Meta-meta point: the fact that you used your own post as an example, rather than a post where you weren't involved, made a very large difference in how I engaged with this Moderating LessWrong article. It prevented me from finishing it. Specifically, when I got to the point where you quoted Benquo, I felt honor-bound to stop reading and go check the source material, which turned out to be voluminous and ambiguous enough that I ran out of time before I came back. This is because, long ago, I made it a bright-line rule to never read someone quoting someone they disagree with in a potentially-hostile manner without checking the original context. That wouldn'tve felt necessary if the examples were drawn from discussions where you weren't involved.
4Duncan_Sabien3yUpvoted; I think this is an entirely reasonable heuristic, and that it probably
validly protects you in somewhere between three-out-of-four and
ninety-nine-out-of-a-hundred instances.
However, I don't think that it's impossible to do the thing correctly/in a
principled manner (obviously, since I attempted it). I'd be interested in a
follow-up from you later in which you share your impression of something like
"this is an example of why I have that heuristic in the first place" versus "my
heuristic fired correctly, given starting conditions, but this ended up not
being an example of why I have that policy."
(Including detail, if you feel like offering it.)
0Duncan_Sabien3yAlthough, it occurs to me that if you're holding to this heuristic in all cases,
this also would've prevented you from reading a significant chunk of benquo's
comments (which also were quoting someone they disagreed with in a
potentially-hostile manner) without first making sure you had my original post
fresh in your mind.
9jimrandomh3yI don't think I have it nailed down quite precisely what triggers the TAP--in
practice it's actually more like a pattern-match with things I was seeing and
thinking about when I established the TAP years ago, and descriptively speaking
I think the trigger condition involves my information about a narrative having
been filtered through a single author. I will say that, when I later made a
serious effort to sort out what happened in detail (partly written up, not yet
posted), it involved a lot of switching back and forth between tabs and little
to no reliance on quotations.
(Meta, in case it's relevant to anyone: it's felt to me like LW has been deluged in the last few weeks by people saying things that seem very clearly wrong to me (this is not at all specific to this post / discussion), but in such a way that it would take a lot of effort on my part to explain clearly what seems wrong to me in most of the cases. I'm not willing to spend the time necessary to address every such thing or even most of them, but I also don't have a clear sense of how to prioritize, so I approximately haven't been commenting at all as a result because it just feels exhausting. I'm making an exception for this post more or less on a whim.)
There's a lot I like about this post, and I agree that a lot of benquo's comments were problematic. That said, when I put on my circling hat, it seems pretty clear to me that benquo was speaking from a place of being triggered, and I would have confidently predicted that the conversation would not have improved unless this was addressed in some way. I have some sense of how I would address this in person but less of a sense of how to reasonably address it on LW.
Endorsed as well, although I think I might have a disagreement re: how reasonable it is for the site to expect people to conform to certain very concrete and grokkable standards even while triggered. I would claim, for instance, that even in the Dragon Army posts, where I was definitely triggered, you can't find any examples of egregious or outright violations of the principles laid out in this essay, and that fewer than 15% of my comments would contain even moderate, marginal violations of them.
I note that this is a prediction that sticks its neck out, if anybody wants to try. I have PDFs of both threads, including all comments, that I share with anyone who requests them.
Highly endorse this. And this is fact might be the entirety of my crux of disagreement.
(expressing wariness about delving into the details here. I am not willing to delve into details that focus on the recent Benquo thread until I've had a chance to talk to Ben in more detail. Interested in diving into details re: past discussions I've had with Duncan, but would probably prefer to do that in a non-public setting because the nature-of-the-thing makes it harder)
[edit: I think I have cruxes separate from this, but they might be similar/entangled]
9Benquo3yI was definitely something like triggered, but as far as I can tell, specific
bounded attempts to tell me what I was doing wrong were easy for me to evaluate,
and comments that attempted to take my perspective seemed pretty far off and
substantially exacerbated this. Notably, I found it comparatively easy to engage
with Duncan's criticisms.
3Said Achmiz3yThat does seem pretty overreaching to me, though. I mean, it would certainly
have the effect of “slowing down the conversation”, in that I wouldn’t want to
converse with someone who used this sort of rhetorical move!
There has to be some way to add the hypothesis that someone is triggered into the conversation. Like, sure, maybe the given example doesn't cut it, and maybe it's hard/tricky/subtle/we won't get it on the first five tries/no one solution will fit all situations. And maybe you're pointing at something like, this isn't really a hypothesis, but is an assertion clothed so as to be sneakily defensible.
But people do get triggered, and LessWrong has got to be the kind of place where that, itself, can be taken as object—if not by the person who's currently in the middle of a triggered state, then at least by the people around them.
There has to be some way to add the hypothesis that someone is triggered into the conversation. … But people do get triggered, and LessWrong has got to be the kind of place where that, itself, can be taken as object—if not by the person who’s currently in the middle of a triggered state, then at least by the people around them.
I don’t agree with this at all. In fact, I’d say precisely the opposite: Less Wrong has got to be exactly the kind of place where “someone is triggered” should not be added into the conversation—neither by the “triggered” person, nor by others.
My emotional state is my own business. If we are having a conversation on Less Wrong, and I do something which violates some norm, by all means confront me about that violation. I will not use being “triggered” as an excuse, a justification, or even a thing that you have any obligation at all to consider; in return, you will not psychoanalyze me and ask things like whether I am “triggered”. That is—or, rather, that absolutely should be—the social contract.
