On what other basis, then, are we to believe any of this stuff about rewiring neurons, “electrical resistance = emotional resistance”, etc. etc.? We’ve beentold that there’s no evidence whatsoever for any of it and that Romeo got the idea for the latter claim, in particular, from literally nowhere at all. So we can’t believe any of this on the basis of evidence, because we’ve been given none, and told that none is forthcoming. And you say we’re not to believe it on Romeo’s word. What’s left?
You both latched onto the least interesting part of the post. The part that is literally just Romeo throwing out some wild speculations. The post would probably have been a bit cleaner to not mention the few wild speculations he mentions, but getting caught up on the tiny details seems to miss the forest from the trees.
The more interesting part is the general framework where he matches up the some of the processes mentioned in Buddhism with some insights from behavioral psychology, psychotherapy, and pop psychology. It gives a framework to start understanding why anecdotally people claim such big effects from extended meditation practice, and gives an insight about how one might begin to test the hypothesis that this is what’s happening, both personally and on a broader scale.
You both latched onto the least interesting part of the post. The part that is literally just Romeo throwing out some wild speculations.
That is not the impression given by either the post or Romeo’s replies. He seems to be making it clear that he is, indeed, making the affirmative claims that I attributed to him, and has thus far given no indication of intending to retract or weaken them.
But let me ask you this: do you think the post would stand on its own, without these “speculations” (as you call them)? By way of attempting to answer this question, I’ve taken the liberty of creating a copy of Romeo’s post on Google Docs, where I’ve crossed out every “speculative” claim for which we’ve not been given evidence.
I was fairly conservative in choosing parts to cross out; I’ve left intact, for example, almost everything that mentions or talks about forms of psychotherapy and other concepts from psychology. (Although, by rights those ought to go as well; we do not accept such claims without evidence ordinarily; why should we accept them when speaking about meditation or Buddhism? But perhaps speculating about the “softer” sciences is a lesser offense, so I mostly let these things stand.)
What is left, nevertheless, is a post that is missing any justifications or explanations of its core claims; a post essentially indistinguishable from any number of similar posts that have appeared on Less Wrong, since its relaunch almost two years ago.
The more interesting part is the general framework where he matches up the some of the processes mentioned in Buddhism with some insights from behavioral psychology, psychotherapy, and pop psychology. It gives a framework to start understanding why anecdotally people claim such big effects from extended meditation practice, and gives an insight about how one might begin to test the hypothesis that this is what’s happening, both personally and on a broader scale.
Allow me to make a suggestion, then, for anyone who writes these sorts of posts in the future:
First, enumerate, explain, and cite the “insights from behavioral psychology, psychotherapy, and pop psychology” which you intend to use as explanations. Be specific in your claims, and be diligent in your references. Consult the latest available sources, to make sure that your insights of choice have survived the replication crisis; provide these sources to your readers.
Then, having established a basis for what follows, make all the interesting connections to Buddhism, meditation, or what have you.
This approach would kill two birds with one stone: it would prevent such distracting digressions as we’re now engaged in, and it would also strengthen your own understanding of the ideas that you are presenting.
It seems to me that doing things in this way would benefit everyone involved.
I don’t mind the speculation, just the wild ones. Putting forward a hypothesis based on your own experience is incredibly useful, especially if it makes novel predictions that people can then go test themselves.
I don’t know how for instance Romeo would have gotten the “electrical resistance” thing from his own experience, but many of the other tools I can see by him noticing similarities between what happens in meditation, and various other psychotherapies he has tried or read about. This is actually one place where I expect phenomenological experience to provide valuable hypotheses, which is why I think this sort of post can be credible enough to want to test/further explore the hypotheses.
The two parts I mentioned are simply the most obviously speculative and unjustified examples. I also don’t have any real reason to believe the vaguer pop psychology claims about building stories, backlogs, etc.
The post would probably have been a bit cleaner to not mention the few wild speculations he mentions, but getting caught up on the tiny details seems to miss the forest from the trees.
It seems to me LW has a big epistemic hygiene problem, of late. We need to collectively stop make excuses for posting wild speculations as if they were fact, just because the same post also contains some interesting ‘insight’. We should be downvoting such posts and saying to the author “go away and write this again, with the parts that you can’t justify removed”.
