I feel like I am one of the most engaged EAs in my local community, but the beliefs Torres ascribes to EA are so far removed from my own
This might have to do with “how local” your local community is. It seems to me that the weirder sides of EA (which I usually consider bad, but others here might not) are common in the EA hubs (Bay Area, Oxbridge, London, and the cluster of large groups in Europe) but not as common in other places (like here in Israel).
But even then, a nuanced engagement with that would require making distinctions, not just going “all EA evil”. Both Torres and Gebru these days are very invested in pushing this label of “TESCREAL” to bundle together completely different groups, from EAs who spend 10% of their income in malaria nets to rationalists who worry about AI x-risk to e/accs who openly claim that ASI is the next step in evolution. I think here there are two problems:
abstract moral philosophy can’t be for the faint of heart, you’re engaging with the fundamental meaning of good or evil, you must be able to realize when a set of assumptions leads to a seemingly outrageous conclusion and then decide what to make of that. But if a moral philosopher writes “if we assume A and B, that leads us to concluding that it would be moral to eat babies”, the reaction can’t be “PHILOSOPHER ENDORSES EATING BABIES!!!!”, because that’s both a misunderstanding of their work and if universalized will have a chilling effect leading to worse moral philosophy overall. Sometimes entertaining weird scenarios is important, if only to realise the contradictions in our assumptions;
for good or bad, left-wing thought and discourse in the last ten-fifteen years just hasn’t been very rational. And I don’t mean to say that the left can’t be rational. Karl Marx’s whole work was based on economics and an attempt to create a sort of scientific theory of history, love it or hate it the man obviously had a drive more akin to those of current rationalists than of current leftists. What is happening right now is more of a fad, a current of thought in which basically rationality and objectivity have been sort of depreciated as memes in the left wing sphere, and therapy-speak that centres the self and the inner emotions and identity has become the standard language of the left. And that pushes away a certain kind of mind very much, it feels worse than wrong, it feels like bullshit. As a result, the rationalist community kind of leans right wing on average for evaporative cooling reasons. Anyone who cares to be seen well by online left wing communities won’t associate. Anyone who’s high decoupling will be more attracted, and among those, put together with the average rationalist’s love for reinventing the wheel from first principles, some will reach some highly controversial beliefs in current society that they will nevertheless hold up as true (read: racism). It doesn’t help when terms like “eugenics” are commonly used to lump together things as disparate as Nazis literally sterilizing and killing people and hypothetical genetical modifications used to help willing parents have children that are on average healthier or live longer lives, obviously very different moral issues.
Honestly I do think the rationalist space needs to confront this a bit. People like Roko or Hanania hold pretty extreme right wing beliefs, to the point where you can’t even really call them rational because they are often dominated by confirmation bias and the usual downfalls of political polarization. Longtermism itself is a pretty questionable proposition in my book, though that argument still lays in the space of philosophy for the most part. I would be all for the rise of a “rational left”, both for the good of the rationalist community and for the good of the left, which is currently really mired into an unproductive circle jerk of emotionalism and virtue signalling. But this “TESCREAL” label if anything risks having the opposite effect, and polarizing people away from these philosophically incompetent and intellectually dishonest representatives of what’s supposed to be the current left wing intelligentsia.
I think the trend you describe is mostly an issue with “progressives”, i.e. “leftists” rather than an issue for all those left of center. And the rationalists don’t actually lean right in my experience. They average more like anti-woke and centrist. The distribution in the 2024 ACX survey below has perhaps a bit more centre-left and a bit less centre and centre-right than the rationalists at large but not by much, in my estimation.
Fair! I think it’s hard to fully slot rationalists politically because, well, the mix of high decoupling and generally esoteric interests make for some unique combinations that don’t fit neatly in the standard spectrum. I’d definitely qualify myself as centre-left, with some more leftist-y views on some aspects of economics, but definitely bothered by the current progressive vibe that I hesitate to define “woke” since that term is abused to hell but am also not sure how to call since they obstinately refuse to give themselves a political label or even recognise that they constitute a noteworthy distinct political phenomenon at all.
How was this survey done, by the way? Self ID or some kind of scored test?
So the thing with self-identification is that I think it might suffer from a certain skew. I think there’s fundamentally a bit of a stigma on identifying as right wing, and especially extreme right wing. Lots of middle class, educated people who perceive themselves as rational, empathetic and science-minded are more likely to want to perceive themselves as left wing, because that’s what left wing identity used to prominently be until a bit over 15 years ago (which is when most of us probably had their formative youth political experiences). So someone might resist the label even if in practice they are on the right half of the Overton window. Must be noted though that in some cases this might just be the result of the Overton window moving around them—and I definitely have the feeling that we now have a more polarized distribution anyway.
Do we actually have hard statistical evidence that rationalists as a group “lean right”? I am highly unsympathetic to right rationalism, as you can see here: https://forum.effectivealtruism.org/posts/kgBBzwdtGd4PHmRfs/an-instance-of-white-supremacist-and-nazi-ideology-creeping?commentId=tNHd9C8ZbazepnDqs And it certainly feels true emotionally to me that “rationalism is right-wing”. (Which is one reason I consider myself an EA but not a rationalist, although that is mostly just because I entered EA through academic philosophy not rationalism and other than reading a lot of SSC/ACX over the years, have only ever interacted with rationalists in the context of doing EA stuff.) Certain high profile individual rationalist seem to hold a lot of taboo/far-right beliefs (i.e. Scott Alexander on race and IQ here: https://www.astralcodexten.com/p/book-review-the-origins-of-woke). Roko and Hannania are of course even more right-wing (and frankly pretty gross in my view), though hopefully they are outliers.
BUT
Over the years, I have observed a general pattern with, what we can call “rationalist-like” groups: i.e. lots of men, mostly straight and white, lots of autism broadly construed, an interest in telling the harsh truth, reverence for STEM and skepticism of the humanities, more self-declared right libertarians than the population average etc.:
1) The group gains a reputation for being right-wing, sexist, bigoted etc.
