Racist IQ Theories are Inextricable from the FTX Story – and the AI Bubble.

The ongoing surge of horrendous racism is coming directly from the tech industry.

Hello and welcome to a belated weekend installment of Dark Markets. The news has yet again reiterated the broader significance of deconstructing Sam Bankman-Fried’s ideology and behavior – this time, in the form of the junk IQ science which meaningfully enabled his fraud. The EA, longtermist, and transhumanist movements are to varying degrees either based on or continuations of ideological efforts to legitimize power-based hierarchies as “natural realities” – one of the main reasons I’m motivated to examine their fundamentally deceptive nature. Their commitment to that unreality has enabled multiple frauds and long cons – first FTX, and now Artificial Intelligence.

Charles Murray, author of “The Bell Curve” and father of modern racialist IQ thinking.

Last week was emotionally taxing.

The unholy trinity of JD Vance, Peter Thiel, and Donald Trump gave us a wave of blood libel against legal Haitian immigrants in Ohio, a stark reminder of the very darkest evils in the American soul. The victims are, again, legal immigrants, recruited by a dying city to revitalize its population. They are now in fear of racial terrorism because of the current Republican candidate for Vice President – who now proudly admits the stories were false, but useful.

This is a dark moment for America. But I was most directly upset by the wave of truly disgusting claims about IQ surfacing on Twitter. This is particularly surreal as someone who actually lived through the first wave of the IQ debate in the 1990s, following the publication of Charles Murray’s The Bell Curve. For much of the 2000s and 2010s, that debate seemed firmly settled against the racists. So it was truly, deeply unsettling to see people like LBRY founder Jeremy Kauffman trying to excuse anti-Haitian hate by making scientifically spurious claims about population-level IQ.

Probably the single most clear rebuttal of a “racial” grounding for IQ comes from adoption studies, where adoption into a wealthy family raises IQ scores a staggering 12-18 points. Similarly, population level IQs rise with economic development – the average U.S. IQ rose 18 points between 1948 and 2002 – a time period, notably, in which the country became substantially less “white”.

I’ve never hesitated to call Trumpism a fascist movement, and that’s never been more clear than it is now. Max Read has a compelling theory that this has intensified on Twitter lately because Elon Musk is desperately trying to shore up his strategically questionable decision to endorse Donald Trump. The good news is the whole thing seems to be absolutely blowing up in their faces – America has ample appetite for straight-up race war shit, but perhaps not quite enough of a constituency to elect a President.

But more to the point for this publication: this moment of racist IQ rhetoric exposes important truths about Sam Bankman-Fried and the ideological conspiracy behind FTX.

“IQ” is Substantially a Product of Environment

First, let’s just dispense with the notion of heritable population-level IQ that underpins Charles Murray’s white supremacist fake science, and which is now being echoed by credulous, hate-fueled rubes like Jeremy Kauffman.

Intelligence obviously varies between individuals, and IQ testing captures at least some of that difference – though as we’ll get to, mistaking IQ test results for an absolute measure of “intelligence” is also mistaken. But the problem comes when Murray and other IQcels try to claim that intelligence as such varies between distinct population groups; and that it is overwhelmingly genetically heritable.

Both of these claims are straightforwardly false, and this short summary by mainstream intelligence researchers in Vox includes the key rebuttals. Overall, they write, “Murray takes the heritability of intelligence as evidence that it is an essential inborn quality, passed in the genes from parents to children with little modification by environmental factors. This interpretation is much too strong — a gross oversimplification.”

Probably the single most clear rebuttal of a “racial” grounding for IQ comes from adoption studies, where adoption into a wealthy family raises IQ scores a staggering 12-18 points. Similarly, population IQs rise with economic development – the average U.S. IQ has rose 18 points between 1948 and 2002 – a time period, notably, in which the country became substantially less “white”. Much like height, intelligence depends at least as much on getting enough to eat and being taught how to think as on the raw genetic material you’re working with.

This matters because the ultimate goal of IQ discourse is to naturalize hierarchies of wealth and power between racial groups. That is, someone like Charles Murray wants to prove that Black people are poorer than whites, both within a nation like the United States and globally, not because of historical injustice, but because they have lower genetic capability – and that no social program or wealth redistribution will change that.

This has the effect of essentially excusing slavery, imperialism, and their more contemporary counterparts, exclusion and disinvestment. This is particularly notable in the case of Haiti, which has suffered centuries of economic discrimination at a global scale because its people successfully rebelled against slavery in 1804, an achievement that if anything marks Haitians out for their historically unique strength as a people.

Population IQ claims also help make the (false) case against social programs and wealth redistribution, which even by the meager logic of neoliberalism are intended to give the most gifted people in a population a chance to thrive regardless of the circumstances of their birth. Charles Murray was explicitly a policy advocate, and his policy stance was essentially that welfare was wasted on Black people, because they as a whole were fundamentally, genetically incapable of thriving and contributing to society.

The racial domination agenda of population IQ research is made much clearer if we look at Murray’s work after The Bell Curve, particularly 2003’s Human Accomplishment. That book purports to rank entire cultures on the basis of their achievement, but also includes such fun claims that no music made since the 1950s has lasting cultural relevance (hmm, what happened to music in the 1950s?); and that no woman has ever been relevant to Western philosophy – a statement that would conveniently devalue Hannah Arendt, the most important philosophical critic of fascism.

If all this abuse of science to the cause of racial dominance doesn’t make your blood boil, you should probably get your own IQ tested.

