Understanding the psychology of HBD/Racial Realism/Dark Enlightenment

The Something Awful Forums > Discussion > Debate & Discussion > Understanding the psychology of HBD/Racial Realism/Dark Enlightenment

«‹›39 »

Zodium: Jun 19, 2004

not sure what you want me to elaborate on here since you didn't really elaborate on why you think they're mutually exclusive, but

'deductive' versus 'inductive' inference is the kind of dichotomy that comes from learning about inference from old cookbooks without learning how to infer. they're artifacts of contrived teaching examples getting to say "the object of study begins here and ends here because that is most convenient for the preconceived point I want to reach," which is exactly what you don't get to do in science--phenomena define their own scopes, and we don't get to redefine them, at least not if we want our inference to be valid. we just get to work out where the boundaries are and what's inside them. in science, there is only inference.

rudatron posted:

Science is a subset of Reason (It is a method of reasoning about the world - in particular, it like Reason proper rejects teleology), deductive & inductive reasoning are not mutually exclusive (inductive reasoning is the extension of deductive logic onto the area of uncertainty), and democracy and fascism do not overlap (fascism is explicitly anti-democratic, because it denies the value of human life proper and sees oppression as a desirable state, it is therefore anti-humanistic/inhuman - Democracy gives power to a majority on the basis of the equal value of human beings, therefore it is humanistic).

if you think about it, science and reason are the same thing. they're both five letter words with another letter on the end. ergo, deductive and inductive reasoning are also the same thing. QED.

:captainpop:

# ? Jan 4, 2016 14:28

Adbot: ADBOT LOVES YOU

# ? May 26, 2024 20:36

Cingulate: Oct 23, 2012; by Fluffdaddy

rudatron posted:

Science is a subset of Reason (It is a method of reasoning about the world - in particular, it like Reason proper rejects teleology), deductive & inductive reasoning are not mutually exclusive (inductive reasoning is the extension of deductive logic onto the area of uncertainty), and democracy and fascism do not overlap (fascism is explicitly anti-democratic, because it denies the value of human life proper and sees oppression as a desirable state, it is therefore anti-humanistic/inhuman - Democracy gives power to a majority on the basis of the equal value of human beings, therefore it is humanistic).

Similarly, Fascism and Stalinism were violently opposed, thus Fascism and Stalinism have nothing in common.

(Your post is stupid.)

Zodium posted:

not sure what you want me to elaborate on here since you didn't really elaborate on why you think they're mutually exclusive, but

'deductive' versus 'inductive' inference is the kind of dichotomy that comes from learning about inference from old cookbooks without learning how to infer. they're artifacts of contrived teaching examples getting to say "the object of study begins here and ends here because that is most convenient for the preconceived point I want to reach," which is exactly what you don't get to do in science--phenomena define their own scopes, and we don't get to redefine them, at least not if we want our inference to be valid. we just get to work out where the boundaries are and what's inside them. in science, there is only inference.

I don't see the point where you actually said anything. You're saying "this viewpoint is naive", but that's very vague.

I counter deduction and induction are reasonably well defined, and refer to Gelman & Shalizis "Philosophy and the practice of Bayesian statistics":

quote:

...A substantial school in the philosophy of science identifies Bayesian inference with inductive inference and even rationality as such, and seems to be strengthened by the rise and practical success of Bayesian statistics. We argue that the most successful forms of Bayesian statistics do not actually support that particular philosophy but rather accord much better with sophisticated forms of hypothetico-deductivism...

# ? Jan 4, 2016 14:47

Zodium: Jun 19, 2004

i never suggested your entirely semantical distinction wasn't well defined, I suggested it was entirely semantical

e: inappropriate

Zodium fucked around with this message at 14:52 on Jan 4, 2016

# ? Jan 4, 2016 14:50

Cingulate: Oct 23, 2012; by Fluffdaddy

Okay I don't understand your point.

# ? Jan 4, 2016 14:50

Zodium: Jun 19, 2004

come, sit on my lap and I will tell you a story about how, when two inferences love each other very much, they get under the blankets and make an estimate about the space of possible factors on which to expend their limited time/resources, then select those members with the highest plausibilities. then, they do analysis about the space of possible tests, and the space of possible interpretations, and so forth. the scope of all these analyses must be bounded by something known a priori to be constant relative to the measure, scale and tolerance of the intended inference, or the inference will not be bounded at all, and therefore meaningless. and that's how baby inferences are made.

as men of science, surely we can calmly and rationally imply that we each are naive morons with world views that rest on nothing more than fairy tales and deeply held unexamined premises without either of us having to explicitly say so. quite frankly, if you don't mind my saying so, it's a little rude. :colbert:

# ? Jan 4, 2016 15:39

Cingulate: Oct 23, 2012; by Fluffdaddy

Zodium posted:

come, sit on my lap and I will tell you a story about how, when two inferences love each other very much, they get under the blankets and make an estimate about the space of possible factors on which to expend their limited time/resources, then select those members with the highest plausibilities. then, they do analysis about the space of possible tests, and the space of possible interpretations, and so forth. the scope of all these analyses must be bounded by something known a priori to be constant relative to the measure, scale and tolerance of the intended inference, or the inference will not be bounded at all, and therefore meaningless. and that's how baby inferences are made.

