You’re in the dangerous position of suffering from confirmation bias on account of having so little exposure to people who are highly accomplished. The people who I’m talking about have mathematical productivity of order ~100,000x that of the average mathematician. Most mathematicians are terrified of talking to them on account of the expectation that they’d come across as really stupid. These are not people who pat people on the back for jumping through hoops.
On an object level: in the course of working on my speed dating project, I rediscovered logistic regression, collaborative filtering, and hierarchical modeling. I rediscovered cross validation and how it can be combined with stepwise regression to identify robustly generalizable patterns in data. This led me to the discovery that principal component analysis greatly reduces concerns about multiple hypothesis testing, and greatly clarifies what’s going on.
The trouble is that I can’t credibly signal that this represents unusually high quality work, because you don’t have the subject matter knowledge that you would need to make an assessment. This is my point above: it’s not clear to me that there’s anything that I can say to change your mind.
The question that you should ask yourself is: if you’re so rational and intelligent, why aren’t you more successful? It’s convenient to attribute it to luck of the draw, but the fact is that you’re actually roughly 1 million times lower in intellectual caliber than the highest intellectual caliber people in the world. Returns to IQ and aesthetic discernment aren’t linear in expectation, they’re exponential. And you have no way of knowing this. Which is why I’m taking the time to explain this to you.
You’re totally misreading the situation: I could be talking to people who are thousands of times more sophisticated than you, and it would be much more interesting to me, and instead I’m talking to you because I care about you. It should make you feel much higher status than you currently are, not like I’m being offensive. But ultimately, if you’re not receptive, I can’t do anything to help. :-(
Whether what you write in the above comment is true or not (and by the way, I should mention that I believe you), it’s an empirical fact about human psychology that taking a “holier than thou” attitude never helps if you want the other person to actually listen. And maybe it doesn’t feel to you like you’re taking a “holier than thou” attitude—or even any attitude at all. Maybe to you, you’re just stating the facts. That’s fine. But you’ve got to take into the account how the other person feels—and speaking for myself, I perceived a lot of condescension from your comment. (And then there’s also the fact that the average person on LW is much less likely to take authority as an argument, anyway.)
I’m not quite sure how to signal greater knowledge without also issuing a status challenge, and I somewhat doubt that there is a way. But you could do a lot better simply by cleaning up your tone a bit. For example, this
You’re totally misreading the situation: I could be talking to people who are thousands of times more sophisticated than you, and it would be much more interesting to me, and instead I’m talking to you becsuse I care about you. It should make you feel much higher status than you currently are, not like I’m being offensive. But ultimately, if you’re not receptive, I can’t do anything to help. :-(
could have been phrased as
You can feel free to disagree about the level of my accomplishments if you want, although I should note that the people I’m talking about aren’t likely to pat you on the back for just “jumping through hoops”. But ultimately, I’m making these posts because I care about helping you. If you don’t want my help or you think my help is suspect, there’s nothing forcing you to take my advice. However, questioning my level of ability is not really productive, in my opinion; if you think I’m not qualified to give advice, just don’t take my advice.
EDIT: I’m not saying Lumifer’s been doing any better. In particular, “your arguments are that of a child” was really poorly phrased, IMO.
Thanks for the feedback, I do really appreciate it – I think that you’re absolutely right. I was showing an empathy deficit there. Consistently showing empathy is difficult, and I’m working on it.
But I shouldn’t face social punishment for spending thousands of hours developing deep subject matter knowledge. I shouldn’t face social punishment for having a deep desire to help people. It shouldn’t be that people who have the stated objectives of being less wrong and overcoming bias are hostile to me for speaking the truth. That’s not a good incentive structure for our culture (whether on LW or in the world) to adopt.
For many years I felt like I couldn’t be open about who I am, even amongst Less Wrong people or mathematicians. I’m not going to hide who I am just so that people don’t have to feel uncomfortable about someone being more sophisticated and empathetic than they are. The sin of underconfidence is just as dangerous as the sin of overconfidence. If people can’t handle knowing the facts about me, it’s because they have psychological issues to work out rather than because there’s something wrong with me.
Edit: I may appear to be exhibiting an empathy deficit here as well – it’s sort of inevitable, I shouldn’t be internalizing perspectives that are fundamentally misguided at the cost of my own mental health.
In particular, “your arguments are that of a child” was really poorly phrased, IMO.
So, I certainly cringed empathetically when I read that—but on reflection I agreed with the assessment, and the issue I saw was that it was said in public, and by someone who doesn’t seem to have established rapport beforehand. So I’m not sure I agree that it’s a phrasing issue.
