Much like Facebook being told no to its crypto plans, the issue here is less about morality,and more about state control. I'm sure military programs are using this technology to further state ends and "super soldiers" or some such approach will be sought. The public just can't be allowed access, and they'll hide that stifling behind moral platitudes.
> Wouldn't super scientists, engineers, businessmen, salesmen or politicians be more useful to a state than a super soldier?
Soldiers are by social role, servants of government. Super scientists, engineers, salesman, and politicians all present different threats and competition to government. Useful to a civilization yes, but government isn't a civilization, it's people who are clinging onto a bucking-bronco for power and safety.
This seems like a flip dismissal of a pretty serious challenge to the ontology of the grandparent comment. Soldiers can also present threats and competition to government (this is literally what a coup is). Soldiers also retire and go into various fields, including science, engineering, sales, and politics. And almost everything the military does is dependent on external contractors and public/private partnerships and, indeed, university research. So it's an awfully broad conspiracy to add value to infantry grunts, who play a shrinking role in the conception of U.S. military power.
And also every other country in the world has to opt to engage in the same conspiracy, otherwise all the benefits accrue to other countries that are willing to allow the technology to be exposed to civil society (e.g. when the average Chinese university graduate has the benefit of CRISPR super brain, it's going to add much more value than if Fort Hood has a bunch of Captain Americas sitting around).
In total this seems to involve a massive and sprawling worldwide conspiracy involving the coordination of opposing actors who have reasons not to coordinate in service of unclear benefits that don't seem to stand up to much scrutiny.
My point isn't that military isn't a threat to politicians, only that in general they don't seem to see it as one (see historical dictatorships fascination with said supersoldiers)
On top of that, it’s not at all obvious how the logistics work out in a world with volunteer armed forces (and other jobs).
Most editing needs to happen during development, which means you’d need to recruit parents. Maybe some overly-moto idiots would agree but anyone who would seems like dubious stock for a super-human! Even if they did, there’s an 18 year lag before the kid maybe enlists. Then you’ve got to covertly move them to whatever speciality can use their new abilities.
The whole process seems slow, low-yield, and involves so many people [0] that it could not possibly remain secret if someone were trying this. And,
of course, that’s all glossing over the fact that we don’t understand or know how to manipulate polygenetic or plieotropic traits…which is all the strategically interesting ones!
[0] For example, the school nurse flags something. Now what?
Looking at the economics, super soldiers are nearly pure loss. Nearly nothing they could ever do would be net positive on the balance sheet. They're even more lossy than just draftees, what with all the R&D, training, whatever sci-fi stuff they do to their spleen, etc. Maybe they could win a war faster? But the integral is likely to be more expensive all the same.
A 'super-X' would be best fitted for economic generators, not even economic savers. A super-garbageman isn't as good as a super-scientist, year over year.
Unlikely to ever happen. The furthest I see it going would be something like curing some types of rare Mendelian diseases, at least for germ line editing.
The problems with “super” whatever is that (1) we really don’t understand biology or genetics all that well, (2) we can’t even agree on what traits like “intelligence” are, (3) we are too shortsighted to know what to optimize for on the timescale of multiple generations.
The children could not consent, and there is a real danger for various side effects while the benefit was very dubious. The method was also sloppy and didn't produce exactly the intended result but slightly different edits, which again is a recipe for more unintended effects.
It still makes a difference for this kind of experiment. There is a difference between experiments and studies on consenting adults and on children, the bar is higher when you are experimenting on children. And it is much clearer that certain types of experiments are unethical when you perform them on children while on consenting adults you could make an argument that the adults do understand the risk and consent.
The arguments look very different if they had tried to cure a severe genetic disease. The risk/benefit calculation would be entirely different compared to the stunt that was actually performed.
My point is that in cases where consent is impossible like for unborn children, the bar is much higher on what kind of experiment could be considered ethical. There must be a clear potential benefit for the child and an acceptable risk compared to that benefit. This experiment fails on both.
Their "stunt" was when they attempted to make the child immune to HIV because both parents had HIV. Did the children consent to be passed HIV from their mother?
On the one hand we have very well validated methods of preventing transmission of HIV from mother to child, on the other we have an experimental gene modification technique attempting to use a blunt instrument to modify a gene that may reduce the chance of contracting HIV. Yes, the stunt was unnecessary, likely ineffective, and incredibly unethical.
(wiki at least also disagrees with you and said the mothers were HIV negative, so plain old in vitro fertilization would have been all that was actually needed)
Every day, we run experiments on hundreds of thousands of future babies - we mash sperms and eggs together and they mix up in wildly unpredictable ways. Plus countless random mutations. This process already has an uncomfortably high defect rate.