Sam Bankman-Fried is a Walking Rebuttal of IQ Theories

The story of Sam Bankman-Fried is a real-world illustration of both why IQ heritability is essentially false, and why IQ rhetoric is dangerous. Both the explosion of FTX and the linked, currently-ongoing collapse of the AI bubble were fueled by misunderstandings of reality premised on the IQ model of intelligence – and, ultimately, of human moral worth. The disasters driven by IQ, as one component of the broader TESCREAL cluster of scientistic, anti-humanist heuristics, show why it’s important for the future of humanity to fight this particular set of theories.

The question of IQ has deep relevance to the FTX fraud. First, it’s materially clear that Effective Altruism as a movement places high value on actual IQ scores. As reported by Vox, the leadership of the Center for Effective Altruism at one point explored creating a metric for individual workers in the EA ecosystem called “Potential Expected Long Term Instrumental Value,” or PELTIV. That metric included tested IQ, and was to be used to decide who got jobs in the (absurdly lavishly funded) EA ecosystem.

A second major connection to IQ was the direction of FTX customer funds to a massive ($500m) investment in Anthropic, an artificial intelligence startup that grew out of the EA movement, and specifically its bizarre preoccupation with artificial intelligence as a source of “existential risk.” As I’ll expand on a bit below, the way current AI leaders think about “intelligence” is inextricably tied up with the flawed model of “IQ,” and with its most racist instrumentalizations.

(There’s also a more direct, and frankly more infuriating connection I uncovered just last week: Sam Bankman-Fried was brought into EA by Oxford’s Will MacAskill, and MacAskill’s ex-wife Amanda Askell still works at Anthropic, purportedly on “AI Alignment”. This is just one of the many ways that EA affiliates appear to have personally and directly benefitted from funds stolen from FTX customers.)

But the most important connection between FTX and IQ, and easily the one that’s most devastating to the TESCREAList’s understanding of the universe and existence, is that Sam Bankman-Fried is incredibly stupid. That is, he probably scored very high on IQ tests, and much of his own endless faith in himself was clearly rooted in his firm belief that he was a singular genius. But when it came to actually doing things in the world, his choices were simply and objectively ill-considered, again and again undone by ignorance, shortsightedness, and misunderstanding.

There are endless examples of SBF’s remarkable stupidity when it came to pretty much anything outside of math. He didn’t understand cryptocurrency deeply, which helped bring down his empire; there’s not even particularly strong evidence that he was a good trader, since Alameda had only a few good years, and relied on stolen FTX customer money as soon as that was available. The clearest examples of Sam’s stupidity are probably his choice to go an an open-ended self-incrimination media tour after the collapse of FTX, which he told Tiffany Fong he did against legal advice because “lawyers don’t know anything”; and his equally stupid and self-destructive decision to take the stand during his own trial, during which testimony he perjured himself multiple times and literally added years to his prison sentence.

There are endless smaller examples, of my favorites include him not knowing who Anna Wintour was, and his convincingly sincere inability to comprehend that FTX and Alameda Research were separate corporate entities. In fact, the essence of Sam Bankman-Fried’s character may be largely summed up as “guy who is so convinced of his own brilliance that he never thinks twice about his own presumptions.” This is ultimately rooted in an understanding of intelligence tied to the framing of IQ as a single numerical “score” that sums up an individuals intellectual value.

The significance of IQ thinking in the story of FTX can also be expanded to the broader glamour of the “genius” myth that ensnared Bankman-Fried’s many fans, stans, allies, and defenders. This powerful myth was perhaps most visible in the weeks after the FTX collapse, when people like Bill Ackman and Mr. Wonderful professed their continued faith in Sam.

Both for Sam and his defenders, the idea of “genius” became disconnected from actual achievement in the world, in a way that only a framework like “IQ” can enable. Defining intelligence as some objectively measurable single statistic, like an attribute in an RPG, allows for all kinds of mathematized delusions. I’ve recently been reading Franco Birardi’s Breathing, which examines Jonathan Franzen’s novel Freedom, and includes this remarkably resonant passage from Franzen:

“Later, as his troubles began to mount, it would seem to him as if his very good luck, which his childhood had taught him to consider his birthright, had been trumped by a stroke of higher-order bad luck so wrong as to not even be real. He kept waiting for its wrongness, its fraudulence, to be exposed”.

IQ and “genius” were, for SBF and his supporters, an unassailable truth beyond any objective evidence. No matter how stupidly he behaved, there was some kernel of inherent superiority to SBF that could not be touched by reality, and which therefore merited the rejection of his guilt, his failure, and his responsibility.

This is the same transcendentalist logic by which IQ underpins white supremacism.

How IQ Undid the Artificial Intelligence Project Before it Started

Finally, and briefly, the entire current wave of “artificial intelligence” development is premised on the misunderstanding of intelligence baked into the reductive concept of “IQ.” The AI bubble is in the process of deflating right now because the essentially racist foundation of IQ thinking led towards a dead-end in AI development. This is precisely the same sort of fetishization of intelligence, followed by a brute collision with the far more complex reality, that played out with FTX – and yet again, it will see IQ rhetoric help destroy billions of dollars in value through malinvestment.

Specifically, the “IQ” framing of intelligence treats human reasoning as a single statistic, like something in an RPG, with each of us having a different, fixed and static “level” of intelligence. This connects directly to the idea that a model like GPT-4 is “better” than its predecessors because it uses more raw computing power to process larger data sets. But of course, as the ignominious release of OpenAI’s “Strawberry” model last week showed quite definitively, raw compute and scale don’t push artificial intelligence models forward after a certain point – and by the same token, measuring human intelligence with a single metric is inevitably, inherently narrow.

There’s much more to be said about AI and IQ in particular, but we’ll leave it at that tease for now. Stay vigilant out there, folks – we are surrounded by malevolent freaks on all sides. Their goal is to destroy democracy and justice, and they have more money and power than at any time since the 1980s.