And if we accept that, the question is, what is the underlying philosophy? Hypothetico-deductive, or what? Because the standard interpretation of Bayesian inference seems to be inductive.

For what it's worth I'm firmly in the frequentist camp for now because I think science is best viewed as a cumulative (iterated, repeated, constantly failing etc) process.

Zodium posted:

as men of science, surely we can calmly and rationally imply that we each are naive morons with world views that rest on nothing more than fairy tales and deeply held unexamined premises without either of us having to explicitly say so. quite frankly, if you don't mind my saying so, it's a little rude.

I'm with you here though.

# ? Jan 4, 2016 15:58

Zodium: Jun 19, 2004

Cingulate posted:

And if we accept that, the question is, what is the underlying philosophy? Hypothetico-deductive, or what? Because the standard interpretation of Bayesian inference seems to be inductive.

For what it's worth I'm firmly in the frequentist camp for now because I think science is best viewed as a cumulative (iterated, repeated, constantly failing etc) process.
I'm with you here though.

there is no the underlying philosophy in the sense I think you mean. the underlying philosophy for what? there are many schools both within "frequentism" and "Bayesianism." pure statisticians often aren't of any particular school because which is more apt depends on the problem, and they study these relationships more generally. applied statisticians tend to be more invested in a particular school because they tend to deal with particular inferential problems, rather than inference more generally.

hail Bayes and Satan, death to the Fisherian apostates and the Neymaphysites

edit: god dammit, how did we get to Bayes/Frequentism? it's like the Godwin of statistics

Zodium fucked around with this message at 16:58 on Jan 4, 2016

# ? Jan 4, 2016 16:42

Cingulate: Oct 23, 2012; by Fluffdaddy

Zodium posted:

there is no the underlying philosophy in the sense I think you mean. the underlying philosophy for what?

Yours, and those of non-statisticians, e.g. Bayes-hyping psychologists.

I'm a bit surprised you're explaining to me "theoretical statisticians can be pragmatic/agnostic and there are multiple views" after I've posted a link to a paper by theoretical statisticians talking about different views and defending their own eclectic pragmatism.

Zodium posted:

edit: god dammit, how did we get to Bayes/Frequentism? it's like the Godwin of statistics

Intentionally, on my part. I wanted some meat to the discussion and know you had an opinion.

# ? Jan 4, 2016 17:10

Unseen: Dec 23, 2006; I'll drive the tanker

Cingulate posted:

What is deductive inductive reasoning? Because I'm fairly sure they're mutually exclusive.

And yes, as I said, reason and science partially overlap. Much like democracy and fascism.

It was a typo. Deductive and inductive reasoning.

# ? Jan 4, 2016 17:58

Zodium: Jun 19, 2004

Cingulate posted:

Yours, and those of non-statisticians, e.g. Bayes-hyping psychologists.

I'm a bit surprised you're explaining to me "theoretical statisticians can be pragmatic/agnostic and there are multiple views" after I've posted a link to a paper by theoretical statisticians talking about different views and defending their own eclectic pragmatism.
Intentionally, on my part. I wanted some meat to the discussion and know you had an opinion.

are you sure you didn't do it because there was no longer any question regarding the mutual exclusivity of deductive and inductive inference? i'm not sure what you want me to say about the paper since you only brought it up as some kind of defense for why deduction/induction is somehow useful dichotomy for anything. i'm familiar with it but i'm not gonna re-read the whole thing because you made a "ACTUALLY, Websters defines ..." style reference to it :shrug:

as a practicing Bayesian, I frequently update my beliefs according to Bayes' Theorem (pbuh), and after years of banging my head against an unyielding wall, I updated my beliefs to be that Bayesianism/frequentism is another pointless part of our inferential vocabulary. if anything it seems to be hindering methodological progress, so now I only speak of unqualified inference. I still believe that a Bayes factor interpreted as the relative change in plausibility for one model against a reference model is the most apt discipline for the problems behavioral science has to solve today for the simple reason that you can count all you want, but we're not at levels of plausibility where we can be confident the events we're counting are meaningful units. got a real Aether/blind guys feeling up an elephant situation brewing here in terms of theoretical commensurability, if you know what i'm saying

I like Gelman because he makes a little progress with his pragmatic eclecticism without making methodologists seem like a bunch of dicks, but it's a hard emphasis on a little there. we mostly are dicks, so it's good that someone puts on a face. it's also been tried before, and it didn't work after a lifetime of work by one of the most celebrated statisticians of the 20th century, and it's not gonna work this time either. inference is a discipline where there's always going to be room for, well, undisciplined behavior. pragmatic eclecticism is fine for people who are genuinely fastidious and highly knowledgeable, and basically terrible for everyone else because it's fundamentally unprincipled. Gelman unfortunately seems determined to be the next Cohen, which is too bad and hopefully ends less tragically because he's a nice guy.

welp, that's my two cent

Zodium fucked around with this message at 19:35 on Jan 4, 2016

# ? Jan 4, 2016 19:27

Abner Cadaver II: Apr 21, 2009; TONIGHT!

none of this helps me understand how people who worship Reason & Science are believing in Atavistic Negroids in the 21st century

# ? Jan 4, 2016 19:35

Dapper_Swindler: Feb 14, 2012; Im glad my instant dislike in you has been validated again and again.