Doing so may reveal that the ‘insight’ either does or does not successfully stand alone when the supporting ‘speculation’ is removed. Either way, we learn something valuable about the alleged insight; and we benefit directly by not spreading claims that lack evidence.
The tacit claim is that LW should be about confirmatory research and that exploratory research doesn’t belong here. But confirmatory, cited research has never been the majority of content going back to LW 1.0.
I cannot agree with your reading of nshepperd’s comment. I concur with you that exploratory research is entirely appropriate for Less Wrong, and indeed that it is a strength of this forum.
However, with respect, I do not think that your post rises to the level of “exploratory research”. The level of scholarship and rigor on display would need to be improved substantially, before we could label the post as “research” of any kind.
I have seen firsthand what exploratory research is like. Even in the relatively “soft” field I’ve had experience with (HCI), the post at hand would not qualify—not by a long shot.
Perhaps, though, we might investigate the nature of our disagreement in another way. What would you say are three of the best examples of “exploratory research” from Less Wrong’s history? To avoid confounding factors, please limit your examples to the period before the relaunch of the new site; and also avoid Eliezer’s posts from the original Sequences period (as citing them as examples would illuminate nothing).
FYI, I started working on a post this morning (which I’ll most likely refactor into a question before publishing) that explores the question of “what combination of norms and incentive gradients will best output useful progress on LessWrong”.
Will probably be at least a few days before posting, but figured I’d mention it here.
Firstly, because the Sequences were written almost entirely on Overcoming Bias, not on Less Wrong. That alone would suffice. (Less Wrong simply did not exist when Eliezer wrote most of the Sequences.)
Secondly, because we already know that Eliezer can (or could, at least) write interesting and useful things, and have interesting and useful ideas. The question is whether anyone else—specifically, anyone from the Less Wrong commentariat—has that ability; and how we should encourage it, and nurture it; and what epistemic standards, and what community norms, encourage good and interesting and useful and correct ideas. We are, after all, talking about what kind of posts are appropriate for Less Wrong today. So asking whether the site’s founder and originally primary contributor wrote anything of value (especially before the community was even founded) is not relevant.
How is your point of view different, if at all, from the one critiqued here?
It seems like you’re saying LessWrong ought to have a closed canon (the Sequences) and expound on but not add to or accept a substantive critique of them.
I’m not saying anything even remotely like that. I… don’t actually know how you got that from what I wrote. The post you linked seems to have nothing at all to do with what I’m saying.
Clearly, there’s been some great miscommunication here, but I am unsure of what could be the source of it…
Can anyone else write interesting and useful things, and have interesting useful ideas?
What do you mean by this, if not that you’re trying to figure out whether other people share some personal specialness Eliezer has?
If you’re not thinking of the past as an uncaused golden age and Eliezer as a legend of yore, what’s the relation between that question and the question of which kinds of post are appropriate here?
If the problem is that some posts, knowably to a group of people competent to implement shared standards, are neither interesting nor useful, and lack interesting useful ideas, then the obvious solution would be to declare those specific posts inappropriate.
What do you mean by this, if not that you’re trying to figure out whether other people share some personal specialness Eliezer has?
It’s not “some personal specialness”; it’s the ability, inclination, wherewithal, knowledge, expertise, habit, etc., etc., to write posts and comments that are useful, interesting, and otherwise desirable to have insofar as they serve the goals of Less Wrong.
These qualities can be encouraged where present, they can be developed where absent, they can be selected for from among a population, and their application can be incentivized.
But it is clearly not the case that said qualities are simply present in anyone who gets it into their head to write a Less Wrong post.
How common are these salutary qualities? We don’t know (but not very common). How common are they among the current Less Wrong commentariat, in particular? We don’t know (hopefully more common than in the general population, but clearly not as common as we’d like). What community norms, what rules, contribute to increasing and maintaining their prevalence among the membership of the site? We don’t know.
If you’re not thinking of the past as an uncaused golden age and Eliezer as a legend of yore, what’s the relation between that question and the question of which kinds of post are appropriate here?