On Less Wrong (or any similar forum), the interface that we implement is “person who does not, in the context of a conversation/debate/discussion/argument, have any
Maybe this "social contract" is a fine thing for LessWrong to uphold.
But rationalists should not uphold it, in all places, at all times. In fact, in places where active truth-seeking occurs, this contract should be deliberately (consensually) dropped.
Double Cruxing often involves showing each other the implementation details. I open up my compartments and show you their inner workings. This means sharing emotional states and reactions. My cruxes are here, here, and here, and they're not straightforward, System-2-based propositions. They are fuzzy and emotionally-laden expectations, movies in my head, urges in my body, visceral taste reactions.
The sixth virtue is empiricism. The roots of knowledge are in observation and its fruit is prediction. What tree grows without roots? What tree nourishes us without fruit? If a tree falls in a forest and no one hears it, does it make a sound? One says, “Yes it does, for it makes vibrations in the air.” Another says, “No it does not, for there is no auditory processing in any brain.” Though they argue, one saying “Yes”, and one saying “No”, the two do not anticipate any different experience of the forest. Do not ask which beli
Double Cruxing often involves showing each other the implementation details.
Then chalk up another reason to disdain Double Cruxing.
If you don’t want to open up your implementation details to me, that is cool. But we’re not going to go to the depths of truth-seeking together without it.
This (“go to the depths of truth-seeking together”) is certainly an attitude that I would not like to see become prevalent on Less Wrong.
(Noting disagreement with your position here, and willingness to expand on that at some other point in time when there are not 5 discussions going on on LW that I feel like I want to participate in more. Rough model outline is something like: "I think close friendships and other environments with high barriers to entry can indeed strongly benefit from people modeling each other's implementation details. Most of the time when environments with a lower standard of selection or trust try to do this, it ends badly, though it's not obvious to me that they always have to go badly, or whether there are hypothetical procedures or norms that allow this conversation to go well even in lower-trust environments, though I haven't yet seen such a set of norms robustly in action.")
9Said Achmiz3ySure; this is certainly a conversation I’m open to having, and I do understand
the limitations of time and attention. What you just outlined is also helpful; I
look forward to when you have the chance to expand on it.
Something that I guess I've never quite gotten is, in your view Said, what is Less Wrong for? In 20 years if everything on LW went exactly the way you think is ideal, what are the good things that would have happened along the way, and how would we know that we made the right call?
(I have my own answers to this, which I think I've explained before but if I haven't done a clear enough job, I can try to spell them out)
That’s a good question, but a tricky one to answer directly. I hope you’ll forgive me if (à la Turing) I substitute a different one, which I think sheds light on the one you asked:
In the time Less Wrong has existed, what has come out of it, what has happened as a result, that is good and positive; and, contrariwise, what has happened that is unfortunate or undesirable?
Here’s my list, which I expect does not match anyone else’s in every detail, but the broad outlines of which seem to me to be clearly reasonable. These lists are in no particular order, and include great and small things alike:
Pros
The Sequences
Certain other posts or sequences, such as many of Scott’s posts, Alicorn’s “luminosity” sequence, and a small handful of others
Interesting/useful new results in, e.g., decision theories
The recruitment of talented mathematicians/etc. to MIRI, and their resulting work on the alignment problem and related topics
The elevation of the AI alignment problem into mainstream consciousness
“Three Worlds Collide” (and, more broadly, the genre of “rationalist fiction”, which certainly includes a lot of dross, but also some gems, and breaks some fruitful new ground in fiction-space)
4. The recruitment of talented mathematicians/etc. to MIRI, and their resulting work on the alignment problem and related topics
5. The elevation of the AI alignment problem into mainstream consciousness
Cons
1. 2. rationalist communities
8. Almost everything that CFAR does and has done
9. The Effective Altruism movement
Just want to note that I think you may be underestimating the extent to which these things on your Cons list have contributed to these things on your Pros list.
For example:
The EA movement funds MIRI and other AI Safety efforts.
More generally, CFAR and the rationalist community have served as a funnel for MIRI.
The Future of Life Institute (which has promoted and helped fund work on the alignment problem) was founded by CFAR alumni who knew each other from the Boston rationalist community.
0Said Achmiz3yIf the point is “the Cons are not all bad; they are partly good, to the extent
that they contribute to (or are perhaps even necessary for?) the Pros”,
then—granted.
If the point is “the Cons are not bad at all, and the reasons for considering
them to be bad do not exist, because of the fact that they also contribute to
the Pros”, then that is revealed to be manifestly incoherent as soon as it’s
made explicit.
If the point is something else entirely, then I reserve judgment until
clarification.
If you found out some of those cons (or some close version of them) were necessary in order to achieve those pros, would anything shift for you?
For instance, if you see people acting to work on/improve/increase the cons... would you see those people as acting badly/negatively if you knew it was the only realistic way to achieve the pros?
(This is just in the hypothetical world where this is true. I do not know if it is.)
Like, what if we just live in a "tragic world" where you can't achieve things like your pros list without... basically feeding people's desire for community and connection? And what if people's desire for connection often ends up taking the form of wanting to live/work/interact together? Would anything shift for you?
(If my hypothetical does nothing, then could you come up with a hypothetical that does?)