Bushiz posted:

new vice needs to keep the lights on and the outrage generated by sex articles written by non hetero dudes does that.

i dont really mind it just gets boring after while. i mean vice is much better then gawker when it comes to the amount of clickbate/ get mad articles.

# ? Jan 4, 2016 19:41

Zodium: Jun 19, 2004

Abner Cadaver II posted:

none of this helps me understand how people who worship Reason & Science are believing in Atavistic Negroids in the 21st century

if you don't understand how unprincipled cherry picking of evidence and cargo cult statistics contribute to both worship of Reason & Science and beliefs in Atavistic Negroids then you may already be infected. there are good treatments but I am going to prescribe you Jaynes' Probability Theory, take one chapter a week and call me when the prescription is out.

there are stronger treatments if we have to go there, but I don't think we should put you on Fisher or Jeffreys just yet until we see what happens. let's just hope it's not multi-resistant and it's not too late for you already, good day and god bless america to you.

e: man look at science, see science man do thing. science man do science, so conclusion must valid. man also do thing, so conclusion must also valid. but not all science man also statistic man. so sometime thing not valid, and conclusion can not valid.

Zodium fucked around with this message at 19:57 on Jan 4, 2016

# ? Jan 4, 2016 19:42

Pope Guilty: Nov 6, 2006; The human animal is a beautiful and terrible creature, capable of limitless compassion and unfathomable cruelty.

Zodium posted:

are you sure you didn't do it because there was no longer any question regarding the mutual exclusivity of deductive and inductive inference? i'm not sure what you want me to say about the paper since you only brought it up as some kind of defense for why deduction/induction is somehow useful dichotomy for anything. i'm familiar with it but i'm not gonna re-read the whole thing because you made a "ACTUALLY, Websters defines ..." style reference to it

as a practicing Bayesian, I frequently update my beliefs according to Bayes' Theorem (pbuh), and after years of banging my head against an unyielding wall, I updated my beliefs to be that Bayesianism/frequentism is another pointless part of our inferential vocabulary. if anything it seems to be hindering methodological progress, so now I only speak of unqualified inference. I still believe that a Bayes factor interpreted as the relative change in plausibility for one model against a reference model is the most apt discipline for the problems behavioral science has to solve today for the simple reason that you can count all you want, but we're not at levels of plausibility where we can be confident the events we're counting are meaningful units. got a real Aether/blind guys feeling up an elephant situation brewing here in terms of theoretical commensurability, if you know what i'm saying

I like Gelman because he makes a little progress with his pragmatic eclecticism without making methodologists seem like a bunch of dicks, but it's a hard emphasis on a little there. we mostly are dicks, so it's good that someone puts on a face. it's also been tried before, and it didn't work after a lifetime of work by one of the most celebrated statisticians of the 20th century, and it's not gonna work this time either. inference is a discipline where there's always going to be room for, well, undisciplined behavior. pragmatic eclecticism is fine for people who are genuinely fastidious and highly knowledgeable, and basically terrible for everyone else because it's fundamentally unprincipled. Gelman unfortunately seems determined to be the next Cohen, which is too bad and hopefully ends less tragically because he's a nice guy.

welp, that's my two cent

I can't tell if this is a c/p from LessWrong or what.

# ? Jan 4, 2016 19:43

Zodium: Jun 19, 2004

hey

there are some things you don't joke about buddy

# ? Jan 4, 2016 19:50

HappyHippo: Nov 19, 2003; Do you have an Air Miles Card?

Abner Cadaver II posted:

none of this helps me understand how people who worship Reason & Science are believing in Atavistic Negroids in the 21st century

People will claim whatever authority has currency in support of their position. Other popular choices are "god" and "tradition."

# ? Jan 4, 2016 20:05

Cingulate: Oct 23, 2012; by Fluffdaddy

Zodium posted:

are you sure you didn't do it because there was no longer any question regarding the mutual exclusivity of deductive and inductive inference? i'm not sure what you want me to say about the paper since you only brought it up as some kind of defense for why deduction/induction is somehow useful dichotomy for anything. i'm familiar with it but i'm not gonna re-read the whole thing because you made a "ACTUALLY, Websters defines ..." style reference to it

as a practicing Bayesian, I frequently update my beliefs according to Bayes' Theorem (pbuh), and after years of banging my head against an unyielding wall, I updated my beliefs to be that Bayesianism/frequentism is another pointless part of our inferential vocabulary. if anything it seems to be hindering methodological progress, so now I only speak of unqualified inference. I still believe that a Bayes factor interpreted as the relative change in plausibility for one model against a reference model is the most apt discipline for the problems behavioral science has to solve today for the simple reason that you can count all you want, but we're not at levels of plausibility where we can be confident the events we're counting are meaningful units. got a real Aether/blind guys feeling up an elephant situation brewing here in terms of theoretical commensurability, if you know what i'm saying

