Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
Genetic Algorithm theory vs. Eugenics

By turmeric in Op-Ed
Thu Feb 21, 2002 at 03:14:45 AM EST
Tags: Culture (all tags)
Culture

The science of Eugenics, or trying to breed human beings for traits that help society, has been around for at least 100 years. It was (and is) extremely popular and influential in the United States, Canada, the UK, and most prominently Nazi Germany. Unfortunately, the theories of 'genetic algorithms' that have come out of theoretical computer science and mathematics totally destroy the fundamental arguments of the entire field of Eugenics.


The Eugenics movement was responsible for some of the most morally reprehensible acts of the 20th century. If you go to a good university library you will be able to look up old journals and articles about the Eugenics movement and its theories. These are replete with the idea that by forcing certain people to stop breeding, that the 'gene pool' can be improved and thus society can be improved. Perhaps the latest and most famous example of the popularity of this theory is the 'Darwin Awards' website, which although on its surface humorous actually has its entire foundation in Eugenics theory. The basic idea of the site is that it is OK to make fun of dead people if they had 'defective genes' that were 'harming society'.

Tragically the entire idea that one can selectively breed people to get an 'optimal human' is fundamentally flawed. Genetic Algorithms theory involves the generation of computer programs or solutions to problems by following a series of 'breeding' steps. They are roughly as follows:

  1. Identify the problem
  2. Figure out how to encode the problem into some algorithm language, like computer code.
  3. Randomly generate a bunch of solutions to the problem.
  4. Test the solutions by running them against the problem.
  5. Figure out which solutions 'best' solved the problem.
  6. Select a few of the 'best' solutions and erase all the other solutions.
  7. Take these new 'best' solutions and put go back to step 4 until you get tired or find an 'optimal' solution.

Ok, now take this theory, and you can see that Eugenics is basically trying to do the same thing. Here are the same steps, but the way a Eugenicists would perform them on civilization and humanity.

  1. The problem is that society is not perfect because some people are criminals.
  2. This criminality is encoded in their genes.
  3. Genes are already randomly generated.
  4. Determine how 'genetically pure' people are by using various disease tests, IQ tests, intelligence tests, SAT scores, ACT scores, school performance, arrest records, personality tests, psychiatric tests, physical tests, racial heritage, etc, etc, etc.
  5. Figure out which genetic makeups of people are 'best suited' to be the 'least criminal'.
  6. Take the 'best suited' and give them lots of resources, money, education, good jobs, etc. Take all the other people and give them a crappy education, sterilize them to prevent them from breeding, kill them, etc.
  7. Take this new 'super-race' and go back to step 4, until society has no crime anymore.

So there we have it, the perfect way to solve all of the problems of the world, in a few simple steps. The problem is that Genetic Algorithms are not that simple. Genetic Algorithm researchers have determined that merely 'weeding out all the bad solutions' will eventually lead you to a sub-optimal solution to your problem. This is 'over selection', where the person working with the problem was 'too eager' to get a good result so he chopped off all the other algorithms too quickly, but in fact these algorithms that seemed 'inferior' in the first few generations of the solution would in fact have led to a more optimal solution in the future. For example, if you were trying to figure out a new algorithm for speeding up a calculation, and you selected only 1 or 2 of the 'best' algorithms after the first stage of 'evolution', then you might end up with a 'less optimal' final algorithm than you would have if you selected 50 or 60 of the initially generated algorithms for further development.

Thus the Eugenics movement is trying to use a flawed methodology. First of all, the problem is mis-stated. Everyone from the president of the USA to the CEO of Enron is capable of being an anti-social criminal, and it has nothing to do with their genetics. Second of all, different factions of eugenicists will have different ideas about what is a 'solution' to this problem. Some people think its 'racial heritage', some people think it is 'physical strength', some people think it is 'how well they did on the PSAT test' These are generally stupid and always leave out some brilliant person like Beethoven, Einstein, Faraday, etc, all of whom were considered 'socially worthless' at one time or another in their lives. Third of all, even if the Eugenics folks did artificially select the people that they wanted they would eventually get a sub-optimal solution to their problem, for you have to let diversity exist in all stages of 'evolution' of your solution for you to get anything close to an optimal solution.

This 'eugenics' mentality, where you can automatically force people to be the way you like them, has been a plague for many hundreds of years. Every great accomplishment of civilization from artwork to science to culture to government, would have been subject to 'eugenic destruction' in its infancy for being a 'sub optimal solution to a problem'. Remember Picasso, who said something like the greatest impediment to great art is a sense of good taste?

None of this innovation or allowance for diversity and deviance is possible under a Eugenic State. Civilization, if reduced to eugenic control, will grind to a halt and stagnate.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Related Links
o Also by turmeric


Display: Sort:
Genetic Algorithm theory vs. Eugenics | 81 comments (68 topical, 13 editorial, 1 hidden)
Eugenics can work. (3.80 / 15) (#4)
by bc on Wed Feb 20, 2002 at 02:08:50 PM EST

It just depends on how you define the problems, and decide to get to the solutions.

For example, human eyes are inferior to octupus eyes because the network of nerves that run from each rod and cone to the optic nerve are placed on top of the retina iself. Therefore they block light, and this decreases the definition, gives us a whopping big blind spot, and so on. In an octupus, the nerves run underneath the retina, meaning the definition is much higher, no light is blocked, and there is no blind spot. We don't evolve towards having better eyes, with a design similar to that of the octupus, because in order to do so our sight would have to degrade significantly first. This is the only design path, and evolution does not follow a far reaching plan.

However, a eugenecist can follow such a path. There is no reason why, given enough time (a lot of time, and many generations), a committed team of eugenecists could not achieve this.

The problem for eugenics is that it often focuses on behavioural traits that are not as well physically defined. It is unclear what the precise effects of, say, getting rid of criminals and selecting only 'good' people will have. Will you end up with a more optimal solution? Who can tell.

This is why any good eugenecist should be tolerant of results, and take careful records, and be willing to accept and follow 'bad' paths. A gene that may appear to improve the chances that someone will become a criminal may also increase the chances that that person become a high risk taking, successful businessman, in other circumstances. Before selecting out genes, and perhaps certain combinations of genes, we must be entirely sure of the average pros and cons of doing so, and the likely future possibilities.