2Said Achmiz3yThis question is incomplete. The corrected version would read:
“If you found out some of those cons (or some close version of them) were
necessary in order to achieve some of those pros, would anything shift for you?”
Given this, the answer is “of course, and it would depend on which Cons were
necessary in order to achieve which Pros”.
Now, you said that your question is a mere hypothetical, but let’s not
obfuscate: clearly, if not you, then at least other folks here, think that your
hypothetical scenario describes reality. But as Ray commented elsethread, this
is hardly the ideal context to hash out the details of this topic. So I won’t. I
will, however, ask you this:
Do you think that some of the Cons on my list are necessary in order to achieve
some of the Pros? (No need to provide details on which, etc.)
2Duncan_Sabien3y(shucks, now I'm kind of ashamed of my own reply to Said above, which is not
nearly as skillful as this)
6ESRogs3yYes, this is the point. (I wouldn't personally put it quite that way, since by
my own evaluation the things I mentioned -- EA, CFAR, rationalist communities --
are much better than "not all bad" makes it sound. But yes, it seems like
someone who values the things on your pros list should at least think that those
things are not all bad.)
For clarity -- when you say, "granted", do you mean, "Yes, I already believed
that, and I stand by my pros and cons list, as written." Or do you mean, "Good
point. You've given me an update, and I would no longer endorse the statement,
'Almost everything CFAR has done belongs on the con side of a pros-and-cons
list.'"?
If the former (such that you would still endorse the "Almost everything..."
statement), I would challenge whether that position is consistent with both 1)
highly valuing the things on your pros list, and also 2) having an accurate view
of the facts on the ground of what CFAR is trying to accomplish and has actually
accomplished.
I could see that position being consistent if you thought CFAR's other actions
were highly negative. But my guess is that you see them being closer to useless
(and widely overvalued), rather than so negative as to make their positive
contributions a rounding error.
In any case, I'm happy to table that debate if you'd like, as has been suggested
in other comments.
4Said Achmiz3yThe middle way, viz.:
Good point. You’ve given me an update, and I would still endorse the statement,
‘Almost everything CFAR has done belongs on the con side of a pros-and-cons
list.’
A reasonable challenge—or, rather, half of one; after all, what CFAR “is trying
to accomplish” is of no consequence. What they have accomplished, of course, is
of great consequence. I allow that I may have an inaccurate view of their
accomplishments. I would love to see an overview, written by a neutral third
party, that summarizes everything that CFAR has ever done.
I’m afraid your guess is mistaken (though I would quibble with the “rounding
error” phrasing—that is a stronger claim than any I have made).
4ESRogs3yThat's fair. I include the "trying" part because it is some evidence about the
value of activities that, to outsiders, don't obviously directly cause the
desired outcome.
(If someone says their goal is to cause X, and in fact they do actually cause X,
but along the way they do some seemingly unrelated activity Y, that is some
evidence that Y is necessary or useful for X, relative to if they had done Y and
also happened to cause X, but didn't have causing X as a primary goal.
In other words, independently of how much someone is actually accomplishing X,
the more they are trying to cause X, the more one should expect them to be
attempting to filter their activities for accomplishing X. And the more they are
actually accomplishing X, the more one should update on the filter being
accurate.)
2Said Achmiz3yI don’t think that I agree with this framing. (Consider the following to be a
sort of thinking-out-loud.)
Suppose that activity Y is, to an outsider (e.g., me), neutral in value—neither
beneficial nor harmful. You come to me and say: “I have done thing X, which you
take to be beneficial; as you have observed, I have also been engaging in
activity Y. I claim that Y is necessary for the accomplishment of X. Will you
now update your evaluation of Y, and judge it no longer as neutral, but in fact
as positive (on account of the fact—which you may take my word, and my
accomplishment of X, as evidence—that Y is necessary for X)?”
My answer can only be “No”. No, because whatever may or may not be necessary for
you to accomplish outcome X, nonetheless it is only X which is valuable to me.
How you bring X about is your business. It is an implementation detail; I am not
interested in implementation details, when it comes to evaluating your output
(i.e., the sum total of the consequences of all your actions).[1]
[#fn-Xo9rJGM8efT6gQRyK-1]
Now suppose that Y is not neutral in my eyes, but rather, of negative value. I
tally up your output, and note: you have caused X—this is to your credit! But,
at the same time, you have done Y—this I write down in red ink. And again you
come to me and say: “I see you take X to be positive, but Y to be negative; but
consider that Y is necessary for X [which, we once again assume, I may have good
reason to trust is the case]! Will you now move Y over to the other side of the
ledger, seeing as how Y is a sine qua non of X?”
And once again my answer is “No”. Whatever contribution Y has made to the
accomplishment of X, I have already counted them—they are included in the value
I place on X! To credit you again for doing Y, would be double-counting.[2]
[#fn-Xo9rJGM8efT6gQRyK-2] But the direct negative value of Y to me—that part has
not already been included in my evaluation of X; so indeed I am correct in
debiting Y’s value from your account.
An
4Raemon3yIt seems like there's a consistent disagreement here about how much
implementation details matter.
And I think it's useful to remember that things _are_ just implementation
details. Sometimes you're burning coal to produce energy, and if you wrap up
your entire thought process around "coal is necessary to produce energy" you
might not consider wind or nuclear power.
But realistically I think implementation details do matter, and if the best way
to get X is with Y... no, that shouldn't lead you to think Y is good
in-and-of-itself, but it should affect your model of how everything fits
together.