I rediscovered logistic regression, collaborative filtering, and hierarchical modeling. I rediscovered cross validation and how it can be combined with stepwise regression to identify robustly generalizable patterns in data.
You rediscovered? You didn’t know logistic regression existed? What exactly did you rediscover?
This led me to the discovery that principal component analysis greatly reduces concerns about multiple hypothesis testing
I suspect you’re wrong about that. Rotating a matrix (which is what PCA does) doesn’t actually reduce concerns about “hidden” degrees of freedom which you use up by trying multiple hypotheses. I actually think that the usefulness of PCA is often overstated—all you’re doing is selecting linear combinations with the highest variance which is not always the right thing to pay attention to.
All in all that just sounds like pretty standard statistics.
The trouble is that I can’t credibly signal that this represents unusually high quality work
Well, yes, you can’t. Speaking of “unusually high quality”, your Github code contains things like
which should be mildly embarassing. Along with values hardcoded as numbers in the body of the function, etc. I can read (and write) R code just fine—what is it that you consider to be “unusually high quality”?
it’s not clear to me that there’s anything that I can say to change your mind.
I don’t have much of a mind to change. I am doubtful of your assertions of great superiority, but that’s a doubt, not a conviction that you’re just an average math geek.
The question that you should ask yourself is: if you’re so rational and intelligent, why aren’t you more successful?
Yeah: if you’re so smart how come you ain’t rich? :-D
Why aren’t I more successful than what?
the fact is that you’re actually roughly 1 million times lower in intellectual caliber than the highest intellectual caliber people in the world.
Which metric are you using? IQ values are ranks and I just don’t know what “1 million times lower in intellectual caliber” even means.
Returns to IQ and aesthetic discernment aren’t linear in expectation, they’re exponential.
Evidence, please. Not to mention that for particular parameters exponential can be pretty close to linear :-)
I could be talking to people who are thousands of times more sophisticated than you, and it would be much more interesting to me, and instead I’m talking to you becsuse I care about you.
I’m sorry, did I stumble into some Christian revival meeting? What is this shit about trying to guilt me into agreement because you sacrifice so much of your highly valuable utils and hedons only because you care?
I think your ego is in dire need of some deflation.
which should be mildly embarassing. Along with values hardcoded as numbers in the body of the function, etc. I can read (and write) R code just fine—what is it that you consider to be “unusually high quality”?
I was just learning R at the time and in a rush to get things to work. The code itself is not high quality.
What is this shit about trying to guilt me into agreement because you sacrifice so much of your highly valuable utils and hedons only because you care?
I see that you don’t know what you’re missing, I know it’s because you didn’t have the environmental advantages that I did, I know that I could have been in your position if not for the luck of the draw, and so I have pangs of sympathy for you, because your situation is in some sense very close to my own. It would make me feel so good if I could help you. That’s why it’s worth it to me in expectation, even if it’s unpleasant in real time.
But you can’t love someone who doesn’t want you to love him/her. I spent ~15 years on that and it helped no one and came at great cost to my myself. So I’ll withdraw from this conversation.
It would make me feel so good if I could help you.
LOL. I don’t know if you’re imitating a Christian missionary or a Jewish mother, but you’re doing it badly.
But you can’t love someone who doesn’t want you to love him/her.
You can, but the relevant thing is that yes, I have no particular desire for you to love me. I suspect the same is true for the great majority of the LW population. And if I ever go looking for unconditional love, Jesus has a much better spiel that you do—and He, at least, died for my sins :-P
LOL. I don’t know if you’re imitating a Christian missionary or a Jewish mother, but you’re doing it badly.
To be blunt, I think that what’s going on here is that you have an empathy deficit and so don’t experience the warm fuzzy feelings around helping people that some people do. There are some people who would immediately understand where I’m coming from, and say “that totally makes sense.” You seem to be Generalizing From One Example. I don’t know whether it’s genetic or environmental, mutable or immutable, but it’s sad.
You can, but the relevant thing is that yes, I have no particular desire for you to love me.
I’m not generalizing—I’m pointing out that you, singular, you personally are doing it badly. And you are putting a lot of effort into not hearing this message. By the way, have you noticed how your last few comments started to focus on me and my shortcoming and deficiencies?
By the way, have you noticed how your last few comments started to focus on me and my shortcoming and deficiencies?
Not exclusively, I also mentioned my many years of failed efforts along these lines. I’m not claiming that my efforts have been useful. It’s possible that you’ve actually helped people more than I have. My comments about you were made with a view toward giving a comprehensive explanation of why I’m bowing out of the conversation.