I like Gelman because he makes a little progress with his pragmatic eclecticism without making methodologists seem like a bunch of dicks, but it's a hard emphasis on a little there. we mostly are dicks, so it's good that someone puts on a face. it's also been tried before, and it didn't work after a lifetime of work by one of the most celebrated statisticians of the 20th century, and it's not gonna work this time either. inference is a discipline where there's always going to be room for, well, undisciplined behavior. pragmatic eclecticism is fine for people who are genuinely fastidious and highly knowledgeable, and basically terrible for everyone else because it's fundamentally unprincipled. Gelman unfortunately seems determined to be the next Cohen, which is too bad and hopefully ends less tragically because he's a nice guy.

welp, that's my two cent

I'm a bit lost here. My point was, there is a clear hypothetico-deductive, agnostic-pragmatic position, and it's a cool and coherent position, which is incompatible with the idea that inductive vs. deductive is pointless.

I'm partially with you in that in most contexts, the Bayes vs. Frequentism thing is much overhyped (almost exclusively by Bayesians!). But deductive, falsificationist science, vs. the alternative we all too commonly do, is not.

Edit: okay deleted a part of this because we probably don't want this discussion again

# ? Jan 4, 2016 20:26

Zodium: Jun 19, 2004

Cingulate posted:

I'm a bit lost here. My point was, there is a clear hypothetico-deductive, agnostic-pragmatic position, and it's a cool and coherent position, which is incompatible with the idea that inductive vs. deductive is pointless.

I'm partially with you in that in most contexts, the Bayes vs. Frequentism thing is much overhyped (almost exclusively by Bayesians!). But deductive, falsificationist science, vs. the alternative we all too commonly do, is not.

Edit: okay deleted a part of this because we probably don't want this discussion again

perhaps you could explain what this "clear hypothetico-deductive, agnostic-pragmatic" position is in your own words, and why it entails making a distinction between inductive versus deductive inference. the last time you ventured an actual position was when you said that you are "firmly in the frequentist camp" because you "think science is best viewed as a cumulative (iterated, repeated, constantly failing etc) process."

but this doesn't necessarily or even preferentially lead to forms of frequentism, it could just as well be used in support of forms of Bayesianism. for example, I think we should compute Bayes factors to estimate the most plausible parameters, and compute metaprobabilistic distributions in order to select the optimal next step in a principled manner because I think science is best viewed as a cumulative (iterated, repeated, constantly failing etc) process, and we need to leverage long-run efficiencies. I think you're inadvertently confusing things with other things because they coincidentally share some terms and other cosmetic similarities, but I have a remark and a question about this whole clear hypothetico-deductive, agnostic-pragmatic position thing:

* I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a statement. :babbage:
* How are you both 'firmly' in the frequentist camp and in the clear hypothetico-deductive, agnostic-pragmatic camp? what exactly are you agnostic-pragmatic towards?

Zodium fucked around with this message at 21:24 on Jan 4, 2016

# ? Jan 4, 2016 21:20

divabot: Jun 17, 2015; A polite little mouse!

Dapper_Swindler posted:

honestly, devils advocate time. 4chan has gotten a better after 8chan came into being because most of the nutjobs/creeps moved there. plus 4chan has /histories and humanities/ board which isnt to bad, and the mods keep it from being /pol/ jr. yeah /pol/ and /b/ still exist, but i don't go on /pol/ so it doesnt matter.

I must say, it was nice to be able to say about Gamergate "too odious for 4chan".

Jack Gladney posted:

Gavin McInnes is definitely on the periphery of the dork enlightenment, though he has zero influence on Vice today. It might be more "punk rock" contrarian than leftist, though.

Ehh, he's a Taki's Mag racist. They aren't really NRx, in that they can write concisely and have senses of humour (at least somewhat). You'll see NRx in the comments on Taki's but never in the bylines, because oh my loving God they're the worst writers ever.

Zodium posted:

perhaps you could explain

Or PERHAPS NOT. Cingulate shat up the other thread with enough multi-page digressions, please don't encourage him here too.

# ? Jan 4, 2016 23:01

The Vosgian Beast: Aug 13, 2011; Business is slow

HappyHippo posted:

People will claim whatever authority has currency in support of their position. Other popular choices are "god" and "tradition."

I like that they literally reconstructed divine right of kings/ma'at/mandate of heaven as "Gnon".

divabot posted:

Or PERHAPS NOT. Cingulate shat up the other thread with enough multi-page digressions, please don't encourage him here too.

It will not stop, because goons are Sideshow Bob and problem posters are the rakes.

