I think I can understand what Tyler is getting at when he accuses Robin of a penchant for "logical atomism". In the present context, I interpret Tyler’s claim as a plea for greater appreciation of the messiness and ambiguity of evidence when looked at closely, more toleration for different modes of consideration, and less eagerness to embrace a few "stylized facts" and use them to draw bold, sweeping, shocking conclusions; and less faith in the fact/value distinction.
The book “Bias is beautiful” by Tolon demonstrates that there is one basic reality in our biology which helps us understand sociological reactions; our belief systems and even our language. We understand and react with the basic format of self/ non-self to our world and to each other. Every single cell being reacts and tries to survive with this mechanism. This again is neither good, nor bad, but a biological reality. It is easy to see the changes in the language by looking at 20 year old newspapers. The constant change is obvious if you listen to the language different generations speak within the same town.
Everyone will accept the fact that she/he looks at “others” living in a different district with a certain “natural” bias. And as no one will be able to react to everyone else and get to know “them” he will have natural prejudices. So no two individuals will be free of bias as no two are alike. People who believe to be stronger or are stronger will exert pressure on the others in contest for better hunting grounds etc. But also because of “cultural” differences and try to put them in pecking order.
As Dr Tolon demonstrates in his book "Bias is beautiful" : Bias is the basic form of differentiation for all living beings from the immunological level upwards resulting from the differentiation self/non-self.This basic standard differentiation which also has the fundemental deviation= bias (being neither good nor bad- just a basic biological fact) in all intellectual- cognitive understanding of all individuals as opposed to common sense which can be articulated by a group of individuals.
Robin, the "average face claim" might be too strong: it is probably not the case that the most beautiful face is the most average face. And which faces you superimpose affects the beauty of the resulting picture, so that you get a better result by superimposing beautiful faces than by superimposing ugly faces. Nevertheless, I think it is true that when superimposing a large number of random faces (of a given age and gender) you get a beautiful face. Also, I think that averaging faces within a given beauty category tends to enhance beauty, albeit perhaps partly by smoothing out skin imperfections.
Tyler's point is at some level compelling, but we start to enter into Hofstadterian territory, because it seems clear that Robin's style is to attempt to personally overcoming his own biases, raterh than leave it up to a larger social process.
Eliezer, I'm not sure that differences in cognitive style can be cached out purely as differences in anticipated experience (even setting aside general issues with verificationism). Cognitive style might include things like how we allocate our attention; what conceptual frameworks we adopt to represent and navigate a given set of facts; the cognitive strategies we use to form our opinions; whether we prefer to get a small thing exactly right or many big things approximately right; how our abstract thought is linked or dissociated from our emotional lives; etc.
I was also meaning to put the question more generally, not just cognitive style but other personality traits. When you superimpose and average out a large number pictures of different faces, you get a picture of a beautiful face. Does the equivalent hold with respect to personality?
(Aaron, there was a formatting glitch which has now been fixed.)
I still don't understand what "logical atomism" and their disagreement is. This possibly pegs me as a logical atomist, and I am not sure why I would want to dispute it. If the capitalists, tories, whigs and queers can all embrace terms invented by their opponents, why not logical atomists?
Tyler and I seem to largely agree in the above two comments, except that Tyler seems to assume that one cannot personally accept the post-filter social consensus, but must instead believe the output of their personal analysis style. Even if most people do suffer this bias, it is possible to do better.
This is of course a very insightful post. But I don't want to make Robin more like I am, even if he were to become more of a truth-seeker on average. I believe Robin is (roughly) optimizing social welfare with his current intellectual strategy. The social gains from his original ideas (which are numerous, with gains in mega-terms from each) far outweigh the social losses from (what I perceive as) various inaccuracies in Robin's thought. By the way, this refers back to my #4. Many people, including Robin, are more productive when they are biased. I don't want Robin to overcome his biases, it would make him less original and less interesting. It's the extremes that matter -- social processes can discard mistaken ideas and adopt good ones -- so we should favor temperaments that perform well at the extremes. Oddly the greater the scope of betting markets, the more effective the social filters on ideas, and thus the higher the social return to intellectual bias!