2) People in the group get very offended about this; I get a bit offended too: most people I have met within the group seem moderate, with a lean towards the centre-left rather than the centre-right. I feel as a person with mild autism that autistic truth-telling and bluntness is getting stigmatised by annoying, overemotive people who can’t defend their views in a fair argument.
3) Gradually thought leaders of the group have a lot of scandals involving some combination of: misogyny, sexual harassment, Islamophobia, racism, eugenics, Western chauvinism etc.
4) I start thinking “ok, maybe [group] actually is right-wing, and I am either a sucker to be involved with it, or self-deceived about my own (self-declared liberal centrist) political preferences (after all, I do get irritated with the left a lot and believe some un-PC things, I think I am pro some transhumanist genetic engineering stuff in principle maybe, am not particularly left-or right-economically, maybe “liberalism” is what angry Marxists on twitter say it is etc. etc.)”
5) Some survey data comes out about the political views of rank-and-file members of [group]. They are overwhelmingly centre-left liberal: Where there is evidence on views about gender specifically, they are also pretty centre-left liberal. I feel even more confused.
Over the years I have seen this pattern to varying degrees with:
-Movement atheism (Can’t find the survey data I once saw highlighted on twitter on this by a surprised critic of right-wing movement “skepticism” so you’ll have to trust me on this one.)
-Analytic philosophy (relatively speaking: seen as a “right-wing” subject in the humanities in relative terms, and a bastion of sexism: both those might be true, but nonetheless, considerably more analytic philosophers endorse socialism than capitalism, and a slight majority are socialists: https://survey2020.philpeople.org/survey/results/5122)
I have seen less hard data for the rationalists, but I do recall about ten years ago Scott Alexander trumpeting that the average LessWrong user had at least as positive a rating of “feminism” on a 1-5 scale as the average American woman. (Though the median American woman politically is probably like an elderly Latina church goer with economically left-wing socially conservative Catholic views?) And that whilst survey data of SSC readers at one point showed most endorsed “race realism” (I remember David Thorstad pointing this out on twitter), and I would not hesitate to describe ACX as “linked to the far-right”, nonetheless I seem to remember than when Scott surveyed the readers on a 1-10 left-right scale, the median reader was a 4.something, i.e. very slightly more left than right identified:
I am not sure what is going on with this, probably a mixture of:
-People being self-deceived about their views and being more right-wing than they think they are, because the right is stigmatized in wider intellectual culture and people don’t want to see themselves as part of it.
-People in these spaces hold mostly left views, but they mostly hold relatively uncontroversial left-wing views, or are not a prime target for the right-wing press for other reasons, whilst they minority of right-wing views they do hold tend to be radioactively controversial so they end up in the media.
-I mostly read centre-leftish media (The Guardian, Yglesias, Vox until the last couple of years) or critics of “wokeness” who are not straightforwardly conservative (Yglesias again, Singal), rather than conventional conservative stuff, so I hear about “woke”/left anger with these groups, but not right-wing anger with them. I also pay less attention to the latter because I just care less about it; it’s not a source of personal angst for me in the same way.
-People who want to/get to become leaders in these sorts of spaces differ in their traits from the median member of the group in ways that make them predictably more right-wing than the average.
-*Becoming* a leader makes you more right-wing, since you like hierarchy more when you’re on top of the local hierarchy.
-People confused a (perceived and/or real) tendency towards sexual bad behaviour amongst autistic nerds with a right-wing political position.
-These groups are well to the left of the median citizens, but they are to the right of the median person with a master degree, so most people in “intellectual” spheres are correctly picking up on them being more right-wing than them and they’re friends, but wrongly concluding that makes them “right-wing” by the standards of the public as a whole.
-Anything stereotypically “masculine” outside of a strike by manual labourers gets coded as “right” these days, facts be damned.
-There is a distinctive cluster of issues around “biodeterminism” on which these groups are very, very right-wing on average-eugenics, biological race and gender differences etc.-but on everything else they are centre-left.
You have some good hypothesis. One other: a lot of left wing activist types (who are disproportionately noisy) have very strong ideological purity preferences, so a person with mainly left wing views but some right wing views can be condemned for the latter, and likewise a movement with mainly left wing people but a few right wing people can be condemned. Any sufficiently public person or movement, unless they are very homogeneous or very PR-conscious, will eventually reveal they have some diversity of views and hence be subject to potential censure.
In my view, what’s going on is largely these two things:
[rationalists etc] are well to the left of the median citizens, but they are to the right of [typical journalists and academics]
Of course. And:
biodeterminism… these groups are very, very right-wing on… eugenics, biological race and gender differences etc.-but on everything else they are centre-left.
Yes, ACX readers do believe that genes influence a lot of life outcomes, and favour reproductive technologies like embryo selection, which are right-coded views. These views are actually not restricted to the far-right, however. Most people will choose to have an abortion when they know their child will have a disability, for example.
Various of your other hypotheses don’t ring true to me. I think:
People aren’t self-deceiving about their own politics very much. They know which politicians and intellectuals they support, and who they vote for.
Rationalist leadership is not very politically different from the rationalist membership.
Sexual misbehaviour doesn’t change perceived political alignment very much.
The high % of male rationalist is at most a minor factor in the difference between perceived and actual politics.
Yes, ACX readers do believe that genes influence a lot of life outcomes, and favour reproductive technologies like embryo selection, which are right-coded views. They’re actually not restricted to the far-right, however.
The problem is that this is really a short step away from “certain races have lower IQ and it’s kinda all there is to it to explain their socio-economic status”, and I’ve seen many people take that step. Roko and Hanania which I mentioned explicitly absolutely do so publicly and repeatedly.
It sounds like you would prefer the rationalist community prevent its members from taking taboo views on social issues? But in my view, an important characteristic of the rationalist community, perhaps its most fundamental, is that it’s a place where people can re-evaluate the common wisdom, with a measure of independence from societal pressure. If you want the rationalist community (or any community) to maintain that character, you need to support the right of people to express views that you regard as repulsive, not just the views that you like. This could be different if the views were an incitement to violence, but proposing a hypothesis for socio-economic differences isn’t that.