I’m a bacteriologist who uses CRISPR every day, and the biggest problem with CRISPR editing in eukaryotes is off-targets effects, ie, cutting a gene you didn’t mean to. In fact, there’s still a lot about CRISPR we don’t understand at all, like why some sequences seem to be great sites for CRISPR to cut and others don’t seem to work reproducibly.
The most straightforward way to answer the question is that off-target is kind of field lingo for “some other gene got targeted” and results in unexpected cell death.
Imagine in bacteria, if I guide the CRISPR components to a gene that metabolizes an exotic sugar and the cell just inexplicably dies—whereas targeting that very gene in a different spot does what you expected.
In eukaryotes, off-target effects likely occur more often (more genes) but we just don’t see them, because the off-target effect wasn’t immediately fatal. Instead, maybe a tumor suppressor gene got rearranged? It might take decades to see the physiological effect.
I think there's a difference between doing it in theory and the actual implementation in this case. He Jiankui did it using relatively imprecise editing technology with unknown potential issues. That's pretty difference than a theoretical method that would prevent a child from having a disease with a guarantee of no side effects - that's the sort of thing people generally support when polled.
It may do that, but it may not. Many important traits involve huge numbers of genes that have small individual impacts and are mediated by epigenetics. The kinds of edits people are talking about now are single genes that control specific diseases. Even trying to make someone taller is probably out of reach for the near future.
It may aslo become really cheap to do this in the future.
This is never a very interesting comment. Older styles of insulin that no one wants to use are dirt cheap. The newer and preferred kind is still under patent, and one can quibble about that but it was very expensive to develop and Eli Lilly & Co. just announced a monthly out of pocket price cap of $35. Other countries have different systems that cap prices, so pharmas try to take profits in the US. It's not a very good system, but it has produced a lot of really impressive drugs in the last 50 years. We can do better of course, but trashing any medical advance with a quip about insulin is just childish and boring.
I agree that selecting an embryo will become inexpensive.
However, the game will be to select the "best" embryo. This becomes a game of numbers where, on average, the quality of the embryo will be a function of the size the pool of embryos and this will be a function of how much the future parents spend.
You're really saying that on a site mostly about software dev? "What's the big deal with editing binaries in production anyways", flipping one bit can have disastrous consequences. Can you guarantee your gene editing won't amount to a genetic defect or illness? That's a pretty big deal.
Many people complain about the distributional effects. What they fail to see if that the alternative technology: selection of embryos has a far greater distributional impact in the limit.
Assuming that (a) we can sequence embryos reliably, (b) we identify a large number of genes that are considered favorable, and (c) we come up with a way of creating large numbers of embryos (eg: artificial ovaries) then the "quality" of the child in such a future will, generally, be dependent upon that amount the child's parent were willing to spend to find their best possible child.
Do you want a one in ten baby, a one in a thousand baby, or a one in a million?
The price will be proportional to the number of embryos sampled, on average. Since this is a competitive process, no matter how low the price per unit decreases the amount of spend, as a percentage of income/wealth will stay constant.
Further devaluation of our holistic humanity. We already define ourselves largely by the functions we perform in society, but at least current class distinctions are still mostly social. Imagine how much worse it would be if people could literally be bred for certain functions. An engineer will always be a engineer because their traits were optimized for that. A ditch digger will always be a ditch digger because their parents couldn’t afford the engineer package at Build-A-Baby.
I can think of few worlds more dystopian than one where there’s a causal, quantifiable relationship between wealth and biological superiority.
Intelligence and ability already have a much weaker correlation with wealth and success than familial wealth and sheer luck do. Also, why would one genetic modification be any more expensive than another? This is going to happen anyway, sooner or later. The "oh no it will make rich people richer" argument could be applied to a good chunk of all the technologies ever invented, from automobiles to the internet.
The fact that there isn't a causal relationship from wealth to biological superiority is the cause of so many of our problems! Imagine if all the people who got positions via nepotism were actually good at their jobs...
If gene editing can make a human smarter & better looking, then you're dooming your own children to a lifetime of inferiority if you have them naturally. Unlike tech or drug enhanced people, there is no way to detect a genetically enhanced person. So they'd be in your midst, better than you in every way, while you'd be stuck with the realization that there is nothing you can do in this life to change your inferiority to these neo-humans.
"Those who make peaceful revolution impossible will make violent revolution inevitable" - 1962, John F. Kennedy
Pandora's box is already unlocked, we just won't let anyone open it. At the same time, technology is inevitable, and it's probably better to set up legal pathways to access this technology than let the grey market go wild.