When you use your cognitive style to consider some question, your style will influence your answer. But then when you consider that other people with other styles come up with different answers, you should combine those answers together into a net conclusion. In that combination process, you shouldn't give your style priority just because it was your style.
As Conchis notes that Page notes, we are better off having different styles and sharing our results, instead of all having the same style, if we can be fair about combining the results of our differing styles.
So I don't mind having a different style from Tyler, as long as that isn't the cause of (factual) disagreements between us.
"I confess myself frankly suspicious of any so-called "disagreement" that can't be cashed out to a difference of anticipated experiences."
So, if I make a really complicated geocentric model that explains all observations, and someone else makes a really simple heliocentric model that explains all observations, and our models don't make any different predictions about anything we can expect to experience in our pre-space shuttle lives (we don't think a space shuttle is even possible), then we don't disagree?
I mean, I'm just parroting my intro to philosophy of science class (cliché example, I know). But I think everyone puts some value on the simplicity/coherence of a theory, as well as its actual predictions. And the minute you put some value on something other than a theory's predictions, you can easily set up a disagreement where nobody disagrees on any observable fact, but there's a real, heated substantive disagreement.
I've been wondering for a while how the this debate between Robin and Tyler fits in with Scott Page's work on the value of diversity in cognitive strategies (and also whether the insights gleaned from his work has implications for the overcoming bias project more generally). In this case, it appears to suggest an additional possibility that none of Nick's options quite seem to cover: taken individually, which of Robin's and Tyler's cognitive strategies performs better will probably vary from circumstance to circumstance, but the best results may arise from combining them.
To be honest, I'm not sure how exactly this would play out in concrete terms. But just thought I'd throw it out there for people smarter than me to pick up on if they feel like it.
The book “Bias is beautiful” by Tolon demonstrates that there is one basic reality in our biology which helps us understand sociological reactions; our belief systems and even our language. We understand and react with the basic format of self/ non-self to our world and to each other. Every single cell being reacts and tries to survive with this mechanism. This again is neither good, nor bad, but a biological reality. It is easy to see the changes in the language by looking at 20 year old newspapers. The constant change is obvious if you listen to the language different generations speak within the same town.
Everyone will accept the fact that she/he looks at “others” living in a different district with a certain “natural” bias. And as no one will be able to react to everyone else and get to know “them” he will have natural prejudices. So no two individuals will be free of bias as no two are alike. People who believe to be stronger or are stronger will exert pressure on the others in contest for better hunting grounds etc. But also because of “cultural” differences and try to put them in pecking order.
As Dr Tolon demonstrates in his book "Bias is beautiful" : Bias is the basic form of differentiation for all living beings from the immunological level upwards resulting from the differentiation self/non-self.This basic standard differentiation which also has the fundemental deviation= bias (being neither good nor bad- just a basic biological fact) in all intellectual- cognitive understanding of all individuals as opposed to common sense which can be articulated by a group of individuals.
Perhaps Robin was confusing the claims about the attractiveness of symmetrical and average faces.
HA:http://www.overcomingbias.c...
Robin, the "average face claim" might be too strong: it is probably not the case that the most beautiful face is the most average face. And which faces you superimpose affects the beauty of the resulting picture, so that you get a better result by superimposing beautiful faces than by superimposing ugly faces. Nevertheless, I think it is true that when superimposing a large number of random faces (of a given age and gender) you get a beautiful face. Also, I think that averaging faces within a given beauty category tends to enhance beauty, albeit perhaps partly by smoothing out skin imperfections.
(Check out this demo and see for yourself: http://www.faceresearch.org...
Does something similar hold true of (some aspects of) personality?