# ? Jan 5, 2016 00:42

Dapper_Swindler: Feb 14, 2012; Im glad my instant dislike in you has been validated again and again.

divabot posted:

I must say, it was nice to be able to say about Gamergate "too odious for 4chan".

Ehh, he's a Taki's Mag racist. They aren't really NRx, in that they can write concisely and have senses of humour (at least somewhat). You'll see NRx in the comments on Taki's but never in the bylines, because oh my loving God they're the worst writers ever.

Or PERHAPS NOT. Cingulate shat up the other thread with enough multi-page digressions, please don't encourage him here too.

its not just GG poo poo, that mostly dead anyway. its more creepy sex nuts and super /pol/ nuts.

# ? Jan 5, 2016 01:01

Maoist Pussy: Feb 12, 2014; by Lowtax

rudatron posted:

(fascism is explicitly anti-democratic, because it denies the value of human life proper and sees oppression as a desirable state, it is therefore anti-humanistic/inhuman - Democracy gives power to a majority on the basis of the equal value of human beings, therefore it is humanistic).

Actually, fascism elevates human life; that is its purpose. The power of a democratic majority can vote to invoke any aspect of any political system- liberal, fascist, communist -all of which do necessarily overlap.

# ? Jan 5, 2016 03:13

Maoist Pussy: Feb 12, 2014; by Lowtax

Abner Cadaver II posted:

none of this helps me understand how people who worship Reason & Science are believing in Atavistic Negroids in the 21st century

If you are an introverted nerd weirdo who feels emasculated by LeBron James, and you are reaching for something that indicates you are automatically superior to huge athletic dudes, you are probably going to land on Superior But Non-Specific Brain Ability.

# ? Jan 5, 2016 03:18

Peztopiary: Mar 16, 2009; by exmarx

Abner Cadaver II posted:

none of this helps me understand how people who worship Reason & Science are believing in Atavistic Negroids in the 21st century

their prior beliefs (which conveniently enough they never bother to examine) lead to Bayesian (they don't actually know what this means, but they're cargo culting hard enough that nobody calls bullshit) certainty. they're ignorant louts who've learned to parrot technobabble. not even wrong, really.

Edit: Not Even Wrong would be a good name for a LW parody.

# ? Jan 5, 2016 07:06

Darth Walrus: Feb 13, 2012

Maoist Pussy posted:

Actually, fascism elevates human life; that is its purpose. The power of a democratic majority can vote to invoke any aspect of any political system- liberal, fascist, communist -all of which do necessarily overlap.

Wait, what? How does this fit with the death-cult elements you see so often in fascist propaganda (see also, the highly common happy ending in Nazi movies of a beautiful young man dying a heroic death)?

Where the gently caress are you getting your definition of fascism from?

# ? Jan 5, 2016 08:40

Cerebral Bore: Apr 21, 2010; Fun Shoe

Darth Walrus posted:

Wait, what? How does this fit with the death-cult elements you see so often in fascist propaganda (see also, the highly common happy ending in Nazi movies of a beautiful young man dying a heroic death)?

Where the gently caress are you getting your definition of fascism from?

Well, first you redefine "elevates human life" into dying a heroic death and then you apply a thin veil of word salad onto your death cult propaganda.

# ? Jan 5, 2016 11:30

Zodium: Jun 19, 2004

divabot posted:

Or PERHAPS NOT. Cingulate shat up the other thread with enough multi-page digressions, please don't encourage him here too.

it's not a digression. if you want to understand the psychology of Dark Enlightenment types, then you must understand the inferential errors people make in practice when they engage with modern science, and especially fields with high uncertainty like behavioral science. DE is just an iceberg example of errors you yourself (or people you trust to be right) in all likelihood make constantly about all sorts of things without realizing, and their errors begin with lax behavioral/medical science methodologies that are only 'safe' if you observe all the unwritten rules.

think of medical/behavioral methodology as a badly written computer program that works okay as long as you use it in exactly the awkward way the designer says it has to be used, for only exactly the things they extensively pre-tested. if you deviate in any way, which it's almost impossible not to because of how awkward and overdesigned it is, the program will invisibly corrupt every file on your hard disk, gradually replacing words in all your text files with swears, swap all your images with Goatse, and send your bank information to random entries on a list of Russian hackers. somehow, however, this program is still highly respected, so people want to get in on that, copy it and make their own terrible open source versions that are, somehow, even worse and even more exploitable. that's the dark enlightenment.

e: if you think DE is bad, wait till open data initiatives really start taking off and mass idiocy gets a chance to pore over the data for spurious relationships. give it ten years, soft science is about to drown what little influence we have left in a torrent of citizen science bullshit. DE is just the canary in the coal mine.

Zodium fucked around with this message at 13:03 on Jan 5, 2016

# ? Jan 5, 2016 12:50

Cingulate: Oct 23, 2012; by Fluffdaddy

I think we should try and turn this around to actually connect with the topic though Zodium.

