want to split up the variables because everything affects everything else in ways we can't comprehend (hence the computer) so it would only really make sense to have a huge, powerful compy to do it all.

But as you said, individual computers could do interactions between two species, but that doesn't really represent how those two would interact in the whole picture.

What could be done, however, is have a giant computer run something, and then when it gets to something that HAS to be done, it send it out to hundreds of computers to think about. Each computer would get the data slightly altered so that really, millions of calculalations could be pooled together. When that data is returned to the mother computer, it then carries on with each scenario.

The amount of space required is enormous. Terabytes wouldn't do it justice. But, as you said, it'd be a fantastic thing to do. And once we finish it, we can glean data that we would never have known before and could see exactly how things might turn out for each actions. We could run it again to see what would happen if we changed these emissions laws, or if we didn't pass that law, or if 8-legged fire breathing creatures landed.

The only (hah) problemis that there's no way to account for human response to situations.

Wow i get it now......and once the computers have done evolving virtual dna it can be converted to actual dna and voila! new super evolved animal!!!

Keep in mind everyone that bacteria and single celled organisms have extremely high rates of reproduction. Evolutionary time is generational time. So if you want to see 'fast' evolution you want 'fast' generations.

Also, in the orignal consideration, what is one looking at? A dog, for example, can't have its evolution artificially 'speeded up'. A dog isn't moving along a pre-determined evolutionary path. Taking it into a lab and, say, increasing, super-drastically, the numbers of generations it can have in a short period of time will technically be speeding up evolution, but what you get after like a million generations isn't going to in anyway resemble a 'control' natural population that goes thru a million generations on its own.

Also, evolution is just a change in gene frequencies within a population. So if you want to evolvea 'new' species of dog from domesticated dogs, well, bio-engineering would be the fastest way to actually accomplish it.

As far as using computer to simulate genes in a population, undergoing mixing and mutation and selection, well, at what point do you stop? You'd want it, perhaps, to have certain characteristics, no? So why go thru the simulation, you'd have an 'end point' in mind already. Better to just create the genome from scratch, so to speak, no?

If you want to watch "super fast evolution", keep your eyes on flu virii.
They are the fastest adapting organism on the planet.
They exist everywhere, including the blood of every human.

of course the evolution of a dog over millions of years would turn out completely different than the experiment.

A virus is exactly what I had in mind with SARS being a perfect example.
Instant evolution just add host.

Simulating a virus would be helpful for disease control, very much so.

Nygdan: Wha? Who suggested taking a dog and forcing generations upon generations on it? Because that's a crappy idea, as you so eloquently put.

As for the goal in mind, well, yes, sorta. You might have a goal in mind, such as wanting to see what happens to a specific population, but you could very well just want to see how things would change if we kept on our current path, or what would happen in the broader picture if we changed to alternative energy sources or changed emissions standards. Stuff like that.

I'm going to breed dire wolves.

The ultimate guard dogs.

It would only take..I dunno..150 million dollars. No biggie.

Seriously.

Sabertooth cats are easy in comparison. They've reappeared something like 6 times over the course of pre-history. The code is probably a snap to isolate and key-up.

But the Dire Wolves..that's my plan. Gonna call my pair Cerebus and Fenrir.

Sit boys.

the dog thing was just an example for explanation.

it would obviously be betta to do sumat more interesting like a komodo dragon. (breed them MASSIVE......with wings)

[edit on 2-5-2005 by Shadow88]

yes breeding a dragon would be kool i think the thing to do is get really really into this so many good things can come of this we would be creating new speicies! not killing them or hunting them to extinction i think our tech will take us soon into the modern equivelent of the industrial evolution "the genetic revolution" at this point im quite surprised no relgious types come here and tell you all off about playing god

go for it! a business in geneetically modifieing existing species to make them more adaptable brilliant idea

note: there would be mutations that could cause them to be disabled somthing like 1/4 from litters like cloned cats etc if it was done natrally by speeding up there generations and ammount of offspring we would'nt cause that so it would put another relegios point down

i say f... religion......it was made hundreds of years ago.......less and less people follow it or rely on it anymore........god is not real unless god is really aliens........if they protest then just say "haha (two fingas up 2 u!) im off sumwhere else! follow me if ya like! am off 2 breed dragons!!!!"

and no offense to any religious folks but stop being killjoys against EVERYTHING "oooo.....cloning is bad......ooooo........GM is bad......"

IF GOD DIDNT WANT US TO DO THIS...... WHY WOULD HE CREATE US WITH THE ABILITY TO BECOME INTELLIGENT ENUFF TO DO THIS STUFF!!!!

man thats naughty