Ask HN: Is synthetic data generation practical outside academia?
4 points by cpard 9h ago 2 comments
Ask HN: Has anybody built search on top of Anna's Archive?
284 points by neonate 3d ago 146 comments
Longevity Is Now a Factor When Picking an Embryo for IVF
35 Bluestein 89 6/5/2025, 3:09:32 PM wsj.com ↗
I mean, if you'd killed me as an embryo, I'd be every bit as dead today as if you'd killed me yesterday. I can't dissociate my adult self from my embryo self enough to ignore that.
And who knows? Maybe I would have made the cut. Maybe I'm as good as my parents could do. But I can't shake my unease about treating rejected embryos like there aren't adults implicit in them that we're never letting see the light of day.
You wouldn't have existed in any meaningful definition of "you" if your embryo had been destroyed, so you wouldn't be dead. Being dead and being not here are not the same thing. I've never had children. That doesn't mean my children are dead.
> When did something so significant happen that I became an "I"
Consciousness. The "think" part of "I think therefore I am". The part that makes humans human at all. Life experience. Presumably many years of it. You aren't even the same person you were yesterday let alone the cluster of cells that never experienced consciousness.
> I'm just the one organism, from all the way back then until now
Stopping there is arbitrary. If some rockslide hadn't happened a million years ago then your parents wouldn't have been born. It's causal chains and forking paths all the way down. How much do you debate the morality of rockslides? We have the power as a species to stop them now. Do we stop them? Do we never stop them? Both paths are genocide, just of different groups of people who will never be. Which group is the one we should choose to support?
I don't think we are talking about a difference of degrees. That's what makes me viscerally uncomfortable about it. If this tech had been commonplace when my parents decided they wanted to have a baby and conceived me, they would have had to look at this graded set of their offspring, one of which was me, and pick me out of the group--the group of...what? The group of other embryos basically just like me. Genetically speaking, my brothers and sisters. And then they pick me (or don't pick me!) and the rest just...die. Or languish, frozen, in case they chose one that didn't work out, or they want more later.
The ordinary process of IVF is weird enough to me. Adding this dystopian layer of trying to select for the best of your offspring based on genetic hypotheses about their future and disposing of the rest makes my skin crawl.
Good. Done. Problem solved. We agree that we don't need to know exactly when it was to know when it definitely wasn't.
> Adding this dystopian layer of trying to select for the best of your offspring based on genetic hypotheses about their future
IMO you should stick with this part and forget about the embryo part. You can be worried about the process and outcomes of eugenics without focusing on any hypothetical unselected individual.
Not remotely. You provide no defense whatsoever of consciousness as a valid distinction between "you" and "not you yet," so it doesn't matter whether it's present at the beginning of your development or not. Whether one arbitrary feature you think is decisive is present at a particular point is completely irrelevant to anybody else, unless you can defend that that feature actually matters. I'm looking forward to hearing your case for it.
The undisputed reality is that every human adult was once a specific human embryo. "Embryo" is one stage of human development, just like "infant," "adolescent," or "adult."
There might be an argument for dismissing and eliminating certain embryonic humans en masse because of some trait they have or don't have, but you'd really need to pick one and make an argument for it, and it'd be a good idea to make sure the distinction doesn't apply to humans at other stages too if you don't want to have to argue that it'd be fine to kill toddlers, adolescents, or adults if they happened to have the same trait. It would probably also be important to make sure it's empirically verifiable, unless you're fine with whatever argument you make being used to justify killing toddlers, adolescents, or adults who might be thought to have/not have that trait.
But if such an argument does exist, you certainly haven't made it yet here.
> what level of consciousness matters?
We can start at least with the obvious distinction between "None at all" vs "Some" even if we don't know where the change happens. Asking "how much" only works if you want to investigate this dilemma within the "some" category, which we at least agree does not include embryos, and that embryos are in the "none at all" category. Investigating the "some" category strikes me as substantially more difficult because of that question about how much is enough.
For now, I say that "some" level of consciousness is required for mattering, which precludes a "none at all" level of consciousness from mattering.
Consider brain death. When a body becomes nothing more than a mass of tissue kept in motion by machine intervention without any brain activity, do we not say that it's ok for the family to decide to pull the plug? Because I say that it's ok to not keep brain dead patients on life support until the sun eventually explodes and the power runs out. But maybe you don't. And I understand that this is talking about an end rather than a beginning, but it's really just the other end of your question about how much consciousness matters. Clearly in that scenario the consensus answer is "at least some".
