The Singularity Q&A

Q: How could genetic engineering be used to improve intelligence?

A: I'm going to declare up front that I know many of the procedures I am about to discuss are highly questionable from an ethical standpoint, and I am in no way suggesting that human genetic engineering for the sake of greater intelligence should be pursued without very good reason.  What circumstances might qualify?  Well, if all other routes to Singularity proved to be untenable -- which I don't see happening -- a careful genetic engineering program might be the only way to secure our long-term future in the face of ultra-advanced technology.  Our existing level of intelligence just isn't safe to trust with this kind of power any longer than we have to.

That said, I should start by distinguishing between genetic engineering, gene therapy, genetic breeding, and neurohacking.

  • Genetic engineering usually refers to customizing the DNA of an organism with the intent to express some specific trait or series of traits; for very pragmatic reasons, this is usually done while the organism is only a single fertilized cell.

  • Gene therapy refers to an attempt to modify the DNA of some or all of an at least partially mature organism's cells; this is usually done with specially modified viruses, and has so far had some limited success in treating certain human congenital deficiencies.

  • Genetic breeding is the use of DNA testing to select or weed-out organisms in a breeding program; this allows a researcher or livestock breeder to make informed decisions on which seeds or embryos to keep, without having to wait for each organism to reach an age when the desired genetic trait(s) would be expressed.

  • Neurohacking refers to non-genetic modification of a functioning brain with intent to improve some aspect of its performance; an upcoming question will discuss this in more detail.

Because human DNA is believed to directly effect overall intelligence only during early brain development, gene therapy is unlikely to be of much use in this area.  Getting in on the action early enough would require genetic engineering or some kind of breeding program.  Were we talking about some other organism, any clear-thinking scientist today would actually say that a breeding program would be the way to go in this case, since the desired trait (high intelligence) occurs naturally in -- and only in -- the subject organism, and is highly heritable (can be easily passed on by the parents).  But human breeding is a sensitive subject tainted by the horrors of genocide, leaving genetic engineering as the only real alternative for biologically improving intelligence.

The kinds of genetic engineering that could be done with today's technology are rather limited, and scientists should expect the particular problem of higher intelligence to be especially tricky.  Because higher intelligence would have contributed significantly to the reproductive fitness of hunter-gatherer humans, it stands to reason that any simple modifications to our DNA capable of increasing our intelligence would have been discovered by evolution already, and become universal characteristics of our species.  Hence, any simple modifications that do improve human intelligence are likely to come with serious side effects that would've have negated the advantages in a primitive setting, but which may be treatable today.  (This principle has been referred to as Algernon's Law, after a fictional mouse engineered for higher intelligence.)

Currently, the specific chain of processes by which DNA gives rise to complicated traits like intelligence is not very well understood.   Hence, to find any particular genes conducive to high intelligence, researchers would first have to locate a collection of especially bright humans, then see if any particular genes seemed to be unusually common among them.  If a candidate gene was identified, the only reliable ways to test it would be to (1) examine the genes and intellect of the general population, (2) modify some human embryos to carry the gene, then wait for them to mature into children and adults and see how smart they are, or (3) find some men and women who share this special gene, and see if they would be interested in having children together.  (2) or (3) would probably not produce any unprecedentedly smart children, but the results could still be impressive if this particular gene had an increased impact on the child when carried by both parents (some genes do), and was sufficiently rare that this double combo was unlikely to ever occur naturally.

The more exciting possibilities of this kind of low-tech program could only be realized if a number of different, rare genes were found that could each independently improve intelligence.  If these were capable of working together in a single person, then a child engineered to carry them all could be the first to a new plateau of intelligence.

Higher technology and a complete understanding of molecular biology could ultimately enable the full potential of genetic engineering:  Total modification or creation of complex life forms.  With this kind of power, the design of the human brain could be completely revisited with the kinds of deliberate complex improvements that evolution is incapable of stumbling on.  And there would be no reason to stop with the brain; why not improve the body in other ways as well?  This changes to mind and body would mark the evolutionary beginnings of so-called "transhumans", new forms of intelligent life that no longer fit into the homo sapiens sapiens category.

Like neural interfacing, genetic engineering is a science that, to reach its most advanced applications for improving intelligence, would first have to decipher the specific structure and patterns of the human brain.  Once again, this gives Artificial Intelligence an advantage on the path to Singularity, since AI researchers with such precise knowledge of how one type of intelligence works would be well equipped to create another.   Artificial systems also make for better experimental subjects; computers, unlike babies, don't need years to mature physically, and the ethical questions of AI research -- which should not be understated -- may not be as hard to answer as those of genetic engineering.

[Back to Futurism] [Back to Main]  ©2002 by Mitchell Howe