22 ms·

> Each genetic variant slightly increases or decreases cognitive ability. Because it is determined by many small additive effects, cognitive ability is normally

> Each genetic variant slightly increases or decreases cognitive ability. Because it is determined by many small additive effects, cognitive ability is normally distributed, following the familiar bell-shaped curve, with more people in the middle than in the tails. How do we know these genes have an additive effect on IQ? The causal relationship asserted by this quote is very strange. They claim: Because these effects are additive, IQ is normal distributed. What...? IQ is a test. It's normal distributed, like most tests are. They're referring to the central limit theorem here, I assume[1], which not a bad insight. However it should be obvious that this normal distribution phenomenon arises out of such tests being scored additively from a number of relatively independent questions (whose answers can be thought of as independent random variables of unknown distribution). In any case, I don't see how they can justify a causal link (in either direction) between some alleged additive effect of genes and the IQ test itself. Personally, I'm mostly interested in this claim on additive intelligence genes, since I'm not an expert on this. It would be fascinating and exciting if there were additive "intelligence genes". However, as someone working on artificial neural network research as a hobby, I'm highly skeptical of this. It seems much more likely that human intelligence is a delicate balance of many interacting factors relating to the architecture and "algorithms" of the brain. [1] http://en.wikipedia.org/wiki/Central_limit_theorem http://en.wikipedia.org/wiki/Central_limit_theorem