Playing devils advocate here; wouldn’t that be a prime example of evolutionary displacement of an inferior species with one that has a higher fitness? Those modified, superior humans would mate, obviously, and spread their gene pool faster than non-modified humans, thus replacing them. Isn’t that a good thing, overall?
But that’s exactly the life you and me already live, and have come to terms with (I hope?). We’re never going to be rich and famous. So nobody had a choice to begin with.
That's not quite true. We are only rendered inferior by capital, or by athletic ability, a skill, a talent. I was responding to the prospect of something all encompassing which to me is what man-made genetic superiority suggests.
It's ironic how much optimism is inherent in these projections of doom. Given our limited knowledge of human biology, way before "gene editing can make a human smarter & better looking", we're likely to see unintended mutations, suffering and extensive litigation/regulation. Tech innovations often have messy beginnings, but this one in particular will be particularly messy (or just plain gruesome).
Yeah, I'm amazed that so many commenters on HN think that the technology, or even our understanding of things like "smarter & better looking" is anywhere near this level of sophistication.
The sentence that really stood out to me was:
> In the years since He’s revelations, researchers have uncovered further concerns about the use of CRISPR–Cas9 in embryos, including that it can shuffle large segments of chromosomes.
There are a number of horrible genetic diseases, including Huntington's Disease, that do not present until later in life; for Huntington's Disease it typically presents when someone is in their 30s or 40s. We don't know what shuffling chromosomes might do. There's a very real possibility that if there was widespread adoption of CRISPR to make children smarter and more beautiful, it would could with the side effect of condemning their bodies and/or minds to fall apart in their teens, 20s, or 30s.
Animal models won't tell us this. We'll only learn the true consequences over time. This is one of those situations where "move fast and break things" is just wrong.
The dawn of infinite power was heralded in with the annihilation of 2 cities in Japan, and the death of many innocents exposed to open radiation in the development process. The industrial revolution began with pollution that worsened the lifespan on the 'winners' for decades before we figured out how to make these spaces safer. Modern medicine started off with bloodletting & modern psychiatry started off with lobotomies.
The first few generations that adopt a revolutionary technology see their lives get worse before they get better.
Problem is, AI, Medicine & Automation are all likely to produce revolutions that are bigger and faster spreading than anything we've seen before. I do not wish to be a 'growing child' in an era of increasing adoption of this tech. Growing up with just the internet was confusing enough as is.
More than all of these, the scariest is a technology that I've started hearing whispers of. We might 'solve aging' in this century. I hope I get to die before the advent of perpetual life seems to be on the horizon.
Traits are advantageous until they are not. Consider the shift of the US economy from blue collar work toward knowledge work over the last 100 years. Those types of work favor different traits and types of people.
I don’t think that would stop people from foolishly altering their germ line to chase trends. But the reduction of entropy / diversity in the gene pool caused by that would represent an existential threat to humanity.
My guess is that the germ lines of people who tried to outsmart evolution would get weeded out pretty fast.
Everyone says germline editing would reduce diversity in the gene pool, and I can see validity to that argument. But, could it not do the opposite when people start altering their children to align with their own values? I could see the loss of some universally "bad" genes, but a lot of the judgement on this sort of things gonna come down to who does the gene editing and how they exactly implement it. When multiple groups start doing this in practice, I could see it fracturing the species more than homogenizing it. Especially when combined with something like iterated embryonic stem cell selection when you can jump several generations at a time. That would result in multiple post-human species derived from baseline humanity with their own specializations, talents, etc. Which would be more diverse.
Ehhh… all you have to do is take a look at instagram or TikTok to yourself convince yourself otherwise. Humans are pack / herd animals and we tend enjoy the safety of crowds.
Having one's morality based on the metaphysics of illiterate desert peasants living two thousand years ago will generally have bad consequences for one's perspective on life, now and here.
Having said this, CRISPR, for all the good it promises, is the wrong technology. There is little to no reason to change the genes, the hardware: we need to find a way to access the software, the inter-cells communication protocols [1].
[1] Michael Levin | Cell Intelligence in Physiological and Morphological Spaces, https://www.youtube.com/watch?v=jLiHLDrOTW8 (The Anatomical Compiler at 17:55: "birth defects, traumatic injury, cancer, aging are problems of information processing")
Never seen a reason, or never agreed with the reasons? The former strains credibility, but if that is really the case you should watch Gattaca. It's quite good.
We understand biology extremely poorly. It’s only been twenty years since we got a decent look at the genome, and like 5 since we could start changing it at all reliably. What we’ve found is that human gene regulation is almost immeasurably complex. There are a few cases where we know exactly what changes to make to fix diseases, but they’re rare. Mostly we just have no idea what would happen if we started messing around with human genes. So, I think a measure of humility and caution is warranted before we start editing human embryos.