Understanding the mechanics of how the world works is how you improve how the
world works. If you abstract away all the lower level details you lose the
ability to reconfigure them.
2Said Achmiz3yI don’t disagree with what you say, but I’m not sure that it’s responsive to my
comments. I never said, after all, that implementation details “don’t matter”,
in some absolute sense—only (here) that they don’t matter as far as evaluation
of outcomes goes! (Did you miss the first footnote of the grandparent comment…?)
Yes, of course. But I am not the one doing any reconfiguring of, say, CFAR, nor
am I interested in doing so! It is of course right and proper that CFAR
employees (and/or anyone else in a position to, and with a motivation to,
improve or modify CFAR’s affairs) understand the implementation details of how
CFAR does the things they do. But what is that to me? Of academic or general
interest—yes, of course. But for the purpose of evaluation…?
6Raemon3yIt seemed like it mattered with regard to the original context of this
discussion, where the thing I was asking was "what would LW output if it were
going well, according to you?" (I realize this question perhaps unfairly implies
you cared about my particular frame in which I asked the question)
If LessWrong's job was to produce energy, and we did it by burning coal,
pollution and other downsides might be a cost that we weigh, but if we thought
about "how would we tell things had gone well in another 20 years?", unless we
had a plan for switching the entire plant over to solar panels, we should
probably expect roughly similar levels of whatever the costs were (maybe with
some reduction based on efficiency), rather than those downsides disappearing
into the mists of time.
[edit: mild update to first paragraph]
2Said Achmiz3ySure, but more importantly, what you asked was this:
[emphasis mine]
Producing energy by burning coal is hardly ideal. As you say upthread, it’s well
and good to be realistic about what can be accomplished and how it can be
accomplished, but we shouldn’t lose track of what our goals (i.e., our ideals)
actually are.
2Raemon3yOkay, coolio.
I'm not too worried about the conversation continuing in the manner is has, but
I'm pretty sure I've now covered everything I had to say before actually
drilling down into the details.
3Duncan_Sabien3yThere may need to be more buckets than "pro" and "con."
I propose "negative," "neutral," "positive," "instrumental," and "detrimental."
Thus you can get things like "negative and yet instrumental" or "positive and
yet detrimental," where the first word is the thing taken reasonably in
isolation and judged against a standard of virtue or quality, and the second
word is the ramifications of the thing's existence in the world in a long-term
consequentialist sense.
(So returning to my favorite local controversy, punching people is Negative, but
it's possible that punch bug might consequentially be Instrumental for societies
filled with good people that are overall on board with nonviolence and personal
sovereignty.)
Other categorizations might do better to clarify cruxes ... this was my attempt
to create a paradigm that would allow you to zero in on the actual substance of
disagreement.
2Said Achmiz3yLet’s not reinvent the wheel here.
You’re talking about means and ends (which, in a consequentialist framework,
are, of course, just “ends” and “other ends”).
(Your example may thus be translated as “punching people is negative ceteris
paribus, as it has direct, immediate, negative effects; however, the knock-on
effects, etc., may result in consequences which, when all aggregated and
integrated over some suitable future period, are net positive”. Of course this
gets us into the usual difficulties with aggregation, both intra-personally and
interpersonally, but these may probably be safely bracketed… at least,
provisionally.)
4Duncan_Sabien3yI'm talking about you and ESRogs zeroing in on where you disagree, because at
least one of you is wrong and has a productive opportunity to update. Sorry if
the example of punch bug was distracting, but I suspect fairly strongly that it
is inappropriate and oversimplified to just have a pros-and-cons list in the
case of these large evaluations you're making—not least because in a
black-or-white dichotomy, you lose resolution on the places where your
assumptions actually differ.
Confused and curious about why you put Kocherga in positives and all other rationalist social/community/meatspace things in negatives. I don't think the difference between the two is that large. (I'm a Bay Area rationalist-type person who has been to a couple of things at Kocherga)
0Said Achmiz3yThe thing is that “rationalist social/community/meatspace things” is a wrong
category. It lumps together things that are very different.
In general, the way that “rationalists” use the word (and concept) “community”
leads them into error, of exactly this sort. That makes it difficult to discuss
this sort of thing productively.
Unfortunately, untangling this is beyond the scope of a comment thread,
especially a borderline-off-topic one like this. At some point I may attempt it,
in top-level-post form.
9Wei_Dai3yI'm really curious about the cons (even 5 where I'm only aware of one scandal).
Can you link to some existing explanations, provide a summary of why you think
each item on your cons list is bad, and/or write up your thoughts in detail at
some point?
8Said Achmiz3yHmm. I’m not sure that would be a great idea. For all that I disagree strongly
(to say the least) with much of what I listed, still it seems to me that most of
the folks involved aren’t bad people; it doesn’t quite feel right to write up a
“this is everything I think is bad about Less Wrong and everyone involved with
it” sort of essay. Part of the reason I hesitate is that I don’t have all that
much skin in the game; I am not really a member of any of these “rationalist
communities”, so in a sense my criticisms will be those of an outsider. How much
value is there, in such a thing? I don’t know. Certainly many of those who are
closer to the things than I am, have had harsh enough things to say, on all the
subjects I listed. It hardly seems necessary for me to add to that.