bhouston 12y ago
>> Each genetic variant slightly increases or decreases cognitive ability. Because it is determined by many small additive effects, cognitive ability is normally distributed, following the familiar bell-shaped curve, with more people in the middle than in the tails. > How do we know these genes have an additive effect on IQ? I came to post on the same topic. Some gene effects are additive but those are the simple ones and there isn't that many of them. In something complex like a human brain, there are going to be tons of interactions between genes, so that maybe genes A+B makes someone smarter and C+D also make someone smarter, but the combo A+C makes someone actually less smart. The interactions are likely to be insanely complex -- eventually solvable but it will be a while. To use a computer metaphor, the optimization landscape is not smooth and it doesn't only have a single optimal peak, it is highly complex with lots of local hills (which are caused by the interactions.) > The breeding of domesticated plants and animals has changed some populations by as much as 30 standard deviations. Broiler chickens, for example, have increased in size more than four times since 1957. A similar approach could be applied to human intelligence, leading to IQs greater than 1,000. This chicken metaphor is very wrong in that that was achieved via artificial selection, not genetic engineering. Artificial selection can make directed fast movement through an optimization landscape, but it does this without actually trying to figure out the interactions at the gene level, rather it looks as the results and selectively breads for those. When you do this to human it is called eugenics, and, to put it mildly, it likely isn't going to come back into fashion: https://en.wikipedia.org/wiki/Eugenics https://en.wikipedia.org/wiki/Eugenics
- readerrrr 12y ago
  When you do this to human it is called eugenics, and, to put it mildly, it likely isn't going to come back into fashion. Surprisingly we do it all the time, on animals, and we call it breeding. It is very successful and if applied to humans could produce real geniuses without any new tech or research in a handful of generations. Too bad the word has such negative association.
  - brohee 12y ago
    Last I checked, most people mated with a partner of their choosing. That you call it love and not "selection of the most desirable traits" doesn't change the facts ;-)
    - readerrrr 12y ago
      You are using the same mistake as most people when they associate the word eugenics with mass forced extermination of certain traits. Because of fear and certain individuals in the past. The word can be used in a modern context, where eugenics is performed voluntarily, on a very small population, and under scientific supervision for causes that benefit the entire humanity. The rest can live and love whoever they want. Genetic manipulation functionally falls in the same category, but it is better accepted because it doesn't have the negative history attached to it.
      - Dewie 12y ago
        It isn't voluntary for the offspring that are created through such efforts. Sure, that's not the case for procreation to begin with... but the difference is that these people would effectively be bred for a specific purpose. Then, to keep it voluntary, I guess we would lightly suggest to them at some tender age to consider going into field which requires high intelligence? Or do we not interfere at all, and just observe them from a distance (since you mention scientific supervision)?
        crazypyro 12y ago
imaginenore 12y ago
It actually doesn't matter if some of them cancel each other out. There still is the best combination of them, and statistically it's very likely to be better than anything evolution has come up with. Human population is just not large enough for the evolution to try many combinations.
hderms 12y ago
I would imagine that there are a number of different ways a particular brain could be more successful than another at performing a particular cognitive task. I do believe you are right in thinking that some of them are better thought of as algorithmic, but some of them are perhaps simple and independent (such as better memory recall). Neither algorithmic structures in the brain nor simple characteristics like recall/speed of language parsing can explain everything if we are to believe that the human brain is as complicated as evidence demonstrates.
nautilus 12y ago
This is a great comment and an important question. We are posting it on the Nautilus article site, credited to you, and ask the author to reply.
jerf 12y ago
One the one hand, I would be deeply stunned if an IQ of 1000 was available just by flipping all the genetic switches to "on". The result of doing such a thing that I would actually expect would be to create a non-viable embryo. On the other hand, it is undeniable that genetic fiddling will someday be able to raise intelligence, and even if all it could do was reliably produce people of the 150-170 range, it could still radically change the world. And that's merely hypothesizing the reliable production of an effect that already exists and therefore can not be impossible. One can imagine that an IQ 1000 may be simply impossible on anything recognizably like the neural substrate we run on. We don't have enough information to know that right now. But we know we can get 150-170.
- chubot 12y ago
  Meh, you also need the motivation component. IQ is a measure of capability, not a personality trait. It's perfectly easy to imagine tons of people with high IQs who either unmotivated (e.g. satisfy their itches by say gambling or playing video games) or destructive (criminals). There are probably hundreds of thousands of people with greater IQ than Bill Gates, Larry Page, or Elon Musk. It doesn't mean they have the inclination to change the world.
  - pawn 12y ago
    This can't be overstated. In college, the smartest guy I knew was also the laziest. He used to tell me different ideas he had for how to be the laziest person imaginable while still living a comfortable life, which were always amusing. He could destroy anybody else on campus at chess without effort and not paying attention, focusing instead on EverQuest or some other videogame he was simultaneously playing.
- gwern 12y ago
  > One the one hand, I would be deeply stunned if an IQ of 1000 was available just by flipping all the genetic switches to "on". The result of doing such a thing that I would actually expect would be to create a non-viable embryo. That makes no sense. All these variants would have been obtained by comparing in the general (living) population people with differing copies. Why would these variants be completely harmless in normal people and suddenly lethal in conjunction? Is there any evidence at all for this idea? For example, do we see extremely intelligent people (who will be highly enriched in the relevant variants) dropping like flies in their teens?
chubot 12y ago
It's also a huge mistake to think that this is the ONLY thing those genes do. It's likely that many of them will increase depression, anxiety, ADD, schizophrenia, drug addiction, etc. Evolution produces genomes within the constraint of the environment. I'm no luddite, but I'd be pretty concerned with genetic engineering without regard for the environment of the individual.
stephenhsu 12y ago
The question of additivity of genetic effects is discussed in more detail in reference [1] above (sections 3.1 and also 4): http://arxiv.org/pdf/1408.3421v2.pdf http://arxiv.org/pdf/1408.3421v2.pdf In plant and animal genetics it is well established that the majority of phenotype variance in complex traits which is under genetic control is additive. (Linear models work well in species ranging from corn to cows; cattle breeding is now done using SNP genotypes and linear models to estimate phenotypes.) There are also direct estimates of the additive / non-additive components of variance for human height and IQ, from twin and sibling studies. Again, the conclusion is the majority of variance is due to additive effects. There is a deep evolutionary reason behind additivity: nonlinear mechanisms are fragile and often "break" due to DNA recombination in sexual reproduction. Effects which are only controlled by a single locus are more robustly passed on to offspring. Fisher's fundamental theorem of natural selection says that the rate of change of fitness is controlled by additive variance in sexually reproducing species under relatively weak selection. Many people confuse the following statements: "The brain is complex and nonlinear and many genes interact in its construction and operation." "Differences in brain performance between two individuals of the same species must be due to nonlinear effects of genes." The first statement is true, but the second does not appear to be true across a range of species and quantitative traits. Final technical comment: even the nonlinear part of the genetic architecture can be deduced using advanced methods in high dimensional statistics (see section 4.2 in [1] and also http://arxiv.org/abs/1408.6583 http://arxiv.org/abs/1408.6583).
- stephenhsu 12y ago
  I just realized I've said all of this already in http://arxiv.org/pdf/1408.3421v2.pdf http://arxiv.org/pdf/1408.3421v2.pdf (p.16): ... The preceding discussion is not intended to convey an overly simplistic view of genetics or systems biology. Complex nonlinear genetic systems certainly exist and are realized in every organism. However, quantitative differences between individuals within a species may be largely due to independent linear effects of specific genetic variants. As noted, linear effects are the most readily evolvable in response to selection, whereas nonlinear gadgets are more likely to be fragile to small changes. (Evolutionary adaptations requiring significant changes to nonlinear gadgets are improbable and therefore require exponentially more time than simple adjustment of frequencies of alleles of linear effect.) One might say that to first approximation, Biology = linear combinations of nonlinear gadgets, and most of the variation between individuals is in the (linear) way gadgets are combined, rather than in the realization of different gadgets in different individuals. Linear models works well in practice, allowing, for example, SNP-based prediction of quantitative traits (milk yield, fat and protein content, productive life, etc.) in dairy cattle. ...
Alex3917 12y ago
Speaking of normal distributions, in order for someone to have an IQ of 1,000, wouldn't there have to be trillions and trillions of people?