An example of successful, ad hoc eugenics might be that of the Jewish people. here is an interesting article explaining how the jews self selected using ad hoc eugenics, such that they now have an average IQ of some 117, compared to the 100 of the average population.

Eugenics needn't be so bad in and of itself. The problem is that it has been slurred by the national socialists of germany, and the fascists. Voluntary Eugenics, that does not involve 'discarding' people, but rather not allowing them to reproduce with the elected target group, may well be an interesting endeavour.

♥, bc.

overoptimism (5.00 / 3) (#13)
by ucblockhead on Wed Feb 20, 2002 at 03:15:45 PM EST

However, a eugenecist can follow such a path. There is no reason why, given enough time (a lot of time, and many generations), a committed team of eugenecists could not achieve this.
Well, yes there is... a eugenecist is forced to rely on the same random mutations that natural selection does. Given that mutations vary according to traits (and that this variance is also selected), a mutation in an important structural trait that has massive negative consequences (i.e. blindness) is going to be exceedingly rare. A hypothetical eugenicist could search the entire population of the earth and not find the needed raw material.

You are making a very common error about genetics. You are assuming that any particular trait you can think of can be bred for and is equally likely. Nothing can be further from the truth. There are many "can't get here from there" situations in genetics. Try for millenia and you are still not going to get a green horse, or a six-legged horse, or a flying horse. Just because you can think of a trait doesn't mean you can breed for it.
-----------------------
This is k5. We're all tools - duxup
[ Parent ]

Bah - racist crap (4.42 / 7) (#14)
by greenrd on Wed Feb 20, 2002 at 03:39:25 PM EST

here is an interesting article explaining how the jews self selected using ad hoc eugenics, such that they now have an average IQ of some 117, compared to the 100 of the average population.

I read the first paragraph and gave up in disgust. It says:

The controversy about intelligence usually contrasts the white average IQ of 103 with the African American average IQ of 85. But why have these two groups become the criteria for analysis and discussion by the press? It seems if we want to look at real differences we should compare the very highest average IQ, that of the Ashkenazi Jew (IQ 117), with the lowest IQ, that of the sub-Saharan Black (70). This disparity amounts to a standard deviation of greater than 3, which makes any argument that the difference can be made up by improving the environment of the lowest intelligent groups ludicrous.

I don't see that. Environment includes a lot of things: heat, access to education, food, clean water, healthcare. And these can affect "intelligence" as measured on an IQ test. For example: sufficiently bad malnutrition in childhood can retard someone's intelligence for life - that's established science. I don't think it's ludicrous at all to suggest that the alleged IQ difference could be all down to environmental factors. Even leaving aside all the other criticisms of the idea that IQ differences justify racism (such as the idea that IQ tests are culturally biased and do not measure intelligence exactly, but more like the ability to pass tests, which can be improved through training).

Having said all that, I am a eugenicist - I believe parents should have the right to know if their prospective baby is likely to have a certain medical condition, the right to select desired DNA (perhaps within certain constraints), and the right to have abortions (within the present legal framework). Unfortunately the most vehement eugenicists tend to be racists, but not all of us are.


"Capitalism is the absurd belief that the worst of men, for the worst of reasons, will somehow work for the benefit of us all." -- John Maynard Keynes
[ Parent ]

Reply (and obligatory insults) (2.58 / 12) (#15)
by ubu on Wed Feb 20, 2002 at 03:46:14 PM EST

So what if it's racist? For the sake of intellectual consistency you shouldn't give a flying shit about racism:

Having said all that, I am a eugenicist - I believe parents should have the right to know if their prospective baby is likely to have a certain medical condition, the right to select desired DNA (perhaps within certain constraints), and the right to have abortions (within the present legal framework).
If race is based upon anything but genetics I would like to know how. If not, then you have just given us your explicit permission to exercise our preferences for racial differences in genetic patterns.

To be sure, you candy-coat your racism with a generic, user-friendly "eugenicist" label (racial discrimination?! [monocle pops out] Well, I never!), but you've just abdicated any right to moral indignation whatsoever.

Wanker. Pisshead. Fucknut.

Ubu


--
As good old software hats say - "You are in very safe hands, if you are using CVS !!!"
[ Parent ]
Sexual preferences OK (5.00 / 5) (#19)
by greenrd on Wed Feb 20, 2002 at 04:21:57 PM EST

So what if it's racist?

Racist theories are usually crap. No wait, they're always crap.

If not, then you have just given us your explicit permission to exercise our preferences for racial differences in genetic patterns.

On an individual level, of course! No-one's saying you have to sleep with a Vietnamese woman if you don't want to! And that's not racism. That's just about sexual preferences. I tend to find Caucasian women more attractive than black women, on the whole, but that's not racism, that's just a preference.

I purposely didn't say anything about the governmental level, because I'm not really sure about that.


"Capitalism is the absurd belief that the worst of men, for the worst of reasons, will somehow work for the benefit of us all." -- John Maynard Keynes
[ Parent ]

Reply (and obligatory insults) (2.00 / 8) (#21)
by ubu on Wed Feb 20, 2002 at 04:36:16 PM EST

I purposely didn't say anything about the governmental level, because I'm not really sure about that.

No problem, just let us know what you decide and we'll make a law or something.

Retard. Canadian. Aborted fetus.

Ubu


--
As good old software hats say - "You are in very safe hands, if you are using CVS !!!"
[ Parent ]
fucknut (2.25 / 4) (#51)
by streetlawyer on Thu Feb 21, 2002 at 03:32:24 AM EST

No-one's saying you have to sleep with a Vietnamese woman if you don't want to! And that's not racism. That's just about sexual preferences. I tend to find Caucasian women more attractive than black women, on the whole, but that's not racism, that's just a preference.

I see that, like so many self-styled leftists who always do a big number about their anti-racism you are a sexist twat.

--
Just because things have been nonergodic so far, doesn't mean that they'll be nonergodic forever
[ Parent ]

not racism (4.33 / 3) (#54)
by kubalaa on Thu Feb 21, 2002 at 05:35:14 AM EST

Would you clarify what your definition of racism is, then, that includes deciding who is smarter, but not who is better in bed?

