We are now at the point where we can realistically expect to see interventions that significantly increase human intelligence.
Some approaches build on our real but limited understanding of how the brain works. For example, by adding an extra copy of the NR2B gene, the Doogie mouse developed better memory. Today there are about 30 genetically modified mouse strains with improved memory, most making use of various tweaks of the same enzyme path influenced by the NR2B gene.
But I am not going to talk about those here. My favored methods lean very heavily on natural selection. The beauty of these approaches is that you don’t have to understand the mechanism. Dog breeders have successfully developed dogs with special behaviors without having any understanding of the neurochemistry of those behaviors. Medieval farmers
developed Guernsey cows without knowing a thing about how milk is produced. In a sense, of course, even the understand-everything approach leans heavily on natural selection, since it tries to comprehend an already-existing complex phenomenon, rather than invent it from scratch.
I should probably address one concern before I go further. Some people might worry that since natural selection optimizes traits, increasing human intelligence would naturally upset some balance, mess up some precise tradeoff, and so such attempts are foredoomed. Forgeddaboutit. The tradeoffs are optimized, all right, but for past environments, not the present. We have a lot more elbow room nowadays; you could say that the trade space is roomier. Certain costs of intelligence that were once crucial are no longer. If a more active brain used up 10% more calories, that’s a sacrifice that most Americans would be willing to make. If higher intelligence requires a larger brain, well, we have C-sections.
I see five feasible approaches: selection, spell-checking the genome, QTLs for intelligence, cloning, and hybrid vigor.
The first approach, old-fashioned selection, is powerful, conceptually simple, but slow. It’s already happened. The gradual increase of human intelligence over the past couple of million years, the existing differences between the races, and the increased intelligence of genetically isolated merchant castes such as the Parsees and Ashkenazi Jews are all products of natural selection. A government with consistent and lasting policies could select for intelligence and achieve striking results in a few centuries, maybe less. But no state ever has, and no existing government seems interested.
Natural selection for intelligence may furnish us with some examples of alleles that noticeably boost intelligence (such as the Ashkenazi sphingolipid mutation). If so, those alleles could used in other, faster approaches.
The second approach, spell-checking the genome, might be also called taking out the trash. The idea is simply to fix most or all of the broken genes. There are a lot of them. The effects of genetic errors do not merely take the form of serious diseases like PKU, which causes severe retardation in untreated individuals with two copies of the defective gene. Carriers of PKU, those with only one copy of the defective gene, also suffer ill effects, mainly a slight depression of IQ. Everyone carries single copies of damaged genes that would be lethal in double dose – in fact, geneticists estimate that the average individual carries as many as five lethal genes. As a carrier, you may produce half the normal level of some enzyme, which usually does not cause severe trouble in itself but does decrease efficiency. The average individual, indeed every individual, is thought to carry many more than five less serious genetic errors, perhaps as many as 100, as these less serious errors are only slowly eliminated by natural selection. In other words, every person has many hidden genetic flaws that reduce performance and efficiency in many ways. The total
impact of these errors is thought to be quite large, and must surely be very significant for human intelligence, which is probably the most complex of all adaptations.
If we could edit the genome, we could fix them all. Probably no one would object to fixing any particular genetic error, and that must mean that there could be no objections to fixing all of them. After all, he who has said A must also say Z! The difference between humans whose DNA has been spell-checked and normals is likely to be large. They would almost certainly be smarter and healthier, and this kind of change is, I think, guaranteed to work and be safe. We know that the unbroken form of any given gene is safe, since almost everyone else already has it. You could compare these spell-checked humans to a Lexus – every little thing is done just right. There is no innovation at all in this kind of genetic intervention, almost no risk, and the resulting advantages are likely to be very significant.
The third approach is to find and take advantage of the genes that are responsible for the fraction of IQ variation that is not due to mutational pressure, not just caused by typos.
These take the form of multiple alleles of a particular gene, one associated with slightly higher intelligence and the other(s) with slightly lower intelligence. Each variant is fairly common and not thought to be a genetic error. If we could identify these polymorphic genes affecting IQ, and were able to edit the human genome, we could choose the variants associated with increased IQ. If we made such changes in most of the genes with significant effects on IQ, we could produce individuals with high IQs, quite possibly higher than any ever observed. Since no single gene seems to account for much of the variation – at least in the populations we have studied so far – this would require fairly heavy editing of the genome, perhaps comparable with that required for spell-checking.