Are you agreeing here or are you arguing that because humans have shown a lack of restraint in the past they should continue to show a lack of restraint?
To add to this, we have poor understanding of the sequential deployment of gene expression patterns over a lifetime. A genetic change that prevents a disease in an adult, might have unintended consequences in a child.
Most gene editing isn't tweaking the regulation of genes, nor is it inserting or removing genes, but rather swapping them out for near-identical copies that have slight differences such as not being associated with chronic illness, which is much, much more predictable.
Certainly, Mendelian diseases are a case where gene editing is currently justifiable. In fact, researchers have already cured “bubble boy” syndrome in several kids so that they can get a normal cold and recover just like anyone else. Remarkable stuff.
We might make humans who don't suffer from ailments or who are beautiful. But we might also create people who have terrible, irreparable problems. Those people might die from the problems.
Small genetic modifications cause awful, devastating illnesses. Artificial genome modifications can be done pretty reliably, but we know it's not foolproof. We also have no way to check our work or debug potential genetic issues. We're flying entirely blind, and our hubris can (and in all likelihood, will) result in terrible suffering and death until we get FAR better at doing this.
Because mistakes would cause a lifetime of consequences and given our limited knowledge, it is impossible not to make them. When considering the possibilities of gene editing, take into account an Elephant-Man type outcome, not just the Brad Pitt type outcome (since at this point the one is more likely than the other).
Because it can/will be used by certain groups of privileged classes to perpetuate said privilege, up to and including things that could cast shadows similar to eugenics.
It's where the edges are that matter here, not just the basic idea.
And when people say "that just couldn't happen" given the last half decade of societal weirdness, is no longer a really acceptable response anymore for many.
89 comments
[ 5.2 ms ] story [ 180 ms ] threadOn the other hand, no word about the status of the crispr-kids
Soldiers are by social role, servants of government. Super scientists, engineers, salesman, and politicians all present different threats and competition to government. Useful to a civilization yes, but government isn't a civilization, it's people who are clinging onto a bucking-bronco for power and safety.
And also every other country in the world has to opt to engage in the same conspiracy, otherwise all the benefits accrue to other countries that are willing to allow the technology to be exposed to civil society (e.g. when the average Chinese university graduate has the benefit of CRISPR super brain, it's going to add much more value than if Fort Hood has a bunch of Captain Americas sitting around).
In total this seems to involve a massive and sprawling worldwide conspiracy involving the coordination of opposing actors who have reasons not to coordinate in service of unclear benefits that don't seem to stand up to much scrutiny.
Most editing needs to happen during development, which means you’d need to recruit parents. Maybe some overly-moto idiots would agree but anyone who would seems like dubious stock for a super-human! Even if they did, there’s an 18 year lag before the kid maybe enlists. Then you’ve got to covertly move them to whatever speciality can use their new abilities.
The whole process seems slow, low-yield, and involves so many people [0] that it could not possibly remain secret if someone were trying this. And, of course, that’s all glossing over the fact that we don’t understand or know how to manipulate polygenetic or plieotropic traits…which is all the strategically interesting ones!
[0] For example, the school nurse flags something. Now what?
Looking at the economics, super soldiers are nearly pure loss. Nearly nothing they could ever do would be net positive on the balance sheet. They're even more lossy than just draftees, what with all the R&D, training, whatever sci-fi stuff they do to their spleen, etc. Maybe they could win a war faster? But the integral is likely to be more expensive all the same.
A 'super-X' would be best fitted for economic generators, not even economic savers. A super-garbageman isn't as good as a super-scientist, year over year.
The problems with “super” whatever is that (1) we really don’t understand biology or genetics all that well, (2) we can’t even agree on what traits like “intelligence” are, (3) we are too shortsighted to know what to optimize for on the timescale of multiple generations.
My point is that in cases where consent is impossible like for unborn children, the bar is much higher on what kind of experiment could be considered ethical. There must be a clear potential benefit for the child and an acceptable risk compared to that benefit. This experiment fails on both.
(wiki at least also disagrees with you and said the mothers were HIV negative, so plain old in vitro fertilization would have been all that was actually needed)
1. What are some examples of off-target effects?
2. Are there any examples of them causing adverse effects in organisms?
Imagine in bacteria, if I guide the CRISPR components to a gene that metabolizes an exotic sugar and the cell just inexplicably dies—whereas targeting that very gene in a different spot does what you expected.