I wrote the grandparent comment in order to communicate, to Ray (and the rest of
the LW team, and any others who may be in a position to affect the future of
Less Wrong), what my views on this matter are. It seems to me that I’ve
succeeded in that. So, as to your question… meaning no disrespect at all, I’d
prefer, if possible, not to turn this discussion into an airing of dirty
laundry.
9Wei_Dai3yCan you (or anyone else) link to such complaints? (For example I don't think
I've ever seen a complaint that almost all “rationalist communities” in the
United States or the concept of “rationalist community” is harmful on net.)
My model of the LW team is that they would disagree with a lot of your cons
prior to seeing your views, and seeing your views (without knowing the reasoning
behind them) wouldn't cause them to make a large update in your direction. Would
you agree with this, and if so how do you plan to cause them to update or
otherwise to accomplish the goal of "the list of Cons has not only not been
added to, but the current entries on it forgotten and passed into misty memory,
left far behind and overshadowed by the Pros"?
Why not? You mentioned "most of the folks involved aren’t bad people" but if
they are actually doing bad things surely it doesn't make sense to let them keep
doing those things just to spare their feelings?
FWIW, I appreciated Said giving a response that was a succinct but comprehensive answer – I think further details might make sense as a top-level post but would probably take this thread in too many different directions. I think there's something useful for people with really different worldviews being able to do a quick exchange of the high level stuff without immediately diving into the weeds.
5Said Achmiz3yTo a first approximation, nothing that anyone ever says to anyone, on any topic
on which the target already has any sort of opinion, causes them to make a large
update in the speaker’s direction. All we can hope for is small updates
(assuming we do not discard the “update” model altogether—which I rather think
it’s time we did; but that is a separate discussion).
If others, who are closer to the matter, have already spoken, and more
specifically, more critically, and more plainly, then what hope do I have of
stopping anyone from doing any bad things? My intent is to do what I can to
nudge Less Wrong toward the direction in which I think it should go. That is
all. I have no greater ambitions for this discussion.
5DanielFilan3yIs there any chance you could let those of us who speak English but not Russian
know what that is?
5Said Achmiz3yGoogle Translated version of the page
[https://translate.google.com/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=https%3A%2F%2Fkocherga-club.ru%2F&edit-text=&act=url]
English-language interview with one of Kocherga’s co-founders
[https://www.hse.ru/en/success/ko4erga/]
4Raemon3yThanks, that was helpful, and I appreciate the question-substitution.
1Duncan_Sabien3yOkay, largely convinced, given stronger norms around what sorts of behavior are
okay and what aren't. I think I was thinking that openly addressing
triggered-ness as object would be a helpful in finding the best way to return a
conversation to normalcy.
I still don't like the idea of having a particular set of hypotheses being
taboo; I can buy an instrumental argument that we might want to make an
exception around triggeredness that's similar to the exceptions around positing
that someone might have a lot of unacknowledged racist biases—
(I think we stay away from that on consequentialist grounds and not because we
don't form and check those hypotheses in subtle ways)
—but in general I think LW should be a place where there's always a correct,
dispassionate, epistemically careful, and socially neutral way to say pretty
much anything.
But overall, I like your ... Turing test? ... argument. "If it's posting like a
LWer, and replying like a LWer, and acting like a LWer, then it's a LWer;
doesn't matter what its internal state is." I'd be willing to give up a small
swath of conversational space, to purchase that. Indeed, other people misreading
me as triggered and playing status games (when that wasn't my experience and the
comments I'm making aren't evidence for that hypothesis except
circumstantially/via pattern-matching) has been a big headache for me.
9Said Achmiz3yIndeed.
Exactly. We can make it even more stark:
“Have you considered that maybe you only think that because you’re just really
stupid? What’s your IQ?”
“Have you considered that maybe you’re a really terrible person and a sociopath
or maybe just evil?”
[to a woman] “You seem angry, is it that time of the month for you?”
etc.
We don’t say these sorts of things. Any of them might be true. But we don’t say
them, because even if they are true, it’s none of our business. Really, the only
hypothesis that needs to be examined for “why person X is saying thing Y” is
“they think that it’s a good idea to say thing Y”.
Note that this is a very broad class of hypotheses. It’s much broader, in
particular, than merely “person X thinks that thing Y is [insofar as it
constitutes any sort of proposition(s)] true”. It excludes only things where you
say something, not because you’re consciously choosing to say it in the service
of some conversational (or other) goal, but because you’re compelled to say it,
by forces outside of your control.
And maybe you are. But to the extent that you do not choose to say a thing, but
are compelled to say it, we—your interlocutors—are not interacting with you.
Rather, we are interacting with the abstract person-interface
[https://meltingasphalt.com/personhood-a-game-for-two-or-more-players/] which
“you” are implementing, which—by specification—chooses to say and do things, and
is not compelled to do anything.
5Duncan_Sabien3yI'll note that, empirically, we do say these things. Or at least, people say
them to me, and they're net upvoted, and no one takes a public stance against
it, mods included. And I'm not just referring to benquo or to the overt troll in
the original Dragon thread, either.
(There's a BIG difference between, e.g., Ray silently private messaging Benquo,
and Ray saying out loud in the thread "I'm privately messaging Benquo about
this.")
