Motus Mentis

From Stephen Hsu’s blog, here’s a video of an hour-long panel discussion with Dr. Hsu, Steven Pinker, and Dalton Conley on the subject of genetic engineering and the heritability of human traits, particularly intelligence.

This topic is a minefield in the West, and so great care is taken, and necessary pieties uttered — and some obvious, elephant-in-the-room topics are completely ignored. There is, however, a clear consensus that (a) almost every human trait is significantly heritable; (b) that intelligence is real, quantifiable, and predictive of life outcomes; (c) that intelligence, like every other human trait, has a significant genetic basis; and (d) that both our understanding of polygenic traits and our ability to edit the genome are advancing exponentially.

Dr. Pinker repeatedly expressed doubt about the likelihood of rapid adoption of eugenic applications, however. He points out that human cloning has been possible for a long time now, but is still illegal everywhere, and that while breeding for intelligence has also been possible forever, no society seems terribly interested in doing it. (One might object that Dr. Pinker himself could arguably be seen as the output of such a process.) He generally seems confident in the power of tradition and taboo to keep a lid on this sort of thing. I think he is very wrong about this.

Dr. Hsu, on the other hand — who is an adviser to the BGI Cognitive Genomics research center in Shenzen, China, that is working hammer and tongs to advance this technology — seems not to share Pinker’s opinion here, and I think that, given the venue (the 92nd Street Y, in New York City), he was holding his cards close to his chest. (I know people who know Dr. Hsu, and my understanding is that he thinks it very likely that we will be able to engineer human IQs hundreds of points higher in fairly short order.)

The social and moral implications of this technology are enormous, probably much more so than most people have really imagined, and progress in this area is accelerating. One deceptive feature about exponentially rising curves is that if you use the slope at any given point to extrapolate future values of Y, you are always vastly underestimating, so our sense of the rate of change here — and our sense of having time to deal with it — is quite certainly wrong. There are in fact three different curves here: there is the rate of technological change, which is soaring into the sky; there is the much slower rate at which we can comprehend and predict, in moral and intellectual terms, what the implications of the technology will be; and slowest of all there is the rate at which the society, and in particular the law, can accommodate those changes. These three curves are peeling apart very quickly now.

Of the three panelists, by far the least interesting, I thought, was the social scientist Dalton Conley, whose most notable contribution was to take a swipe at the work of Charles Murray and Richard Herrnstein. Indeed, he made claims about the trending effects of assortative mating that struck me as flat-out wrong, but I will do some homework before I comment further about that.

Anyway, have a look. Your comments are welcome as always.