All of which is to say that I spend a decent chunk of the time being the guy in the room who is mostaware of the fuckery swirling around me, and therefore the guy who is most bothered by it.
I feel like this claim is being used as evidence to bolster a call to action, without being well supported. Unsupported evidence is always worse than supported, but it’s especially grating in this case because providing counter-evidence can reasonably be predicted to be interpreted as a social attack, which people have inhibitions against making (and then when do they do get made, come out in really unproductive ways). I would prefer a norm where, if you’re going to claim raw or relative intelligence as a reason to believe you, you need to provide at least a link to evidence.
That’s a fine norm. What sort of evidence would you be interested in, in this case?
I can more easily provide copious evidence of me being aware of fuckery that others aren’t (e.g. 100 FB posts over the last year, probably) than of high-enough-intelligence-to-justify “decent chunk.” But I could e.g. dig up my SAT scores or something.
I expect that people are sort of adversarially misconstruing “decent chunk” as a claim of something like “the vast majority of the time,” or whatever, which I grant I left myself open to but is not a claim I would make (since it isn’t true).
I should note that my interest in sharing the FB posts is that they render “Duncan’s claim to rare insight” discussable. I don’t expect them to move my opinion much because we’re FB friends, and I see many of them already.
Now that you’ve acknowledged that this is fair to assess in its own right, I’d like to share my current assessment of your insight levels:
You have at least once said things I found extremely novel and useful, although still had major disagreements with. I shared the first part with you privately, although did not update you when I developed more concerns about the model.
You have at least once waged a major campaign alone at great social cost, that I appreciate a great deal and respected a lot. I already shared this one privately with you but it seems worth noting in public.
Your FB is a mix of things I agree with and disagree with. Some of the disagreements I might change my mind on if we could have a good discussion, and under other circumstances I would be really excited to have those discussions, because I care about this a lot too and people paying sufficient attention are rare. But I (almost?) never do, because just the thought makes me feel weary. I expect the conversations to be incredibly high friction with no chance of movement on your part. I expect this feeling to be common, and for that lack of feedback to be detrimental to your model building even if you start out far above average.
One counterargument is that there are lots of comments on those posts. That’s true, and maybe they’re covering all the bases, but I would be surprised.
A reasonable question here is if losing those discussions represents lost value to you. I can point you to public twitter threads of mine I think you would find interesting, and to a pseudonymous blog if you agree to limited anonymity we can discuss out of band.
It is only in writing this up that I realized how much lost value this represents to me: I do think if friction was low enough we’d have really interesting discussions we both learned from.
Things that contribute to that feeling of friction
In one private interaction where we disagreed on a norm, your frame was “you [Elizabeth’s] frame is impossibly bad and meta-uncooperative”
You’ve repeatedly deleted posts on LW and FB, in what sure looked like anger.
General sense from reading your LW and FB posts/comments, although I haven’t done a quantified assessment here.
I remember on FB post claiming you were good at receiving feedback, which given my priors means you have so thoroughly discouraged feedback you’re not aware of the scope of the problem.
I guess my overall claim here is not that you don’t have useful insight- you do- but that your certainty in the superiority of that insight reduces your ability to iterate and correct, and nobody can get everything right on the first try.
I do not have, and do not believe I have claimed to have, anything like “certainty in the superiority of my insight.” Happy to just state here explicitly: I don’t have anything like certainty in the superiority of my insight.
What I have is confidence that, when I’m perceiving that something is going sideways, something is, in fact, going sideways.
That’s a far cry from always knowing what it is, which is itself a far cry from having any idea how to fix it.
I’m confused as to how I’m perceived as claiming superiority of insight when e.g. all I could come up with in the above essay was a set of ideas that I myself identified as terrible and insufficient.
My comment is low context both because I don’t think I’ve seen you and Elizabeth talk before and also because I only skimmed the parent comments.
When you say
I’m confused as to how I’m perceived as claiming superiority of insight when e.g. all I could come up with in the above essay was a set of ideas that I myself identified as terrible and insufficient.
This doesn’t seem to me evidence against you claiming to have superior insight under Elizabeth’s usage of the term. My reading is that she uses the term relatively, I.e. that she believes that you believe your claims about the world are right while others’ claims are wrong (or more likely to be true than others’). Terrible, as you used it in the essay, I took to be in absolute terms, as in “will these interventions help? Idk”
I’m more confident in the quote not providing evidence against her usage of “superior insight” than I am in defining her intended meaning.