Well, it’s complicated. I think in theory these things should be open to discussion (see my point on moral philosophy). But now suppose that hypothetically there was incontrovertible scientific evidence that Group A is less moral or capable than Group B. We should still absolutely champion the view that wanting to ship Group A into camps and exterminate them is barbaric and vile, and that instead the humane and ethical thing to do is help Group A compensate for their issues and flourish at the best of their capabilities (after all, we generally hold this view for groups with various disabilities that absolutely DO hamper their ability to take part in society in various ways). But to know that at all can be also construed as an infohazard: just the fact itself creates the condition for a Molochian trap in which Group A gets screwed by nothing other than economic incentives and everyone else acting in their full rights and self-interest. So yeah, in some way these ideas are dangerous to explore, in the sense that they may be a case where truth-finding has net negative utility. That said, it’s pretty clear that people are way too invested in them either way to just let sleeping dogs lie.
*”Something like”= if you substitute “all there is” with “a major cause, which makes some standard albeit controversial ways of targeting racial inequality fail a cost/benefit test that they might otherwise pass.
**Full quote: ’Everyone is so circumspect when talking about race that I can never figure out what anyone actually knows or believes. Still, I think most people would at least be aware of the following counterargument: suppose you’re the math department at a college. You might like to have the same percent black as the general population (13%). But far fewer than 13% (let’s say 2%) of good math PhDs are black. So it’s impossible for every math department to hire 13% black math professors unless they lower their standards or take some other drastic measure.
Okay, says our hypothetical opponent. Then that means math grad programs are discriminating against blacks. Fine, they’re the ones we should be investigating for civil rights violations.
No, say the math grad programs, fewer than 13% of our applicants are black too.
Fine, then the undergrad programs are the racists. Or if they can prove they’re not, then the high schools are racist and we should do busing. The point is, somebody somewhere along the line has to be racist, right?
I know of four common, non-exclusive answers to this question.
Yes, the high schools (or whatever) are racist. And if you can present a study proving that high schools aren’t racist, then it’s the elementary schools. And if you have a study there too, it’s the obstetricians, giving black mothers worse pregnancy care. If you have a study disproving that too, why are you collecting all these studies? Hey, maybe you’re the racist!
Maybe institutions aren’t too racist today, but there’s a lot of legacy of past racism, and that means black people are poor. And poor people have fewer opportunities and do worse in school. If you have a study showing that black people do worse even when controlled for income, then maybe it’s some other kind of capital, like educational capital or social capital. If you have studies about those too, see above.
Black people have a bad culture. Something something shoes and rap music, trying hard at school gets condemned as “acting white”. They should hold out for a better culture. I hear nobody’s using ancient Sumerian culture these days, maybe they can use that one.
White people have average IQ 100, black people have average IQ 85, this IQ difference accurately predicts the different proportions of whites and blacks in most areas, most IQ differences within race are genetic, maybe across-race ones are genetic too. I love Hitler and want to marry him.
None of these are great options, and I think most people work off some vague cloud of all of these and squirm if you try to make them get too specific. I don’t exactly blame Hanania for not taking a strong stand here. It’s just strange to assume civil rights law is bad and unnecessary without having any opinion on whether any of this is true, whether civil rights law is supposed to counterbalance it, and whether it counterbalances it a fair amount.
A cynic might notice that in February of this year, Hanania wrote Shut Up About Race And IQ. He says that the people who talk about option 4 are “wrong about fundamental questions regarding things like how people form their political opinions, what makes for successful movements, and even their own motivations.” A careful reader might notice what he doesn’t describe them as being wrong about. The rest of the piece almost-but-not-quite-explicitly clarifies his position: I read him as saying that race realism is most likely true, but you shouldn’t talk about it, because it scares people.
(*I’m generally against “calling people out” for believing in race realism*. I think people should be allowed to hide beliefs that they’d get punished for not hiding. I sympathize with some of these positions and place medium probability on some weak forms of them. I think Hanania is open enough about where he’s coming from that this review doesn’t count as a callout.)
I think you are onto something—and I think there is a distinction here between “elites” and “rank and file”, so to speak. Not too surprisingly since these are people from very distinct backgrounds often anyway! I kind of shudder when I see high profile rationalists casually discussing betting or offering prizes of tens of thousands of dollars over small internet arguments, because it’s fairly obvious these people live in a completely different world than mine (where my wife would rightfully have my head if I spaffed half of my year’s salary for internet points). And having different material interests is fairly likely to skew your politics.
One more thing is that often the groups that you describe are most attracted to being libertarian—which is kind of a separate thing, but more right than left coded usually (though it’s the “laissez fair capitalism” kind of right, not the “round up the ethnic minorities and put them in camps” one).
There is a distinctive cluster of issues around “biodeterminism” on which these groups are very, very right-wing on average-eugenics, biological race and gender differences etc.-but on everything else they are centre-left.
This is kind of a key point because there’s also two dimensions to this. One is, “which statements about biodeterminism are true, if any?”, and the other is “what should we do about that?”. The first is a scientific question, the latter a political and moral one. But the truth is that because the right wing has offered some very awful answers to the latter, it has become an important tenet on the left to completely deny that any such statements could be true, which kind of cuts the problem at its roots. This is probably correct anyway for vastly disproved and discredited theses like “black people have lower IQ”, but it gets to the point of denying that IQ is inheritable or correlates at all with anything worth calling “intelligence”, which to me feels a bit too hard to believe (and even if it was—ok, so what is a better measure of intelligence? There has to exist one!).
And well, a community of high decoupling, high intelligence, science minded autists is probably the one that’s most likely to take issue with that. Though again it should be very wary of the risk of going down the lane of self-aggrandizement in which you fall for any supposed “study” more or less flawed that says that group so-and-so is just constitutionally stupid, no need to think any harder about why they do badly.
rationalist community kind of leans right wing on average
Seems false. It leans right compared to the extreme left wing, but right compared to the general population? No. Its too libertarian for that. I bet rightists would also say it leans left, and centrists would say its too extreme. Overall, I think its just classically libertarian.
