I don’t think we are at the point where we can adequately assess the risks involved. It’s known that higher IQ is correlated with major depression, bipolar disorder, and schizophrenia. What use is having a super-intelligent child if they have to spend most of their teenage and early adult years away from society, in a medicated stupor?
There may also be other genetic side effects to increased intelligence, such as increased risk of alcohol dependence and substance abuse.
I think I remember a study saying that over an IQ of 130, there is no correlation between increased intelligence and success/happiness.
It would probably be far more worthwhile to focus on having children of moderate-to-high IQ score (120-130 range), and put more emphasis on better upbringing, instilling values such as the importance of socializing and putting effort into one’s goals. The focus that some transhumanists seem to have on raw intelligence seems a bit childish and naive.
If James_Miller meant ‘genetic basis of intelligence’ (and I think he did) then I am pointing out that that may not be predictive of actual intelligence when measured in the real world after development. You could just as well say I’m ‘optimizing for intelligence’. I am simply making it clear that I’m not optimizing for at-birth intelligence.
Nutrition, intellectually stimulating environments, presence of both parents, and existence of other children to play with have all been shown to be positively correlated with doing better at school, for one. I’m sure there are many other factors.
Another point, not directly related to your question, but related to OP’s question, is that an IQ of, say, 130 may not be that high (and definitely not that high compared to the LW average) but it is 2 standard deviations above the mean… if everyone reached that average level of intelligence it would be a vast improvement in average intelligence over what it is now.
if everyone reached that average level of intelligence it would be a vast improvement in average intelligence over what it is now.
I agree, but this isn’t actionable information for transhumanists. In contrast, a few transhumanist couples could, perhaps, in a decade create a biological super-intelligence. I would love to get an 18-year-old reader of LW to start thinking about doing this.
It’s certainly possible to use simple selective breeding techniques to increase intelligence beyond what would ever likely occur naturally. Modern experience in selective breeding of, for example, cattle for milk production has resulted in herds of cows that produce far more milk than even the most extreme natural outlier ever produced. And furthermore there are statistical tools that can take as input various traits (various intelligence scores and also factors relating to general health and well-being) and produce, as outputs, pairings that would result in optimal intelligence increase. Going further, modern genomics techniques (like sperm sorting and prediction of traits from embryonic gene sequences) could make the process even more rapid.
But it could never be done in a decade. Modern techniques require a minimum of around ~5 generations to properly maximize traits beyond what would be found in the natural population (this varies hugely depending on the trait, of course, but 5 generations is a commonly-used ballpark estimate). Assuming impregnation starts as soon as reproductive viability is achieved, that gives a figure of 75 years.
The only thing that could shorten this would be designer baby technology. A simple method could be using embryonic stem cells to go directly to gametes without having to go through birth, development, and maturation. The downside to this is that prediction of intelligence based on just embryonic DNA is flimsy; much more generations would probably be required, and a few ‘interim’ individuals would probably have to at least reach school age for model calibration. Assuming, say, three interim stages, that gives 24 years. Even this would require a huge amount of resources—and not to mention the sacrifice and enormous ethical issues involved.
I can’t see even modern genetics technology achieving biological superintelligence any shorter than that, unless you are willing to throw trillions of dollars at it.
We identify a bunch of genes that either increase or decrease intelligence and then use CRISPR to edit the genomes of embryos to create super-geniuses. Just eliminating mutational load from an embryo might do a lot.
The reason this approach won’t work is that genes aren’t linear factors that can added up together in that way. Even in something as simple as milk production, you need to do selection over multiple generations and evaluate each generation separately, building up small genetic changes over time.
If you could construct an actual model relating various genes to intelligence, in a way that took into account genetic interactions, then you could do what you propose in a single generation, but we are very very far from being able to construct such a model at present.
As it stands today, if you just carried out that naive approach you would end up with a non-viable embryo or, in the best-case scenario, a slightly-higher-than-average intelligence person. Not a super-genius.
When researching my book I was told by experts that the intelligence genes which vary throughout the human population probably are linear. Consider President Obama who has a very high IQ but who also has parents who are genetically very different from each other. If intelligence genes worked in a non-additive complex way people with such genetically diverse parents would almost always be very unintelligent. We don’t observe this.
HLS students of any skin color have high IQs as measured by standardized tests. The school’s 25th percentile LSAT score is 170, which is 97.5th percentile for the subset of college graduates who take the LSAT. 44% of HLS students are people of color.
There are decades of studies of the heritability of IQ. Some of them measure H², which is full heritability and some of them measure h², “narrow sense heritability”; and some measure both. Narrow sense heritability is the linear part, a lower bound for the full broad sense heritability. A typical estimate of the nonlinear contribution is H²-h²=10%. In neither case do they make any assumptions about the genetic structure. Often they make assumptions about the relation between genes and environment, but they never assume linear genetics. Measuring h² is not assuming linearity, but measuring linearity.