I would indeed be interested in the FB posts, with the caveat that those then become debatable, and it seems quite possible that that will overwhelm comments on the topic of this post. I also think it would help to declare how you intend to handle people bringing up statements and actions (public or private) not of your choosing. I think deluks handled his comment extremely poorly and I would have done it very differently,* but I do think your claim makes “every bad decision you ever made” relevant, and the overall track record of threads on every bad decision a person ever made is extremely bad.
I think the FB posts are better evidence of “being the one to speak up” than “the only one noticing”. That’s not a criticism: speaking up is really important, and doing it when it matters is a huge service. People who notice but don’t act aren’t very helpful. But it is a different claim
*I can get into how if you are curious but am worried deluks has poisoned the well on reasonable discussion of that problem.
I assume by “very differently,” you mean “would not have included outright falsehoods.”
I don’t see the connection between:
(approximately) “I’m pretty smart and I am pretty conscientious and also I’ve hyperfocused on this domain for a long time, so I notice stuff in this domain a lot more often than most people”
and
[whatever claim you and deluks think I made that is tantamount to an assertion of perfection, or something]
Like, it seems that you’re asking me to defend some outlandish claim that I have not made. That’s the only situation in which “every bad decision I ever made” would be relevant—if I had claimed to always or overwhelmingly often be competent, or something. It sounds like “this guy claimed no white ravens! All we gotta do is find one!”
I clearly did not make that claim, as you can see by scrolling up and just reading. I claimed that I twitch over this stuff, and frequently observe fuckery that other people do not even notice.
But anyway, in addition to the-corpus-of-all-of-my-essays, such as the handful I’ve published this month and things like In Defense of Punch Bug and It’s Not What It Looks Like and Invalidating Imaginary Injury and Common Knowledge and Miasma and any number of my Admonymous answers, I also spent five minutes scrolling back through my most recent FB posts and here are the first twenty or so that seemed to be about noticing or caring about fuckery (not organized by impressiveness):
I was able to see 5-6 in firefox without incognito, and then it asked me to log in (both on ones I already saw and ones I didn’t). Seems like some sort of “You have 3 more articles this month” tactic but without telling you.
Are you opening them in incognito browsers? They seem to work straightforwardly for me in non-logged-in browsers and don’t know what might be different for you.
So far, this problem has replicated on every browser on every platform I’ve tried it on, in both regular and private windows. Chrome, Firefox, Opera, on Mac, Windows, Linux… I have not been able to view any of the given posts in any way at all.
I assume by “very differently,” you mean “would not have included outright falsehoods.”
The first of many differences, yes. I also would have emphasized the part where I thought the evidence something was wrong was obvious and you didn’t (in ways that were visible to me), not the part where you proactively coordinated evidence sharing when it was personally costly to you, which was a social good.
I did not think you were claiming perfection, but I think “It’s like being a native French speaker and dropping in on a high school French class in a South Carolina public school” is a very strong claim of superiority, far beyond “I notice stuff in this domain a lot more often than most people”. Native speakers can be wrong, but in a disagreement with a disengaged high schooler you will basically always take the word of the native speaker. I additionally think the problems in iteration I outlined in a sister thread really put a ceiling on your insights, although admittedly that affects analysis and improvements much more than noticing.
Also, re: evidence that something was wrong was obvious
I dunno. This sounds like an excuse, and an excuse is all that many people will hear, but:
My current model is that Brent, whether consciously or unconsciously/instinctively, did in fact do something resembling cultivating me as a shield, by never egregiously misbehaving in my sight. And many of the other people around me, seeing egregious misbehavior somewhat often, assumed (reasonably) that I must be seeing it, too, and not minding.
But after it all started to come out, there were something like a dozen fully dealbreaking anecdotes handed to me by not-necessarily-specifically-but-people-in-the-reference-class-of Rob, Oli, Nate, Logan, Nick, Val, etc., any one of which would have caused me to spring into action, except they just never mentioned it and I was never in the room to see it.
FWIW: I believe you that Brent cultivated you, and I think you talking about that has been really useful in educating people (including me) about how toxic people do that. I do think it had to be some damn strong cultivation to overcome the baseline expectations set by his FB posts, and I’d be interested in hearing you talk about what he did to overcome that baseline- not because I think you were especially susceptible, but because whatever he did worked on a lot of people, and that makes it useful to understand.
Well, for starters, I had unfollowed him on FB by about 2016 as a result of being just continually frustrated by his relentless pessimism. So I probably missed a whole lot of what others saw as red flags.
I think “It’s like being a native French speaker and dropping in on a high school French class in a South Carolina public school” is a very strong claim of superiority, far beyond “I notice stuff in this domain a lot more often than most people”.