I'd say, quite simply, racism is making any generalization/judgement/statement based purely on race. This is not only common, it's unavoidable. Your brain is all about making generalizations -- it's a generalization to think, for example, that water is always wet. By doing so, you risk marginalizing all the dry water. Obviously this is an exaggeration, but racism has the same basis; from someone's personal experience, they have recognized a set of attributes that seem to generally apply to a race. Their generalization may be statistically accurate or not, it doesn't matter.

The point is to be flexible and open-minded. It's not "wrong" to say that, in your experience, sub-saharan blacks are, on average, stupider. It is wrong to allow this generalization to become an expectation, which colors your future perceptions, your opinions of individuals, and your ability to admit when you are mistaken.

The part of "racism" which makes it bad in the usual definition is close-minded, stubborn conservativism. There's nothing wrong with forming opinions based on race -- if you say there is you have to explain how this is different than any other allowable opinion based on experience.

[ Parent ]

Sheer numbers are against you (3.20 / 5) (#18)
by bc on Wed Feb 20, 2002 at 04:05:05 PM EST

Certainly, environmental factors certainly play a role in IQ results. But you must remember that the IQ test is designed to be environment independent. It does not measure knowledge or education, but rather native intelligence. This means that educational and socioeconomic factors can be dismissed, where those factors are 'soft', or social.

Your other point was that of 'hard' factors, such as malnutrition, or lack of heat, which certainly would effect IQ test results. However, if you are seriously claiming that poor people in America today are suffering from malnutrition and lack of heat, I am afraid you are mistaken. Certainly, they may be relatively poor, but they are not at baseline, objective povery levels that would make such factors effect IQ results.

I encourage you to take a look at these statistics demonstrating the amazing genetic differences in IQ across race. Such differences cannot be environmental, they are indicative of fundamental genetic differences, and show that eugenics can make a difference.

♥, bc.
[ Parent ]

I have to disagree... (4.80 / 5) (#20)
by CokeBear on Wed Feb 20, 2002 at 04:32:08 PM EST

I have to disagree with you there.

As a person who has personally experienced wide variations in class during my lifetime, I can say with some authority that the lower classes of people in North America have substantially lower quality of nutrition, education, and housing, then the upper classes.

In particular, when discussing food, wealthy people have access to much higher quantity and quality of food than poor people.

Examples: Middle and lower-middle class people consume far more fast food than rich people.
Poor people, especially those on welfare or food stamps, simply can't afford to make sure their meals are balanced.

When it comes to housing, as a middle class person, you might take central heating and air conditioning for granted, but these are luxuries that many poor people don't have access to. As a result, their attention is more focused on their environment when they are young, in such a way that might get in the way of learning. (And I don't just mean learning facts, but learning how to learn... building the connections in the brain that will help them to score well on an IQ test later.

[ Parent ]

Hmm, that's strange. (3.14 / 7) (#22)
by bc on Wed Feb 20, 2002 at 04:53:00 PM EST

If they are malnourished, why are 50% of blacks obese?

If they are malnourished, taller on average than whites?

Are you saying that this malnutrition only affects the brain? What a very odd statement.

♥, bc.
[ Parent ]

vitamins (3.66 / 3) (#26)
by ucblockhead on Wed Feb 20, 2002 at 05:35:16 PM EST

Because maltrition is a result of a lack of vitamins, not of a lack of calories.

The thing about height is that nutrition has a multiplicative effect. Different population groups have different heights due to genetics.
-----------------------
This is k5. We're all tools - duxup
[ Parent ]

Wait a minute, you are being idealogical (4.00 / 4) (#31)
by bc on Wed Feb 20, 2002 at 06:39:51 PM EST

You blithely state that height is genetic, whilst stridently arguing that intelligence is not.

Why one, and not the other? Doesn't every argument you have lobbed against intelligence being genetic also apply to height?

Could it be that you think intelligence cannot be determined by genetics for idealogical reasons, rather than hard evidence?

Also, I find it hard to believe that people can get fat, these days, and yet be vitamin deprived. Perhaps if they were eating caviar and fois gras this would be true, but cheap foods are pcked with vitamin & mineral supplements. Hell, tapwater is these days. Just look at the side of a cornflakes box.

Also, nomatter the genetics, blacks could not possibly be taller, on average, than whites, if they were nutritionally deprived.

Please excuse any spelling & grammer mistakes in this and other posts by me in this thread, I have had rather too much Glenmorangie.

♥, bc.
[ Parent ]

Nope (5.00 / 4) (#34)
by ucblockhead on Wed Feb 20, 2002 at 07:18:04 PM EST

I'm not saying intelligence is not genetic.

Intelligence is genetic. So is height.

Yet both vary trememdously according to environment. Five hundred years ago, the average European man was 5'6". Now he's 5'10". That's because genetics controls potential, not actual.

Same with intelligence.

The trouble is that you are suffering from the common misconception that "genetic" means "exactly determined". It doesn't.

And yes, blacks could easily be taller than average than whites for the simple reason that there are genetic differences in height. It should be pretty obvious that a child with parents who are 6'6" is going to be taller than a child with parents who are 5'6", even if the child with taller parents is malnourished. He'll just not be as much taller as he potentially could have been.

As far as nutrition goes, quite frankly, you are ignorant. Talk to a nutritionist. Or look at prenatal vitamins and do some math.

But more to the point: It is well known that alcohol use during pregnancy causes babies with lower intelligence. This is not controlled for in the adoptions studies you site. Hence they are worthless.

(And that's just one of many, many prenatal environment factors that can effect intelligence.)
-----------------------
This is k5. We're all tools - duxup
[ Parent ]

Indeed (none / 0) (#46)
by fhotg on Thu Feb 21, 2002 at 01:37:23 AM EST

If they are malnourished, why are 50% of blacks obese?
Because obesity is caused by malnutrition duh.
If they are malnourished, taller on average than whites?
If that is true, it likely has to do with a similar cause that makes them darker than whities. Something in the genes maybe ?