The past is not an uncaused golden age, but whatever its causes were, they cannot possibly include “the norms and rules of Less Wrong”, because Less Wrong did not exist. Given that “what should be the norms and rules of Less Wrong” is, in fact, what we are discussing, said golden age (i.e., Eliezer writing the Sequences) is irrelevant.
As for Eliezer being a “legend of yore”… well, consider the following analogy, inspired by the very post you linked.
Suppose Einstein, late in life, founds a school for aspiring brilliant theoretical physicists. Some time passes, and I inquire of the administration whether their school has, in fact, produced any brilliant physics theories; indeed, can it? Are their teaching methods any good, even? “What makes you suspect otherwise?”, the administrators ask me. “Well,” say I, “name some brilliant theoretical phycisists who’ve come from your school.” “Why,” comes the reply, “there was Einstein!”
Articles that explore new ideas are harder to write productively than articles that don’t.
Eliezer, uniquely on LessWrong, wrote productive articles exploring new ideas.
Therefore, in order to cultivate productivity, we should not attempt to imitate Eliezer by exploring new ideas, but instead write the other sorts of articles, which are easier to write productively.
If that’s what you mean, then that just isa proposal for closure of primary canon, allowing only carefully explicated commentaries and analyses and extensions of existing ideas.
None of that is even in the vicinity of what I meant.
I do not, quite frankly, know how you got any of that from what I wrote in this thread. As far as I can see, your attempted summary of my points is simply one big non sequitur.
Are you sure you’re not just rounding off my comments to the nearest cached criticism?
If you reread what I’ve written and still believe the provided summary is fair, I’ll attempt to re-explain, I guess…
I don’t have any other good hypotheses for why, when asking for examples of what good exploratory research would look like, you specified that we should exclude Eliezer. I can generate one more, bad hypothesis, which is that you want to make sure we’re hitting a sufficiently low standard, since we should grade on a curve and not hold ourselves to the standard of matching Eliezer. But that seems wildly in conflict with a bunch of other stuff you’ve said here.
ETA: On reflection it’s plausible that I was misreading you and the latter really is what you meant the whole time. If so, oops, not sure how I made that mistake. Thanks for asking me to reread!
As you took the time to reread my comments, it seems only fair that I should take the time to attempt another explanation, as perhaps a rewording will help to dispel any remaining confusion. I hope you’ll excuse my using your earlier comment as a jumping-off point, though I know you no longer endorse this interpretation of my view:
Articles that explore new ideas are harder to write productively than articles that don’t.
This is true. However, as I wrote in this comment, I believe “exploratory research” to be a (perhaps not unique, but certainly unusual) strength of Less Wrong. That such articles are harder to write only means that it is more important—given how few places on the internet have any capability to produce such writing—that we do these things well.
Eliezer, uniquely on LessWrong, wrote productive articles exploring new ideas.
First, again, I do not think that it is sensible to view the Sequences as having been written on Less Wrong—not least because they, in fact, weren’t! (You will note, by the way, that I specified Eliezer’s writings from the Sequences period for exclusion—not all his writings!)
That aside, I do not think this quoted bit is true either; Eliezer’s contributions were not uniquely excellent. I can easily come up with examples of good “exploratory research” articles written on Less Wrong by people who aren’t Eliezer. I asked Romeo to provide examples of his own because I thought (and still think) that seeing what he considers to be good “exploratory research” from Less Wrong’s past would help to illuminate the substance of our disagreement.
Yes, we all agree (presumably) that the Sequences are great; that is, more or less, why we’re all here. But the fact that Eliezer wrote the Sequences, and we saw that they were good, doesn’t help us very much. That we all agree on that is all well and good, but on what do we disagree? Something, clearly, but in what details?—that’s the question.
Therefore, in order to cultivate productivity, we should not attempt to imitate Eliezer by exploring new ideas, but instead write the other sorts of articles, which are easier to write productively.
I do not think this is true either. As mentioned above, I think that “exploratory research” is something Less Wrong can do well. It is, in fact, one of the few forums that has demonstrated this capacity. That is why it’s important that we preserve and nurture that rare and precious quality; that is why it’s important that we do “exploratory research” right.
