Kawoomba comments on Rationality Quotes November 2012 - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (898)
Except that it is. You don't dissuade a racist or a misogynist or whatever with brain surgery. You just show them that their model of minorities/women/homosexuals/whatever is flawed. You don't alter their brain to terminally value, say, preventing slavery, you just show them that to satisfy their existing terminal value of avoiding human suffering they should prevent slavery. That has no effect on Coherent Extrapolated Volition.
Very idealistic. What if putting down other humans is an actual terminal value for some people?
Something in their mind would most likely be broken for this to happen, according to data I've seen. Even the worst polypaths (sociopath + psychopath + whatever pathy you want to throw in for the most despicable example of human possible) apparently still see themselves as heroes "saving the world", or at least as some kind of "good guy in the long run". Not that I'm implying that the "All Evil people are actually Good because they just know that the world needs balance and humans need a common enemy" myth is true, because that's been shown false even more clearly.
More importantly, CEV aims are reflectively coherent values. If they have a terminal value of killing people, this value is extremely likely to conflict with other of their values, maybe even with their own wishes regarding self-values ("I wish I didn't enjoy killing humans so much"), and would definitely not cohere with most other humans' values unless there's four billion people out there who secretly desire very very much to kill humans all the time but live in utter misery thanks to a global conspiracy that successfully chains them down or some other just-as-unlikely factor (e.g. some undiscovered freeloader's problem or tragedy of the commons that we've somehow never noticed).
Overall, my current odds are very very low that any such value would survive extrapolation when you attempt to have a reflectively coherent system where someone wishes they didn't value X, but do value it.
Every (neurotypical) human I have ever observed, even indirectly, terminally valued human life. Have you ever met a racist? They're not evil mutants, they simply don't think minorities are people (and may think they are an active threat to "real" humans.) Of course, a mind that terminally values killing humans is possible. I simply haven't seen any evidence that it exists in real life, and plenty of evidence for minds with a stable CEV that terminally values human life (among other things.)
Just not true. Not all racist people are confused on matters of fact in such a convenient way.
In my experience, they are. Could you provide a counterexample? Bearing in mind that I was using "people" to exclude evil mutants.
Part of the problem here may be that you are using a strong notion of racist. So the issue may be definitional. For example, I have a close relative who says nasty things about other racial minorities (especially that people from a certain racial group are stupid, lazy, fat, etc.). I'm pretty sure that person has no desire to kill people of that racial group. There are different degrees of racism.
Yes, I was using a "strong" racist as an example. But that close relative's claims that the minorities are lazy/fat/whatever is either correct or incorrect as a point of fact, and has no effect on their terminal values.
You could call me a proto-racist in that I think that some races are more intelligent than others, more civilized, more violent, while still very much human.
There are no confusions of fact between me and someone who hates people of "lesser" races, only a lack of compassion.
And there I was thinking civilisation was memetic, not genetic.
No one said it was genetic.
People always assume that acknowledgeing a trait in a person requires you to have an explanation for it. And then they note that all possible explanations are politically controverisal, so they conclude that the trait does not actually exist. This is bad logic, as far as I can tell.
The fact is, race is a good predictor of things like civilization, intelligence, violence, etc. I offer no explanations.
Eh? What is this thing you call "race," Earth Monkey?
We used to think the answer was obvious. You know, it's obvious what "race" someone is, isn't it? Until you start to look at the details.
Race is a cultural convention. There is a science of population genetics, and it isn't about "race." Rather, people use population genetics to infer the social marker called "race."
I adopted an African girl. What "race" is she? What determines this? She has tribal markings on her eyes -- or the scars from tribal medicine for conjunctivitis, hard to tell -- but the markings are characteristic of her region and tribe, so someone who knows could tell where she comes from, as to the region.
I once had a friend tell me that my Chinese daughter was, of course, going to be more intelligent than the Ethiopian girl. The Chinese daughter is no slouch, intellectually, but her younger sister is definitely smart as hell. My friend was a racist. Lots of people are racist. That is, they believe that race is a biological or even a "spiritual" reality. He wasn't being mean, he was just being ignorant.