This paper finds a lower bound for h² of 0.4 and 0.5 for crystallized and fluid intelligence, respectively, in childhood. I say lower bound because it only uses SNP data, not full genomes. It mentions earlier work giving a narrow sense heritability of 0.6 at that age. That earlier work probably has more problems disentangling genes from environment, but is unbiased given its assumptions.
We fitted a linear mixed model y = µ + g + e, where y is the phenotype, m is the mean term, g is the aggregate additive genetic effect of all the SNPs and e is the residual effect.
If you have 3511 individuals and 549692 SNPs you won’t find any nonlinear effects.
3511 observations of 549692 SNPs is already overfitted 3511 observations of 549692 * 549691 gene interactions is even more overfitted and I wouldn’t expect that the four four principal components they calculate to find an existing needle in that haystack.
Apart from that it’s worth noting that IQ is g fitted to a bell curve. You wouldn’t expect a variable that you fit to a bell curve to behave fully linearly.
No, they didn’t try to measure non-linear effects. Nor did they try to measure environment. That is all irrelevant to measuring linear effects, which was the main thing I wanted to convey. If you want to understand this, the key phrase is “narrow sense heritability.” Try a textbook. Hell, try wikipedia.
That it did well on held-back data should convince you that you don’t understand overfitting.
Actually, I would expect a bell curve transformation to be the most linear.
That it did well on held-back data should convince you that you don’t understand overfitting.
They didn’t do well on the gene level: Analyses of individual SNPs and genes did not result in any replicable genome-wide significant association
No, they didn’t try to measure non-linear effects. Nor did they try to measure environment. That is all irrelevant to measuring linear effects, which was the main thing I wanted to convey.
No, the fact that you can calculate a linear model that predicts h_2 in a way that fits 0.4 or 0.5 of the variance doesn’t mean that the underlying reality is structured in a way that gene’s have linear effects.
To make a causal statement that genes work in a linear way the summarize statistic of is not enough.
I would not recommend making confident pronouncements which make it evident you have no clue what you are talking about.
While I haven’t worked with the underlying subjects in the last few years I did take bioinformatics courses by people who had a clue what they were talking about and the confident pronouncement I make are what I learned there.
No, it was assumed that genes controlling milk production were linear, because it was much easier to study them that way, and unfortunately over time many people came to simply accept that fact as true, when it has never been proven (in fact it’s been proven conclusively otherwise).
I don’t think we are at the point where we can adequately assess the risks involved. It’s known that higher IQ is correlated with major depression, bipolar disorder, and schizophrenia. What use is having a super-intelligent child if they have to spend most of their teenage and early adult years away from society, in a medicated stupor?
There may also be other genetic side effects to increased intelligence, such as increased risk of alcohol dependence and substance abuse.
I think I remember a study saying that over an IQ of 130, there is no correlation between increased intelligence and success/happiness.
It would probably be far more worthwhile to focus on having children of moderate-to-high IQ score (120-130 range), and put more emphasis on better upbringing, instilling values such as the importance of socializing and putting effort into one’s goals. The focus that some transhumanists seem to have on raw intelligence seems a bit childish and naive.
What are you optimizing for?
The optimal mix of intelligence and ability to make use of intelligence.
You just shifted all the meaning to the word “optimal”.
Optimal when maximizing for what?
No I did not.
If James_Miller meant ‘genetic basis of intelligence’ (and I think he did) then I am pointing out that that may not be predictive of actual intelligence when measured in the real world after development. You could just as well say I’m ‘optimizing for intelligence’. I am simply making it clear that I’m not optimizing for at-birth intelligence.
I still don’t understand you.
Is there any measurable value that you are optimizing for? What is it?
What do you mean specifically with that sentence?
Nutrition, intellectually stimulating environments, presence of both parents, and existence of other children to play with have all been shown to be positively correlated with doing better at school, for one. I’m sure there are many other factors.
Another point, not directly related to your question, but related to OP’s question, is that an IQ of, say, 130 may not be that high (and definitely not that high compared to the LW average) but it is 2 standard deviations above the mean… if everyone reached that average level of intelligence it would be a vast improvement in average intelligence over what it is now.
I agree, but this isn’t actionable information for transhumanists. In contrast, a few transhumanist couples could, perhaps, in a decade create a biological super-intelligence. I would love to get an 18-year-old reader of LW to start thinking about doing this.
It’s certainly possible to use simple selective breeding techniques to increase intelligence beyond what would ever likely occur naturally. Modern experience in selective breeding of, for example, cattle for milk production has resulted in herds of cows that produce far more milk than even the most extreme natural outlier ever produced. And furthermore there are statistical tools that can take as input various traits (various intelligence scores and also factors relating to general health and well-being) and produce, as outputs, pairings that would result in optimal intelligence increase. Going further, modern genomics techniques (like sperm sorting and prediction of traits from embryonic gene sequences) could make the process even more rapid.
But it could never be done in a decade. Modern techniques require a minimum of around ~5 generations to properly maximize traits beyond what would be found in the natural population (this varies hugely depending on the trait, of course, but 5 generations is a commonly-used ballpark estimate). Assuming impregnation starts as soon as reproductive viability is achieved, that gives a figure of 75 years.