> There are severely disabled adults who have never been as "conscious" as some of the more advanced other animals. Do they just not count?
Question needed for me to establish a more shared understanding: Are you vegan? (I am not.)
If so, most people in the world are not, and that means if we're comparing to animals then clearly, at least en masse, we do discount lesser animal consciousness even if good vegans don't. But at least this framework would be consistent with a view that a threshold of human-level consciousness shouldn't matter.
If not, do you have a basis other than degrees of sapience for believing that a human embryo is worth worrying about but a chicken embryo is not that doesn't stem from egocentric instinctive queasiness? (I mean "egocentric" neutrally here, not negatively) Because the only other option I know of is "the <religious_book> says so", which is thought-terminating and we won't be able to move past that.
> If this tech had been commonplace when my parents decided they wanted to have a baby and conceived me, they would have had to look at this graded set of their offspring, one of which was me, and pick me out of the group
I get what you're saying but people actually already do this decision-making and have been doing so for a very long time. They just do it ahead of time by choosing who they reproduce with. Couples all over the world make decisions literally all the time about passing on negative potentially-heritable traits, from mild Autism to Scoliosis to Tay-Sachs.
I mean that is great news when you get into the job market, right? Instead of 400 applicants for that job you need, only competing against 300 — dead babies don’t need to work!!!
We can do that, sure, but what do we gain? More heartache, harder decisions, more prison time? What's the benefit?
I think it’s the attempt to be moral by a simplistic definition of ‘doing no harm’ that leads to the contortions to the definition of harm. In this context we’re talking about if it’s ok to harm a small set of cells, and again it depends on context. Consider the act of assaulting a pregnant woman to harm the fetus, it’s the same fetus that has come to the same harm the only thing that is different is who made the choice.
Yes. But I don't wander around saying that harvesting it doesn't harm the plant.
Not to mention the obvious difference between something that has lived experience and something that has none.
Second, personhood is a philosophical thing so the medical and scientific communities' opinions are not relevant
Third, I looked up lived experience on wiki and it is talking about knowledge gained from the experience. A baby who was born yesterday doesn't have lived experiences so are they not a person?
Non-persons cannot be me, a person. A precursor is a precursor is a precursor.
> Are you suggesting there are human nonpersons?
We quite often use the word "human" as an indicator of future or past potential or characteristic without meaning that the thing itself is a human. My dad's ashes are "human remains", but they're definitely not a person.
Just like how a sapling is a tree even if it is still in a container. You don't have to wait for the tree to be transplanted to the ground for it to be a tree.
A fetus is part of human development different from toddlers and adults which are also parts of human development different from each other. A sperm is part of human development. A protein molecule is part of human development. A second glass of wine on a rainy evening is part of human development.
> A second glass of wine on a rainy evening is part of human development.
You are just being obtuse. We aren't talking about that sort of development and you know it. We are talking about a toddler becoming an adolescent and a teenager becoming an adult.
If you want to make a falsifiable statement, go for it, but don’t make a metaphysical statement and then assume it’s well established physics.
If you kill a dog then the dog died. If you kill an embryo then the embryo is dead. It is irrelevant if either of those two are persons.
I'm not taking a position on if an embryo has personhood.
> Third, destroying an embryo means you ceased to exist, not that you never existed like the original person I responded to said.
Something that eventually became me ceased to exist. It wasn't anywhere close to being me yet, just a precursor, like Napoleon's chest hair.
I find it hard to believe you really aren't familiar with this basic truth of biology, which makes the rest of your argument seem disingenuous.
It sounds like you might actually care more about the intention behind the selection than the fact that selection has occurred. That someone might have looked at you specifically and said "no thanks". That objection is completely orthogonal to age. Someone looks at you every day right now and says "no thanks", whether it's for a job or for a sexual partner. Someone every day is deciding that they do not want your sperm or your egg, which means that your hypothetical child together is prevented from being born, the same as it will be prevented by this. And it may feel bad to think about it, but the selection is that person's to make. Life is full of making selections and being or not being selected. This isn't special.
Of course there is. The embryo is the same organism as an adult, just at a different stage of its own life. All it needs is time, nutrition, and safety and it will become an adult. It's the same life form at different stages. It takes an intervention to stop an embryo from becoming an adult.
For any other precursor, something fundamental has to change. For either of the gametes (the immediate precursors), they have to merge with the other gamete to form a whole new organism. That's extremely different. It's even more different the farther back you go.