Mostly agree and have found your post insightful, but am not too sure about the ‘confront this a bit’ part. I feel both most EAs and most Rationalists are very solidly on the left (not the radical, SJW fringe, but very clearly left of center, Democratic-leaning). I vaguely remember having read somewhere Tyler Cowen describing EA as ‘what SJW should be like’. Still, I feel that political partisanship and accepting labels is such a generally toxic and counterproductive affair that it is best avoided. And I think there’s probably some inevitable tension inside EA between people who prioritize the search for veracity and effectiveness, and a high degree of respect for the freedom to explore unconventional and inconvenient truth, and others who might lean more towards prioritizing more left-coded practices and beliefs.
So I am actually perhaps less familiar with the distribution of political beliefs in EAs specifically and I’m thinking about rationalist-adjacent communities more at large, and there’s definitely some people more comfortable around some pretty racist stuff than you’d find elsewhere (as someone else quoted—ACX just published a review of Hanania’s book “The origins of woke”, and the book is apparently a big screed against civil rights law. And knowing Hanania, it’s not hard to guess what he’s driving at). So at least there’s a certain tendency in which open-mindedness and willingness to always try to work everything out from first principles can let in some relatively questionable ideas.
I do agree about the problem with political labels. I do worry about whether that position will be tenable if the label of “TESCREAL” takes off in any meaningful way. Labels or not, if the rationalist community writ large gets under sustained political attack from one side of the aisle, natural alliances will be formed and polarization will almost certainly occur.
So I am actually perhaps less familiar with the distribution of political beliefs in EAs specifically and I’m thinking about rationalist-adjacent communities more at large
The results of the ACX survey just came out and allow us to examine political affiliation and alignment both across the whole sample and based on LW / EA ID.
First, the overall sample.
This is a left-right scale: nearly 70% were on the left side of the spectrum.
Political affiliation: mostly liberal, social democratic and libertarian (in that order).
Now looking at LW ID to assess rationalist communities:
Quite similar, but LW ID’d people lean a bit more to the left than the general readership.
In terms of political affiliation, LWers are substantially less conservative, less neo-reactionary, less alt-right and much more both liberal and libertarian.
Now looking at EA ID (though I would not expect EA ID’s ACX respondents to reflect the EA community as a whole: they should be expected to be more ACX and rationalist leaning):
EAs are more left, 16.4% are on the right end of the spectrum, though 9.7% are in the category immediately right of centre and 13.5% in one of the two most centre-right categories, only 2.9% are more right-leaning than that. (That’s still more right-leaning that the general EA Survey sample, which I would expect to be less skewed, which found 2.2% center-right, 0.7% right.
In terms of political affiliation, EAs are overwhelmingly liberal (almost 50% of the sample) followed by social democratic (another 30.5%), with 15.3% libertarians. There are 4% Conservatives and <1% for each of alt-right or neo-reactionary (for context, 3 and 5 respondents respectively), so definitely into lizardman territory.
Thanks, that’s useful! I guess the surprising thing is maybe just that there still are some fairly prominent names in the rationalist space that express obviously very right wing views and that they are generally almost not seen as such (for example Scott Alexander just wrote a review of Hanania’s new book in which I’d say he almost ends up sounding naive by how much he doesn’t simply acknowledge “well, clearly Hanania is barely stopping shy of saying black people are just stupider”, something that Hanania has said openly elsewhere anyway, so it’s barely a mystery that he believes it).
I would need to dig up specific stuff, but in general I’d suggest to just check out his Twitter/X account https://twitter.com/RichardHanania and see what he says. These days it’s completely dominated by discourse on the Palestine protests so it’s hard to dig out anything on race. Mind you, he’s not one to hold a fully stereotypical GOP-aligned package of ideas—he has a few deviations and is secular (so for example pro-choice on abortion; also he’s definitely not antisemitic, in fact he explicitly called himself prosemitic, as he believes Jews to be smarter). But on race I’m fairly convinced he 100% believes in scientific racism from any time he’s talked about it. I don’t want to link any of the opinion pieces around that argue for this (but there’s a fair deal if you want to check them out and try to separate fact from fiction—many point out that he’s sort of switched to some more defensive “bailey” arguments lately, which he seems to do and explicitly advocate for as a strategy in his latest book “The Origins of Woke” too, again see the ACX review). But for some primary evidence, for example, here’s a tweet about how crime can only be resolved by more incarceration and surveillance of black people:
He used to write more explicitly racist stuff under the pseudonym Richard Hoste until a few years ago. He openly admitted this and wrote an apology blog post in which he basically says that he was young and went a bit too far. Now whether this corresponds to a genuine moderation (from extremely right wing to merely strongly socially conservative and anti-woke) is questionable, because it could just as well be a calculated retreat from a motte to a bailey. It’s not wild to consider this possibility given that, again, he explicitly talks about how certain arguments would scare normies too much so it’s better to just present more palatable ones. And after all that is a pretty sound strategy (and one Torres accused EAs of recently re: using malaria bednets as the bailey to draw people into the motte of AI safety, something that of course I don’t quite see as evil as he implies it to be since I think AI safety absolutely is a concern, and the fact that it looks weird to the average person doesn’t make it not so).
At this point from all I’ve seen my belief is that Hanania mostly is a “race realist” who thinks some races are inherently inferior and thus the correct order of things has them working worse jobs, earning less money etc. and all efforts in the opposite direction are unjust and counterproductive. I don’t think he then moves from that to “and they should be genocided”, but that’s not a lot. He still thinks they should be an underclass and for now thinks that the market left to its own devices would make them so, which would be the rightful order of things. That’s the model of him I built, and I find it hard to believe that Scott Alexander for example hasn’t seen all the same stuff.
This might have to do with “how local” your local community is. It seems to me that the weirder sides of EA (which I usually consider bad, but others here might not) are common in the EA hubs (Bay Area, Oxbridge, London, and the cluster of large groups in Europe) but not as common in other places (like here in Israel).
