Artificial Evolution

[David Sklansky]

Suppose you had available to you a billion chimpanzees, state of the art DNA evaluation equipment, all the help you need and a billion years to conduct your experiment. Which is to produce a human. By mating exactly those chimpanzees who you choose with the help of your equipment along with their children and their childrens children. If you are an expert biologist would you succed?

[SNOWBALL]

Are you working on the assumption that chimps are our ancestors? They're not. We're cousins that descended from the same ancestor.

[Phil153]

Interesting question. I don't know if this would be possible from chimpanzees, because they've evolved along a different branch. They may have developed traits that prevent devolution to pre-human form.

For example, I don't think you could ever turn a horse into a echidna, even though the same two could develop from the correct pre-horse, pre-echnida ancestor.

The extreme similarity between chimp and human DNA would lead me to say it's doable, but I don't know.

[Sephus]

i don't think you could succeed to the extent that you would end up with a population that other expert biologists would find impossible to distinguish from present day humans.

can't say that's a terribly educated guess though.

[PairTheBoard]

Maybe to avoid technical difficulties, instead of asking for a result identical to the human genome, ask for a human-like result with a neo-cortex and human speaking and reasoning ablilities.

On what grounds could anyone hesitate to say Yes to that proposition?

PairTheBoard

[vhawk01]

I think you'd probably be a pretty huge dog to create anything anyone would ever confuse with a human. That includes on, say, a Turing test.

[vhawk01]

[ QUOTE ]
Maybe to avoid technical difficulties, instead of asking for a result identical to the human genome, ask for a human-like result with a neo-cortex and human speaking and reasoning ablilities.

On what grounds could anyone hesitate to say Yes to that proposition?

PairTheBoard

[/ QUOTE ]

I would definitely hesitate.

[PairTheBoard]

[ QUOTE ]
[ QUOTE ]
Maybe to avoid technical difficulties, instead of asking for a result identical to the human genome, ask for a human-like result with a neo-cortex and human speaking and reasoning ablilities.

On what grounds could anyone hesitate to say Yes to that proposition?

PairTheBoard

[/ QUOTE ]

I would definitely hesitate.

[/ QUOTE ]

Really? On what grounds? I'm puzzled because I find myself hesitating as well. Yet it seems like it should be a slam dunk Yes if you are confident in Evolutionary Theory.

PairTheBoard

[Phil153]

http://img241.imageshack.us/img241/1475/australopithecuserectusev6.jpg
This bit doesn't look too hard.

And given the extreme malleability of the human brain, the existing advanced social skills of apes, and their existing intelligence (said to be equivalent to a 5 year old), I see no reason why a simple brain size increase wouldn't make them more intelligence and conscious.

[NotReady]

[ QUOTE ]

By mating exactly those chimpanzees who you choose with the help of your equipment along with their children and their childrens children. If you are an expert biologist would you succed?


[/ QUOTE ]

Hmm, it just occurred to me. Instead of trying to turn computers into AI why don't scientists start this project? Shouldn't take a million years since we're already almost god.

[TomCowley]

I have no doubt that a scientist who could predict the reproductive viability of each DNA combination could get there. It's a slam dunk that walking back up to a common ancestor, then down to human, is possible in a billion years. A scientist who had to randomly try conversion paths might run out of time (because the "selection pressure" could well be nonexistent-to-counterproductive)

[soon2bepro]

First of all, let me say that different human beings are different from each other, we call all of us "human" because there's no other alive type of animal that resembles us enough. If hom0 erectus and others were still kicking, we may have some trouble there. So for this experiment we need to set a standard for how much like today's various kinds of humans should the evolved chimp(s) be.

Second, humans didn't evolve from chimps. I don't know enough about evolution theory, but I believe I read/heard here and there that some scientists believe evolution only goes towards more complex DNA, so that mutations "backwards" wouldn't work, or something. It sounds a little wacky to me, but if this was true then you probably couldn't get a human from a chimp.

Leaving those two points aside, on to your question, which is very interesting.

Something that would really help answering this question is finding out what causes each kind of mutation (which as of today we call random mutations if I'm not mistaken).

If these can be isolated, then the answer would be yes I presume. If not, then the answer would probably be that there would be a probability of success. Given the number of specimens and the vast time available, I'd say it'd be a rather large one, but that would depend mostly on the chance of the correct "random" mutations happening.

Selection (be it natural or artificial) can only do so much. If extremely unlikely mutations need to take place, then there's definitely a chance that it couldn't be done without somehow making them happen.