What determines it? Ancestry. Race is basically a way of asking "who were your ancestors?" and accepting a blurry answer because, well, each person has a lot of ancestors! That version of race is obviously a biological reality, because people have different ancestries, even going back long distances, and the ancestry distribution can be geographically plotted. If you go back thirty generations for me, I would need to have about a billion distinct ancestors for there to be no inbreeding; the entire world didn't have that many people! Europe, the probable source for most of my ancestry, only had about 50 million people thirty generations ago, and even then it's unlikely that all of them are my ancestors- for one, many of them didn't have any children! I'd estimate somewhere less than 10% of the total world population at any point since 1000 AD is in my ancestry, and the distribution of their contribution to my ancestry is pretty localized. It's probable there's many people out there who share none of my ancestry for a full thirty generations back, and there's one who (probably) shares it completely.
Knowing she was adopted from Africa, odds are good that she's mostly African. That's only one step more informative than "human," since it only gives you the archaic racial category- Negroid- which tells you as much as "Caucasoid" or "Mongoloid." Ethnicity would give a much narrower picture- about one person in six is African, but only about one person in four thousand is Gurage.
Adding on the data that she's Ethiopian muddies the picture- due to its northeastern position, Ethiopia has been the site of significant mixing, and there's quite a bit of ethnic diversity: the primary ethnicity, Oromo, is only a third of the population- your Chinese daughter, though, most likely has significant Han ancestry (92% of the population of mainland China).
So, using the archaic terms and assuming she's from one of the more prevalent ethnicities, your daughter probably has about 60% Caucasoid ancestry and 40% Negroid ancestry.
So, good IQ estimates in Africa are generally hard to come by, but Ethiopia supposedly has the world's lowest average IQ, at 63 (administered in 1991, sample size of 250), and China is estimated to have an average IQ of 100. Working off that data (and assuming both groups have a standard deviation of 15), that gives a 96% chance that the Chinese daughter is smarter. Now, the Ethiopian data is spotty, especially the normality assumption- one of the pitfalls of historic IQ testing is that 0 scores are treated as 0s, dragging down the average, instead of an separate number of "people who didn't understand the concept of the test." It's also not clear what selection effects adoption has; children that get adopted out are likely to not be representative of the country as a whole, and it's hard to say if that would be a positive or negative effect. If we use the African American average IQ of 85 instead of the estimated Ethiopian averaged IQ, and still assume that we should use the Chinese average, we get a 76% chance that the Chinese daughter is cleverer.
Of course, given that they're your daughters, there's not much reason to guess; you could just get them both tested, which would be way cheaper and more informative than sponsoring another test of Ethiopian national IQ.
Genetically differentiated human populations defined by phenotype.
A quote from wikipedia:
"Forensic physical anthropologist and professor George W. Gill has said that the idea that race is only skin deep "is simply not true, as any experienced forensic anthropologist will affirm" and "Many morphological features tend to follow geographic boundaries coinciding often with climatic zones. This is not surprising since the selective forces of climate are probably the primary forces of nature that have shaped human races with regard not only to skin color and hair form but also the underlying bony structures of the nose, cheekbones, etc. (For example, more prominent noses humidify air better.)" While he can see good arguments for both sides, the complete denial of the opposing evidence "seems to stem largely from socio-political motivation and not science at all". He also states that many biological anthropologists see races as real yet "not one introductory textbook of physical anthropology even presents that perspective as a possibility. In a case as flagrant as this, we are not dealing with science but rather with blatant, politically motivated censorship".
Better than, say, poverty? Source please.
Make sure you're distinguishing between the claim that P(intelligence = x|income = i) = P(intelligence = x|race = r,income = i), which would be that poverty screens off the effects of race, and the claim that P(intelligence = x|race = r) = P(intelligence = x), which is the claim that intelligence and race are unconditionally independent. The first claim is only relevant to nyan_sandwich's claim if by "good" you mean "better than income" rather than "worth knowing."
As it stands, both of those claims are pretty obviously false if you take a look an unbiased look at the data. Life is not fair.
What do you mean by "race?" I notice a lot of discussion below on this topic already, but the term is unclear to me, and I don't see how anyone can usefully disagree or agree without this information. Some people use "race" to indicate loose groupings based around skin color, whereas others mean much more strictly a specific genetic group.