But even then, a nuanced engagement with that would require making distinctions, not just going “all EA evil”. Both Torres and Gebru these days are very invested in pushing this label of “TESCREAL” to bundle together completely different groups, from EAs who spend 10% of their income in malaria nets to rationalists who worry about AI x-risk to e/accs who openly claim that ASI is the next step in evolution. I think here there are two problems:
abstract moral philosophy can’t be for the faint of heart, you’re engaging with the fundamental meaning of good or evil, you must be able to realize when a set of assumptions leads to a seemingly outrageous conclusion and then decide what to make of that. But if a moral philosopher writes “if we assume A and B, that leads us to concluding that it would be moral to eat babies”, the reaction can’t be “PHILOSOPHER ENDORSES EATING BABIES!!!!”, because that’s both a misunderstanding of their work and if universalized will have a chilling effect leading to worse moral philosophy overall. Sometimes entertaining weird scenarios is important, if only to realise the contradictions in our assumptions;
for good or bad, left-wing thought and discourse in the last ten-fifteen years just hasn’t been very rational. And I don’t mean to say that the left can’t be rational. Karl Marx’s whole work was based on economics and an attempt to create a sort of scientific theory of history, love it or hate it the man obviously had a drive more akin to those of current rationalists than of current leftists. What is happening right now is more of a fad, a current of thought in which basically rationality and objectivity have been sort of depreciated as memes in the left wing sphere, and therapy-speak that centres the self and the inner emotions and identity has become the standard language of the left. And that pushes away a certain kind of mind very much, it feels worse than wrong, it feels like bullshit. As a result, the rationalist community kind of leans right wing on average for evaporative cooling reasons. Anyone who cares to be seen well by online left wing communities won’t associate. Anyone who’s high decoupling will be more attracted, and among those, put together with the average rationalist’s love for reinventing the wheel from first principles, some will reach some highly controversial beliefs in current society that they will nevertheless hold up as true (read: racism). It doesn’t help when terms like “eugenics” are commonly used to lump together things as disparate as Nazis literally sterilizing and killing people and hypothetical genetical modifications used to help willing parents have children that are on average healthier or live longer lives, obviously very different moral issues.
Honestly I do think the rationalist space needs to confront this a bit. People like Roko or Hanania hold pretty extreme right wing beliefs, to the point where you can’t even really call them rational because they are often dominated by confirmation bias and the usual downfalls of political polarization. Longtermism itself is a pretty questionable proposition in my book, though that argument still lays in the space of philosophy for the most part. I would be all for the rise of a “rational left”, both for the good of the rationalist community and for the good of the left, which is currently really mired into an unproductive circle jerk of emotionalism and virtue signalling. But this “TESCREAL” label if anything risks having the opposite effect, and polarizing people away from these philosophically incompetent and intellectually dishonest representatives of what’s supposed to be the current left wing intelligentsia.
I think the trend you describe is mostly an issue with “progressives”, i.e. “leftists” rather than an issue for all those left of center. And the rationalists don’t actually lean right in my experience. They average more like anti-woke and centrist. The distribution in the 2024 ACX survey below has perhaps a bit more centre-left and a bit less centre and centre-right than the rationalists at large but not by much, in my estimation.
Fair! I think it’s hard to fully slot rationalists politically because, well, the mix of high decoupling and generally esoteric interests make for some unique combinations that don’t fit neatly in the standard spectrum. I’d definitely qualify myself as centre-left, with some more leftist-y views on some aspects of economics, but definitely bothered by the current progressive vibe that I hesitate to define “woke” since that term is abused to hell but am also not sure how to call since they obstinately refuse to give themselves a political label or even recognise that they constitute a noteworthy distinct political phenomenon at all.
How was this survey done, by the way? Self ID or some kind of scored test?
This was just a “where do you rate yourself from 1-10” type question, but you can see more of the questions and data here.
So the thing with self-identification is that I think it might suffer from a certain skew. I think there’s fundamentally a bit of a stigma on identifying as right wing, and especially extreme right wing. Lots of middle class, educated people who perceive themselves as rational, empathetic and science-minded are more likely to want to perceive themselves as left wing, because that’s what left wing identity used to prominently be until a bit over 15 years ago (which is when most of us probably had their formative youth political experiences). So someone might resist the label even if in practice they are on the right half of the Overton window. Must be noted though that in some cases this might just be the result of the Overton window moving around them—and I definitely have the feeling that we now have a more polarized distribution anyway.
Do we actually have hard statistical evidence that rationalists as a group “lean right”? I am highly unsympathetic to right rationalism, as you can see here: https://forum.effectivealtruism.org/posts/kgBBzwdtGd4PHmRfs/an-instance-of-white-supremacist-and-nazi-ideology-creeping?commentId=tNHd9C8ZbazepnDqs And it certainly feels true emotionally to me that “rationalism is right-wing”. (Which is one reason I consider myself an EA but not a rationalist, although that is mostly just because I entered EA through academic philosophy not rationalism and other than reading a lot of SSC/ACX over the years, have only ever interacted with rationalists in the context of doing EA stuff.) Certain high profile individual rationalist seem to hold a lot of taboo/far-right beliefs (i.e. Scott Alexander on race and IQ here: https://www.astralcodexten.com/p/book-review-the-origins-of-woke). Roko and Hannania are of course even more right-wing (and frankly pretty gross in my view), though hopefully they are outliers.
BUT
Over the years, I have observed a general pattern with, what we can call “rationalist-like” groups: i.e. lots of men, mostly straight and white, lots of autism broadly construed, an interest in telling the harsh truth, reverence for STEM and skepticism of the humanities, more self-declared right libertarians than the population average etc.:
1) The group gains a reputation for being right-wing, sexist, bigoted etc.
2) People in the group get very offended about this; I get a bit offended too: most people I have met within the group seem moderate, with a lean towards the centre-left rather than the centre-right. I feel as a person with mild autism that autistic truth-telling and bluntness is getting stigmatised by annoying, overemotive people who can’t defend their views in a fair argument.