I find this helpful, and I think it’s a fair and reasonable reading that I should have ruled out.
What I meant by choosing that example in particular was that French contains a lot of sounds which English speakers literally can’t perceive at first, until they practice and build up some other background knowledge. That’s … not entirely different from a claim of superiority, but I tried to defuse the sense of superiority by noting that a lot of it comes from just relentlessly attending to the domain—”it’s not that I’m doing anything magic here, many of the people I’m hanging out with are smarter or conscientiouser, it’s just that I happen to have put in more reps is all.”
Ah, this makes sense and is helpful, and now that you’ve spelled it out I can see how it connects to other things in the post in ways I didn’t before. It also makes cases of failure much less relevant, since no one has all phenomes.
Worth noting that I noticed the kerning example seemed very different than the native speaker example, but the “native speaker in a room full of bored teenagers” claim felt so strong I resolved in that direction.
(I agree that there’s an issue here with bucketing counter-evidence with social attack, and that a norm of providing evidence is preferable to a silent bucket error; also, it seems likely that there are belief-like claims that (1) are useful to make, even though at least by default they’re mixed in with social moves, and that (2) are difficult (costly, say) to provide evidence for. If such claims are common, it might be worth having an epistemic status that can contain that nuance, something like “this claim is acknowledged to be too costly for many pairs of people to reach Aumann agreement on, and shouldn’t yet function as a common knowledge belief among groups containing many such pairs, but it’s still what I think for what that’s worth”. Maybe there’s a short phrase that already means this, like “from my perspective...”, though sadly such things are always diluting.)
I feel like this claim is being used as evidence to bolster a call to action, without being well supported. Unsupported evidence is always worse than supported, but it’s especially grating in this case because providing counter-evidence can reasonably be predicted to be interpreted as a social attack, which people have inhibitions against making (and then when do they do get made, come out in really unproductive ways). I would prefer a norm where, if you’re going to claim raw or relative intelligence as a reason to believe you, you need to provide at least a link to evidence.
That’s a fine norm. What sort of evidence would you be interested in, in this case?
I can more easily provide copious evidence of me being aware of fuckery that others aren’t (e.g. 100 FB posts over the last year, probably) than of high-enough-intelligence-to-justify “decent chunk.” But I could e.g. dig up my SAT scores or something.
I expect that people are sort of adversarially misconstruing “decent chunk” as a claim of something like “the vast majority of the time,” or whatever, which I grant I left myself open to but is not a claim I would make (since it isn’t true).
I should note that my interest in sharing the FB posts is that they render “Duncan’s claim to rare insight” discussable. I don’t expect them to move my opinion much because we’re FB friends, and I see many of them already.
Now that you’ve acknowledged that this is fair to assess in its own right, I’d like to share my current assessment of your insight levels:
You have at least once said things I found extremely novel and useful, although still had major disagreements with. I shared the first part with you privately, although did not update you when I developed more concerns about the model.
You have at least once waged a major campaign alone at great social cost, that I appreciate a great deal and respected a lot. I already shared this one privately with you but it seems worth noting in public.
Your FB is a mix of things I agree with and disagree with. Some of the disagreements I might change my mind on if we could have a good discussion, and under other circumstances I would be really excited to have those discussions, because I care about this a lot too and people paying sufficient attention are rare. But I (almost?) never do, because just the thought makes me feel weary. I expect the conversations to be incredibly high friction with no chance of movement on your part. I expect this feeling to be common, and for that lack of feedback to be detrimental to your model building even if you start out far above average.
One counterargument is that there are lots of comments on those posts. That’s true, and maybe they’re covering all the bases, but I would be surprised.
A reasonable question here is if losing those discussions represents lost value to you. I can point you to public twitter threads of mine I think you would find interesting, and to a pseudonymous blog if you agree to limited anonymity we can discuss out of band.
It is only in writing this up that I realized how much lost value this represents to me: I do think if friction was low enough we’d have really interesting discussions we both learned from.
Things that contribute to that feeling of friction
In one private interaction where we disagreed on a norm, your frame was “you [Elizabeth’s] frame is impossibly bad and meta-uncooperative”
You’ve repeatedly deleted posts on LW and FB, in what sure looked like anger.
General sense from reading your LW and FB posts/comments, although I haven’t done a quantified assessment here.
I remember on FB post claiming you were good at receiving feedback, which given my priors means you have so thoroughly discouraged feedback you’re not aware of the scope of the problem.
I guess my overall claim here is not that you don’t have useful insight- you do- but that your certainty in the superiority of that insight reduces your ability to iterate and correct, and nobody can get everything right on the first try.