Incidentally, there is no canonical "race," just generally-agreed upon loose labels that vary from person to person. Because of this, it is generally not useful for predicting anything, and should be avoided, I think. A "white person" from Sicily and a "white person" from Iceland do not have much more in common with each other than they might with a disparate other range of people, so it's not a meaningful grouping (except perhaps when speaking of historical things). It is wiser to be more exact.
There's the additional danger that you will be misunderstood, and that someone will (very reasonably) think that you are advocating simple-minded racism of a common sort. Saying "race is a good predictor of things like civilization, intelligence," etc. is a fairly specific sort of social code, and if you don't actually mean that "black people are dumb" or "Asians can't drive," (and I'm not saying that you necessarily do) then you should find another sort of phrasing.
yes, "race" as normally used is woefully underdefined.
Woah there. To the extent you can agree on a test for race, it will be useful for prediction. Obviously some tests (actual genetic heritage) will be more interesting that others "lol what's ur skin color". As you say:
Yes, agree. Let's be specific enough so that we all agree which set of people we are talking about, and agree that that is a meaningful grouping. Then I think we will find that membership in that set will predict many things.
Saying that race is a good predictor of such things is roughly equivalent to saying "black people are dumb" or whatever (with suitable disclaimers of probabilisticness that really should not be needed on this site). Call that simple minded if you like; I'd rather be right than high-minded.
(and just-so it's clear, racially-based (or anything-based, really) hatred is stupid and unproductive. compassion is so much nicer.)
I thought genes had something to do with race, but feel free to clarify.
Across cultures?
Currently unknown, since they are strongly correlated anyway; race is also a good predictor of cultures.
In case you're not just trolling, you should be aware that your comments might be better received if you were a bit less abrupt and toned down the snark.
While I'm going to have to disagree with you on a point of fact there, (is that inherently less civilized? Whatever, we're getting off topic.) I note that someone who agrees with you but argues for, say, racial warfare or whatever, is likely under the impression that these natural differences between races are likely to jeopardise White civilization, and/or believes that these differences are more drastic than you do.
not necessarily. Sometimes it seems that way and would take quite the conspiracy to make it not true, but it could be either way and the point would be the same.
Oh, it's entirely beside the point. I'm just curious.
... which races are we talking about here? On reflection, I can't think of any race that seems less civilized. I could just about buy a difference in intelligence, and I can see how you might think violent inclinations are affected by race (although I'm pretty sure that's sociological,) but I don't really see how you can plausibly make the claim that civilization has a racial component.
let's take civilization off the table for now.
I'm under the impression that controlled studies have found race-intelligence links. (maybe this is my confusion of basic facts). Seems plausible.
Sociology is pretty strongly corellated with this race thing, and race is more visible. Violence is definately linked to race in the USA (even if it's only through social class, but I've seen arguments that race is an even better predictor).
It would be nice to have some solid literature surveys on this. Unfortunately, given the controversy, it's hard to just trust what the academics are saying.
Care to name names?
Names of racists who believe other races are inferior? Racists who think other races are an active threat? Commenters who have claimed racists are evil mutants? I'm not sure what you could be asking that couldn't be answered with, at most, a quick Google search or the phrase "all of them?"
Sorry for being too vague (I tend to be so when posting from my smartphone). What I was asking for, specifically, is for names of racists who "simply don't think minorities are people". My priors are that this is a straw-man, but if not, please excuse my ignorance of racist doctrine.
While I'm having a surprisingly hard time finding any quotes by racists (I'm mostly getting quotes about racism, which is of course very different,) I've definitely seen quotes from slave-owners and British Empire types talking about how the savages cannot control their impulses, have low inteligence and so on. Hitler would be the obvious answer, but it's always hard to be sure what he actually believed - we still don't know what religion he was, for instance.
Luckily, there are sites dedicated to calling out racists. Unluckily, they are often taken in by trolls, and of course the more intelligent/charismatic/educated racists are often filtered out.
Nevertheless:
source
Some other examples:
source
source
source
source
source
source
source
source
source
source
source
source
source
source
source
What do you mean by "people"?
Satisfying my criteria for moral consideration, I suppose. I'm not entirely sure what the details of these criteria are, but CEV would presumably extract this metric, along with the other components of my utility function.
People tend not to value killing one-of-us, where "us" can be defined very broadly or very narrowly, or anywhere inbetween. Is that one terminal value or many?