In eukaryotes, off-target effects likely occur more often (more genes) but we just don’t see them, because the off-target effect wasn’t immediately fatal. Instead, maybe a tumor suppressor gene got rearranged? It might take decades to see the physiological effect.
It may aslo become really cheap to do this in the future.
OTOH, selection of embryos, by it's nature, will always be expensive.
However, the game will be to select the "best" embryo. This becomes a game of numbers where, on average, the quality of the embryo will be a function of the size the pool of embryos and this will be a function of how much the future parents spend.
Most people complain since they like to complain.
Many people complain about the distributional effects. What they fail to see if that the alternative technology: selection of embryos has a far greater distributional impact in the limit.
... but they like to complain.
Can you ELI5?
Do you want a one in ten baby, a one in a thousand baby, or a one in a million?
The price will be proportional to the number of embryos sampled, on average. Since this is a competitive process, no matter how low the price per unit decreases the amount of spend, as a percentage of income/wealth will stay constant.
I can think of few worlds more dystopian than one where there’s a causal, quantifiable relationship between wealth and biological superiority.
If gene editing can make a human smarter & better looking, then you're dooming your own children to a lifetime of inferiority if you have them naturally. Unlike tech or drug enhanced people, there is no way to detect a genetically enhanced person. So they'd be in your midst, better than you in every way, while you'd be stuck with the realization that there is nothing you can do in this life to change your inferiority to these neo-humans.
Pandora's box is already unlocked, we just won't let anyone open it. At the same time, technology is inevitable, and it's probably better to set up legal pathways to access this technology than let the grey market go wild.I get where you’re coming from, but whether the barrier is 20 meters or infinitely high doesn’t matter in practice, because I can’t jump neither…
The sentence that really stood out to me was: > In the years since He’s revelations, researchers have uncovered further concerns about the use of CRISPR–Cas9 in embryos, including that it can shuffle large segments of chromosomes.
There are a number of horrible genetic diseases, including Huntington's Disease, that do not present until later in life; for Huntington's Disease it typically presents when someone is in their 30s or 40s. We don't know what shuffling chromosomes might do. There's a very real possibility that if there was widespread adoption of CRISPR to make children smarter and more beautiful, it would could with the side effect of condemning their bodies and/or minds to fall apart in their teens, 20s, or 30s.
Animal models won't tell us this. We'll only learn the true consequences over time. This is one of those situations where "move fast and break things" is just wrong.
The dawn of infinite power was heralded in with the annihilation of 2 cities in Japan, and the death of many innocents exposed to open radiation in the development process. The industrial revolution began with pollution that worsened the lifespan on the 'winners' for decades before we figured out how to make these spaces safer. Modern medicine started off with bloodletting & modern psychiatry started off with lobotomies.
The first few generations that adopt a revolutionary technology see their lives get worse before they get better.
Problem is, AI, Medicine & Automation are all likely to produce revolutions that are bigger and faster spreading than anything we've seen before. I do not wish to be a 'growing child' in an era of increasing adoption of this tech. Growing up with just the internet was confusing enough as is.
More than all of these, the scariest is a technology that I've started hearing whispers of. We might 'solve aging' in this century. I hope I get to die before the advent of perpetual life seems to be on the horizon.
I take what you're saying to mean that you don't think we should cure aging?
Do you believe that we should cure cancer?
Do you believe that we should not have cured smallpox?
Why or why not?
I don’t think that would stop people from foolishly altering their germ line to chase trends. But the reduction of entropy / diversity in the gene pool caused by that would represent an existential threat to humanity.
My guess is that the germ lines of people who tried to outsmart evolution would get weeded out pretty fast.
Having said this, CRISPR, for all the good it promises, is the wrong technology. There is little to no reason to change the genes, the hardware: we need to find a way to access the software, the inter-cells communication protocols [1].
[1] Michael Levin | Cell Intelligence in Physiological and Morphological Spaces, https://www.youtube.com/watch?v=jLiHLDrOTW8 (The Anatomical Compiler at 17:55: "birth defects, traumatic injury, cancer, aging are problems of information processing")
Even so, https://pubmed.ncbi.nlm.nih.gov/35723296/
https://pubmed.ncbi.nlm.nih.gov/30995372/
Small genetic modifications cause awful, devastating illnesses. Artificial genome modifications can be done pretty reliably, but we know it's not foolproof. We also have no way to check our work or debug potential genetic issues. We're flying entirely blind, and our hubris can (and in all likelihood, will) result in terrible suffering and death until we get FAR better at doing this.
It's where the edges are that matter here, not just the basic idea.
And when people say "that just couldn't happen" given the last half decade of societal weirdness, is no longer a really acceptable response anymore for many.