7habryka3y(I am curious of examples of this, either here or via PM. I think I basically
agree with you that there were multiple period in which we haven't been able to
reliably moderate all content on LW, but I also care about setting the
historical record straight, and right now we have a bunch more resources for
moderation available than we had over the last few weeks, so it might still be
the correct call to add mod annotations to those threads, saying that these
things are over the line. I have somewhat complicated feelings about writing
publicly that we are in a private conversation with someone, since that does
tend to warp expectations a bunch, but I am still pretty open to making it a
policy that when we ping users about infractions in private, that we also make a
relevant note on the thread, and that the benefits might just reliably outweigh
the costs here.)
2Duncan_Sabien3yThe vast majority of the examples are in the two Dragon Army threads, one from
LW1 and the other from LW2, which are now Gone. I am willing to share the PDFs
with you (Ray already has them), but in this case there's no useful retroactive
action.
(The rest are in the thread quoted in this essay, and at my last check (last
night) are still un-addressed.)
9habryka3yI don't remember anything in the second Dragon Army thread that fit this
pattern, but it's been a while and I was pretty busy at the time and don't think
I was able to read everything before the thread got removed, so I would be
curious about the pdf.
Agree that there are things unresolved in the thread quoted here. I definitely
plan to address them, but currently want to wait until the private conversations
we are having with people come to a natural stop.
1Unreal3yi find this idea very distasteful
6habryka3yInterested in more input on this. It seems obvious to me that future readers of
the original Dragon Army thread should not think that writing stuff like the
numbers guy did, will not result in a ban or punishment. And since I want
LessWrong to be a timeless archive, it's important for historic discussion to be
kept similarly curated as present discussion.
2Unreal3yIf you only plan on annotating past discussions that have long-since died, I
mind a lot less. But for a discussion that is still live or potentially live, it
feels like standing on a platform and shouting through a loudspeaker. I'd
advocate for only annotating comments without any activity within the past X
months.
2habryka3yAh, yes. I was thinking of all the old stuff that is much older than that (such
as the original DA thread). Anything that's still active should have different
policies.
2Benquo3yCan you give an example of where I said this?
0Duncan_Sabien3yI suspect all mods would prefer that you and I not directly engage just yet,
until there's structure in place for a facilitated and non-weaponized
conversation.
3Benquo3yThe comment I was responding to was attributing an opinion to me. A norm (even a
temporary one) in which you can do that, but I can't ask for evidence, seems
like it ends up allowing whichever of us is more interested in the exercise to
snipe at the other unchallenged pretty much indefinitely.
I'm not interested in sniping at you right now, I'm just interested in people
parsing the literal comment of my comments (and your posts) and not attributing
to me things that I did not in fact say.
A norm (even a temporary one) in which you can do that, but I can't ask for evidence, seems like it ends up allowing whichever of us is more interested in the exercise to snipe at the other unchallenged pretty much indefinitely.
To be clear on my view (as a mod), it is fine for you to ask for evidence (note that habryka did as well, earlier), and also fine for Duncan to disengage. I suspect that the world where he disengages is better than the one where he responds, primarily because it seems to me like handling things in a de-escalatory way often requires not settling smaller issues until more fundamental ones are addressed.
I do note some unpleasantness here around the question of who gets "the last word" before things are handled a different way, where any call to change methods while a particular person is "up" is like that person attempting to score a point, and I frown on people making attempts to score points if they expect the type of conversation to change shortly.
As a last point, the word "indefinitely" stuck out to me because of the combination with "temporary" earlier, and I note that the party who is more interested in repeatedly doing the 'disengage until facilitated conversation' move is also opening themselves up to sniping in this way.
4Duncan_Sabien3yIn particular, there is something happening here that I notice myself wanting to
narrativize as weaponized disingenuousness (which is probably not Ben's
intention) that's like ...
... politely following the rules, over here in this thread, and by example of
virtuous action making me seem unreasonable for not wanting to reply ...
... whereas over in the other thread, I get the impression that this exact rule
is the one he was breaking (e.g. when he explicitly asserted that I want to
ghettoize people, when what I said was that we could treat people who found
punch bug norms highly costly in a manner analogous to how we treat people with
peanut allergies (to the best of my knowledge, there is no ghetto in which we
confine people with peanut allergies)).
It reminds me of the phrase peace treaties are not suicide pacts
[https://extranewsfeed.com/tolerance-is-not-a-moral-precept-1af7007d6376]. In
fact the norm Ben is pushing for here is one I already follow, the vast majority
of the time, except in cases where I see the other person as having already
repeatedly demonstrated that they don't hold themselves to the same standard. I
don't like being made to look bad for having a superseding principle prevent me
from proving, in this case, that I am in fact principled in this way, too.
My favorite world would be one in which someone else would reliably make points
such as this one, and so I could disengage in this particular
likely-to-be-on-tilt case, while also feeling that all the things which "need"
to be said will be taken care of.
8Benquo3yDuncan's comment here persuaded me to go search for cases where my use of
"ghetto" was ambiguous between quoting Duncan and making a claim about what his
proposal implied. I've added clarifying notes in the cases that seemed possibly
ambiguous to me. If anyone (including but not limited to Duncan) points out
cases I've missed, and I agree that they're potentially ambiguous, I'll be happy
to correct those as well.
I still stand by the claim, but it's important to distinguish that claim from a
false impression that Duncan said that he envisioned ghettoes for people who
don't want to play punchbug. He didn't say that.
