A new genetic technology that can quickly change entire populations gets broader | Hacker News Reader

A new genetic technology that can quickly change entire populations gets broader

nautil.us

56 pointsdnetesn44 comments

How long until someone releases a gene-drive that dramatically reduces the yield of a staple food crop? Humanity is nowhere near responsible enough for this technology - we should be terrified of it.

We'll have to deal with it the same way we have dealt with nuclear weapons. By the threat of MAD. Since we do not have the maturity as a species not to weaponise in the first place.

There’s a major visibility difference between a nuclear bomb and gene editing.

I'm not convinced MAD actually works. Once you see the bombs flying, the only logical thing to do is to give up your pre-war resolve of retaliation and completely submit - start printing Soviet flags so to speak. Of course, pre-war you'll try to convince the enemy that it isn't so, but they are not stupid.

The only thing that saved us IMO was the power of the political elites (in East and West), that despite all saber-rattling rhetorics, never wanted a war. Now that that power is waning, I'm not sure what will save us this time around. Maybe religious elites can instill a taboo against using this kind of technology for harm?

MAD works until bombs start flying. But if you let the enemy know you won't retaliate after the bombs fall, MAD breaks down and they're no longer deterred from dropping bombs in the first place.

MAD works by creating a clear, indisputable casual relationship between you attacking and you getting destroyed.

Vulcan's might do the logical thing, but Humans will fight back. MAD is one case where human emotion is both the cause and solution of a problem.

>Once you see the bombs flying, the only logical thing to do is to give up your pre-war resolve of retaliation and completely submit

Not necessarily -- it depends on one's ethical philosophy. And it is well-known that humans do sometimes seek revenge even at the risk of their own lives. Plus, there are procedures in place for retaliating responding to a nuclear attack, so all that is required is for military personnel to act in accordance with the established procedures.

The difference is that I can't enrich Uranium in my garage and no country can do that undetected.

A bio-weapon like this is much simpler to develop and the fewer resources you can dedicate to keep it, the higher the odds it leaks into the environment with catastrophic effects.

Nukes cause localized damage - a blast zone, some fallout and that's it. These weapons are not that easily contained.

In this example, humans would have to willingly propagate whatever crops have been altered to include the gene drive and have reduced yield. It would be trivially easy to just plan a different cultivar of whatever transgenic crop was affected or just use the same CRISPR-Cas9 technology to remove the gene drive from the genome. I think that the point missed in this and many of the other comments is that gene drives don't follow typical mendelian inheritance patterns - they propagate in all progeny but can't 'infect' other individuals in a population.

No willing propagation is required - if you're unlucky and your crops get pollinated with the useless-for-food gene, then all their progeny will carry the gene as well. How would you stop this gene from spreading throughout the population? Cultivate everything in hermetically sealed greenhouses?

The offspring of the useless-for-food plants won't make it to reproduction themselves if humans don't nurture them along. Farm plants are completely dependent on constant human intervention to survive, that's why there are no natural cornfields. What's being proposed would have to be a "superweed," able to outcompete all of the plants that are already out there and trying to grow in cornfields.

So just go through your cornfield, examine each stalk of corn if it's carrying the gene, and if so, pull it out? It's not competing the same way a weed competes - all it has to do is pollinate the other plant. A weed can't do that.

You can't just jump straight to the end state of the air being completely packed with compromised pollen. The gene would have to work its way through the seed distribution system, and that's where it would be stopped.

Here's how a country could control an attack: If a field contains the gene, burn it (don't bother discriminating between carriers and non-carriers). Don't plant seeds that came from fields that had the gene. As a result, the initial attack will last only for one generation, no worse than if the attacker killed the plants. This is practical because the initial attack will be limited to a small number of fields that the country will be able to afford to loose.

How do you know when the attack has begun? If < 1% of a field carries the gene, you won't know it's infected. You re-use the seeds, and the wind blows the pollen, into multiple fields, so now you have many fields, each with a tiny proportion of infected. What do you do when 10% of the fields has 1% infected plants, and the remaining 90% has an unknown, but smaller % of infected?

In order for that to work, the gene would have to be designed so that it spread across the nation before it became significant in any one cornfield. Pollen gets spread inside of a single field a lot more than it gets spread from field to field, so that would be difficult. It's similar to how an extremely deadly virus won't spread as far as a strain of the common cold.

We have had the ability to release bio weapons and or chemical weapons for a long time now with almost no real world use outside of the middle east. Virtually no individuals have done so (there are a few cases). Someone with the scientific know how to produce a gene drive edited crop as you suggest would likely also have had the ability to produce a bio weapon and they have not. So the odds are probably very low. The production of ricin in large volumes is even easier and there have been only a handful of real world cases and those have been small cases.