3) Gradually thought leaders of the group have a lot of scandals involving some combination of: misogyny, sexual harassment, Islamophobia, racism, eugenics, Western chauvinism etc.
4) I start thinking “ok, maybe [group] actually is right-wing, and I am either a sucker to be involved with it, or self-deceived about my own (self-declared liberal centrist) political preferences (after all, I do get irritated with the left a lot and believe some un-PC things, I think I am pro some transhumanist genetic engineering stuff in principle maybe, am not particularly left-or right-economically, maybe “liberalism” is what angry Marxists on twitter say it is etc. etc.)”
5) Some survey data comes out about the political views of rank-and-file members of [group]. They are overwhelmingly centre-left liberal: Where there is evidence on views about gender specifically, they are also pretty centre-left liberal. I feel even more confused.
Over the years I have seen this pattern to varying degrees with:
-Movement atheism (Can’t find the survey data I once saw highlighted on twitter on this by a surprised critic of right-wing movement “skepticism” so you’ll have to trust me on this one.)
-Analytic philosophy (relatively speaking: seen as a “right-wing” subject in the humanities in relative terms, and a bastion of sexism: both those might be true, but nonetheless, considerably more analytic philosophers endorse socialism than capitalism, and a slight majority are socialists: https://survey2020.philpeople.org/survey/results/5122)
-”Tech” itself: https://www.noahpinion.blog/p/silicon-valley-isnt-full-of-fascists
-EA itself (Compare our bad reputation on the left-as far as I can tell, with the fact that more EAs identify as “left” than “centre” “centre-right” “right” “libertarian” or “other” put together even when “centre-left” is also an option: https://forum.effectivealtruism.org/posts/AJDgnPXqZ48eSCjEQ/ea-survey-2022-demographics#Politics).
I have seen less hard data for the rationalists, but I do recall about ten years ago Scott Alexander trumpeting that the average LessWrong user had at least as positive a rating of “feminism” on a 1-5 scale as the average American woman. (Though the median American woman politically is probably like an elderly Latina church goer with economically left-wing socially conservative Catholic views?) And that whilst survey data of SSC readers at one point showed most endorsed “race realism” (I remember David Thorstad pointing this out on twitter), and I would not hesitate to describe ACX as “linked to the far-right”, nonetheless I seem to remember than when Scott surveyed the readers on a 1-10 left-right scale, the median reader was a 4.something, i.e. very slightly more left than right identified:
I am not sure what is going on with this, probably a mixture of:
-People being self-deceived about their views and being more right-wing than they think they are, because the right is stigmatized in wider intellectual culture and people don’t want to see themselves as part of it.
-People in these spaces hold mostly left views, but they mostly hold relatively uncontroversial left-wing views, or are not a prime target for the right-wing press for other reasons, whilst they minority of right-wing views they do hold tend to be radioactively controversial so they end up in the media.
-I mostly read centre-leftish media (The Guardian, Yglesias, Vox until the last couple of years) or critics of “wokeness” who are not straightforwardly conservative (Yglesias again, Singal), rather than conventional conservative stuff, so I hear about “woke”/left anger with these groups, but not right-wing anger with them. I also pay less attention to the latter because I just care less about it; it’s not a source of personal angst for me in the same way.
-People who want to/get to become leaders in these sorts of spaces differ in their traits from the median member of the group in ways that make them predictably more right-wing than the average.
-*Becoming* a leader makes you more right-wing, since you like hierarchy more when you’re on top of the local hierarchy.
-People confused a (perceived and/or real) tendency towards sexual bad behaviour amongst autistic nerds with a right-wing political position.
-These groups are well to the left of the median citizens, but they are to the right of the median person with a master degree, so most people in “intellectual” spheres are correctly picking up on them being more right-wing than them and they’re friends, but wrongly concluding that makes them “right-wing” by the standards of the public as a whole.
-Anything stereotypically “masculine” outside of a strike by manual labourers gets coded as “right” these days, facts be damned.
-There is a distinctive cluster of issues around “biodeterminism” on which these groups are very, very right-wing on average-eugenics, biological race and gender differences etc.-but on everything else they are centre-left.
You have some good hypothesis. One other: a lot of left wing activist types (who are disproportionately noisy) have very strong ideological purity preferences, so a person with mainly left wing views but some right wing views can be condemned for the latter, and likewise a movement with mainly left wing people but a few right wing people can be condemned. Any sufficiently public person or movement, unless they are very homogeneous or very PR-conscious, will eventually reveal they have some diversity of views and hence be subject to potential censure.
In my view, what’s going on is largely these two things:
Of course. And:
Yes, ACX readers do believe that genes influence a lot of life outcomes, and favour reproductive technologies like embryo selection, which are right-coded views. These views are actually not restricted to the far-right, however. Most people will choose to have an abortion when they know their child will have a disability, for example.
Various of your other hypotheses don’t ring true to me. I think:
People aren’t self-deceiving about their own politics very much. They know which politicians and intellectuals they support, and who they vote for.
Rationalist leadership is not very politically different from the rationalist membership.
Sexual misbehaviour doesn’t change perceived political alignment very much.
The high % of male rationalist is at most a minor factor in the difference between perceived and actual politics.
The race stuff is much more right-coded than some of the other genetic/disability stuff.
The problem is that this is really a short step away from “certain races have lower IQ and it’s kinda all there is to it to explain their socio-economic status”, and I’ve seen many people take that step. Roko and Hanania which I mentioned explicitly absolutely do so publicly and repeatedly.
It sounds like you would prefer the rationalist community prevent its members from taking taboo views on social issues? But in my view, an important characteristic of the rationalist community, perhaps its most fundamental, is that it’s a place where people can re-evaluate the common wisdom, with a measure of independence from societal pressure. If you want the rationalist community (or any community) to maintain that character, you need to support the right of people to express views that you regard as repulsive, not just the views that you like. This could be different if the views were an incitement to violence, but proposing a hypothesis for socio-economic differences isn’t that.