[ Parent ]
design (3.00 / 1) (#23)
by ucblockhead on Wed Feb 20, 2002 at 05:02:20 PM EST

Not every design meets its goals. The idea that you can design an IQ test culture has no effect on is pure conjecture. No study has ever shown any IQ test to be culturally unbiased. However, many, many different IQ tests have been shown to be culturally biased despite designer's claims to the contrary.
-----------------------
This is k5. We're all tools - duxup
[ Parent ]
So, in that case (3.66 / 3) (#24)
by bc on Wed Feb 20, 2002 at 05:18:42 PM EST

In that case, one might expect black children adopted into white families to get the same IQ scores, on average.

But they don't.

If your claim where true, than IQ tests, if designed by black people, would have a reversed bias. And the discrepancy I just pointed out would not exist.

♥, bc.
[ Parent ]

Not necessarily (none / 0) (#25)
by ucblockhead on Wed Feb 20, 2002 at 05:33:21 PM EST

"Environment" starts at conception.

Some of the strongest environmental factors that effect intelligence are prenatal. Alcohol use is only the most well known.

The "studies" you point to do not take this into account and are thus worthless.
-----------------------
This is k5. We're all tools - duxup
[ Parent ]

And it's well known... (none / 0) (#39)
by scanman on Wed Feb 20, 2002 at 11:41:13 PM EST

...that alcoholism runs in families.

"[You are] a narrow-minded moron [and] a complete loser." - David Quartz
"scanman: The moron." - ucblockhead
"I prefer the term 'lifeskills impaired'" - Inoshiro

[ Parent ]

Runs in families (5.00 / 2) (#52)
by PresJPolk on Thu Feb 21, 2002 at 04:15:35 AM EST

Religion runs in families, too.

[ Parent ]
blah (5.00 / 1) (#55)
by mikeliu on Thu Feb 21, 2002 at 05:54:40 AM EST

However, whereas religion only runs in families culturally, alcoholism runs in families gentically. I am unaware of any studies correlating religious leanings of adopted children with religions of their genetic parents. However, the genetic link of alcoholism is well established.

[ Parent ]
That is incorrect (4.00 / 1) (#69)
by ucblockhead on Thu Feb 21, 2002 at 11:30:22 AM EST

susceptibility to alcoholism runs in families.

There's a profound difference.

How different cultures (and subcultures) deal with alcohol abuse is left to an exercise for the reader. It is interesting to note, though, that self-medication is often inspired by negative occurances in life. Like "being poor".
-----------------------
This is k5. We're all tools - duxup
[ Parent ]

Intra-uterine environment? (2.50 / 2) (#29)
by pattern on Wed Feb 20, 2002 at 05:57:14 PM EST

Proper nutrition in utero is at least as important as during childhood. Adoption doesn't make a sufficient control in that research.



[ Parent ]
No they weren't (none / 0) (#61)
by Obvious Pseudonym on Thu Feb 21, 2002 at 08:36:27 AM EST

Certainly, environmental factors certainly play a role in IQ results. But you must remember that the IQ test is designed to be environment independent. It does not measure knowledge or education, but rather native intelligence. This means that educational and socioeconomic factors can be dismissed, where those factors are 'soft', or social. (your emphasis)

IQ tests were certainly not designed to be environmentally independent. They were designed to prove what the designers already 'knew' - i.e. that white males were superior to females and males of other races.

I suggest you read Stephen J Gould's 'The Mismeasure Of Man' for a full history of IQ testing and a detailed analysis of why an IQ score shows nothing other than how good you are at doing IQ tests.

(I realise that not everyone here agrees with Gould's 'Puncuated Equilibrium' evolutionary theories, but this is an excellently researched book on the history of science that should be required reading for any Psychology or Biology student and has nothing to do with his work as an evolutionary scientist.)

Obvious Pseudonym

I am obviously right, and as you disagree with me, then logically you must be wrong.
[ Parent ]

Little changes have large results. (none / 0) (#73)
by Alarmist on Thu Feb 21, 2002 at 03:02:54 PM EST

However, if you are seriously claiming that poor people in America today are suffering from malnutrition and lack of heat, I am afraid you are mistaken. Certainly, they may be relatively poor, but they are not at baseline, objective povery levels that would make such factors effect IQ results.

It's been said already that we know malnutrition early in life can cause impairment that is permanent in nature.

What was missing from that is that the level of malnutrition needed is very low, and that malnutrition isn't simply not having enough to eat; it's not eating enough of the right things. A diet rich in fats, carbohydrates, and mineral-leeching acids makes poor nutritional sense, and a person who subsists on such things as their primary foods runs a good chance of having some sort of developmental difficulties if this regiment dominates their early diet.

Fun fact: in the United States, we consume mainly fats, carbohydrates, and mineral-leeching acids. Just as improved nutrition and medical care has made the average person in the United States taller, poor diet is now making the average person duller.


[ Parent ]

Regarding intelligence tests... (5.00 / 1) (#77)
by Alfie on Thu Feb 21, 2002 at 10:02:38 PM EST

Keep in mind that IQ tests measure one's ability to take IQ tests. How this relates to real life depends on how similar living life is to taking IQ tests.



[ Parent ]
Eugenics vs. Selection (4.00 / 1) (#66)
by dennis on Thu Feb 21, 2002 at 10:48:42 AM EST

Self-selection as you describe the Jews doing, and as others have described everyone doing (eg. "he's got too much hair on his back, forget it"), is different from eugenics in one crucial way: with self-selection, everyone's selecting their own mates. With eugenics, some group in power is making the decisions. This power will inevitably be abused, and that's why what happened before shouldn't have been surprising, and why eugenics as a social strategy is and will always be an unmitigated Bad Thing. IMHO.

[ Parent ]
mother's decisions (4.66 / 9) (#5)
by deadplant on Wed Feb 20, 2002 at 02:19:41 PM EST

"Master plans" aside, the individuals involved in the human procreative process have to make decisions. Obviously this usually starts with stuff like "hmm, that guy's got too much hair in his back. Ooooh, look at that hunk over there!" Then after this mate selection phase most responsible adults consider whether or not to have children.