Oh by the way (and I know it's not the point), given a billion years of progressive scientific research with the means and manpower you say, I believe it is extremely likely that you could get advanced enough to create human DNA by combining carbon and the other elements that make it up.

[SNOWBALL]

[ QUOTE ]

I think you'd probably be a pretty huge dog to create anything anyone would ever confuse with a human. That includes on, say, a Turing test.



[/ QUOTE ]

no way. With this kind of timeline and the most advanced artificial selection techniques, I think you could easily come up with a species much smarter than humans, and therefore a species capable of imitating us.

[Piers]

Chimps not being our ancestor makes it particularly tricky, but a billion years and expert guidance yes, but only if the biologist make use of the imminent lifespan enhancing DNA upgrades /images/graemlins/crazy.gif

If you allow tinkering with the DNA than its almost trivial /images/graemlins/wink.gif

[MidGe]

[ QUOTE ]
Suppose you had available to you a billion chimpanzees, state of the art DNA evaluation equipment, all the help you need and a billion years to conduct your experiment. Which is to produce a human. By mating exactly those chimpanzees who you choose with the help of your equipment along with their children and their childrens children. If you are an expert biologist would you succed?

[/ QUOTE ]

Gee, I'd do better than merely humans! Evolution is not that efficient after all... it simply works!

[Rduke55]

[ QUOTE ]
I see no reason why a simple brain size increase wouldn't make them more intelligence and conscious.

[/ QUOTE ]

There's nothing simple about increasing brain size. First you have to alter the skull and jaw to fit it in there. The hips would have to change to allow the larger head to get through during childbirth(and if you want this animal to walk the hips have to change in a specific way - and then you'll have to change the brain for walking - and if you want this critter to be able to manipulate objects with fine motor control of the fingers that's another big change to certain parts of the brain). If humans are any indication, the different birth canal isn't enough, the gestation period would have to change and you'd have to give birth to a more altricial offspring. That would require a radical shift in the social organization since the offspring would wean much later and you'd probably need biparental care instead of the harem situations in the great apes (more brain changes - in fact many neuroscientists see the relationship between sociality and brain size one of the strongest).
Then the metabolic demands would require a shortened GI tract. Also because of the increased metabolic demands (nervous tissue is the most metabolically demanding tissue) you'd have to change the diet significantly - that means everything from teeth to enzymes to organs. Also, if we are talking meat (and we probably are) you need to change a lot of behaviors because occasional hunting (like the apes currently do) is way different that hunting for sustinence. Following herds, greater group cooperation, etc.
Not to mention the problems inherent in increasing the size of the brain. We've discussed some of these before. Things like connectivity and processing will change dramatically requiring new modules or areas of the brain that would need to be connected in brand new ways.
All that's just for increasing brain size.

Also, we've had a similarly sized brain for a lot longer than we've had complex language or culture so it's not just getting more material in there.

[Rduke55]

My knee-jerk reaction is that it would be more unlikely than many people would think because of reasons like constraints that Phil153 and others were alluding to, among other reasons.

But a billion years is a mind-numbing amount of time so who knows?

An important point is that do we have the ability to change genes? Because the way you phrased your question it sounded like we have to work with what we already have or genetic changes that crop up.

[vhawk01]

[ QUOTE ]
[ QUOTE ]

By mating exactly those chimpanzees who you choose with the help of your equipment along with their children and their childrens children. If you are an expert biologist would you succed?


[/ QUOTE ]

Hmm, it just occurred to me. Instead of trying to turn computers into AI why don't scientists start this project? Shouldn't take a million years since we're already almost god.

[/ QUOTE ]

Shouldn't take a million....just like 850,000? You know tough it is to get funding for 10,000 year projects, much less 850,000 year ones? Think of the overhead!

[vhawk01]

[ QUOTE ]
I have no doubt that a scientist who could predict the reproductive viability of each DNA combination could get there. It's a slam dunk that walking back up to a common ancestor, then down to human, is possible in a billion years. A scientist who had to randomly try conversion paths might run out of time (because the "selection pressure" could well be nonexistent-to-counterproductive)

[/ QUOTE ]

Right, and I don't think we have enough understanding of what genes cause what to make this anywhere near maximally efficient. In fact, I think it would be closer to random guessing than perfect efficiency. Not even that far removed from dog breeding. So, we'd get a bunch of chimps with bigger heads, and I'm less confident than Phil is that this automatically means language, culture and exponential expansion of intelligence.