bhouston 12y ago

>> Each genetic variant slightly increases or decreases cognitive ability. Because it is determined by many small additive effects, cognitive ability is normally distributed, following the familiar bell-shaped curve, with more people in the middle than in the tails. > How do we know these genes have an additive effect on IQ? I came to post on the same topic. Some gene effects are additive but those are the simple ones and there isn't that many of them. In something complex like a human brain, there are going to be tons of interactions between genes, so that maybe genes A+B makes someone smarter and C+D also make someone smarter, but the combo A+C makes someone actually less smart. The interactions are likely to be insanely complex -- eventually solvable but it will be a while. To use a computer metaphor, the optimization landscape is not smooth and it doesn't only have a single optimal peak, it is highly complex with lots of local hills (which are caused by the interactions.) > The breeding of domesticated plants and animals has changed some populations by as much as 30 standard deviations. Broiler chickens, for example, have increased in size more than four times since 1957. A similar approach could be applied to human intelligence, leading to IQs greater than 1,000. This chicken metaphor is very wrong in that that was achieved via artificial selection, not genetic engineering. Artificial selection can make directed fast movement through an optimization landscape, but it does this without actually trying to figure out the interactions at the gene level, rather it looks as the results and selectively breads for those. When you do this to human it is called eugenics, and, to put it mildly, it likely isn't going to come back into fashion: https://en.wikipedia.org/wiki/Eugenics https://en.wikipedia.org/wiki/Eugenics

readerrrr 12y ago
When you do this to human it is called eugenics, and, to put it mildly, it likely isn't going to come back into fashion. Surprisingly we do it all the time, on animals, and we call it breeding. It is very successful and if applied to humans could produce real geniuses without any new tech or research in a handful of generations. Too bad the word has such negative association.
- brohee 12y ago
  Last I checked, most people mated with a partner of their choosing. That you call it love and not "selection of the most desirable traits" doesn't change the facts ;-)
  - readerrrr 12y ago
    You are using the same mistake as most people when they associate the word eugenics with mass forced extermination of certain traits. Because of fear and certain individuals in the past. The word can be used in a modern context, where eugenics is performed voluntarily, on a very small population, and under scientific supervision for causes that benefit the entire humanity. The rest can live and love whoever they want. Genetic manipulation functionally falls in the same category, but it is better accepted because it doesn't have the negative history attached to it.
    - Dewie 12y ago
      It isn't voluntary for the offspring that are created through such efforts. Sure, that's not the case for procreation to begin with... but the difference is that these people would effectively be bred for a specific purpose. Then, to keep it voluntary, I guess we would lightly suggest to them at some tender age to consider going into field which requires high intelligence? Or do we not interfere at all, and just observe them from a distance (since you mention scientific supervision)?
      - crazypyro 12y ago

imaginenore 12y ago

It actually doesn't matter if some of them cancel each other out. There still is the best combination of them, and statistically it's very likely to be better than anything evolution has come up with. Human population is just not large enough for the evolution to try many combinations.

hderms 12y ago

I would imagine that there are a number of different ways a particular brain could be more successful than another at performing a particular cognitive task. I do believe you are right in thinking that some of them are better thought of as algorithmic, but some of them are perhaps simple and independent (such as better memory recall). Neither algorithmic structures in the brain nor simple characteristics like recall/speed of language parsing can explain everything if we are to believe that the human brain is as complicated as evidence demonstrates.

nautilus 12y ago

This is a great comment and an important question. We are posting it on the Nautilus article site, credited to you, and ask the author to reply.

jerf 12y ago

One the one hand, I would be deeply stunned if an IQ of 1000 was available just by flipping all the genetic switches to "on". The result of doing such a thing that I would actually expect would be to create a non-viable embryo. On the other hand, it is undeniable that genetic fiddling will someday be able to raise intelligence, and even if all it could do was reliably produce people of the 150-170 range, it could still radically change the world. And that's merely hypothesizing the reliable production of an effect that already exists and therefore can not be impossible. One can imagine that an IQ 1000 may be simply impossible on anything recognizably like the neural substrate we run on. We don't have enough information to know that right now. But we know we can get 150-170.