Tyler's point is at some level compelling, but we start to enter into Hofstadterian territory, because it seems clear that Robin's style is to attempt to personally overcoming his own biases, raterh than leave it up to a larger social process.
Hopefully, maybe I heard wrong.
Robin,That assertion isn't currently acknowledged on wikipedia:
http://en.wikipedia.org/wik...
Nick, I've heard that average face claim is a myth.
Eliezer, I'm not sure that differences in cognitive style can be cached out purely as differences in anticipated experience (even setting aside general issues with verificationism). Cognitive style might include things like how we allocate our attention; what conceptual frameworks we adopt to represent and navigate a given set of facts; the cognitive strategies we use to form our opinions; whether we prefer to get a small thing exactly right or many big things approximately right; how our abstract thought is linked or dissociated from our emotional lives; etc.
I was also meaning to put the question more generally, not just cognitive style but other personality traits. When you superimpose and average out a large number pictures of different faces, you get a picture of a beautiful face. Does the equivalent hold with respect to personality?
(Aaron, there was a formatting glitch which has now been fixed.)
I still don't understand what "logical atomism" and their disagreement is. This possibly pegs me as a logical atomist, and I am not sure why I would want to dispute it. If the capitalists, tories, whigs and queers can all embrace terms invented by their opponents, why not logical atomists?
Tyler and I seem to largely agree in the above two comments, except that Tyler seems to assume that one cannot personally accept the post-filter social consensus, but must instead believe the output of their personal analysis style. Even if most people do suffer this bias, it is possible to do better.
This is of course a very insightful post. But I don't want to make Robin more like I am, even if he were to become more of a truth-seeker on average. I believe Robin is (roughly) optimizing social welfare with his current intellectual strategy. The social gains from his original ideas (which are numerous, with gains in mega-terms from each) far outweigh the social losses from (what I perceive as) various inaccuracies in Robin's thought. By the way, this refers back to my #4. Many people, including Robin, are more productive when they are biased. I don't want Robin to overcome his biases, it would make him less original and less interesting. It's the extremes that matter -- social processes can discard mistaken ideas and adopt good ones -- so we should favor temperaments that perform well at the extremes. Oddly the greater the scope of betting markets, the more effective the social filters on ideas, and thus the higher the social return to intellectual bias!
When you use your cognitive style to consider some question, your style will influence your answer. But then when you consider that other people with other styles come up with different answers, you should combine those answers together into a net conclusion. In that combination process, you shouldn't give your style priority just because it was your style.
As Conchis notes that Page notes, we are better off having different styles and sharing our results, instead of all having the same style, if we can be fair about combining the results of our differing styles.
So I don't mind having a different style from Tyler, as long as that isn't the cause of (factual) disagreements between us.
"I confess myself frankly suspicious of any so-called "disagreement" that can't be cashed out to a difference of anticipated experiences."
So, if I make a really complicated geocentric model that explains all observations, and someone else makes a really simple heliocentric model that explains all observations, and our models don't make any different predictions about anything we can expect to experience in our pre-space shuttle lives (we don't think a space shuttle is even possible), then we don't disagree?
I mean, I'm just parroting my intro to philosophy of science class (cliché example, I know). But I think everyone puts some value on the simplicity/coherence of a theory, as well as its actual predictions. And the minute you put some value on something other than a theory's predictions, you can easily set up a disagreement where nobody disagrees on any observable fact, but there's a real, heated substantive disagreement.
I've been wondering for a while how the this debate between Robin and Tyler fits in with Scott Page's work on the value of diversity in cognitive strategies (and also whether the insights gleaned from his work has implications for the overcoming bias project more generally). In this case, it appears to suggest an additional possibility that none of Nick's options quite seem to cover: taken individually, which of Robin's and Tyler's cognitive strategies performs better will probably vary from circumstance to circumstance, but the best results may arise from combining them.
To be honest, I'm not sure how exactly this would play out in concrete terms. But just thought I'd throw it out there for people smarter than me to pick up on if they feel like it.