[Piers]

[ QUOTE ]
and I don't think we have enough understanding of what genes cause what to make this anywhere near maximally efficient

[/ QUOTE ]

Sure, but we will know a lot more in fifty years. And given a billion year project you don’t need to be that efficient.

[ QUOTE ]
and I'm less confident than Phil is that this automatically means language, culture and exponential expansion of intelligence.

[/ QUOTE ]

Language? Intelligence? I thought we were just trying to create a chick I could get pregnant.

Making chimps super intelligence is completely different and more straightforward. Just let the bright ones breed (IQ tests), take a lot less than a million years. But you wont get humans of course. The only major problem is going to be keeping the specimens happy once they reach human intelligence.

[Rduke55]

[ QUOTE ]
[ QUOTE ]

I think you'd probably be a pretty huge dog to create anything anyone would ever confuse with a human. That includes on, say, a Turing test.



[/ QUOTE ]
I think you could easily come up with a species much smarter than humans, and therefore a species capable of imitating us.

[/ QUOTE ]

Why do you think we could get smarter than humans?

Also, smarter in what way?

An interesting thing would be to see how smarter humans could get. I think we're pretty close right now.

[ QUOTE ]

no way. With this kind of timeline and the most advanced artificial selection techniques

[/ QUOTE ]

I'm not sure what people mean when they say "the most advanced artificial selection techniques" it's just which ones you let breed. Look what it has done for dogs.

[Rduke55]

[ QUOTE ]
I have no doubt that a scientist who could predict the reproductive viability of each DNA combination could get there.

[/ QUOTE ]

What does reproductive viability have to do with what we are talking about here?

[Piers]

[ QUOTE ]
I think you could easily come up with a species much smarter than humans, and therefore a species capable of imitating us.


[/ QUOTE ]

Yes, much easier than creating humans.

[ QUOTE ]
and therefore a species capable of imitating us.

[/ QUOTE ]

Difficult to persuade the super intelligent chimps to downgrade themselves to humans though.

[Piers]

[ QUOTE ]
Why do you think we could get smarter than humans?

[/ QUOTE ]

Take IQ test of all the chimps. Probably does not matter much how as long as intelligence based. Only let the most intelligent breed, (say top 0.01% in the first generation.) So how much will the chimps average IQ increase per generation? How many generations to an average IQ of 150. When will the chimps decide they have enough and ‘break out’.

[ QUOTE ]
An interesting thing would be to see how smarter humans could get. I think we're pretty close right now.

[/ QUOTE ]

It becomes much easier if you are allowed to tamper with the DNA directly. A hundred years before we will all be having IQ boosting DNA upgrades. However I did not think DS was allowing us to do that – too easy.

No smart arse comments as to how one biologist however experienced is going to IQ test a billion chimps.

[Rduke55]

[ QUOTE ]
Only let the most intelligent breed, (say top 0.01% in the first generation.) So how much will the chimps average IQ increase per generation? How many generations to an average IQ of 150. When will the chimps decide they have enough and ‘break out’.


It becomes much easier if you are allowed to tamper with the DNA directly. A hundred years before we will all be having IQ boosting DNA upgrades. However I did not think DS was allowing us to do that – too easy.


[/ QUOTE ]

1) I think there would be severe constraints on how smart a chimp (or human, or dolphin, elephant, etc. etc.) could get.

2) Even so, I keep thinking about the Ashkenazi jews in this regard. Most people wouldn't argue that not only is ther avg. intelligence high but some of the smartest people on the planet are Ashkenazi. The problem is that they are susceptible to some hereditary diseases. These hereditary diseases have neurological symptoms and I don't think that's a coincidence. Once you start selecting for fast brain stuff it's probably easy for it to go wrong. In addition, you would really have to worry about inbreeding (but we do have a billion chimps)

[Phil153]

These Jews aside, there are plenty of smart people that don't have these afflictions. Is there any sign of a correlation between 120-140 IQ and genetic illnesses? Because if not, Ashkenazi jews are kind of irrelevant to the argument.

[vhawk01]

[ QUOTE ]
[ QUOTE ]
I have no doubt that a scientist who could predict the reproductive viability of each DNA combination could get there.

[/ QUOTE ]

What does reproductive viability have to do with what we are talking about here?

[/ QUOTE ]
Piers brought up the obvious point...we've created humans the second he is able to get one of them pregnant.