The general pattern is “I tried to help people and they misconstrued it because they didn’t have enough empathy to have a visceral understanding of the fact that I wanted to help them, so rather than being touched, they just found it irritating, and I made sacrifices when it should have been a priori clear that they were doomed to failure.” I finally get it now.
The problem is that you believe that your internal motivation justifies your expectations of other people.
Because your intentions are virtuous you expect that other people be “touched”, be grateful, help you by steelmanning your arguments, etc. And it’s not a matter of empathy, it’s a matter of whether your state of mind imposes obligations on other people. It looks reasonable to you because from your point of view you only want to teach and it’s reasonable that other people help you teach them. But try taking an external view (and try being more consequentialist, too).
Christian missionaries appeared in this subthread not by accident—they also care and also want to help and also make sacrifices to teach what they teach.
We’re not in disagreement! :-) What I’m saying is that after many years, I finally came around to understanding what you’re telling me right now. Your remarks are a useful update further in the same direction. I’ve genuinely benefitted from this interaction.
You’re in the dangerous position of suffering from confirmation bias on account of having so little exposure to people who are highly accomplished. The people who I’m talking about have mathematical productivity of order ~100,000x that of the average mathematician. Most mathematicians are terrified of talking to them on account of the expectation that they’d come across as really stupid. These are not people who pat people on the back for jumping through hoops.
On an object level: in the course of working on my speed dating project, I rediscovered logistic regression, collaborative filtering, and hierarchical modeling. I rediscovered cross validation and how it can be combined with stepwise regression to identify robustly generalizable patterns in data. This led me to the discovery that principal component analysis greatly reduces concerns about multiple hypothesis testing, and greatly clarifies what’s going on.
The trouble is that I can’t credibly signal that this represents unusually high quality work, because you don’t have the subject matter knowledge that you would need to make an assessment. This is my point above: it’s not clear to me that there’s anything that I can say to change your mind.
The question that you should ask yourself is: if you’re so rational and intelligent, why aren’t you more successful? It’s convenient to attribute it to luck of the draw, but the fact is that you’re actually roughly 1 million times lower in intellectual caliber than the highest intellectual caliber people in the world. Returns to IQ and aesthetic discernment aren’t linear in expectation, they’re exponential. And you have no way of knowing this. Which is why I’m taking the time to explain this to you.
You’re totally misreading the situation: I could be talking to people who are thousands of times more sophisticated than you, and it would be much more interesting to me, and instead I’m talking to you because I care about you. It should make you feel much higher status than you currently are, not like I’m being offensive. But ultimately, if you’re not receptive, I can’t do anything to help. :-(
Whether what you write in the above comment is true or not (and by the way, I should mention that I believe you), it’s an empirical fact about human psychology that taking a “holier than thou” attitude never helps if you want the other person to actually listen. And maybe it doesn’t feel to you like you’re taking a “holier than thou” attitude—or even any attitude at all. Maybe to you, you’re just stating the facts. That’s fine. But you’ve got to take into the account how the other person feels—and speaking for myself, I perceived a lot of condescension from your comment. (And then there’s also the fact that the average person on LW is much less likely to take authority as an argument, anyway.)
I’m not quite sure how to signal greater knowledge without also issuing a status challenge, and I somewhat doubt that there is a way. But you could do a lot better simply by cleaning up your tone a bit. For example, this
could have been phrased as
EDIT: I’m not saying Lumifer’s been doing any better. In particular, “your arguments are that of a child” was really poorly phrased, IMO.
Thanks for the feedback, I do really appreciate it – I think that you’re absolutely right. I was showing an empathy deficit there. Consistently showing empathy is difficult, and I’m working on it.
But I shouldn’t face social punishment for spending thousands of hours developing deep subject matter knowledge. I shouldn’t face social punishment for having a deep desire to help people. It shouldn’t be that people who have the stated objectives of being less wrong and overcoming bias are hostile to me for speaking the truth. That’s not a good incentive structure for our culture (whether on LW or in the world) to adopt.
For many years I felt like I couldn’t be open about who I am, even amongst Less Wrong people or mathematicians. I’m not going to hide who I am just so that people don’t have to feel uncomfortable about someone being more sophisticated and empathetic than they are. The sin of underconfidence is just as dangerous as the sin of overconfidence. If people can’t handle knowing the facts about me, it’s because they have psychological issues to work out rather than because there’s something wrong with me.
