Apr 11, 2023·edited Apr 11, 2023Liked by Ives Parr
Progressives will embrace IVG as a technology as it will allow single women and LGBT couples to have children without needing to involve men. IVG will be marketed as a "reproductive technology" that expands "women's rights" (though hopefully, once the cat is out of the bag, the eugenic benefits will follow soon after).
Single women (especially highly-educated, urbanized, professional single women) and LGBT couples are two major constituenties of the Democratic coalition. Coincidentally, both of these groups suffer massive fertility gaps: the discrepency between the number of the children they would have in an "ideal" world and the number of children they actually end up having.
The source of the fertility gaps between the two groups mostly differ. For highly-educated single women, the fertility gap is mostly caused by a combination of lifestyle choices and hypergamy. For LBGT couples, the source of the fertility gap is simply biology.
Currently, if single women or LBGT couples want to have children, they have to go to a sperm bank. However, here is where we throw in an extra wrinkle: there are racial differences in sperm bank donation. Being more specific, there is a shortage of black male sperm donors.
This is extremely problematic for black women as it compounds their already-dire dating market situation. As mentioned before, women are hypergamous and have a strong preference to marry men as educated (or more educated) than themselves. But there are *large* discrepencies in educational attainment between black men and black women.
"Among Black students in higher education, women are more likely than men to earn degrees: Black women get 64.1% of bachelor’s degrees, 71.5% of master’s degrees and 65.9% of doctoral, medical, and dental degrees."
So college-educated black women outnumber college-educated black men almost two-to-one. And it gets worse. Not only are black women competing for a very limited supply of "viable" black men, but black men are seen as more desirable than black women across the racial spectrum (https://gwern.net/doc/psychology/okcupid/raceandattraction20092014.html). So educated black men can (and do) marry out.
Will progressives really have the heart to tell black lesbian couples that they can't have children? I don't think so.
There really is no track record of religious people opposing IVF. Even with Dobbs gone IVF clinics in Texas will tell you it hasn't affected them.
The bigger change would be that if you have artificial wombs and artificial eggs...the inherent value of woman goes way way down. While this would free them from the trials of childbirth, it would also mean they weren't necessary for childbirth.
Catholics are supposed to oppose IVF because it separates reproduction from the conjugal act. I think discarding embryos is also immoral. But most don't follow the rules that closely, and there's no major pushback on stuff like IVF. I think even many pro-lifers have way less concern for a couple of days old embryos.
I'm not sure what the implications of that would be. I figure heterosexual men will still want wives to raise children with. One benefit of artificial wombs would be that people might be more willing to be careful about unintentional pregnancy and be more willing to use polygenic screening. That would be an overall good thing.
I know one Catholic couple that won't do IVF for religious purposes, though they also suspect that the woman uterus can't support a child and they have had multiple chemical pregnancies. If that is the case, IVF wouldn't help anyway. I suppose they could do surrogacy, but lots of people object to surrogacy even on non-religious grounds. They also don't have the money for surrogacy.
They have zero interest whatsoever on the government regulating IVF and don't really consider people who do it to be committing some big sin (unlike abortion).
Personally, I think that all the pro-life/pro-choice stuff is really just a proxy for more primary values like what is the measure of a good life, what is the purpose of sex and marriage, etc. There is a pretty strong correlation between abortion views and TFR, at least amongst the high IQ. In general I think its just people who like kids versus people who don't.
Because people are squeamish about fighting directly over those values they channel it into abortion. But IVF doesn't really have anything to do with abortion. Abortion kills children. IVF makes children. People who are undergoing IVF are pro-natal, so basically aligned with the pro-life movement.
How does this shift the underlying cost effectiveness of PGP-P embryo selection?
For example, my understanding is that, with current technology, we could significantly reduce the risk of certain multigenomic (?) diseases and/or get 2-3 increased points of IQ using PGP-P embryo selection but the cost would be somewhere between $30k-$50k and it's somewhere between extremely inconvenient and painful for the mother. So, unsurprisingly, most people aren't really interested because that's, well, a lot of money for kinda marginal returns, unless you're at high risk for passing on a nasty genetic disorder
So, not in the theoretical, but for someone interested in the practical implications of this over a 5 year time frame, does this significantly increase the benefits, lower the cost, or make it easier on the mother?
Just spitballing, I think I remember Gwern's article on PGP-P selection indicating moderately declining marginal returns to order of magnitude increases in the number of embryos but I can't recall with clarity, so I'm not sure how much of a qualitative improvement this will be. It does sound a lot easier on the mother but it also sounds a lot more expensive.