The difference is that when you have a big red destroy-the-species button, "virtually no individuals", "almost no real world use" and "probably very low" and "handful of real world cases" just isn't good enough. It only takes one person doing this once to affect the whole world. Can you think of a single offensive technology that hasn't been used even once?

I hear what you are saying and generally agree with your points. Like most dual use technologies though, nuclear for example, their are pro's and cons. I do think that there are enough benefits to pursue the technology though such as the potential to eliminate malaria, sickle cell, Huntingtons etc. I completely agree that there are risks but I think the technology is likely simple enough that those risks exist now regardless of institutional support so we may as well seek to get the most out of it.

Also I think the risk here is likely overstated, the genetically edited organisms still need to breed and populate to spread the edited gene. You cannot utilize this technology in weaponized form in something like a gas attack or wide spread targeting. If its an edited person, it would take dozens of generations to affect the general population and would likely be stopped in the mean time. So I am not really seeing the major risk.

is this useful beyond improved breeding for genetic experiments? I understand there is scientific value in being able to quickly and cheaply breed in fewer generations a mouse with specific sets of genes as a tool for downstream research.

I have a hard time imagining gene drive is useful for improving some species: how does one determine the utility of one set of genes versus another? we can not accurately reflect the real world in a lab environment to determine if some change is a net advantage or net disadvantage over all environments, situations a species faces. I agree it will be possible to assess the performance of traits in a lab benchmark, but then it's obvious we are over fitting to the genetic fitness of pulling a lever, or running in a wheel etc, and disregarding the cost of the change on all other functions the organism would face in a representative test. The real world with millions or billions of creatures is a much faster and parallellized test, i.e. a lab with space for 1000 mice, could in theory face 1000 micedays per day of artificial selection, while those in the wild are performing giga micedays per day of natural selection over the whole challenge set without overfitting...

Again, I do see the utility for breeding selective traits and downstream learning about the functions of the genome in a cheaper way compared to fifty fifty coin flip breeding...

In some cases, genetic variation in humans has performed a natural experiment for us. It's absolutely crucial to understand the potential impact of a proposed genetic edit in the realm of messy reality. As an example, Ccr5 delta32 is present in a small fraction of Europeans, and noted for granting HIV immunity but potentially greater risk to West Nile virus. We don't know enough about all the real life consequences of Ccr5 delta32 at the moment, but natural experiments are a rich enough source of information to get us there, at least in theory.

There are quite a few useful applications. Sickle Cell anemia is caused by a recessive gene, "Recessive means that two copies of the gene are necessary to have the trait, one inherited from the mother, and one from the father." If we code the gene out, no more Sickle Cell. There are also a plethora of degenerative diseases such as Huntington's that are caused by a defect in a single gene that this could potentially help eliminate. I would think essentially any disease that can be genetically tested for could be potentially cured by the application of gene drives.

How likely are these single-gene diseases to arise from a point mutation instead of being inherited?

From my limited understanding, Sickle cell at least is always inherited.

what keeps sickle cell anemia in the gene pool? is there some rare beneficial aspect?

or is it because natural selection acts largely on chromosomes instead of genes, and the other variants of the chromosomes have similar negative imperfections? or is it because the chromosome variants that propagate sickle cell anemia contain a different gene that is comparatively better than the equivalent gene on some of the non SCA carrying chromosomes?

i.e. chromosome binding of genes forces natural selection to largely take the good with the bad genes, without cherry picking genes within chromosomes, and mostly operates on chromosome fitness as opposed to gene fitness?

Being heterozygous for the sickle cell trait gives some protection against malaria, which explains why the two tend to be common in the same parts of the world:

https://en.wikipedia.org/wiki/Heterozygote_advantage#Sickle-...

(Nicely guessed, by the way!)

thanks for pointing this out!

does this mene using gene drive to expell SCA alleles from the gene pool would result in a higher malaria rate? it would seem that if one could put both SCA and non-SCA alleles in the same chromosome the advantage would be unconditional?

This is one of the many existential risk vectors we face as a species.

It's also disturbing that the same people who brought us widespread hormone-like chemicals leeching into our food are arguing that this, too, is good for us.

Scientists are definitely batting low on responsible R+D and not causing society-wide or global damage through their amoral-to-immoral, ego-stroking work.

But sure. It's about mosquitoes and not some post-doc trying to be alpha nerd to the point they're willing to endanger the species.

> Yet what first drew the developmental biologist Kimberly Cooper to dabble in gene drive technology for mice was not conservation but evolution. In her laboratory at UCSD, she and her colleagues study the evolution of the jerboa, a long-legged rodent with a jumping, bipedal gait, which diverged from the mouse lineage tens of millions of years ago. Cooper and her colleagues are trying to find the unknown combination of genetic changes responsible for that evolutionary transformation by making lab mice carry incrementally more jerboa traits.