One thing that makes Duncan's criticisms comparatively easy to evaluate here is
that he's grounding things in the object level text with a fairly high degree of
precision. I don't always agree with the criticisms, and sometimes strongly
dispute his characterization of what I meant (though that's at least evidence
that something I wrote was unclear), of course.
6Duncan_Sabien3yUpvoted, and appreciated on a visceral, emotional level.
There's a set of moderation-challenges that the post doesn't delve into, which are the ones I struggle most with – I don't have a clear model of what it'd mean to solve these, whereas the challenges pointed to in the OP seem comprehensible, just hard. I'm interested in thoughts on this.
1. Difficulty with moderating just-over-the-line comments, or non-legibly-over-the-line comments
The most common pattern I run into, where I'm not sure what to do, is patterns of comments from a given user that are either just barely over the line, or where each given comment is under the line, but so close to a line that repetition of it adds up to serious damage – making LW either not fun, or not safe feeling.
The two underlying generators I'm pointing at here seem to be:
Not actually approaching a discussion collaboratively. Sometimes this is legit bad faith, and sometimes it's just... not enough willingness to do interpretive labor, or a vague underlying sense that you're not trying to be on the same team.
Not being up-to-speed enough to contribute to a discussion. This feels harshest to take action on, even when I'm putting on my maximally friendly h
The most common pattern I run into, where I’m not sure what to do, is patterns of comments from a given user that are either just barely over the line, or where each given comment is under the line, but so close to a line that repetition of it adds up to serious damage – making LW either not fun, or not safe feeling.
What I used to do on the #lesswrong IRC was put every time I see someone make a comment like this into a journal, and then once I find myself really annoyed with them I open the journal to help establish the pattern. I'd also look at peoples individual chat history to see if there's a consistent pattern of them doing the thing routinely, or if it's a thing they just sometimes happen to do.
I definitely agree this is one of the hardest challenges of moderation, and I pretty much always see folks fail it. IMO, it's actually more important than dealing with the egregious violations, since those are usually fairly legible and just require having a spine.
My most important advice would be don't ignore it. Do not just shrug it off and say "well nothing I can do, it's not like I can tell someone off for being annoying". You most certainly can and should for many kinds of 'annoying'. The alternative is that the vigor of a space slowly gets sucked out by not-quite-bad-actors.
Not actually approaching a discussion collaboratively.
Not being up-to-speed enough to contribute to a discussion.
Yeah, these are two of the things that have been turning me off from trying to keep up with comments the most. I don't really have any ideas short of incredibly aggressive moderation under a much higher bar for comments and users than has been set so far.
6Zvi3yWorth noting that we currently have a lot of discussion going on, and the more
we have, the more inclined I am to support moderation raising its standards for
what is acceptable, and getting more aggressive. A shift from 'not generating
discussion' to 'generating lots of discussion some of which may be bad' changes
the situation a lot.
This article gave me a bunch of food for thought. I don't think it addresses my main cruxes re: previous disagreements I've had with Duncan, but it definitely gave me some new ideas and new vantage points to view old ones.
(Note 1: I won't be commenting on Duncan's comments on Benquo's comments because I'm still in the process with chatting with Benquo about it. I have a number of relevant disagreements with both Ben and Duncan, hope to resolve those disagreements at some point, but meanwhile don't have the bandwidth I'd require to engage with both of them at once)
Some thoughts so far:
1. Hierarchy of Goals
The hierarchy of "purposes of LessWrong" that Duncan describes is roughly the same one I'd describe. A concern or difference in framing I have here is that several the of the stages reinforce each other in a cyclical fashion.
I'm not quite sure you can cleanly prioritize truth over truthseeking culture.
If our culture isn't outputting useful accumulation of knowledge, then it's failing at our core mission. Definitely. But in the situations where truthseeking-culture vs truth seem to be in conflict, I think it... (read more)
5Duncan_Sabien3yI should note that I do think this post contains all of my cruxes re:
disagreement; i.e. that in every case where you've strongly disagreed with me
about norms or policies or how-a-given-conversation-should-go, the principle I
was acting on was among those laid out here.
(Most specifically: when it's justified to punch back, what counts as
not-meeting-the-standard-of-rationality, and how much leadership is obligated to
actively defend the right-but-unpopular.)
From my own past experience as a mod and admin, I predict that scale concerns
are miscalibrated. It's a lot at first, but just as some teachers wage an
unending losing battle against student misbehavior while others basically see no
problems at all, ever ... once the standard is set and enforcement is clear,
consistent, and credible, problems 90% stop occurring.
(I acknowledge that I have not fully responded to all of your points; I wanted
to register these things quickly but other stuff is probably worth responding to
later in other comments.)
8Raemon3yYeah, I can definitely imagine this being the case. I don't have strong opinions
on this concept, although I'm in part worried that doing it right involves a lot
of skill, and failing to do it right may make things worse.
4Raemon3yYeah, it makes sense that this is the parts that seem like most salient points
of disagreement. But I think it's fairly important (and has been the last couple
times we talked about this), that I agree with most of the cruxes listed here,
and yet disagree with the conclusion. So it's important that whatever is causing
the disagreement isn't actually covered here (or at least, not covered
sufficiently)
-3Duncan_Sabien3y... I note that there's a point in the near future at which continued lack of
any public action by the LW team stops being "we want to take our time and get
this right and not add fuel to the fire" and starts being a de facto endorsement
and a taking of sides (since the comments I claim are objectionable remain
visible to all, net upvoted, more than a week later, sans any moderating
influence or perspective).