Since I assume you are not suggesting that we all start having babies with random partners off the street then you have to accept that we DO make decisions about how and when to make new people. I for instance would probably not have children with someone who has AIDS(which is not to say i wouldn't love/have protected sex with an AIDS sufferer). This is a reasonable and natural decision to make. We can and must make decisions about the genetic makup of our children by judging our potential mates.

We do not have the resources to explore every possible genetic path. Since each of us has only a few tries it's reasonable to expect that we will try to make our own children come out reasonably healthy so that they don't suffer too much and so that we have a good chance at grandchildren. I would argue that direct genetic analysis of our mates is merely a progression of the existing mate selection tools we use now. We are less likely to have sex with a person covered in lesions, genetic analysis allows us to 'see' several more diseases/risk factors that don't show up on the surface.


p.s. forget about 'state sponsored eugenics', nobody is advocating that seriously. What we are all talking about is individual's decisions about their own children, people already make these decisions themselves, we are just talking about what other/new factors they might consider.



You should read Lewontin (2.62 / 8) (#6)
by medham on Wed Feb 20, 2002 at 02:23:58 PM EST

At Harvard, for a true view of what modern eugenics is about.

Lewontin, like most intelligent eugenicists, understands the dialectical relationship between class structure and breeding.

The real 'medham' has userid 6831.

I rated you one... (4.00 / 6) (#40)
by scanman on Wed Feb 20, 2002 at 11:49:04 PM EST

...because you did not contribute to the discussion whatsoever. You did not discuss Lewontin's theories, or even provide links. Your pithy display of intellectual snobbery does not impress me.

"[You are] a narrow-minded moron [and] a complete loser." - David Quartz
"scanman: The moron." - ucblockhead
"I prefer the term 'lifeskills impaired'" - Inoshiro

[ Parent ]

And (1.00 / 5) (#43)
by medham on Thu Feb 21, 2002 at 12:26:19 AM EST

I've used my automod script to rate all of your comments to zero, for that cheeky display.

Lewontin was also responsible for Not in Our Genes, which I think everyone stopping to take the time to read would contribute greatly to this here discussion.

The real 'medham' has userid 6831.
[ Parent ]

Can you summarize? (2.33 / 3) (#48)
by scanman on Thu Feb 21, 2002 at 03:05:32 AM EST

or would you rather just be a snob?

"[You are] a narrow-minded moron [and] a complete loser." - David Quartz
"scanman: The moron." - ucblockhead
"I prefer the term 'lifeskills impaired'" - Inoshiro

[ Parent ]

Summaries (5.00 / 1) (#70)
by ucblockhead on Thu Feb 21, 2002 at 11:31:38 AM EST

In order to summarize, he'd have to know what the fuck he was talking about.
-----------------------
This is k5. We're all tools - duxup
[ Parent ]
I agree, you're only adding noise (5.00 / 1) (#64)
by Teukels on Thu Feb 21, 2002 at 09:35:04 AM EST

For you fail to explain what this writer you love adds to this discussion. I too can shout: "you should read this and this and this and you will probably be better informed"

Luuk

[ Parent ]
We are all eugenicists (4.12 / 8) (#7)
by medham on Wed Feb 20, 2002 at 02:28:05 PM EST

The word means "good genes," of course, and everyone is looking for them. Scientific EP research has demonstrated conclusively that mate selection is based upon traits that confer fitness and fertility; your mate is as fit and fertile as you could find, and it's always been so.

Exceptios, done from "love" or "intellect," are just that: aberrations. Evolution is a theory of populations, not individuals.

The real 'medham' has userid 6831.

error (4.75 / 4) (#9)
by ucblockhead on Wed Feb 20, 2002 at 02:44:33 PM EST

Mate selection is based on phenotype, not genotype. This introduces a level of indirection that cannot be ignored.

It is also, by the way, an important distinction in that "classical" eugenicists used phenotype as well, and subjective evaluations of phenotypes at that. Indeed, they had little choice. Their modern equivalents are not so restricted.

Whether that is a "good" thing or a "bad" thing is an interesting question. Gattaca, blah, blah, blah.
-----------------------
This is k5. We're all tools - duxup
[ Parent ]

How's your gene-detector? (4.66 / 3) (#11)
by medham on Wed Feb 20, 2002 at 02:48:47 PM EST

Of course phenotype is the only thing available to the senses, and of course there's a determinative relationship.

Lewontin, again, provides a good introduction to these issues from a eugenicist standpoint.

The real 'medham' has userid 6831.
[ Parent ]

determinative relationships (4.50 / 2) (#12)
by ucblockhead on Wed Feb 20, 2002 at 02:53:05 PM EST

Is there? Always?

Yes, I know, some people feel it convenient to believe so, but...
-----------------------
This is k5. We're all tools - duxup
[ Parent ]

ahhhhhh (none / 0) (#50)
by streetlawyer on Thu Feb 21, 2002 at 03:29:01 AM EST

it's OK, he's one of us. Lewontin a eugenicist? BC, you fucker.

--
Just because things have been nonergodic so far, doesn't mean that they'll be nonergodic forever
[ Parent ]
Not exactly (5.00 / 2) (#27)
by jolly st nick on Wed Feb 20, 2002 at 05:35:29 PM EST

It's more like there are certain invariants cross culturally that are correlated with fitness for child bearing on the male judging femalse side of the equation: clear skin, large hip to waist ration, symmetrical features. Items like breast size, skin color, nose shape are culturally ingrained but have no such correlation. Furthermore the fitness traits that are involved with female attractiveness have more to do with fertility than eugenic fitness.

On the distaff side, cultural invariants female selection of male mates are more complex as befits the higher biological "investment" females make in pregnancy. Males are (to some degree) judged attractive for the resources they control and their likely willingness to devote these resources to child rearing. Again this has little to do with eugenic fitness and more to do with having immediate success in rearing children to reproductive age.



[ Parent ]

health (4.33 / 3) (#28)
by ucblockhead on Wed Feb 20, 2002 at 05:51:19 PM EST

It is also important to note that many of the "attractiveness" traits are essentially signs of general health. It should be obvious that "general health" is only partly correlated with genetics. (And much less so, today with modern healthcare, then it was in neolithic times.)
-----------------------
This is k5. We're all tools - duxup
[ Parent ]
wrong, wrong and wrong (none / 0) (#49)
by streetlawyer on Thu Feb 21, 2002 at 03:27:51 AM EST

Evolutionary psychology isn't scientific, it hasn't done anything in the direction that you suggest which could be counted as "research" and it hasn't established your conclusion. Other than that, only grammatical errors, well done.