And in a discussion of whether we, today, are doing something well, it makes no sense at all to reply that our forum’s founder, over a decade ago, before the forum even existed, did that thing well!
Thus my question to Romeo (and, I suppose, to anyone else who agrees with his view, but disagrees with mine) stands:
What are three of the best examples of good “exploratory research” articles from Less Wrong’s history? (The Sequences, and other posts from that period, excluded.)
You both latched onto the least interesting part of the post. The part that is literally just Romeo throwing out some wild speculations. The post would probably have been a bit cleaner to not mention the few wild speculations he mentions, but getting caught up on the tiny details seems to miss the forest from the trees.
The more interesting part is the general framework where he matches up the some of the processes mentioned in Buddhism with some insights from behavioral psychology, psychotherapy, and pop psychology. It gives a framework to start understanding why anecdotally people claim such big effects from extended meditation practice, and gives an insight about how one might begin to test the hypothesis that this is what’s happening, both personally and on a broader scale.
That is not the impression given by either the post or Romeo’s replies. He seems to be making it clear that he is, indeed, making the affirmative claims that I attributed to him, and has thus far given no indication of intending to retract or weaken them.
But let me ask you this: do you think the post would stand on its own, without these “speculations” (as you call them)? By way of attempting to answer this question, I’ve taken the liberty of creating a copy of Romeo’s post on Google Docs, where I’ve crossed out every “speculative” claim for which we’ve not been given evidence.
I was fairly conservative in choosing parts to cross out; I’ve left intact, for example, almost everything that mentions or talks about forms of psychotherapy and other concepts from psychology. (Although, by rights those ought to go as well; we do not accept such claims without evidence ordinarily; why should we accept them when speaking about meditation or Buddhism? But perhaps speculating about the “softer” sciences is a lesser offense, so I mostly let these things stand.)
What is left, nevertheless, is a post that is missing any justifications or explanations of its core claims; a post essentially indistinguishable from any number of similar posts that have appeared on Less Wrong, since its relaunch almost two years ago.
Allow me to make a suggestion, then, for anyone who writes these sorts of posts in the future:
First, enumerate, explain, and cite the “insights from behavioral psychology, psychotherapy, and pop psychology” which you intend to use as explanations. Be specific in your claims, and be diligent in your references. Consult the latest available sources, to make sure that your insights of choice have survived the replication crisis; provide these sources to your readers.
Then, having established a basis for what follows, make all the interesting connections to Buddhism, meditation, or what have you.
This approach would kill two birds with one stone: it would prevent such distracting digressions as we’re now engaged in, and it would also strengthen your own understanding of the ideas that you are presenting.
It seems to me that doing things in this way would benefit everyone involved.
Thanks for the clear suggestions/feedback.
I don’t mind the speculation, just the wild ones. Putting forward a hypothesis based on your own experience is incredibly useful, especially if it makes novel predictions that people can then go test themselves.
I don’t know how for instance Romeo would have gotten the “electrical resistance” thing from his own experience, but many of the other tools I can see by him noticing similarities between what happens in meditation, and various other psychotherapies he has tried or read about. This is actually one place where I expect phenomenological experience to provide valuable hypotheses, which is why I think this sort of post can be credible enough to want to test/further explore the hypotheses.
The two parts I mentioned are simply the most obviously speculative and unjustified examples. I also don’t have any real reason to believe the vaguer pop psychology claims about building stories, backlogs, etc.
It seems to me LW has a big epistemic hygiene problem, of late. We need to collectively stop make excuses for posting wild speculations as if they were fact, just because the same post also contains some interesting ‘insight’. We should be downvoting such posts and saying to the author “go away and write this again, with the parts that you can’t justify removed”.
Doing so may reveal that the ‘insight’ either does or does not successfully stand alone when the supporting ‘speculation’ is removed. Either way, we learn something valuable about the alleged insight; and we benefit directly by not spreading claims that lack evidence.
The tacit claim is that LW should be about confirmatory research and that exploratory research doesn’t belong here. But confirmatory, cited research has never been the majority of content going back to LW 1.0.