IVF is expensive in the United States and it can be quite costly to go through a cycle. Other countries can do a cycle at a much lower cost. China is now subsidizing IVF [1]. PGT-P seems like a smart move if you are already going through IVF and especially if you are already biopsying the embryos to look for aneuploidy. The cost on top would be paying for the sequencing and the information from Genomic Prediction or whatever other company. The costs of sequencing depend on the type of sequencing [2]. The cost of sequencing is falling faster than exponential and I believe it is around $200 [3].
The expected return for selection from 5 embryos ~3 disability-adjusted life years and from 10 embryos is ~4 disability-adjusted life years. [4] That is average to top. There are cases where an embryo could live a significantly shorter life, and that could be prevented which seems particularly worthwhile. The gain from IQ selection using information available in 2019 was around 2.5 IQ points for average to top embryos. As the genomic databases grow, the predictive ability will grow as well and the expected return will grow higher. But the major bottleneck is eggs.
I think IVF is uncomfortable and this would make it easier on the mother. Taking skin cells or blood cells would be easier. It might be less costly as well. It would probably only require one blood draw, rather than multiple extractions of eggs. I'm sure the information is available but I do not know the cost of reprogramming a somatic to an induced pluripotent stem cell, nor do I know how much it will cost to turn that into an egg cell. If it is trivial, then the main cost would be sequencing and biopsying, I would guess. If trends continue, sequencing will not be very expensive at all pretty soon and so even whole genome sequencing hundreds of embryos may not be that costly.
This would significantly increase the gains to genetic enhancement because the major bottleneck is the number of embryos. The expected gain for a normally distributed trait is proportional to the square root of the logarithm of the number of embryos [5]. Gwern guesses the cost $5k->$100k and the gain will be 1 - 5 SD for IQ depending on the budget [7]. I'm not sure where the returns would make it not worthwhile...it would depend on a lot of assumptions. IVG would make way for iterated embryo selection as well, which would allow for gains of many standard deviations...but it would possibly be quite costly. When we start getting into "smarter than anyone who has ever lived" the value of such people is probably tremendously large in expectation. How much is it worth to have someone much smarter than Newton, Einstein or von Neumann? I think it would be extroardinarily worthwhile.
We obviously care about multiple traits and it's better to optimize by considering a bunch of different traits so we could have larger gains in "betterness." Rather than pushing to the limits of smarts, we might want to select for higher resistance to depression because expected welfare gains would be larger at some point. It gets pretty complicated. I hope that helps some.
So, first off, thanks for writing all that, fantastic stuff.
Second, I'm not sure what you said, here's what I heard. I heard that it's likely to be much easier for the mother, for the cost to increase substantially, potentially around 2x right now, and for the IQ benefit to increase dramatically, ie going from 2-3 IQ points to 1 standard deviation. That feels like a really big deal, because in normal human terms, 2-3 IQ points is not noticeable while a gain of 1 standard deviation is, almost by definition, immediately obvious to observers.
Those alone feel like a big deal. That sounds like a much more expensive program but with much, much higher rewards and I want to make sure I've got that right. For a simple explanation, if the total cost was $50-$70k for a 1 SD improvement in IQ, that's a major bill but it compares extremely favorably to, say, the financial returns of a bachelor's degree.
There's also another part where I heard that iterated selection may be possible. That's...extremely exciting, especially if we're looking to reasonably increase human IQ in a 20-30 year time frame to solve AI alignment, but it sounds more theoretical than actionable at this point and...I'd be concerned it violates the "not my kid" part of parenting. Eg, it would not be hard to dramatically increase the IQ of a nation, or your individual kid, by using the sperm of a genius, there was an entire business model built around this and...no one really did it. People want to have their own kids. I'm not sure whether having the equivalent of your great-great-great grandson will violate that deep intuitive desire for your own kids.
Again, thanks for the effort, just trying to make sure I understand these developments given limited time to follow them.
On your point about IES, I agree. I think boosting cognitive ability for preventing existential risks is extremely important, but I agree with your point about the kids being unrelated.
I'm not sure about the cost increasing after IVG. It may decrease the cost.
It's hard to give an exact number because it's a function of polygenic score too. With enough eggs, it would be more than 1 SD. I will need to look into this more. Apologies. Gwern has more number crunching here (https://gwern.net/embryo-selection#limiting-step-eggs-or-scores).
Baby mice have been created from two male mice:
https://denovo.substack.com/p/of-mice-and-men-and-eggs
You might be interested in this: https://www.nationalacademies.org/our-work/in-vitro-derived-human-gametes-as-a-reproductive-technology-scientific-ethical-and-regulatory-implications
Thank you!
Progressives will embrace IVG as a technology as it will allow single women and LGBT couples to have children without needing to involve men. IVG will be marketed as a "reproductive technology" that expands "women's rights" (though hopefully, once the cat is out of the bag, the eugenic benefits will follow soon after).