You're either woefully ignorant of basic biology or just completely philosophically disingenuous. Do you sincerely not know how organisms work at all? What they are? That they go through stages? That they have a clear beginning, a definite first stage of their lifecycle, before which they don't exist and after which they have to die in order to stop existing?
This is some real elementary school science class stuff. It's not just humans. This is basic basic stuff.
Taking intervention isn't as useful a selection criterion as you make it out to be. It also takes intervention to stop teenagers from making babies with each other. And all they need is time. They often don't even need nutrition or safety. That teenagers left to their own devices will make babies is a fundamental and incontrovertible truth. It may as well be considered a force of nature. It's the entire history of all of biology from the very first multi-cellular organism until today. And yet modern society usually succeeds in preventing them from doing it because we recognize the harms caused by it.
> For any other precursor, something fundamental has to change.
This is both arbitrary and covers a lot more than you admit. For a child to become you today it had to change in very fundamental ways, almost exclusively by external social influence. You are not the same as the child. You are dramatically and fundamentally changed. Your ideas are dramatically and fundamentally different. Your feelings are dramatically and fundamentally different. Your behavior is dramatically and fundamentally different.
This notion of fundamental change only holds up if you arbitrarily hold onto one particular kind of change while blithely ignoring all the countless rest.
> For either of the gametes (the immediate precursors), they have to merge with the other gamete.
So? That's not any more fundamental than the fact that external social influences have merged with your frame of mind to change your behaviors, emotions, and views as you've aged. You are as fundamentally changed from the child that became you as the child is from the sperm and egg it came from. You are changing constantly by merging with external influence. Your ideas are as different as ideas can be, your emotions are as different as emotions can be, your behaviors are as different as behaviors can be. You are the culmination of a lot of merging, not alone in a vacuum.
It used to be that most people died young. If your genes weren't strong enough, disease would've knocked you out before you became a teenager.
It's only relatively recently that we've made it possible for people of all genetic types to survive into adulthood.
This is much more akin to the old practice of leaving babies to die by exposure if they seemed sickly/weak, which still turns my stomach.
I think it would even be a different thing to me if we were screening sperm and ova for their genes and only combining the best sets than screening the embryos.
Okay, two factors.
What if ... both?
PS. Lots of "exponential" all-or-nothing, intelligence related "arms races" around lately ...
... interesting times.-
We can barely manipulate simple organisms without causing negative second order effects to cascade through our systems.
Your approach assumes that the distribution of those SNPs at the population level is immaterial.
What makes an allele attractive at one frequency can expose new liabilities at a different frequency.
No it doesn't.
So there is this thing called statistics...There is a false assumption in behavioral genetics that behavior and intelligence are weakly linked to 1000s of SNPs which would make it rather difficult to stay within normal bounds, but in reality there are a few rare SNPs that dominates high intelligence. The RCCX genes of TNXB, CYP21A2, and C4 being the strongest. Of these I would focus on only two TNXB SNPs and there are many people walking around today who have them.
Veritasium had a good primer on IQ scores and what they actually mean for your life
TLDW: IQ score seems to really only correlates to being too dumb to be conscripted and effective in war.
Going on Derek's work there, even if other countries do somehow manage to genetically engineer their children for higher IQ scores, and even if they manage to get all the environmental hazards (lead paint, nutrition, quality sleep, etc) out of the way for those children, and even if they don't then do some other draconian nonsense to those kids in the ~19 years until they become productive, then, even then, there seems to be absolutely no guarantee that those kids will go on to do any dominating of other countries in social, political, or military contexts.
Outside of being too dumb to be trusted with a firearm, IQ seems to have no real bearing on measurable life outcomes. As the current science stands, from my understanding of it, it would be incredibly foolish to pursue genetic engineering/breeding as a method of international relations.
Similar to putting weapons on drones, it used to be unthinkable and now it's unthinkable not to.
Heck, for argument, "incels" would argue the females of the species have been doing this forever, at least in very broad strokes.-
(To be clear, I think that's an important part of what makes me so uncomfortable with it. I don't mean that as a justification for why it should be fine to do it.)
I’m not sure that gene edited super intelligence is in the near future.
I don't have any trouble saying this is a race I'm fine with "losing" because I refused to play, because the prospect of living in such a world sounds horrible to me.
From my perspective, it's not that different from the reason I don't own a remote bunker anywhere. I wouldn't want to survive a nuclear holocaust, because living in a post-nuclear-holocaust world sounds horrible.