I do not have, and do not believe I have claimed to have, anything like “certainty in the superiority of my insight.” Happy to just state here explicitly: I don’t have anything like certainty in the superiority of my insight.
What I have is confidence that, when I’m perceiving that something is going sideways, something is, in fact, going sideways.
That’s a far cry from always knowing what it is, which is itself a far cry from having any idea how to fix it.
I’m confused as to how I’m perceived as claiming superiority of insight when e.g. all I could come up with in the above essay was a set of ideas that I myself identified as terrible and insufficient.
My comment is low context both because I don’t think I’ve seen you and Elizabeth talk before and also because I only skimmed the parent comments.
When you say
This doesn’t seem to me evidence against you claiming to have superior insight under Elizabeth’s usage of the term. My reading is that she uses the term relatively, I.e. that she believes that you believe your claims about the world are right while others’ claims are wrong (or more likely to be true than others’). Terrible, as you used it in the essay, I took to be in absolute terms, as in “will these interventions help? Idk”
I’m more confident in the quote not providing evidence against her usage of “superior insight” than I am in defining her intended meaning.
I would indeed be interested in the FB posts, with the caveat that those then become debatable, and it seems quite possible that that will overwhelm comments on the topic of this post. I also think it would help to declare how you intend to handle people bringing up statements and actions (public or private) not of your choosing. I think deluks handled his comment extremely poorly and I would have done it very differently,* but I do think your claim makes “every bad decision you ever made” relevant, and the overall track record of threads on every bad decision a person ever made is extremely bad.
I think the FB posts are better evidence of “being the one to speak up” than “the only one noticing”. That’s not a criticism: speaking up is really important, and doing it when it matters is a huge service. People who notice but don’t act aren’t very helpful. But it is a different claim
*I can get into how if you are curious but am worried deluks has poisoned the well on reasonable discussion of that problem.
I assume by “very differently,” you mean “would not have included outright falsehoods.”
I don’t see the connection between:
(approximately) “I’m pretty smart and I am pretty conscientious and also I’ve hyperfocused on this domain for a long time, so I notice stuff in this domain a lot more often than most people”
and
[whatever claim you and deluks think I made that is tantamount to an assertion of perfection, or something]
Like, it seems that you’re asking me to defend some outlandish claim that I have not made. That’s the only situation in which “every bad decision I ever made” would be relevant—if I had claimed to always or overwhelmingly often be competent, or something. It sounds like “this guy claimed no white ravens! All we gotta do is find one!”
I clearly did not make that claim, as you can see by scrolling up and just reading. I claimed that I twitch over this stuff, and frequently observe fuckery that other people do not even notice.
But anyway, in addition to the-corpus-of-all-of-my-essays, such as the handful I’ve published this month and things like In Defense of Punch Bug and It’s Not What It Looks Like and Invalidating Imaginary Injury and Common Knowledge and Miasma and any number of my Admonymous answers, I also spent five minutes scrolling back through my most recent FB posts and here are the first twenty or so that seemed to be about noticing or caring about fuckery (not organized by impressiveness):
https://www.facebook.com/duncan.sabien/posts/4946318165402861
https://www.facebook.com/duncan.sabien/posts/4924756570892354
https://www.facebook.com/duncan.sabien/posts/4920978771270134
https://www.facebook.com/duncan.sabien/posts/4918223831545628
https://www.facebook.com/duncan.sabien/posts/4918136751554336
https://www.facebook.com/duncan.sabien/posts/4911478975553447
https://www.facebook.com/duncan.sabien/posts/4907166399318038
https://www.facebook.com/duncan.sabien/posts/4907029615998383
https://www.facebook.com/duncan.sabien/posts/4889568061077872
https://www.facebook.com/duncan.sabien/posts/4888310204536991?comment_id=4888448571189821
https://www.facebook.com/duncan.sabien/posts/4877212198980125?comment_id=4877320285635983
https://www.facebook.com/duncan.sabien/posts/4874111979290147
https://www.facebook.com/duncan.sabien/posts/4859938947374117
https://www.facebook.com/duncan.sabien/posts/4855597157808296
https://www.facebook.com/duncan.sabien/posts/4840078599360152
https://www.facebook.com/duncan.sabien/posts/4834628186571860
https://www.facebook.com/duncan.sabien/posts/4832098550158157
https://www.facebook.com/duncan.sabien/posts/4825568070811205
https://www.facebook.com/duncan.sabien/posts/4822568564444489
https://www.facebook.com/duncan.sabien/posts/4821518481216164
… and I note that those twenty are all just within the past six weeks. A six-week period in which I also wrote six LW essays about social dynamics.