chubot 12y ago
Meh, you also need the motivation component. IQ is a measure of capability, not a personality trait. It's perfectly easy to imagine tons of people with high IQs who either unmotivated (e.g. satisfy their itches by say gambling or playing video games) or destructive (criminals). There are probably hundreds of thousands of people with greater IQ than Bill Gates, Larry Page, or Elon Musk. It doesn't mean they have the inclination to change the world.
- pawn 12y ago
  This can't be overstated. In college, the smartest guy I knew was also the laziest. He used to tell me different ideas he had for how to be the laziest person imaginable while still living a comfortable life, which were always amusing. He could destroy anybody else on campus at chess without effort and not paying attention, focusing instead on EverQuest or some other videogame he was simultaneously playing.
gwern 12y ago
> One the one hand, I would be deeply stunned if an IQ of 1000 was available just by flipping all the genetic switches to "on". The result of doing such a thing that I would actually expect would be to create a non-viable embryo. That makes no sense. All these variants would have been obtained by comparing in the general (living) population people with differing copies. Why would these variants be completely harmless in normal people and suddenly lethal in conjunction? Is there any evidence at all for this idea? For example, do we see extremely intelligent people (who will be highly enriched in the relevant variants) dropping like flies in their teens?

chubot 12y ago

It's also a huge mistake to think that this is the ONLY thing those genes do. It's likely that many of them will increase depression, anxiety, ADD, schizophrenia, drug addiction, etc. Evolution produces genomes within the constraint of the environment. I'm no luddite, but I'd be pretty concerned with genetic engineering without regard for the environment of the individual.

stephenhsu 12y ago

The question of additivity of genetic effects is discussed in more detail in reference [1] above (sections 3.1 and also 4): http://arxiv.org/pdf/1408.3421v2.pdf http://arxiv.org/pdf/1408.3421v2.pdf In plant and animal genetics it is well established that the majority of phenotype variance in complex traits which is under genetic control is additive. (Linear models work well in species ranging from corn to cows; cattle breeding is now done using SNP genotypes and linear models to estimate phenotypes.) There are also direct estimates of the additive / non-additive components of variance for human height and IQ, from twin and sibling studies. Again, the conclusion is the majority of variance is due to additive effects. There is a deep evolutionary reason behind additivity: nonlinear mechanisms are fragile and often "break" due to DNA recombination in sexual reproduction. Effects which are only controlled by a single locus are more robustly passed on to offspring. Fisher's fundamental theorem of natural selection says that the rate of change of fitness is controlled by additive variance in sexually reproducing species under relatively weak selection. Many people confuse the following statements: "The brain is complex and nonlinear and many genes interact in its construction and operation." "Differences in brain performance between two individuals of the same species must be due to nonlinear effects of genes." The first statement is true, but the second does not appear to be true across a range of species and quantitative traits. Final technical comment: even the nonlinear part of the genetic architecture can be deduced using advanced methods in high dimensional statistics (see section 4.2 in [1] and also http://arxiv.org/abs/1408.6583 http://arxiv.org/abs/1408.6583).

stephenhsu 12y ago
I just realized I've said all of this already in http://arxiv.org/pdf/1408.3421v2.pdf http://arxiv.org/pdf/1408.3421v2.pdf (p.16): ... The preceding discussion is not intended to convey an overly simplistic view of genetics or systems biology. Complex nonlinear genetic systems certainly exist and are realized in every organism. However, quantitative differences between individuals within a species may be largely due to independent linear effects of specific genetic variants. As noted, linear effects are the most readily evolvable in response to selection, whereas nonlinear gadgets are more likely to be fragile to small changes. (Evolutionary adaptations requiring significant changes to nonlinear gadgets are improbable and therefore require exponentially more time than simple adjustment of frequencies of alleles of linear effect.) One might say that to first approximation, Biology = linear combinations of nonlinear gadgets, and most of the variation between individuals is in the (linear) way gadgets are combined, rather than in the realization of different gadgets in different individuals. Linear models works well in practice, allowing, for example, SNP-based prediction of quantitative traits (milk yield, fat and protein content, productive life, etc.) in dairy cattle. ...

Alex3917 12y ago

Speaking of normal distributions, in order for someone to have an IQ of 1,000, wouldn't there have to be trillions and trillions of people?