Zodium posted:

perhaps you could explain what this "clear hypothetico-deductive, agnostic-pragmatic" position is in your own words, and why it entails making a distinction between inductive versus deductive inference. the last time you ventured an actual position was when you said that you are "firmly in the frequentist camp" because you "think science is best viewed as a cumulative (iterated, repeated, constantly failing etc) process."

but this doesn't necessarily or even preferentially lead to forms of frequentism, it could just as well be used in support of forms of Bayesianism. for example, I think we should compute Bayes factors to estimate the most plausible parameters, and compute metaprobabilistic distributions in order to select the optimal next step in a principled manner because I think science is best viewed as a cumulative (iterated, repeated, constantly failing etc) process, and we need to leverage long-run efficiencies. I think you're inadvertently confusing things with other things because they coincidentally share some terms and other cosmetic similarities, but I have a remark and a question about this whole clear hypothetico-deductive, agnostic-pragmatic position thing:

* I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a statement. :babbage:
* How are you both 'firmly' in the frequentist camp and in the clear hypothetico-deductive, agnostic-pragmatic camp? what exactly are you agnostic-pragmatic towards?

Well what do you want - long-run coverage, or Bayes? Because if the answer is "both", the onus of explanation is on you.

Philosophically, I'm convinced that frequentist probability is more important than Bayesian probability, but pragmatically, I assume Bayesian methods are in many contexts the best tool - at least in those situations where priors make coefficients identifiable! And all of ML, which really doesn't care about frequentist probability, or, actually, usually, probability. (I can go into more detail here if that's unclear.)

My best bet at bringing this back to the main question is that at the very least, people need to understand science is about uncertainty, not certainty. Inferential statistics are about admitting you don't know. p values/Bayes factors are not telling you you're right, they're telling you how careful you should be.

And then, whenever you're building up a political ideology or concrete plan, or whenever you're making judgements about e.g. groups, you need to understand that when you're building on data, you're building on uncertainty, not certainty.

Zodium posted:

e: if you think DE is bad, wait till open data initiatives really start taking off and mass idiocy gets a chance to pore over the data for spurious relationships. give it ten years, soft science is about to drown what little influence we have left in a torrent of citizen science bullshit. DE is just the canary in the coal mine.

The only influence social science has right now is giving people a misleading graph and a science journalism headline that says the opposite of the real research to back up their near-unshakeable preconceptions.

# ? Jan 5, 2016 14:13

rudatron: May 31, 2011; by Fluffdaddy

Cingulate posted:

The only influence social science has right now is giving people a misleading graph and a science journalism headline that says the opposite of the real research to back up their near-unshakeable preconceptions.

The point I think is that the push on 'big data' is going to lead to a lot of spurious poo poo, on account of the people pushing it not adequately applying degrees of freedom, which can actually manifest itself in very unusual ways.

Maoist Pussy posted:

Actually, fascism elevates human life; that is its purpose. The power of a democratic majority can vote to invoke any aspect of any political system- liberal, fascist, communist -all of which do necessarily overlap.

Yes yes, you have a fascism fetish, I get it, doesn't change that what I said was true. The ethical basis for democracy lies in the inherent value of people and their right to choose their own destiny (government by the people, for the people), which is totally at odds with fascist ideology that values violence above all (government by the strong, for the strong). It only values a democratic system in so far as the majority supports fascist thought, if they don't then that just means they're wrong. That is, fascism is willing to use democracy, but it has no respect for it, and is in fact at odds with it.

tl;dr democracy is more than just 'elections'.

rudatron fucked around with this message at 15:09 on Jan 5, 2016

# ? Jan 5, 2016 15:07

Zodium: Jun 19, 2004

Cingulate posted:

I think we should try and turn this around to actually connect with the topic though Zodium.
Well what do you want - long-run coverage, or Bayes? Because if the answer is "both", the onus of explanation is on you.

Philosophically, I'm convinced that frequentist probability is more important than Bayesian probability, but pragmatically, I assume Bayesian methods are in many contexts the best tool - at least in those situations where priors make coefficients identifiable! And all of ML, which really doesn't care about frequentist probability, or, actually, usually, probability. (I can go into more detail here if that's unclear.)

My best bet at bringing this back to the main question is that at the very least, people need to understand science is about uncertainty, not certainty. Inferential statistics are about admitting you don't know. p values/Bayes factors are not telling you you're right, they're telling you how careful you should be.

And then, whenever you're building up a political ideology or concrete plan, or whenever you're making judgements about e.g. groups, you need to understand that when you're building on data, you're building on uncertainty, not certainty.
The only influence social science has right now is giving people a misleading graph and a science journalism headline that says the opposite of the real research to back up their near-unshakeable preconceptions.

hello my friend, i see you have inadvertently quoted my post without replying to it. let me helpfully quote the relevant sections for your convenience.

Zodium posted:

perhaps you could explain what this "clear hypothetico-deductive, agnostic-pragmatic" position is in your own words, and why it entails making a distinction between inductive versus deductive inference. the last time you ventured an actual position was when you said that you are "firmly in the frequentist camp" because you "think science is best viewed as a cumulative (iterated, repeated, constantly failing etc) process."