Single women (especially highly-educated, urbanized, professional single women) and LGBT couples are two major constituenties of the Democratic coalition. Coincidentally, both of these groups suffer massive fertility gaps: the discrepency between the number of the children they would have in an "ideal" world and the number of children they actually end up having.
The source of the fertility gaps between the two groups mostly differ. For highly-educated single women, the fertility gap is mostly caused by a combination of lifestyle choices and hypergamy. For LBGT couples, the source of the fertility gap is simply biology.
Currently, if single women or LBGT couples want to have children, they have to go to a sperm bank. However, here is where we throw in an extra wrinkle: there are racial differences in sperm bank donation. Being more specific, there is a shortage of black male sperm donors.
https://www.phillytrib.com/news/health/america-has-a-black-sperm-donor-shortage-black-women-are-paying-the-price/article_8da455ef-fec8-56e5-be6e-167e8b0bbc6c.html
This is extremely problematic for black women as it compounds their already-dire dating market situation. As mentioned before, women are hypergamous and have a strong preference to marry men as educated (or more educated) than themselves. But there are *large* discrepencies in educational attainment between black men and black women.
https://www.aauw.org/resources/article/fast-facts-woc-higher-ed/
"Among Black students in higher education, women are more likely than men to earn degrees: Black women get 64.1% of bachelor’s degrees, 71.5% of master’s degrees and 65.9% of doctoral, medical, and dental degrees."
So college-educated black women outnumber college-educated black men almost two-to-one. And it gets worse. Not only are black women competing for a very limited supply of "viable" black men, but black men are seen as more desirable than black women across the racial spectrum (https://gwern.net/doc/psychology/okcupid/raceandattraction20092014.html). So educated black men can (and do) marry out.
Will progressives really have the heart to tell black lesbian couples that they can't have children? I don't think so.
Very interesting data. Thank you.
There really is no track record of religious people opposing IVF. Even with Dobbs gone IVF clinics in Texas will tell you it hasn't affected them.
The bigger change would be that if you have artificial wombs and artificial eggs...the inherent value of woman goes way way down. While this would free them from the trials of childbirth, it would also mean they weren't necessary for childbirth.
Catholics are supposed to oppose IVF because it separates reproduction from the conjugal act. I think discarding embryos is also immoral. But most don't follow the rules that closely, and there's no major pushback on stuff like IVF. I think even many pro-lifers have way less concern for a couple of days old embryos.
I'm not sure what the implications of that would be. I figure heterosexual men will still want wives to raise children with. One benefit of artificial wombs would be that people might be more willing to be careful about unintentional pregnancy and be more willing to use polygenic screening. That would be an overall good thing.
I know one Catholic couple that won't do IVF for religious purposes, though they also suspect that the woman uterus can't support a child and they have had multiple chemical pregnancies. If that is the case, IVF wouldn't help anyway. I suppose they could do surrogacy, but lots of people object to surrogacy even on non-religious grounds. They also don't have the money for surrogacy.
They have zero interest whatsoever on the government regulating IVF and don't really consider people who do it to be committing some big sin (unlike abortion).
Personally, I think that all the pro-life/pro-choice stuff is really just a proxy for more primary values like what is the measure of a good life, what is the purpose of sex and marriage, etc. There is a pretty strong correlation between abortion views and TFR, at least amongst the high IQ. In general I think its just people who like kids versus people who don't.
Because people are squeamish about fighting directly over those values they channel it into abortion. But IVF doesn't really have anything to do with abortion. Abortion kills children. IVF makes children. People who are undergoing IVF are pro-natal, so basically aligned with the pro-life movement.
Interesting thoughts
How does this shift the underlying cost effectiveness of PGP-P embryo selection?
For example, my understanding is that, with current technology, we could significantly reduce the risk of certain multigenomic (?) diseases and/or get 2-3 increased points of IQ using PGP-P embryo selection but the cost would be somewhere between $30k-$50k and it's somewhere between extremely inconvenient and painful for the mother. So, unsurprisingly, most people aren't really interested because that's, well, a lot of money for kinda marginal returns, unless you're at high risk for passing on a nasty genetic disorder
So, not in the theoretical, but for someone interested in the practical implications of this over a 5 year time frame, does this significantly increase the benefits, lower the cost, or make it easier on the mother?
Just spitballing, I think I remember Gwern's article on PGP-P selection indicating moderately declining marginal returns to order of magnitude increases in the number of embryos but I can't recall with clarity, so I'm not sure how much of a qualitative improvement this will be. It does sound a lot easier on the mother but it also sounds a lot more expensive.