--
Just because things have been nonergodic so far, doesn't mean that they'll be nonergodic forever
[ Parent ]
A rather heated assertion (none / 0) (#76)
by medham on Thu Feb 21, 2002 at 06:39:50 PM EST

With so little content. Since I'm only vain about my grammar, I'll point out that, while there is a typo, there are no further solecisms.

Perhaps you need to visit the world of Tooby and Cosmides for some updates about what is actually happening in the world of EP.

The real 'medham' has userid 6831.
[ Parent ]

Einstein (4.16 / 6) (#8)
by claudius on Wed Feb 20, 2002 at 02:30:39 PM EST

I'm curious. Which aspect of Einstein's life do you deem as being "socially worthless?" The apocryphal (and inaccurate) story about his failing grades in school, his influential letter to Roosevelt which instigated the Manhattan Project, or his refusal to accept public office in the then newly formed state of Israel?

"were considered 'socially worthless'" (5.00 / 2) (#10)
by cyberformer on Wed Feb 20, 2002 at 02:48:09 PM EST

It says that Einstein and others "were considered 'socially worthless'", not that they actually were worthless. This is certainly accurate, though the reason was general anti-semitism, not really anything to do with his own personality or intelligence.

[ Parent ]
Not well thought out. (4.66 / 3) (#16)
by SIGFPE on Wed Feb 20, 2002 at 04:00:16 PM EST

Everyone from the president of the USA to the CEO of Enron is capable of being an anti-social criminal, and it has nothing to do with their genetics
Are you sure genetics is irrelevant? Every single president of the US and every single CEO of any company shares about 99.9% of their DNA with every other.

Every great accomplishment of civilization from artwork to science to culture to government, would have been subject to 'eugenic destruction' in its infancy for being a 'sub optimal solution to a problem'
I doubt that this is true. But even if it were - how do you know they wouldn't in fact have been replaced by even greater achievemnts had eugenics been implemented?
SIGFPE
Jeez, (2.00 / 4) (#37)
by jungleboogie on Wed Feb 20, 2002 at 09:42:59 PM EST

Two completely bogus articles in one sitting. You just want to see your name in lights, don't you ?

they wouldnt be bogus if i had a library (1.00 / 1) (#42)
by turmeric on Thu Feb 21, 2002 at 12:14:32 AM EST

its kind of hard to research anything over the fucking dial-up internet when you have no money.

[ Parent ]
You are whining. (none / 0) (#74)
by Alarmist on Thu Feb 21, 2002 at 03:10:50 PM EST

its kind of hard to research anything over the fucking dial-up internet when you have no money.

I don't have any money either, yet I can do scads of research, even when I don't have access to a library. How? I use the same "fucking dial-up internet" that you do.

It seems, though, that I learned how to do research with a variety of tools, and that a library, while fantastic, is not the only one out there.

I think a better explanation is either laziness or lack of available time. You can find out more or less anything you want to know if you're willing to invest the time in it, so please don't try to make me pity you with a half-assed sob story about the woes of dial-up.


[ Parent ]

interesting if trollish (4.80 / 5) (#38)
by mikpos on Wed Feb 20, 2002 at 10:18:08 PM EST

I voted +1 FP even though the article was obviously trollish and had some serious problems. Heated unproductive debate will probably follow, though. It's at least entertaining.

I'll just bring up one problem I had with it (though there were many). Consider you're trying to solve a problem with a genetic algorithm. Initially you create say 1000 algorithms. Instead of just picking the best 1 or 2 like you suggest, what if you were to pick the best 999? You seem to suggest that picking 1 or 2 would yield a less optimal result than picking 50 or 60, so logically picking 999 would be even better, right?

The point is that it's not the degree to which you whittle which determines how "eugenic" it is. Eugenics is not a matter of degree. Choosing 99.9% of the best reproducing objects should maintain diversity, and you say yourself it leads to a pretty good solution, and yet it's still very obviously eugenics.

Weak arguments make the world what it is today (4.00 / 2) (#45)
by autonomous on Thu Feb 21, 2002 at 01:11:59 AM EST

Its pretty impressive how you can compare an algorithm to a body of work done by racist people before most of us were born. Its also impressive how the concept of adding chlorine to the gene pool can be discounted simply because racist people once were behind the idea. Think of all the ideas that have made our world what it is today, then think of the horrible people who came up with some of them. Given the logic that no good idea has bad people promoting it for the wrong reasons, we can eliminate every idea. Just because you don't like to admit that perhaps you or I are not the best persons to be continuing the human race doesn't mean that cleaning it up would be evil. Its just hard to admit
-- Always remember you are nothing more than a collection of complementary chemicals worth not more than $5.00
Flawed description, argument (5.00 / 4) (#53)
by Chmrr on Thu Feb 21, 2002 at 04:23:09 AM EST

I've got several bits of beef with your argument. First off, you're describing GAs confusingly at best, wrongly at worst. In GAs, "individuals" are generally fixed-length binary strings, with different locations coding for different parts of solutions to a problem. You also forgot to mention the critcal steps of mutation and crossover -- it's those little bits which just happen to give you evolution. The way you've described it above, evolution would never happen.

That said, you're also misrepresenting the problem of overselection. Overselection is a problem when the population size is small, and there is a "good enough" solution which is likely to be arrived at fast. The biggest problem with overselection is it recudes the diversity of the population. Because of this, you lose access to possibly useful bits of material which are not currently useful. It's the loss of diversity which kills you, not the overselection itself; there are many ways to lose diversity, and overselection is just one of them.

Furthermore, overselection is mostly a problem in early generations of GA -- and, frankly, I'd like to imagine that humans arn't "randomly generated." The complexity of the human being hints that it's been a long and arduous process to get where we are today. It can also be argued that natural selection is no longer having an effect -- in GA terms, it would be the same as starting to let in those "less fit" individuals into the next generation. Which does nothing to improve the final result.