* How are you both 'firmly' in the frequentist camp and in the clear hypothetico-deductive, agnostic-pragmatic camp? what exactly are you agnostic-pragmatic towards?

i can understand wanting to escape the obvious dead end you've walked yourself into here. it's fine if this is as much as you're willing to invest, but there's no backsies if you want this line of discussion to continue. if you don't, that's cool too.

rudatron posted:

The point I think is that the push on 'big data' is going to lead to a lot of spurious poo poo, on account of the people pushing it not adequately applying degrees of freedom, which can actually manifest itself in very unusual ways.

it's this, also it's actually not that different from what's happening now, only the conclusions reached will be much crazier because the sanity checks you learn from actual study don't come with the autodidact package

# ? Jan 5, 2016 15:39

Cingulate: Oct 23, 2012; by Fluffdaddy

rudatron posted:

The point I think is that the push on 'big data' is going to lead to a lot of spurious poo poo, on account of the people pushing it not adequately applying degrees of freedom, which can actually manifest itself in very unusual ways.

More spurious poo poo than what we already have?

Zodium posted:

hello my friend, i see you have inadvertently quoted my post without replying to it. let me helpfully quote the relevant sections for your convenience.

i can understand wanting to escape the obvious dead end you've walked yourself into here. it's fine if this is as much as you're willing to invest, but there's no backsies if you want this line of discussion to continue. if you don't, that's cool too.

Are you sure that level of smug is justified?
I've asked you a question because I'm not sure how to talk to you about this before you explain yourself.

But on the quick, I'm pragmatic about what tools to use (in many situations, you can't model something with frequentist tools, in others not with Bayesian tools, and a lot of the interesting stuff is increasingly hybrid or agnostic - and many Bayesian tools have decent Frequentist properties et vice versa) and I believe you can get decent stuff done within any framework, but philosophically, I think science as a project (instead of a specific act of conducting One Piece of Science) should be more concerned with its long-term properties (= Frequentist concerns) than with satisfying certain desires for results that are easily interpretable from a Bayesian view on probability.

Edit: and I'm fully with Popper on the question of what science can and should be doing - that is, falsifying hypotheses, not conducting induction wrt. the probability of hypotheses.

Cingulate fucked around with this message at 16:50 on Jan 5, 2016

# ? Jan 5, 2016 16:48

Peztopiary: Mar 16, 2009; by exmarx

I'm interested in what you guys are talking about, but it'd be more interesting in it's own thread.

# ? Jan 5, 2016 17:12

Zodium: Jun 19, 2004

i know, i've been telling myself to write it for years. i've fallen into a comfy MO of dropping into other threads for a couple of pages whenever inferential errors come up and tbh I'm not seeing myself break out of it.

Cingulate, PM me if you want to continue this, but please address my concerns if so. (please also note that, as per my previous publication re: Kolmogorov complexity based terminological equivalencies, address is different from allude)

# ? Jan 5, 2016 17:46

Munin: Nov 14, 2004

Darth Walrus posted:

Wait, what? How does this fit with the death-cult elements you see so often in fascist propaganda (see also, the highly common happy ending in Nazi movies of a beautiful young man dying a heroic death)?

Where the gently caress are you getting your definition of fascism from?

The "beautiful young man dying a heroic death" is part of a more general German cultural thing and that motif is most closely associated with "Sturm und Drang" in particular. "Sturm und Drang" was a reaction against the rationalism of the enlightenment which several young male writers in particular felt lacked blood and passion. Goethe's "The Sorrows of Young Werther" was the most successful work of the movement and became a breakthrough hit across Europe and led to young men across the continent topping themselves in dramatic ways or adopt the dress and style of the hero of the novel. Goethe himself ended up slightly pissed at the fact that despite all the other great works he created he often was only popularly known as that guy who wrote Werther.

The reason why the nazis glommed onto it in particular is the manly man nature of it and also that it was one of the first pan-Germanic cultural movements which achieved prominence across Europe.

# ? Jan 5, 2016 17:48

divabot: Jun 17, 2015; A polite little mouse!

Zodium posted:

e: if you think DE is bad, wait till open data initiatives really start taking off and mass idiocy gets a chance to pore over the data for spurious relationships. give it ten years, soft science is about to drown what little influence we have left in a torrent of citizen science bullshit. DE is just the canary in the coal mine.

ehh, I'm a skeptic, dismantling that sort of BS pseudoscience is what I do already for fun. The approach to doing so would be the same as the counterattack on everyone who starts with their conclusion and shores it up with whatever spurious BS they can find to prop it up.

I concur that there will be an onslaught of this BS as you describe, and it'll be pretty terrible. But it won't actually be worse than what we routinely deal with already; skeptics will start addressing the bad thinking in question in proportion to its public notice, and e.g. a well-written RationalWiki article (for all RW's hideous gaping flaws) does get picked up by the press as a source in practice. (At which point they're sourcewashed enough to go into Wikipedia, and the initial heavy lifting's done. Then it's mostly the slog of publicity.)