I cannot agree with your reading of nshepperd’s comment. I concur with you that exploratory research is entirely appropriate for Less Wrong, and indeed that it is a strength of this forum.
However, with respect, I do not think that your post rises to the level of “exploratory research”. The level of scholarship and rigor on display would need to be improved substantially, before we could label the post as “research” of any kind.
I have seen firsthand what exploratory research is like. Even in the relatively “soft” field I’ve had experience with (HCI), the post at hand would not qualify—not by a long shot.
Perhaps, though, we might investigate the nature of our disagreement in another way. What would you say are three of the best examples of “exploratory research” from Less Wrong’s history? To avoid confounding factors, please limit your examples to the period before the relaunch of the new site; and also avoid Eliezer’s posts from the original Sequences period (as citing them as examples would illuminate nothing).
I feel like this needs it’s own post and discussion. There’s definitely a difference of opinion here worth clarifying.
FYI, I started working on a post this morning (which I’ll most likely refactor into a question before publishing) that explores the question of “what combination of norms and incentive gradients will best output useful progress on LessWrong”.
Will probably be at least a few days before posting, but figured I’d mention it here.
Why?
Firstly, because the Sequences were written almost entirely on Overcoming Bias, not on Less Wrong. That alone would suffice. (Less Wrong simply did not exist when Eliezer wrote most of the Sequences.)
Secondly, because we already know that Eliezer can (or could, at least) write interesting and useful things, and have interesting and useful ideas. The question is whether anyone else—specifically, anyone from the Less Wrong commentariat—has that ability; and how we should encourage it, and nurture it; and what epistemic standards, and what community norms, encourage good and interesting and useful and correct ideas. We are, after all, talking about what kind of posts are appropriate for Less Wrong today. So asking whether the site’s founder and originally primary contributor wrote anything of value (especially before the community was even founded) is not relevant.
ETA: tentatively deprecated, see here.
How is your point of view different, if at all, from the one critiqued here?
It seems like you’re saying LessWrong ought to have a closed canon (the Sequences) and expound on but not add to or accept a substantive critique of them.
I’m not saying anything even remotely like that. I… don’t actually know how you got that from what I wrote. The post you linked seems to have nothing at all to do with what I’m saying.
Clearly, there’s been some great miscommunication here, but I am unsure of what could be the source of it…
ETA: Tentatively deprecated, see here.
What do you mean by this, if not that you’re trying to figure out whether other people share some personal specialness Eliezer has?
If you’re not thinking of the past as an uncaused golden age and Eliezer as a legend of yore, what’s the relation between that question and the question of which kinds of post are appropriate here?
If the problem is that some posts, knowably to a group of people competent to implement shared standards, are neither interesting nor useful, and lack interesting useful ideas, then the obvious solution would be to declare those specific posts inappropriate.
It’s not “some personal specialness”; it’s the ability, inclination, wherewithal, knowledge, expertise, habit, etc., etc., to write posts and comments that are useful, interesting, and otherwise desirable to have insofar as they serve the goals of Less Wrong.
These qualities can be encouraged where present, they can be developed where absent, they can be selected for from among a population, and their application can be incentivized.
But it is clearly not the case that said qualities are simply present in anyone who gets it into their head to write a Less Wrong post.
How common are these salutary qualities? We don’t know (but not very common). How common are they among the current Less Wrong commentariat, in particular? We don’t know (hopefully more common than in the general population, but clearly not as common as we’d like). What community norms, what rules, contribute to increasing and maintaining their prevalence among the membership of the site? We don’t know.
The past is not an uncaused golden age, but whatever its causes were, they cannot possibly include “the norms and rules of Less Wrong”, because Less Wrong did not exist. Given that “what should be the norms and rules of Less Wrong” is, in fact, what we are discussing, said golden age (i.e., Eliezer writing the Sequences) is irrelevant.
As for Eliezer being a “legend of yore”… well, consider the following analogy, inspired by the very post you linked.
Suppose Einstein, late in life, founds a school for aspiring brilliant theoretical physicists. Some time passes, and I inquire of the administration whether their school has, in fact, produced any brilliant physics theories; indeed, can it? Are their teaching methods any good, even? “What makes you suspect otherwise?”, the administrators ask me. “Well,” say I, “name some brilliant theoretical phycisists who’ve come from your school.” “Why,” comes the reply, “there was Einstein!”