Finally, there is a counter-argument. It has been shown (by Fukushima 1990, at least) that sometimes problems (especially those with large population sizes) are best solved by a process known as greedy over-selection. Essentially, this is a process which intentinoally over-selects individuals. It is much more effective than tournament or roulette-style selection.

The above should not be taken as an argument for eugenics, but rather a debunking of some rather ugly hand-waving and incorrect statements on your part.


Networking -- only one letter away from not working.
The "Criminality" Gene .... (3.00 / 1) (#56)
by craigtubby on Thu Feb 21, 2002 at 06:23:46 AM EST

Aha, well statistics in the UK show that there is one group of people that have over 4 times more offenders than the other group.

I say we should remove this group from the Gene pool, we should make sure they can't breed more criminals.

Oh, you want to know what this group is? Well it's males aged between the 10 & 60.

Source http://www.homeoffice.gov.uk/rds/digest4/chapter3.pdf

Avergae numbers of males found guilty or cautioned for indictable offences per 100 population - 2.44

Average numbers of females found guilty or cautioned for indictable offences per 100 population - 0.55

Largest group of males found guilty or cautioned for indictable offences per 100 population - 18 yr old with 9.04

Largest group of females found guilty or cautioned for indictable offences per 100 population - 15 yr old with 1.78



try to make ends meet, you're a slave to money, then you die.

* Webpage *

Eat your own spew (1.75 / 4) (#57)
by Hopfrog on Thu Feb 21, 2002 at 07:11:10 AM EST

Look at the friggin chart and explain the 10 point IQ difference between adopted white kids and biological kids.

Hop.

Mistake (none / 0) (#67)
by Hopfrog on Thu Feb 21, 2002 at 11:29:10 AM EST

This was a falsly placed reply to a previous post. Somebody delete it.

Hop.

[ Parent ]
Shortsightedness of the Eugenics program (4.00 / 1) (#58)
by Hopfrog on Thu Feb 21, 2002 at 07:22:50 AM EST

I like to be open minded, and refuse to just throw away the eugenics theory without examination, much as it might offend me.
Let us assume that these charts are all right, and the IQ tests are unbiased, and so on. What these people are saying is that IQ should be the sole trait the determines selection. What about sport? If certain populations have genes that make them more intelligent, then other populations have genes that make them react quicker at sport games, that make them better able to perform physically strenous activities, and that make them have certain other forms of wit (e.g trash talking).
If these population groups are selected out, these genes go lost. If the populations are merged, the group IQ will drop in the short term, the group fitness will drop in the short term, but the genetic combination possibility will increase.
This means that the human would be more adaptable. It also means that after the merging, a group can be isolated to increase intelligence, and this will happen without losing the other potentially useful genes.

Hop.

Newbie question (3.00 / 2) (#59)
by EriKZ on Thu Feb 21, 2002 at 07:36:50 AM EST

OT
Hi, I've recently disovered GA and I'm trying to learn more and experiment with it. It's difficult to do so however, it seems to still be a pretty new field.

The main site that many people seem to love IlliGAL is about two years out of date. Any suggestions on where else I should go? Or a better way to learn this stuff besides browsing the net?
/OT

GA (5.00 / 1) (#65)
by cyberdruid on Thu Feb 21, 2002 at 10:38:05 AM EST

First of all, if you are interested in GA, I would advise you to check out Genetic Programming (GP), which IMHO is the most interesting variation on evolutionary computation. Traditionally GP has been done in LISP, because it is a good language for meta-programming. Nowadays, GPs that write directly to machine code is becoming more common. I have, for example, written a machine code GP myself. It is not so hard if you have some basic knowledge of assembler.

Another interesting variation, with good potential for complexity is to let a GA change the weights in a recurrent neural network. The problem with recurrent NNs (these are mostly used when you want some sort of short-term memory, like when training some an autonomous agent) is normally that the standard learning method (back propagation of the error) does not work. This is why it so interesting to let a GA do the learning and just use the NN as a way of transforming a few (perhaps 20 - 50) weights to complex behaviour. We once used this method together with the "30 monkeys in a bus"-algorithm to make a robot (made out of lego mindstorm) learn how to seek out lightbulbs and avoid walls, entirely on its own. It was damn cool to watch.

For some reason, there are very few good AI tutorials on the web. I have been thinking about writing something on the subject myself, but I haven't gotten around to it.

A few links:
Generation 5 This is perhaps the best site for AI news. It also carries quite a few essays and tutorials on basic AI-concepts.
Google, of course, has a nice directory on AI If you haven't been there already, check it out. Plenty of introductory material there.
Eliezer's seed-AI A good introduction to some far-out AI thinking :). I am working on a sort of seed-AI, but it is not anything like his ideas (mine is better ;).
Ray Kurzweil's homepage More, mostly fringe science, stuff. Very fun and interesting reading, from one of the most famous visionaries.

Wanna know more? Don't hesitate to ask - I love AI.

[ Parent ]

Cool (none / 0) (#75)
by EriKZ on Thu Feb 21, 2002 at 06:28:26 PM EST


What you're doing sounds very interesting Cyberdruid. Got an email so I can talk to you some more?

[ Parent ]
sure (none / 0) (#79)
by cyberdruid on Fri Feb 22, 2002 at 06:17:58 AM EST

f96dafe _at_ dd. chalmers. se



[ Parent ]
Newbie question (3.00 / 1) (#60)
by EriKZ on Thu Feb 21, 2002 at 07:37:11 AM EST

OT
Hi, I've recently disovered GA and I'm trying to learn more and experiment with it. It's difficult to do so however, it seems to still be a pretty new field.

The main site that many people seem to love IlliGAL is about two years out of date. Any suggestions on where else I should go? Or a better way to learn this stuff besides browsing the net?
/OT

Yeah, leave the house... (none / 0) (#72)
by Jack of Hearts on Thu Feb 21, 2002 at 02:26:34 PM EST

and look in your local library. There are *tons* of books on the subject...

[ Parent ]
ARgh (none / 0) (#81)
by olethros on Mon Jan 20, 2003 at 10:57:21 AM EST

Newbie question. Argh. Google search. Argh. Look for an author named "Holland" and someone else named "Koza".