The only thing that could shorten this would be designer baby technology. A simple method could be using embryonic stem cells to go directly to gametes without having to go through birth, development, and maturation. The downside to this is that prediction of intelligence based on just embryonic DNA is flimsy; much more generations would probably be required, and a few ‘interim’ individuals would probably have to at least reach school age for model calibration. Assuming, say, three interim stages, that gives 24 years. Even this would require a huge amount of resources—and not to mention the sacrifice and enormous ethical issues involved.
I can’t see even modern genetics technology achieving biological superintelligence any shorter than that, unless you are willing to throw trillions of dollars at it.
We identify a bunch of genes that either increase or decrease intelligence and then use CRISPR to edit the genomes of embryos to create super-geniuses. Just eliminating mutational load from an embryo might do a lot.
The reason this approach won’t work is that genes aren’t linear factors that can added up together in that way. Even in something as simple as milk production, you need to do selection over multiple generations and evaluate each generation separately, building up small genetic changes over time.
If you could construct an actual model relating various genes to intelligence, in a way that took into account genetic interactions, then you could do what you propose in a single generation, but we are very very far from being able to construct such a model at present.
As it stands today, if you just carried out that naive approach you would end up with a non-viable embryo or, in the best-case scenario, a slightly-higher-than-average intelligence person. Not a super-genius.
When researching my book I was told by experts that the intelligence genes which vary throughout the human population probably are linear. Consider President Obama who has a very high IQ but who also has parents who are genetically very different from each other. If intelligence genes worked in a non-additive complex way people with such genetically diverse parents would almost always be very unintelligent. We don’t observe this.
Evidence?
Harvard Law Review
Counter-evidence: affirmative action.
In any case, it’s interesting that Obama’s SAT (or ACT) scores are sealed as are his college grades, AFAIK.
HLS students of any skin color have high IQs as measured by standardized tests. The school’s 25th percentile LSAT score is 170, which is 97.5th percentile for the subset of college graduates who take the LSAT. 44% of HLS students are people of color.
When I see funny terms like “people of color” (or, say, “gun deaths”), I get suspicious. A little bit of digging, and...
Black students constitute 10-12% of HLS students. Most of the “people of color” are Asians.
No, actually, genetic studies of both milk production and IQ show them to be mainly linear.
That selective breeding has to be done slowly has nothing to do with genetic structure.
What kind of study do you think shows IQ to be mainly linear?
I would guess that you confuse assumptions that the researchers behind a study make to reduce the amount of factors with finding of the study.
There are decades of studies of the heritability of IQ. Some of them measure H², which is full heritability and some of them measure h², “narrow sense heritability”; and some measure both. Narrow sense heritability is the linear part, a lower bound for the full broad sense heritability. A typical estimate of the nonlinear contribution is H²-h²=10%. In neither case do they make any assumptions about the genetic structure. Often they make assumptions about the relation between genes and environment, but they never assume linear genetics. Measuring h² is not assuming linearity, but measuring linearity.
This paper finds a lower bound for h² of 0.4 and 0.5 for crystallized and fluid intelligence, respectively, in childhood. I say lower bound because it only uses SNP data, not full genomes. It mentions earlier work giving a narrow sense heritability of 0.6 at that age. That earlier work probably has more problems disentangling genes from environment, but is unbiased given its assumptions.
The linked paper says:
If you have 3511 individuals and 549692 SNPs you won’t find any nonlinear effects. 3511 observations of 549692 SNPs is already overfitted 3511 observations of 549692 * 549691 gene interactions is even more overfitted and I wouldn’t expect that the four four principal components they calculate to find an existing needle in that haystack.
Apart from that it’s worth noting that IQ is g fitted to a bell curve. You wouldn’t expect a variable that you fit to a bell curve to behave fully linearly.
No, they didn’t try to measure non-linear effects. Nor did they try to measure environment. That is all irrelevant to measuring linear effects, which was the main thing I wanted to convey. If you want to understand this, the key phrase is “narrow sense heritability.” Try a textbook. Hell, try wikipedia.
That it did well on held-back data should convince you that you don’t understand overfitting.
Actually, I would expect a bell curve transformation to be the most linear.
They didn’t do well on the gene level:
Analyses of individual SNPs and genes did not result in any replicable genome-wide significant associationNo, the fact that you can calculate a linear model that predicts h_2 in a way that fits 0.4 or 0.5 of the variance doesn’t mean that the underlying reality is structured in a way that gene’s have linear effects.
To make a causal statement that genes work in a linear way the summarize statistic of is not enough.
I would not recommend making confident pronouncements which make it evident you have no clue what you are talking about.
While I haven’t worked with the underlying subjects in the last few years I did take bioinformatics courses by people who had a clue what they were talking about and the confident pronouncement I make are what I learned there.
OK, let’s try a simpler piece of advice: first, stop digging.
No, it was assumed that genes controlling milk production were linear, because it was much easier to study them that way, and unfortunately over time many people came to simply accept that fact as true, when it has never been proven (in fact it’s been proven conclusively otherwise).