Would you find this reply persuasive?
No.
Obviously LessWrong does not produce writers of similar productivity to Eliezer.
You didn’t explain why that means we should discourage some but not other articles by non-Eliezers.
Huh? What are you referring to…?
Could you quote what thing I said in this thread, that you are summarizing as “we should discourage some but not other articles by non-Eliezers”?
I now think I may have misread you quite badly—see this comment—if so, thanks for your patience.
It seems like you’re arguing:
Articles that explore new ideas are harder to write productively than articles that don’t.
Eliezer, uniquely on LessWrong, wrote productive articles exploring new ideas.
Therefore, in order to cultivate productivity, we should not attempt to imitate Eliezer by exploring new ideas, but instead write the other sorts of articles, which are easier to write productively.
If that’s what you mean, then that just is a proposal for closure of primary canon, allowing only carefully explicated commentaries and analyses and extensions of existing ideas.
None of that is even in the vicinity of what I meant.
I do not, quite frankly, know how you got any of that from what I wrote in this thread. As far as I can see, your attempted summary of my points is simply one big non sequitur.
Are you sure you’re not just rounding off my comments to the nearest cached criticism?
If you reread what I’ve written and still believe the provided summary is fair, I’ll attempt to re-explain, I guess…
I don’t have any other good hypotheses for why, when asking for examples of what good exploratory research would look like, you specified that we should exclude Eliezer. I can generate one more, bad hypothesis, which is that you want to make sure we’re hitting a sufficiently low standard, since we should grade on a curve and not hold ourselves to the standard of matching Eliezer. But that seems wildly in conflict with a bunch of other stuff you’ve said here.
ETA: On reflection it’s plausible that I was misreading you and the latter really is what you meant the whole time. If so, oops, not sure how I made that mistake. Thanks for asking me to reread!
The latter is closer to what I meant, certainly.
As you took the time to reread my comments, it seems only fair that I should take the time to attempt another explanation, as perhaps a rewording will help to dispel any remaining confusion. I hope you’ll excuse my using your earlier comment as a jumping-off point, though I know you no longer endorse this interpretation of my view:
This is true. However, as I wrote in this comment, I believe “exploratory research” to be a (perhaps not unique, but certainly unusual) strength of Less Wrong. That such articles are harder to write only means that it is more important—given how few places on the internet have any capability to produce such writing—that we do these things well.
First, again, I do not think that it is sensible to view the Sequences as having been written on Less Wrong—not least because they, in fact, weren’t! (You will note, by the way, that I specified Eliezer’s writings from the Sequences period for exclusion—not all his writings!)
That aside, I do not think this quoted bit is true either; Eliezer’s contributions were not uniquely excellent. I can easily come up with examples of good “exploratory research” articles written on Less Wrong by people who aren’t Eliezer. I asked Romeo to provide examples of his own because I thought (and still think) that seeing what he considers to be good “exploratory research” from Less Wrong’s past would help to illuminate the substance of our disagreement.
Yes, we all agree (presumably) that the Sequences are great; that is, more or less, why we’re all here. But the fact that Eliezer wrote the Sequences, and we saw that they were good, doesn’t help us very much. That we all agree on that is all well and good, but on what do we disagree? Something, clearly, but in what details?—that’s the question.
I do not think this is true either. As mentioned above, I think that “exploratory research” is something Less Wrong can do well. It is, in fact, one of the few forums that has demonstrated this capacity. That is why it’s important that we preserve and nurture that rare and precious quality; that is why it’s important that we do “exploratory research” right.
And in a discussion of whether we, today, are doing something well, it makes no sense at all to reply that our forum’s founder, over a decade ago, before the forum even existed, did that thing well!
Thus my question to Romeo (and, I suppose, to anyone else who agrees with his view, but disagrees with mine) stands:
What are three of the best examples of good “exploratory research” articles from Less Wrong’s history? (The Sequences, and other posts from that period, excluded.)
Thank you, this makes more sense to me.