Dominic Cummings wants to bring science and technology into the heart of government. As our world is transformed by machine learning, quantum computing, phenotypic sequencing and evolutionary biology, Cummings believes the institutions of democracy are stuck, stifled by what he calls “the forces of entropy”. Ideally there would be no need “to persuade people” to “fix broken institutions”, he writes, but alas “this option is not available in politics”.
For Cummings the answer is to turn No 10 (and now No 11) into Britain’s own “Singularity Hub”, in which a “high-performance team” makes decisions on the basis of “facts and quantitative models”. The first hire for this team, Andrew Sabisky, was revealed to be more of a gadfly than a scientist, but this should not perhaps have come as a surprise. Sabisky embodies a mindset that is reshaping conservative politics around the world.
I first encountered this mindset at Facebook a few years ago. At the end of an informal reading group, in which computer scientists discussed books from history, political theory and philosophy, someone suggested Jordan Peterson’s 12 Rules for Life. The discussion changed as elementary statistics and crude evolutionary biology were invoked to support a libertarian techno-determinism. Since then I have met hundreds more Andrew Sabiskys at conferences on artificial intelligence, bioethics, and genetic sequencing. Most are big fans of Jordan Peterson and Peter Thiel. A cadre of smart, young and middle-aged men fervently believes that politically correct liberals are suppressing the “uncomfortable conclusions” of modern science.
Like Cummings, they speak at length about genetics and intelligence and the more daring, like Sabisky, voice opinions about race. Statistics about group disparities in IQ scores are used to provoke progressives into denouncing IQ tests as constructs. For Peterson’s acolytes this is unscientific, the “postmodern fallacy”. Stories from the animal kingdom are combined with nebulous references to artificial intelligence and a legion of statistics to describe an impending world of unprecedented control. To fail to agree with these men – and they are almost all men – is to deny the science, and the impending future.
Cummings is a talented marketer, and he knows the power of this approach. To paint Sabisky as the victim of “political pundits who don’t know what they’re talking about” is to cast him as a martyr to unscientific and backward thinking. It takes time and patience to expose bogus science and fallacious logic, particularly when it permeates almost everything they say – and by the time it has been properly explained, the news cycle has moved on.
But Sabisky’s claims are worth refuting. Take his claim, couched in the smug language of provocation, that the disparity (which he exaggerates) in IQ testing scores to Americans who identify as black and Americans who identify as white can “parsimoniously explain” why black people are disproportionately likely to be classified as having an “intellectual disability”, which he revealingly describes as “mental retardation”. This logic is profoundly flawed. A difference in means, the average rate across two groups, simply reveals that at present there is a correlation between a social category defined by humans and performance on a test designed by humans. There is nothing natural about this observation. Differences in historic disadvantage, such as three centuries of enslavement and systemic prejudice, is more likely to explain the disparity than some undefined underlying essence.
This faulty logic is employed to prop up the underlying assumption, which Cummings himself made explicit in a paper to Michael Gove in 2014, that intelligence is heritable. It is – almost all differences in behavioural traits are somewhat heritable – but Cummings misconstrues the idea of heritability. Heritability is not a special property of specific traits that are determined by genes, it is a description of the basic biological realities with which we are born. Our genes interact with our environments, chance, and our own choices to shape our lives.
Recent analyses of DNA structures have shown that complex human traits, such as intelligence or criminality, are almost never associated with a single, determinative gene. There is no gene for IQ. Hundreds of genetic variants are associated with tiny variations in intelligence. Together they predict a fraction of someone’s IQ score. Anyone who claims we will soon understand IQ at the level of individual genes is either abusing the science or does not understand it.
What’s more, heritable traits can change. Height, for example, is about 80 to 90 per cent heritable, but the average height of British men and women has increased by about 11cm in the last century. IQ is still less heritable and more modifiable. The adoption of a child by a family wealthier than the one into which they were born has been associated with an IQ difference of between 12 and 18 points.
To be clear, then, there is no evidence that a significant proportion of the observed IQ differences between socially defined groups, including races, are caused by persistent genetic differences.
More broadly, correlations that show some traits are partly inherited do nothing to reduce the effectiveness of environmental interventions. This is not the science Cummings wants to criticise programmes such as Sure Start, however; the demonstrable effect of early intervention programs on children’s’ prospects is a more complicated, less provocative story than a simplistic argument based on genetics.
[See also: The pseudoscience of hate]
Jordan Peterson shows what happens when this thinking becomes part of larger cultural feuds. His fans attend conferences wearing T-shirts printed with lobsters, because his book begins by comparing humans to lobsters, which monitor their status in terms of hierarchies of dominance. “Dominance hierarchies”, Peterson writes, “are older than trees”, as if social inequalities are natural, inevitable, and beyond our power to control.
This is a fundamental misreading of evolutionary biology. Yes, we could breed strong men who run faster or jump higher. But the point of evolutionary biology is that we cannot know in advance whether running faster and jumping higher will prove, in the long run, to be useful. This is why eugenics is junk science – because it presupposes that it is possible to know what is right for a species. People like Jordan Peterson, Dominic Cummings, and Andrew Sabisky have no unique foresight into the process of natural selection. Their pronouncements about which traits are best for the human race are prejudice, not science.
Ugly politics has long worn bad science as a mask, and when democracy swallows up Cummings, as it has Sabisky, it will continue to do so. But if we take the time to understand the science, we give the politics nowhere to hide.
Josh Simons is a EJ Safra Fellow in Ethics and an Affiliate at the Berkman Klein Centre at Harvard University, a former Research Scientist at Facebook and Policy Advisor for the Labour Party.