[Borodog]

[ QUOTE ]
[ QUOTE ]
Only let the most intelligent breed, (say top 0.01% in the first generation.) So how much will the chimps average IQ increase per generation? How many generations to an average IQ of 150. When will the chimps decide they have enough and ‘break out’.


It becomes much easier if you are allowed to tamper with the DNA directly. A hundred years before we will all be having IQ boosting DNA upgrades. However I did not think DS was allowing us to do that – too easy.


[/ QUOTE ]

1) I think there would be severe constraints on how smart a chimp (or human, or dolphin, elephant, etc. etc.) could get.

2) Even so, I keep thinking about the Ashkenazi jews in this regard. Most people wouldn't argue that not only is ther avg. intelligence high but some of the smartest people on the planet are Ashkenazi. The problem is that they are susceptible to some hereditary diseases. These hereditary diseases have neurological symptoms and I don't think that's a coincidence. Once you start selecting for fast brain stuff it's probably easy for it to go wrong. In addition, you would really have to worry about inbreeding (but we do have a billion chimps)

[/ QUOTE ]

As an aside, has anyone written on what the selection pressures on the Ashkenazi were? I have my own theories, but I'd like to know if anyone has written on this in the past.

[Rduke55]

[ QUOTE ]
These Jews aside, there are plenty of smart people that don't have these afflictions. Is there any sign of a correlation between 120-140 IQ and genetic illnesses? Because if not, Ashkenazi jews are kind of irrelevant to the argument.

[/ QUOTE ]

Well, there are plenty of Ashkenazi jews who don't either. I think talking about selection pressures for high intelligence having detrimental side effects is very relevant.

Also, there are studies suggesting some mental illnesses, personality disorders, or lesser mental or emotional deficits in the highest strata of intelligent folks. I would guess it's the same idea. Can't get something for nothing.

[Rduke55]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
I have no doubt that a scientist who could predict the reproductive viability of each DNA combination could get there.

[/ QUOTE ]

What does reproductive viability have to do with what we are talking about here?

[/ QUOTE ]
Piers brought up the obvious point...we've created humans the second he is able to get one of them pregnant.

[/ QUOTE ]

I think it's more likely we get a smart chimp than something we could breed with.

[Rduke55]

[ QUOTE ]

As an aside, has anyone written on what the selection pressures on the Ashkenazi were? I have my own theories, but I'd like to know if anyone has written on this in the past.

[/ QUOTE ]

I remember reading a few papers on this. One of the popular ideas is something about selection pressure for intelligence because of the types of jobs they were regulated to in the middle ages. Individuals that were good at them became financially well off, the rich raised more children to adulthood so there was some differential reproduction there.

Of course there most likely was a bottleneck or some restriction on gene flow that would make these traits even more frequent in the population along with the bad ones.

[Neuge]

I'm not up-to-date on current research on human evolution, but I just came across this (http://www3.interscience.wiley.com/cgi-bin/abstract/114250882/ABSTRACT?CRETRY=1&SRETRY=0). What are the implications of this on the topic of this thread? Could you direct the physical adaptations required as pointed out by rduke by artificially inducing selective pressures such as this? Do you think this a cause or consequence of human brain evolution?

[TomCowley]

[ QUOTE ]
What does reproductive viability have to do with what we are talking about here?

[/ QUOTE ]

I'm trying to creat a human, not a chimp with human intelligence. To create a human from a chimp requires some number of DNA mutations. Since they aren't all going to happen at once, the "simplest plan" is to test every chimp for mutations that make it more human-like and selectively breed them. If life were easy, you'd end up with a human after a little while. However, there may reach a point where further chimp->human mutation results in something that can't reproduce, meaning you've reached a dead end.

Doing the direct DNA mutation scheme not only requires knowledge of the starting and ending DNA (which we do know), but ALSO a conversion path such that "chimps" can continue to reproduce at every point.

[wacki]

[ QUOTE ]
But a billion years is a mind-numbing amount of time so who knows?

[/ QUOTE ]

A billion years ago multicellular life was just starting. Birds didn't get started until 150 million years ago. Looking at the impact that selective farming had on crops in the last 10,000 years alone I think many people in this thread are underestimating what you can do in a billion years. The real question, and this is a big one, is how many chimps do you have to work with? If the answer is a billion chimps then evolving them into a human should be easy. I mean look how much humans vary on this planet at this very second. If you only have a few thousand chimps then the task becomes much more difficult. An artist needs a variety of paint to work with when creating a masterpiece. Time can only do so much.