> It's also disturbing that the same people who brought us widespread hormone-like chemicals leeching into our food are arguing that this, too, is good for us.

I am not sure this argument works, because it's continuation is that "they also brought us autism inducing vaccines, cancer inducing wifi, and potato fish hybrids that are evil in some vague way".

Whether you like it or not eliminating mosquitos, colorado beetle, etc. will help billions of people to live better lives, and your ad hominem at "some post-doc trying to be alpha nerd" is out of place, because no matter their motives, the work of people like this is what allows us to write things on screen, instead of sitting in a cave.

> I am not sure this argument works, because it's continuation is that "they also brought us autism inducing vaccines, cancer inducing wifi, and potato fish hybrids that are evil in some vague way".

Just because wifi doesn't induce cancer doesn't mean it's a good idea to have plastic leeching into our water. I don't see the correlation.

Also, I should add that messing with the balance of an ecosystem can have drastic unforeseen consequences. China decided to get rid of sparrows (or some other small bird) mid-20th century, and for the next few years they were struck with famine because there were no birds left to control the locust population, which then multiplied out of control.

not only is the comment about China accurate it elicits a great point.

to add in a Hayek quote

>The curious task of economics is to demonstrate to men how little they really know about what they imagine they can design.'

there is a great podcast from Econtalk [0] talking about this in detail that I highly recommend a listen if you are not familiar with it.

back to the point, these changes have unforeseen consequences, some may be good but some could cause a new type of famine or disease that could have devastating consequences. I do not think this point can be overstated.

[0] https://www.econtalk.org/frank-dikotter-on-maos-great-famine...

There can be unforeseen consequences, that's why nobody advocates to just randomly create gene drives for all the insects. The targeted species of mosquito is not the only one, it is an invasive species in many places, and everywhere there are other mosquitos that can take its place.

Should we use gene drives to eliminate species with caution? Of course! But does that mean we should not even think about it just because there may be unforeseen consequences (or as the grandparent comment says: because it is invented by scientists who also invented plastic, which other people used as cheap water bottles, and there is an unconfirmed hypothesis about it being bad)?

I think the answer to this question is clearly no, and the way the grandparent comment is written is simply antiscience fear-mongering.

nobody banned thinking or communicating about technological possibilities, and it is healthy to communicate concerns

I just think that there needs to be significantly more caution - almost a pathologically high amount of caution.

Think of it from a software engineering perspective. All the unit tests that are done to ensure 100% code coverage, all the testing in identical staging environments, and yet we still have to revert sometimes, once the code is in production.

And this all still happens in environment where the code is something we wrote ourselves, is properly and thoroughly documented, with all dependencies and relationships clearly defined. And still, major problems happen but it's nice that we can just go to Git and try again.

With genetic engineering, the genetic code was not written by us, has zero documentation, the relationships are left for us to figure out, and there's no proper staging environment with a working ecosystem for us to test things in. We basically deploy straight to production after a few unit tests. So, yes, definitely I don't think there is currently enough caution being taken.

Of course, the reality is that genetic engineering won't stop, not until we shoot ourselves in the foot a few times and people realize the seriousness of it.

>the work of people like this is what allows us to write things on screen, instead of sitting in a cave.

This is a pet-peeve of mine in arguments about the future. Why is it unthinkable that someone could support the development of one technology, and not the other? Better yet, why is unthinkable that someone could prefer that one technology gets developed _before_ another? Yes, implicit in technological development is a dependency chain, but that does not imply every technological advancement is predicated on every other technological advancement.

Pretending that all technology/research has the same social utility is ridiculous (not that you are specifically doing so). Unfortunately in this case, it appears we are developing a great method for extincting species without also having a method prevent that.

I was really hoping that someone had made an (obviously improbable) potato fish hybrid. I imagine that would be tasty (and full of unintended side effects of crossing a starch and a protein).

To save everyone the ducking of it, I think it is a reference to Mona the Vampire, in which potato jellyfish hybrids do... things. See for yourself: https://youtu.be/do4e7AJKafE?t=634

Society is ultimately responsible for protecting itself, not individuals. You can't blame individuals for their compartmentalization, irreverence, and sociopathy when we reward people for these things. The post-doc only produces work that other people want to publish, after all.

These new biotechnologies are eerily reminiscent of what's in the novel "Oryx and Crake" by Margaret Atwood. In a subplot, an inner circle of biohackers compete to drive species to complete extinction for fun. It's worth a read

https://en.wikipedia.org/wiki/Oryx_and_Crake

Reply on news.ycombinator.com