Getting back to the DE case, it isn't in fact shored up at all well by science or history, because in practice everything they write is pronouncement ex culo. Moldbug in particular is a bloody terrible historian. This is clearer in his earlier pieces, which were quite short! He used to write posts of about 1000-2000 words, with a clear thesis and falsifiable claims. But his commenters kept rudely pointing out his bullshit sources and ludicrous logical leaps, so he shifted to his more famous style in which he makes an amazing claim, spends 1000-2000 words redefining English, then shows that his sentence is not provably wrong if you use his special meanings of words. The bad logic and bad sources are as bad as ever, but nobody who doesn't already ~~hate the ni~~ agree with his broad outlines has the patience to check any more. Job done!

# ? Jan 5, 2016 20:24

Dapper_Swindler: Feb 14, 2012; Im glad my instant dislike in you has been validated again and again.

Munin posted:

The "beautiful young man dying a heroic death" is part of a more general German cultural thing and that motif is most closely associated with "Sturm und Drang" in particular. "Sturm und Drang" was a reaction against the rationalism of the enlightenment which several young male writers in particular felt lacked blood and passion. Goethe's "The Sorrows of Young Werther" was the most successful work of the movement and became a breakthrough hit across Europe and led to young men across the continent topping themselves in dramatic ways or adopt the dress and style of the hero of the novel. Goethe himself ended up slightly pissed at the fact that despite all the other great works he created he often was only popularly known as that guy who wrote Werther.

The reason why the nazis glommed onto it in particular is the manly man nature of it and also that it was one of the first pan-Germanic cultural movements which achieved prominence across Europe.

but that trope has been around since humans, it was big in victorian times. (much ealier too obviously) example being, Young james dies for queen and country against the uncivilized hoard from _________ which will be brought to heel for queen and country and be made loyal subjects. alot of the other countries bought into that poo poo too. the Victorian ideal of the martyr of war poo poo died during world war 1. it still floats around today, in various flavors.

# ? Jan 5, 2016 21:40

Woolie Wool: Jun 2, 2006

The Vosgian Beast posted:

I like that they literally reconstructed divine right of kings/ma'at/mandate of heaven as "Gnon".

It will not stop, because goons are Sideshow Bob and problem posters are the rakes.

Is reinventing old ideas from principles, giving them stupid names, and passing them off as his own work Moldbug's schtick or something?

# ? Jan 5, 2016 23:39

Darth Walrus: Feb 13, 2012

Munin posted:

The "beautiful young man dying a heroic death" is part of a more general German cultural thing and that motif is most closely associated with "Sturm und Drang" in particular. "Sturm und Drang" was a reaction against the rationalism of the enlightenment which several young male writers in particular felt lacked blood and passion. Goethe's "The Sorrows of Young Werther" was the most successful work of the movement and became a breakthrough hit across Europe and led to young men across the continent topping themselves in dramatic ways or adopt the dress and style of the hero of the novel. Goethe himself ended up slightly pissed at the fact that despite all the other great works he created he often was only popularly known as that guy who wrote Werther.

The reason why the nazis glommed onto it in particular is the manly man nature of it and also that it was one of the first pan-Germanic cultural movements which achieved prominence across Europe.

It was a thing for the other fascists as well, though. The Italian Futurist Manifesto was pretty much 'live fast, die young, and leave a good-looking corpse', and specifically praised 'the beautiful ideas worth dying for', and the Spanish Falangists infamously had the battle-cry of 'Viva la muerte!'. I'd say the pre-existing death-cultist elements in German culture, if they existed, were simply a helpful boon for fascism's popularity, not something that shaped the ideology as it was.

# ? Jan 5, 2016 23:58

The Vosgian Beast: Aug 13, 2011; Business is slow

Woolie Wool posted:

Is reinventing old ideas from principles, giving them stupid names, and passing them off as his own work Moldbug's schtick or something?

There's a reason he caught on in the Less Wrong sphere.

# ? Jan 6, 2016 01:26

Adbot: ADBOT LOVES YOU

# ? May 26, 2024 20:36

InediblePenguin: Sep 27, 2004; I'm strong. And a giant penguin. Please don't eat me. No, really. Don't try.

Dapper_Swindler posted:

but that trope has been around since humans, it was big in victorian times. (much ealier too obviously) example being, Young james dies for queen and country against the uncivilized hoard from _________ which will be brought to heel for queen and country and be made loyal subjects. alot of the other countries bought into that poo poo too. the Victorian ideal of the martyr of war poo poo died during world war 1. it still floats around today, in various flavors.

"but it was big in Victorian times and WWI" is in no way any kind of a counter to the argument about the Germanic Sturm und Drang movement. The Sorrows of Young Werther was published in 1774 and therefore predates your examples quite neatly

# ? Jan 6, 2016 01:27

The Something Awful Forums > Discussion > Debate & Discussion > Understanding the psychology of HBD/Racial Realism/Dark Enlightenment

«‹›39 »