Well, it’s complicated. I think in theory these things should be open to discussion (see my point on moral philosophy). But now suppose that hypothetically there was incontrovertible scientific evidence that Group A is less moral or capable than Group B. We should still absolutely champion the view that wanting to ship Group A into camps and exterminate them is barbaric and vile, and that instead the humane and ethical thing to do is help Group A compensate for their issues and flourish at the best of their capabilities (after all, we generally hold this view for groups with various disabilities that absolutely DO hamper their ability to take part in society in various ways). But to know that at all can be also construed as an infohazard: just the fact itself creates the condition for a Molochian trap in which Group A gets screwed by nothing other than economic incentives and everyone else acting in their full rights and self-interest. So yeah, in some way these ideas are dangerous to explore, in the sense that they may be a case where truth-finding has net negative utility. That said, it’s pretty clear that people are way too invested in them either way to just let sleeping dogs lie.
Scott seems not unsympathetic to something like* that step here**, though he stops short of clear endorsement: https://www.astralcodexten.com/p/book-review-the-origins-of-woke I think this is a dangerous path to go down.
*”Something like”= if you substitute “all there is” with “a major cause, which makes some standard albeit controversial ways of targeting racial inequality fail a cost/benefit test that they might otherwise pass.
**Full quote:
’Everyone is so circumspect when talking about race that I can never figure out what anyone actually knows or believes. Still, I think most people would at least be aware of the following counterargument: suppose you’re the math department at a college. You might like to have the same percent black as the general population (13%). But far fewer than 13% (let’s say 2%) of good math PhDs are black. So it’s impossible for every math department to hire 13% black math professors unless they lower their standards or take some other drastic measure.
Okay, says our hypothetical opponent. Then that means math grad programs are discriminating against blacks. Fine, they’re the ones we should be investigating for civil rights violations.
No, say the math grad programs, fewer than 13% of our applicants are black too.
Fine, then the undergrad programs are the racists. Or if they can prove they’re not, then the high schools are racist and we should do busing. The point is, somebody somewhere along the line has to be racist, right?
I know of four common, non-exclusive answers to this question.
Yes, the high schools (or whatever) are racist. And if you can present a study proving that high schools aren’t racist, then it’s the elementary schools. And if you have a study there too, it’s the obstetricians, giving black mothers worse pregnancy care. If you have a study disproving that too, why are you collecting all these studies? Hey, maybe you’re the racist!
Maybe institutions aren’t too racist today, but there’s a lot of legacy of past racism, and that means black people are poor. And poor people have fewer opportunities and do worse in school. If you have a study showing that black people do worse even when controlled for income, then maybe it’s some other kind of capital, like educational capital or social capital. If you have studies about those too, see above.
Black people have a bad culture. Something something shoes and rap music, trying hard at school gets condemned as “acting white”. They should hold out for a better culture. I hear nobody’s using ancient Sumerian culture these days, maybe they can use that one.
White people have average IQ 100, black people have average IQ 85, this IQ difference accurately predicts the different proportions of whites and blacks in most areas, most IQ differences within race are genetic, maybe across-race ones are genetic too. I love Hitler and want to marry him.
None of these are great options, and I think most people work off some vague cloud of all of these and squirm if you try to make them get too specific. I don’t exactly blame Hanania for not taking a strong stand here. It’s just strange to assume civil rights law is bad and unnecessary without having any opinion on whether any of this is true, whether civil rights law is supposed to counterbalance it, and whether it counterbalances it a fair amount.
A cynic might notice that in February of this year, Hanania wrote Shut Up About Race And IQ. He says that the people who talk about option 4 are “wrong about fundamental questions regarding things like how people form their political opinions, what makes for successful movements, and even their own motivations.” A careful reader might notice what he doesn’t describe them as being wrong about. The rest of the piece almost-but-not-quite-explicitly clarifies his position: I read him as saying that race realism is most likely true, but you shouldn’t talk about it, because it scares people.
(*I’m generally against “calling people out” for believing in race realism*. I think people should be allowed to hide beliefs that they’d get punished for not hiding. I sympathize with some of these positions and place medium probability on some weak forms of them. I think Hanania is open enough about where he’s coming from that this review doesn’t count as a callout.)
″
I think you are onto something—and I think there is a distinction here between “elites” and “rank and file”, so to speak. Not too surprisingly since these are people from very distinct backgrounds often anyway! I kind of shudder when I see high profile rationalists casually discussing betting or offering prizes of tens of thousands of dollars over small internet arguments, because it’s fairly obvious these people live in a completely different world than mine (where my wife would rightfully have my head if I spaffed half of my year’s salary for internet points). And having different material interests is fairly likely to skew your politics.
One more thing is that often the groups that you describe are most attracted to being libertarian—which is kind of a separate thing, but more right than left coded usually (though it’s the “laissez fair capitalism” kind of right, not the “round up the ethnic minorities and put them in camps” one).
This is kind of a key point because there’s also two dimensions to this. One is, “which statements about biodeterminism are true, if any?”, and the other is “what should we do about that?”. The first is a scientific question, the latter a political and moral one. But the truth is that because the right wing has offered some very awful answers to the latter, it has become an important tenet on the left to completely deny that any such statements could be true, which kind of cuts the problem at its roots. This is probably correct anyway for vastly disproved and discredited theses like “black people have lower IQ”, but it gets to the point of denying that IQ is inheritable or correlates at all with anything worth calling “intelligence”, which to me feels a bit too hard to believe (and even if it was—ok, so what is a better measure of intelligence? There has to exist one!).
And well, a community of high decoupling, high intelligence, science minded autists is probably the one that’s most likely to take issue with that. Though again it should be very wary of the risk of going down the lane of self-aggrandizement in which you fall for any supposed “study” more or less flawed that says that group so-and-so is just constitutionally stupid, no need to think any harder about why they do badly.
Seems false. It leans right compared to the extreme left wing, but right compared to the general population? No. Its too libertarian for that. I bet rightists would also say it leans left, and centrists would say its too extreme. Overall, I think its just classically libertarian.