The fact that other people might do things I would find terrible, and that I might rather let them kill me than willingly participate in the world their activity created, seems obvious to me, but I can at least imagine how someone else might see it differently.
Perhaps it's that many things seem worse to me than death?
As a society, we have power if we want to use it. It's never perfect, but it can be very effective. Saying it's impossible is just a means - intentionally or not - of preventing people from using that power.
you have no power over China, which is what the GP implies
It's not like human variability would be lost forever, we can make and keep copies.
Gattaca's core message is that there is no gene for the human spirit, but in reality there probably is, behavior is highly genetic. Also consider that Vincent could have had a heart attack during the mission, that would put his actions into a rather selfish light.
My optimistic view is that an improved understanding of the mechanisms would lead to direct medical interventions, in such a world Vincent would have been able to have his heart properly treated.
I have ME/CFS from a genetic condition and I treat it by suppressing IL-1B cytokines with diet and medication. I was lucky enough to be able to figure this out and I would like more people to have this knowledge.
Biomolecular condensates: https://pmc.ncbi.nlm.nih.gov/articles/PMC9974629/
Take for example that we can read DNA and therefore have a list of every bioactive peptide that the body produces, we also have WGS so we can identify people with changes to these peptides. By using algorithms on these populations we can understand quite a lot about what these peptides do without understanding the mechanism by which they do it.
Basically I think humans taking our genetic code into our own hands intentionally is inevitable and the question shouldn't be whether it should be allowed or not but instead how do we do this wisely and when are we ready for which steps.
> Nucleus also provides probability assessments for the embryos’ IQ, height and eye color. Sadeghi said that Nucleus informs customers of the strength of the various predictions. IQ predictions are limited in accuracy, according to the company.
On top of that you have recessive genetic traits as well which means you can get stuff like skin, hair and eye color that "pause" sometimes over generations. Used to be quite a source of infidelity allegations until Mendel's inheritance research, and frankly it's still something that pops up even today.
If we take 100 embryos and we both predict their heights and IQs at 18, you with only biological sex and me additionally with the heights and IQs of the parents, I will wager a large sum that my predictions will be better than yours.
Are you interested in taking the other side of that bet for even money? I’ll even give you +110 odds to ensure you have a sportsbook-sized edge if there’s no predictive value.
Even twins raised in totally separate households will have very similar IQs later in life.
This epigenetic stuff is invariably copium from people who don’t want to admit their intellectual gifts are in fact gifts and as a corollary, all the downstream economic rewards of intelligence are also mere gifts from the lottery of being born to high-IQ parents.
The explanation seems to be that whatever random genes you got to have an off-average IQ score cancel out with your partner and your offspring then regresses to the mean.
What will be interesting is there's a lot of recent work to suggest IQ/personality differences can be seen through physiognomy, and it'll be interesting to see if as this tech gets better, the people who go through this process start having similar looking children.
What work are you referring to?
https://x.com/hamptonism/status/1929702249439412248
I'm not saying it's very exact, but I think the general premise holds true: there will be certain options we can choose from, and those options will lead to a certain convergence of traits or features. How strong the convergence will be is yet to be seen (if it happens at all)
While a lot of physical characteristics are indicative of some sort of underlying health defect and those defects can have effects on IQ/personality and humans are very evolved to detect these things, that doesn't mean that the absence of these defects through genetic modification will necessarily lead to similar looking people.
For an example of this take a look at the picture of The Martians[0], a group of individuals considered to be astoundingly intelligent. Despite coming from the same time and place and small ethnic group they bear a passing resemblance and many of them look quite different.
Now this doesn't get into the possibility that humans will use genetic modification to impart superficial aesthetic choices on their children that will be subject to the fleeting whims of fashion as we're already seeing with cosmetic surgery and injectable fillers/botox that kind of stuff will be subject to the cyclical nature of fashion/style and not intrinsic to the any sort of physiognomy.
If anything we're likely to see people who look radically different than any traditional expectations of beauty and the traditional correlations between appearance and intelligence will be totally out the window. Sooner or later we'll be able to create people who have feathers instead of hair and someone (probably the kind of person who currently has gauge earrings, split tongues, and eyeball tattoos) will decide that this is what they want for their kids.
There isn't going to be clean mapping between intelligence and someone who has feathers for hair, or purple eyes, or three breasts.
[0] https://en.wikipedia.org/wiki/The_Martians_(scientists)