FYI, every single one of these posts (yes, I tested all the links) is inaccessible to me, because they require logging into Facebook.
(I’m posting this to note that this isn’t a problem specific to that one other post, but seems to be a general problem.)
I was able to see 5-6 in firefox without incognito, and then it asked me to log in (both on ones I already saw and ones I didn’t). Seems like some sort of “You have 3 more articles this month” tactic but without telling you.
Are you opening them in incognito browsers? They seem to work straightforwardly for me in non-logged-in browsers and don’t know what might be different for you.
This is groundhog day Ray; we just found out that it doesn’t work on Opera and Firefox.
(And apparently Chrome Incognito on Windows? I’m confused about the exact line there, because it works on my Chrome Incognito on Mac.)
So far, this problem has replicated on every browser on every platform I’ve tried it on, in both regular and private windows. Chrome, Firefox, Opera, on Mac, Windows, Linux… I have not been able to view any of the given posts in any way at all.
The first of many differences, yes. I also would have emphasized the part where I thought the evidence something was wrong was obvious and you didn’t (in ways that were visible to me), not the part where you proactively coordinated evidence sharing when it was personally costly to you, which was a social good.
I did not think you were claiming perfection, but I think “It’s like being a native French speaker and dropping in on a high school French class in a South Carolina public school” is a very strong claim of superiority, far beyond “I notice stuff in this domain a lot more often than most people”. Native speakers can be wrong, but in a disagreement with a disengaged high schooler you will basically always take the word of the native speaker. I additionally think the problems in iteration I outlined in a sister thread really put a ceiling on your insights, although admittedly that affects analysis and improvements much more than noticing.
Also, re: evidence that something was wrong was obvious
I dunno. This sounds like an excuse, and an excuse is all that many people will hear, but:
My current model is that Brent, whether consciously or unconsciously/instinctively, did in fact do something resembling cultivating me as a shield, by never egregiously misbehaving in my sight. And many of the other people around me, seeing egregious misbehavior somewhat often, assumed (reasonably) that I must be seeing it, too, and not minding.
But after it all started to come out, there were something like a dozen fully dealbreaking anecdotes handed to me by not-necessarily-specifically-but-people-in-the-reference-class-of Rob, Oli, Nate, Logan, Nick, Val, etc., any one of which would have caused me to spring into action, except they just never mentioned it and I was never in the room to see it.
FWIW: I believe you that Brent cultivated you, and I think you talking about that has been really useful in educating people (including me) about how toxic people do that. I do think it had to be some damn strong cultivation to overcome the baseline expectations set by his FB posts, and I’d be interested in hearing you talk about what he did to overcome that baseline- not because I think you were especially susceptible, but because whatever he did worked on a lot of people, and that makes it useful to understand.
Well, for starters, I had unfollowed him on FB by about 2016 as a result of being just continually frustrated by his relentless pessimism. So I probably missed a whole lot of what others saw as red flags.
This indeed changes my opinion a fair bit, and I should have had it as a more active hypothesis.
I find this helpful, and I think it’s a fair and reasonable reading that I should have ruled out.
What I meant by choosing that example in particular was that French contains a lot of sounds which English speakers literally can’t perceive at first, until they practice and build up some other background knowledge. That’s … not entirely different from a claim of superiority, but I tried to defuse the sense of superiority by noting that a lot of it comes from just relentlessly attending to the domain—”it’s not that I’m doing anything magic here, many of the people I’m hanging out with are smarter or conscientiouser, it’s just that I happen to have put in more reps is all.”
It didn’t work.
Ah, this makes sense and is helpful, and now that you’ve spelled it out I can see how it connects to other things in the post in ways I didn’t before. It also makes cases of failure much less relevant, since no one has all phenomes.
Worth noting that I noticed the kerning example seemed very different than the native speaker example, but the “native speaker in a room full of bored teenagers” claim felt so strong I resolved in that direction.
(I agree that there’s an issue here with bucketing counter-evidence with social attack, and that a norm of providing evidence is preferable to a silent bucket error; also, it seems likely that there are belief-like claims that (1) are useful to make, even though at least by default they’re mixed in with social moves, and that (2) are difficult (costly, say) to provide evidence for. If such claims are common, it might be worth having an epistemic status that can contain that nuance, something like “this claim is acknowledged to be too costly for many pairs of people to reach Aumann agreement on, and shouldn’t yet function as a common knowledge belief among groups containing many such pairs, but it’s still what I think for what that’s worth”. Maybe there’s a short phrase that already means this, like “from my perspective...”, though sadly such things are always diluting.)