Also look for "genetic programming" or "evolutionary algorithms"

In general, EAs are very slow methods for solving problems. If nothing else works you can try an EA. You might get something. But it requires fine-tuning of the fitness function and of your data/program representation.
-- Homepage| Music
I miss my rubber keyboard.
[ Parent ]

Galton was thinking too simplistic. (4.00 / 1) (#62)
by Teukels on Thu Feb 21, 2002 at 09:17:56 AM EST

Sir Francis Galton (1822-1911), a younger cousin and friend of Charles Darwin is the inventor of Eugenics. His ideas are since long considered simplistic, his 'successor' Sir Cyril Burt was a fraudulent researcher, his twin studies about intelligence turned out to be fictitious.

Sir Francis Galton was a test-a-holic. London's International Health Exhibition of 1884 featured an exhibit called an Antropometric Laboratory. Every attendant could have him/herself measured in this lab and thus add to 'science'. (Francher, R. E.; Pioneers of Psychology, 1996, pp. 216-245)

In 1865 Galton envisioned competitive Eugenics examinations to be held at prime marriageble age. Ten women of top class should be given away to similar men by the queen at a state marriage.
(Francher, 1996 pp.236)

The bell curve (Herrnstein and Murray, 1990) does support the 'intelligence is innate' view (allthough 'somewhat' better fundamented) but the point is _which_ construct of a human being do _we_ find 'Good' and which do we regard 'Bad'.

I remember Skinner having rats who had to get through a maze. After breeding his best rats for a generation or seven, he had fine top notch maze rats. However, for any other tasks these rats were found to be just as smart (or even dumber) as the rats who were not bred for this task (Tolman experiments).

Longivity of our breed is best preserved by having a multitude of characteristics so our pool of innate capabilities allways serves us best in the end. So I consider the debate on Eugenics dead on facts first and second on ethics.

Luuk


ps. Other things to think about: 'What does an intelligence scale actually measure?', 'is everyone with a criminal record unwanted?' (I remember some heroes here), 'Won't good and bad just re-establish after you have parted good from bad?', 'Just how much entropy does DNA/RNA tolerate?'

Bah, intelligence (none / 0) (#80)
by olethros on Mon Jan 20, 2003 at 10:54:59 AM EST

Intelligence is linked to nutrition as well.
Proteins, vitamins and minerals are essential for the brain's proper growth. I am not talking about a link between intelligence and starvation here. Sufficient calories just keep you alive (and even make you fat), but do not sustain proper growth and development.
-- Homepage| Music
I miss my rubber keyboard.
[ Parent ]
Ancient Eugenics (4.50 / 2) (#63)
by Baldrson on Thu Feb 21, 2002 at 09:32:42 AM EST

The science of Eugenics, or trying to breed human beings for traits that help society, has been around for at least 100 years.

Human culture of humans has been going on at least since the neolithic. The imposition of civilization on global ecosystems is one of the most powerful selective forces imaginable and now, with technological civilization's multipliers on that power affects virtually all populations in profound ways.

The fact that you don't want to take responsibility for your technologies and political policies doesn't mean you aren't conducting eugenics pogroms.

Even so, it is silly to presume that no conscious human culture of humans has been going on except the last 100 years:

Animal husbandry has been a known art for millenia -- probably even before the neolithic -- and slavery is just as ancient. There were "people shows" being held in Turkey as recently as the 20th century, for crying out loud!

-------- Empty the Cities --------


Why diversity is hella important in a GA (5.00 / 3) (#68)
by Skwirl on Thu Feb 21, 2002 at 11:30:21 AM EST

In a Genetic Algorithm, genomes are represented by bit strings, and sex/survival is represented by a fitness function. The fitness function determines the probability of a single genome surviving to mate. Imagine the fitness function works like a roulette wheel, and the most-fit genomes get the largest slice of the roulette wheel, and therefore have the largest chance of survival. However, every genome has a non-zero chance of survival.

You spin the wheel x times. You take the x bit strings you selected and combine them. This is usually done by choosing a random crossover point and truncating off the end of one string at that point and concatenating the end of another one in it's place.

e.g. "10101010" and "01010101" decide to mate. If the crossover point is in the middle, their children are "1010_0101" and "0101_1010". If the crossover point is the last two bits, then their children are "101010_01" and "010101_10".

The x strings you've created by crossover are the next generation, and you start over again with them.

A very simple fitness function might be "favor strings with the most number of 1's." Therefore, your optimum solution would be 1111 1111.

You start with a totally random population. Over several generations, your selection process will cause your most-fit randomly generated genomes to dominate the population. After a while, your population might look something like this:

  1. 1111 1101
  2. 1111 1100
  3. 1111 1100
  4. 1111 1001
  5. 1111 1001
  6. 1111 1001
  7. 0010 0010

Genome #1 is very, very fit, but still below optimal. Genome #7 sucks, and has only survived thus far through pure luck. However, if genome #7 dies, genome #1 will never evolve into the optimal solution, because no other combination of genomes in this population will provide a 1 in the 7th bit place.



--
"Nothing in the world is more distasteful to a man than to take the path that leads to himself." -- Herman Hesse
far too simplistic (4.50 / 2) (#78)
by aigeek on Thu Feb 21, 2002 at 10:39:05 PM EST

Genetic algorithms are a means to search a solution space, with a goal of finding a pretty good answer under tough constraints such as a poorly defined domain or limited computing resources. If you're worried about premature convergence, you just reduce the selection pressure using any of a number of methods. It's not rocket science, but it's not nearly as limited as this article makes it out to be.

The same goes for eugenics. If you manipulate selection pressure in evil ways (forced sterilization, for example) or if you have a stupid fitness function like an IQ test or an ethnic heritage, it's not going to end well. The problem, of course, is choosing a fitness function that will actually work. It's a lot easier for the relatively small and well-defined problems genetic algorithms are usually applied to. In the case of humanity, good luck defining your criteria. Better yet, please don't bother.

The poor information about GAs presented here is truly disappointing though. Don't form any opinions based on this article. If you're interested in this sort of thing, learn about it somewhere else.

Genetic Algorithm theory vs. Eugenics | 81 comments (68 topical, 13 editorial, 1 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!