At the excellent Integrating Genetics and the Social Sciences conference on Thursday and Friday, a late discussion raised an awkward issue: when you say you study the genetics of behaviour, eyebrows get raised.
> The dark idea is that natural selection might be like a bicycle: if you don’t keep moving, you fall off. That is, if natural selection doesn’t operate in a population, then deleterious mutations accumulate by chance.
I think this is equivalent to the idea of genetic load ( https://en.wikipedia.org/wiki/Genetic_load ): in a sexually reproducing population, harmful but non-fatal mutations are added by errors in gene copying during reproduction, & removed by natural selection. Usually, these processes will reach equilibrium at some average number of harmful mutations, but if one of these processes changes, the equilibrium will itself be changed, e.g. if natural selection removes harmful mutations less quickly due to improved medicine then the average number of harmful mutations at equilibrium will increase.
> A dilemma of human progress, then, is that we might have to choose between remaking ourselves so completely that we are no longer recognizably human; …
This would require far more drastic changes, & better technology, than just editing out or screening for harmful mutations. Just removing mutations that harm physical health or intelligence &c. seems more likely to produce a population of unusually healthy, smart, genetically un-diverse people, though I would expect that it may cause some problems if what was thought to be a pointlessly harmful mutation turns out to have had undiscovered benefits. I think you are correct, though, in the longer term (assuming science continues long enough to reach that point).
One of the issues is that right now people want to try screening on polygenic scores. But by definition those scores sum thousands or millions of genetic effects, so it seems very likely that they are pleiotropic (have multiple effects on different outcomes). So to me, this seems like a bad idea.
However, the long run problem isn’t that this kind of artificial selection won’t work — it’s that it will work too well! As you suggest, we might end up with a population of “Midwich Cuckoo“-like people: unnaturally perfect, and all very similar.
> The dark idea is that natural selection might be like a bicycle: if you don’t keep moving, you fall off. That is, if natural selection doesn’t operate in a population, then deleterious mutations accumulate by chance.
I think this is equivalent to the idea of genetic load ( https://en.wikipedia.org/wiki/Genetic_load ): in a sexually reproducing population, harmful but non-fatal mutations are added by errors in gene copying during reproduction, & removed by natural selection. Usually, these processes will reach equilibrium at some average number of harmful mutations, but if one of these processes changes, the equilibrium will itself be changed, e.g. if natural selection removes harmful mutations less quickly due to improved medicine then the average number of harmful mutations at equilibrium will increase.
> A dilemma of human progress, then, is that we might have to choose between remaking ourselves so completely that we are no longer recognizably human; …
This would require far more drastic changes, & better technology, than just editing out or screening for harmful mutations. Just removing mutations that harm physical health or intelligence &c. seems more likely to produce a population of unusually healthy, smart, genetically un-diverse people, though I would expect that it may cause some problems if what was thought to be a pointlessly harmful mutation turns out to have had undiscovered benefits. I think you are correct, though, in the longer term (assuming science continues long enough to reach that point).
One of the issues is that right now people want to try screening on polygenic scores. But by definition those scores sum thousands or millions of genetic effects, so it seems very likely that they are pleiotropic (have multiple effects on different outcomes). So to me, this seems like a bad idea.
However, the long run problem isn’t that this kind of artificial selection won’t work — it’s that it will work too well! As you suggest, we might end up with a population of “Midwich Cuckoo“-like people: unnaturally perfect, and all very similar.