IVF is expensive in the United States and it can be quite costly to go through a cycle. Other countries can do a cycle at a much lower cost. China is now subsidizing IVF [1]. PGT-P seems like a smart move if you are already going through IVF and especially if you are already biopsying the embryos to look for aneuploidy. The cost on top would be paying for the sequencing and the information from Genomic Prediction or whatever other company. The costs of sequencing depend on the type of sequencing [2]. The cost of sequencing is falling faster than exponential and I believe it is around $200 [3].
The expected return for selection from 5 embryos ~3 disability-adjusted life years and from 10 embryos is ~4 disability-adjusted life years. [4] That is average to top. There are cases where an embryo could live a significantly shorter life, and that could be prevented which seems particularly worthwhile. The gain from IQ selection using information available in 2019 was around 2.5 IQ points for average to top embryos. As the genomic databases grow, the predictive ability will grow as well and the expected return will grow higher. But the major bottleneck is eggs.
I think IVF is uncomfortable and this would make it easier on the mother. Taking skin cells or blood cells would be easier. It might be less costly as well. It would probably only require one blood draw, rather than multiple extractions of eggs. I'm sure the information is available but I do not know the cost of reprogramming a somatic to an induced pluripotent stem cell, nor do I know how much it will cost to turn that into an egg cell. If it is trivial, then the main cost would be sequencing and biopsying, I would guess. If trends continue, sequencing will not be very expensive at all pretty soon and so even whole genome sequencing hundreds of embryos may not be that costly.
This would significantly increase the gains to genetic enhancement because the major bottleneck is the number of embryos. The expected gain for a normally distributed trait is proportional to the square root of the logarithm of the number of embryos [5]. Gwern guesses the cost $5k->$100k and the gain will be 1 - 5 SD for IQ depending on the budget [7]. I'm not sure where the returns would make it not worthwhile...it would depend on a lot of assumptions. IVG would make way for iterated embryo selection as well, which would allow for gains of many standard deviations...but it would possibly be quite costly. When we start getting into "smarter than anyone who has ever lived" the value of such people is probably tremendously large in expectation. How much is it worth to have someone much smarter than Newton, Einstein or von Neumann? I think it would be extroardinarily worthwhile.
We obviously care about multiple traits and it's better to optimize by considering a bunch of different traits so we could have larger gains in "betterness." Rather than pushing to the limits of smarts, we might want to select for higher resistance to depression because expected welfare gains would be larger at some point. It gets pretty complicated. I hope that helps some.
[1] https://infoproc.blogspot.com/2023/02/bing-vs-bard-us-china-stem-competition.html
[2] https://learningbioinformatics.substack.com/p/intro-genome-sequencing
[3] https://3billion.io/blog/whole-genome-sequencing-cost-2023
[4] https://www.nature.com/articles/s41598-022-22637-8
[5] https://www.cell.com/cell/fulltext/S0092-8674(19)31210-3
[6] https://gwern.net/embryo-selection
So, first off, thanks for writing all that, fantastic stuff.
Second, I'm not sure what you said, here's what I heard. I heard that it's likely to be much easier for the mother, for the cost to increase substantially, potentially around 2x right now, and for the IQ benefit to increase dramatically, ie going from 2-3 IQ points to 1 standard deviation. That feels like a really big deal, because in normal human terms, 2-3 IQ points is not noticeable while a gain of 1 standard deviation is, almost by definition, immediately obvious to observers.
Those alone feel like a big deal. That sounds like a much more expensive program but with much, much higher rewards and I want to make sure I've got that right. For a simple explanation, if the total cost was $50-$70k for a 1 SD improvement in IQ, that's a major bill but it compares extremely favorably to, say, the financial returns of a bachelor's degree.
There's also another part where I heard that iterated selection may be possible. That's...extremely exciting, especially if we're looking to reasonably increase human IQ in a 20-30 year time frame to solve AI alignment, but it sounds more theoretical than actionable at this point and...I'd be concerned it violates the "not my kid" part of parenting. Eg, it would not be hard to dramatically increase the IQ of a nation, or your individual kid, by using the sperm of a genius, there was an entire business model built around this and...no one really did it. People want to have their own kids. I'm not sure whether having the equivalent of your great-great-great grandson will violate that deep intuitive desire for your own kids.
Again, thanks for the effort, just trying to make sure I understand these developments given limited time to follow them.
On your point about IES, I agree. I think boosting cognitive ability for preventing existential risks is extremely important, but I agree with your point about the kids being unrelated.
I'm not sure about the cost increasing after IVG. It may decrease the cost.
It's hard to give an exact number because it's a function of polygenic score too. With enough eggs, it would be more than 1 SD. I will need to look into this more. Apologies. Gwern has more number crunching here (https://gwern.net/embryo-selection#limiting-step-eggs-or-scores).