Mostly agree and have found your post insightful, but am not too sure about the ‘confront this a bit’ part. I feel both most EAs and most Rationalists are very solidly on the left (not the radical, SJW fringe, but very clearly left of center, Democratic-leaning). I vaguely remember having read somewhere Tyler Cowen describing EA as ‘what SJW should be like’. Still, I feel that political partisanship and accepting labels is such a generally toxic and counterproductive affair that it is best avoided. And I think there’s probably some inevitable tension inside EA between people who prioritize the search for veracity and effectiveness, and a high degree of respect for the freedom to explore unconventional and inconvenient truth, and others who might lean more towards prioritizing more left-coded practices and beliefs.
So I am actually perhaps less familiar with the distribution of political beliefs in EAs specifically and I’m thinking about rationalist-adjacent communities more at large, and there’s definitely some people more comfortable around some pretty racist stuff than you’d find elsewhere (as someone else quoted—ACX just published a review of Hanania’s book “The origins of woke”, and the book is apparently a big screed against civil rights law. And knowing Hanania, it’s not hard to guess what he’s driving at). So at least there’s a certain tendency in which open-mindedness and willingness to always try to work everything out from first principles can let in some relatively questionable ideas.
I do agree about the problem with political labels. I do worry about whether that position will be tenable if the label of “TESCREAL” takes off in any meaningful way. Labels or not, if the rationalist community writ large gets under sustained political attack from one side of the aisle, natural alliances will be formed and polarization will almost certainly occur.
The results of the ACX survey just came out and allow us to examine political affiliation and alignment both across the whole sample and based on LW / EA ID.
First, the overall sample.
This is a left-right scale: nearly 70% were on the left side of the spectrum.
Political affiliation: mostly liberal, social democratic and libertarian (in that order).
Now looking at LW ID to assess rationalist communities:
Quite similar, but LW ID’d people lean a bit more to the left than the general readership.
In terms of political affiliation, LWers are substantially less conservative, less neo-reactionary, less alt-right and much more both liberal and libertarian.
Now looking at EA ID (though I would not expect EA ID’s ACX respondents to reflect the EA community as a whole: they should be expected to be more ACX and rationalist leaning):
EAs are more left, 16.4% are on the right end of the spectrum, though 9.7% are in the category immediately right of centre and 13.5% in one of the two most centre-right categories, only 2.9% are more right-leaning than that. (That’s still more right-leaning that the general EA Survey sample, which I would expect to be less skewed, which found 2.2% center-right, 0.7% right.
In terms of political affiliation, EAs are overwhelmingly liberal (almost 50% of the sample) followed by social democratic (another 30.5%), with 15.3% libertarians. There are 4% Conservatives and <1% for each of alt-right or neo-reactionary (for context, 3 and 5 respondents respectively), so definitely into lizardman territory.
Thanks, that’s useful! I guess the surprising thing is maybe just that there still are some fairly prominent names in the rationalist space that express obviously very right wing views and that they are generally almost not seen as such (for example Scott Alexander just wrote a review of Hanania’s new book in which I’d say he almost ends up sounding naive by how much he doesn’t simply acknowledge “well, clearly Hanania is barely stopping shy of saying black people are just stupider”, something that Hanania has said openly elsewhere anyway, so it’s barely a mystery that he believes it).
Could you provide links to those statements by Hanania?
Not a gotcha, I just have barely heard of this guy and from what you say I expect all discourse around him to be a cesspool.
I would need to dig up specific stuff, but in general I’d suggest to just check out his Twitter/X account https://twitter.com/RichardHanania and see what he says. These days it’s completely dominated by discourse on the Palestine protests so it’s hard to dig out anything on race. Mind you, he’s not one to hold a fully stereotypical GOP-aligned package of ideas—he has a few deviations and is secular (so for example pro-choice on abortion; also he’s definitely not antisemitic, in fact he explicitly called himself prosemitic, as he believes Jews to be smarter). But on race I’m fairly convinced he 100% believes in scientific racism from any time he’s talked about it. I don’t want to link any of the opinion pieces around that argue for this (but there’s a fair deal if you want to check them out and try to separate fact from fiction—many point out that he’s sort of switched to some more defensive “bailey” arguments lately, which he seems to do and explicitly advocate for as a strategy in his latest book “The Origins of Woke” too, again see the ACX review). But for some primary evidence, for example, here’s a tweet about how crime can only be resolved by more incarceration and surveillance of black people:
https://twitter.com/RichardHanania/status/1657541010745081857?lang=en-GB
His RationalWiki article has obviously opinions about him, but also a bunch of links to primary sources in the bibliography:
https://rationalwiki.org/wiki/Richard_Hanania
He used to write more explicitly racist stuff under the pseudonym Richard Hoste until a few years ago. He openly admitted this and wrote an apology blog post in which he basically says that he was young and went a bit too far. Now whether this corresponds to a genuine moderation (from extremely right wing to merely strongly socially conservative and anti-woke) is questionable, because it could just as well be a calculated retreat from a motte to a bailey. It’s not wild to consider this possibility given that, again, he explicitly talks about how certain arguments would scare normies too much so it’s better to just present more palatable ones. And after all that is a pretty sound strategy (and one Torres accused EAs of recently re: using malaria bednets as the bailey to draw people into the motte of AI safety, something that of course I don’t quite see as evil as he implies it to be since I think AI safety absolutely is a concern, and the fact that it looks weird to the average person doesn’t make it not so).
At this point from all I’ve seen my belief is that Hanania mostly is a “race realist” who thinks some races are inherently inferior and thus the correct order of things has them working worse jobs, earning less money etc. and all efforts in the opposite direction are unjust and counterproductive. I don’t think he then moves from that to “and they should be genocided”, but that’s not a lot. He still thinks they should be an underclass and for now thinks that the market left to its own devices would make them so, which would be the rightful order of things. That’s the model of him I built, and I find it hard to believe that Scott Alexander for example hasn’t seen all the same stuff.