Edit: I may appear to be exhibiting an empathy deficit here as well – it’s sort of inevitable, I shouldn’t be internalizing perspectives that are fundamentally misguided at the cost of my own mental health.
So, I certainly cringed empathetically when I read that—but on reflection I agreed with the assessment, and the issue I saw was that it was said in public, and by someone who doesn’t seem to have established rapport beforehand. So I’m not sure I agree that it’s a phrasing issue.
You rediscovered? You didn’t know logistic regression existed? What exactly did you rediscover?
I suspect you’re wrong about that. Rotating a matrix (which is what PCA does) doesn’t actually reduce concerns about “hidden” degrees of freedom which you use up by trying multiple hypotheses. I actually think that the usefulness of PCA is often overstated—all you’re doing is selecting linear combinations with the highest variance which is not always the right thing to pay attention to.
All in all that just sounds like pretty standard statistics.
Well, yes, you can’t. Speaking of “unusually high quality”, your Github code contains things like
which should be mildly embarassing. Along with values hardcoded as numbers in the body of the function, etc. I can read (and write) R code just fine—what is it that you consider to be “unusually high quality”?
I don’t have much of a mind to change. I am doubtful of your assertions of great superiority, but that’s a doubt, not a conviction that you’re just an average math geek.
Yeah: if you’re so smart how come you ain’t rich? :-D
Why aren’t I more successful than what?
Which metric are you using? IQ values are ranks and I just don’t know what “1 million times lower in intellectual caliber” even means.
Evidence, please. Not to mention that for particular parameters exponential can be pretty close to linear :-)
I’m sorry, did I stumble into some Christian revival meeting? What is this shit about trying to guilt me into agreement because you sacrifice so much of your highly valuable utils and hedons only because you care?
I think your ego is in dire need of some deflation.
I was just learning R at the time and in a rush to get things to work. The code itself is not high quality.
I see that you don’t know what you’re missing, I know it’s because you didn’t have the environmental advantages that I did, I know that I could have been in your position if not for the luck of the draw, and so I have pangs of sympathy for you, because your situation is in some sense very close to my own. It would make me feel so good if I could help you. That’s why it’s worth it to me in expectation, even if it’s unpleasant in real time.
But you can’t love someone who doesn’t want you to love him/her. I spent ~15 years on that and it helped no one and came at great cost to my myself. So I’ll withdraw from this conversation.
LOL. I don’t know if you’re imitating a Christian missionary or a Jewish mother, but you’re doing it badly.
You can, but the relevant thing is that yes, I have no particular desire for you to love me. I suspect the same is true for the great majority of the LW population. And if I ever go looking for unconditional love, Jesus has a much better spiel that you do—and He, at least, died for my sins :-P
To be blunt, I think that what’s going on here is that you have an empathy deficit and so don’t experience the warm fuzzy feelings around helping people that some people do. There are some people who would immediately understand where I’m coming from, and say “that totally makes sense.” You seem to be Generalizing From One Example. I don’t know whether it’s genetic or environmental, mutable or immutable, but it’s sad.
Yes, so I’ll stop and withdraw.
I’m not generalizing—I’m pointing out that you, singular, you personally are doing it badly. And you are putting a lot of effort into not hearing this message. By the way, have you noticed how your last few comments started to focus on me and my shortcoming and deficiencies?
Not exclusively, I also mentioned my many years of failed efforts along these lines. I’m not claiming that my efforts have been useful. It’s possible that you’ve actually helped people more than I have. My comments about you were made with a view toward giving a comprehensive explanation of why I’m bowing out of the conversation.
The general pattern is “I tried to help people and they misconstrued it because they didn’t have enough empathy to have a visceral understanding of the fact that I wanted to help them, so rather than being touched, they just found it irritating, and I made sacrifices when it should have been a priori clear that they were doomed to failure.” I finally get it now.
The problem is that you believe that your internal motivation justifies your expectations of other people.
Because your intentions are virtuous you expect that other people be “touched”, be grateful, help you by steelmanning your arguments, etc. And it’s not a matter of empathy, it’s a matter of whether your state of mind imposes obligations on other people. It looks reasonable to you because from your point of view you only want to teach and it’s reasonable that other people help you teach them. But try taking an external view (and try being more consequentialist, too).
Christian missionaries appeared in this subthread not by accident—they also care and also want to help and also make sacrifices to teach what they teach.
We’re not in disagreement! :-) What I’m saying is that after many years, I finally came around to understanding what you’re telling me right now. Your remarks are a useful update further in the same direction. I’ve genuinely benefitted from this interaction.