Rare, regional IQ-boosting variants may also be valuable, just as the genes of wild relatives of domesticated crops can be useful, or the genes of regional landraces.
We know that it is possible for healthy individuals to have a higher-than-average numbers of alleles favoring intelligence, but we do not know the consequences of having the high-IQ version of every possible gene. There has never been any such person. So, guided by experience, we might feel safe in imitating the QTL constellation of a healthy, successful, and stable individual with high IQ, but as for going all the way.. It’s a bit scary. The kid might go into convulsions, or explode in a flash of green light. Better to be cautious.
The knowledge that a particular high-IQ QTL setting works could be very useful, and the only real test is whether it has worked before. Thus, knowledge of the genotypes of highly intelligent and creative individuals would be valuable. Robbing the graves of the great might become very lucrative. Before we’re done with this, we may have armed guards protecting Feynman’s tomb. And of course someone is going to steal (and sequence) Einstein’s brain (1).
This leads us to the fourth approach, old-fashioned cloning. The clone of a very intelligent person will also be very intelligent. Many people say that this is not necessarily true, and, strictly speaking, they’re right – we could accidentally drop the kid on his head – but the IQs of identical twins are, on average, very similar. There are not a huge number of high-achieving twins, let alone twin pairs, perhaps in part because twinhood itself is slightly disadvantageous, but consider the odds of a pair of identical twins both being sought-after string theorists or astrophysicists (not mere rocket scientists). It happens. Another strength of cloning is that it duplicates complex non-additive interactions between genes, which we don’t know enough to design on purpose. Cloning by itself will not produce individuals that are smarter than anyone has ever been before; still, a society in which John Von Neumann was the norm would be … interesting. This approach is relatively low-tech, since we are able to clone some mammalian species already, nor does it require any genetic editing.
Fifth, we might able to harness hybrid vigor. Sometimes hybrids of two populations are considerably more vigorous and productive than members of either parent population. The best-known example is hybrid corn, which is genetics’s biggest practical success. This can mean being smarter, as well. For example, mules are stronger, hardier, have more endurance and are more intelligent than either of their parental species.
Hybrid vigor is unpredictable, in that some pairs of parental populations produce improved offspring and some don’t. Two populations whose exhibit hybrid vigor are said to ‘nick’. It may be that some human populations ‘nick’. If so, we would have a practical zero-tech way of producing enhanced humans. Of course, with our usual luck, one or both of the two nicking parental populations would be rare, obnoxious and only available by paying inordinate sums to some third-world dictator. Worse yet, within a generation or two, prominent enhanced members of our society would be spouting eloquent bullshit about why we have to intervene in some pointless geopolitical dispute involving either or both of their gormless parental populations.
Mixtures of these approaches are possible. Take someone highly intelligent, spell-check his genome, and then clone him. It should work.
Having no design innovations, humans enhanced via any of my five preferred approaches would be fully back-compatible. Unions with ordinary humans would be fully
fertile. The children of spell-checked people and normals would have about half the usual amount of genetic load and would still be mighty impressive.
A country or tribe that did this first, on a large scale, would be competitively superior to everyone else. This might be the case even if they had been constitutionally competitively inferior before. In fact I see a bunch of losers as the most likely to take such a radical step – not to invent it, but to take it. After someone did it, the rest of the world would have to take some kind of drastic catch-up measure or sleep with the Neanderthals. Frankly, I don’t see why they wouldn’t just go ahead and do some spell-checking themselves. But they could also, say, make a pre-emptive thermonuclear strike, unleash a tailored virus, fire up the time machine and nip the whole thing in the pre-bud, go cyborg (Mechs versus Shapers), or escape temporarily through space colonization.
1. Westminster Abbey would be a prime target. Shakespeare would be more
valuable than Einstein – he could write scripts. Of course, there’s a curse.
‘ Good friend, for Iesus’ sake forbear
To dig the dust beneath these stones.
Blest be he who leaves me there,
And curst be he who moves my bones.’