Separately, I also note that benquo's made a comment here that I really really
really want to reply to directly, but that my model of the LW leadership prefers
no engagement until there's facilitation in place. I'm not clear on whether or
not there's a plan to make that happen.
Hm. I am in favor of high standards of discourse but I am remarkably resistant to Duncan imposing his high standards of discourse because he has remarkably different social/discourse intuitions from me.
5tcheasdfjkl3yAlso worried about selective/biased policing and missing out on important points
because the person making the point had trouble phrasing it well. (That's
sometimes the right tradeoff, but not always. Based on the examples in the post
I don't like where Duncan draws the line.)
1Duncan_Sabien3yI'm curious if you predict that I would endorse your description/summary of
where I draw the line (i.e. whether you predict that you can pass my ITT).
I'm also curious for your cruxes, and/or your description of how you would
recommend the example comments be moderated (including not at all if that's your
view).
(It seems useful to simply make the assertion, especially since it provides a
flag for other people to vote on, giving a rough sense of how popular your
stance is. But I think it would be even more useful if it led toward a
reciprocal sharing of your model.)
4tcheasdfjkl3yApologies for the lack of response here; a good and full response would take a
fair bit of time and energy and I'm rather short on both. I'm not sure if I'll
get around to responding at all, though I will if I can. I understand that what
I've written so far is vague and arguably hostile which is not a great
combination, I just think it's probably better than not saying the thing at all.
In the meantime, I would invite people who upvoted my above comments and have
more spoons than me to answer Duncan's questions in my place.
As I understand it, while "In defense of Punch Bang" in some parts argues people should not spend huge amount of attention on basically random noise, this calls for very high attention to detail on the part of moderators, like
"be attentive enough to be the one to catch the slipped-in hidden assumption in the middle of the giant paragraph, and to point out the uncharitable summary even when it’s carefully and politely phrased, and to follow the subthread all the way down to its twentieth reply".
This sounds surprisingly similar as a call why people should be diligently watching for microaggressions: so I would just point to the In Defense of Punch Bang for a reasonable counterargument.
4Duncan_Sabien3yThe difference is, In Defense of Punch Bug is a prescription for general
behavioral norms in society at large, whereas the moderator post is about how to
correctly execute a very specific leadership role within a very specific
subculture. I don't think there's any contradiction; they're advice for
different domains.
Sort of like if one person wrote an essay saying "don't sweat the small stuff"
but then also wrote a memoir about their work as a neurosurgeon and spent a LOT
of time talking about attention to detail. I think this is not incoherent.
(Also note that the inherent chillness of what I labeled collegiate culture
requires the relaxed approach to micro advocated in the punch bug post, rather
than an easily-offended or highly-charged or knee-jerk narrativized reaction to
perceived transgressions.)
6Jan_Kulveit3yI also don't think it is incoherent, just that it seems to me there is some
tension.
I understand both the general norms and the moderation as some kind of
bounded-optimization problems. In case of general norms, as I see it (which may
be different from DPB), big reason of why not to care about micro-things is,
people are limited in attention and cognitive capacity. If they were orders of
magnitude less limited, maybe they should care about subtler effects.
I assume the current LessWrong moderators are also attention and time
constrained. What I quoted seems to me like a call to invest a lot of effort
into improving details, mostly by creating negative feedback loops. I would
agree this may be what should be done, given order or two increase in moderation
resources. With the current capabilities IMO what is more important is high
level steering and creating positive feedback loops.
-2Duncan_Sabien3yWould you argue that current moderation is ideal (or at least very close), given
constraints and tradeoffs?
4Jan_Kulveit3yI think the current moderation is within 30% to the constrained optimum I can
imagine, and I also think there is cheaper room of improvement is in different
directions than you point toward.
0Duncan_Sabien3yCruxes and/or models? The assertion is good on its own, since it allows people
to vote their agreement or disagreement, but it doesn't really move us toward
convergence.
5Jan_Kulveit3yI'll try to write something explicitly but 1) it may take some time 2) I'm
afraid part of my models is now in "intuition"
[https://www.lesswrong.com/posts/yeADMcScw8EW9yxpH/a-sketch-of-good-communication#C277P8Ei66ZC54hTk]
black-box form. Generally my background in this is my past self participating in
translating, trying to design and implement norms in cs.wikipedia community
[https://www.lesswrong.com/posts/kAgJJa3HLSZxsuSrf/arbital-postmortem#b3NT5ks3AHq8GQR4x]
, a long time ago.
What is my baseline ... there actually is a place rather unlike Facebook or
4chan or other places you mention, which was quite successful in its stated
purpose - building highly usable body of aggregated NPOV knowledge, and
community around it: Wikipedia and it's community. IMO there is actually a lot
to learn from their rules and norms, so whenever community norms of WP and LW
differ, IMO it's worth to look into more details. Some differences arise because
different purposes, some because rationality, some are random broken-symmetry
effects, but I suspect significant part is just because LW is orders of
magnitude smaller and did not have time to develop something, or LW is getting
it wrong. (Part of WP norms and culture is in turn based on knowledge developed
previously on MeatBall)
Ok...back to the original question, one of the important directions where LW is
IMO suboptimal is
* Please do not bite the newcomers
[https://en.wikipedia.org/wiki/Wikipedia:Please_do_not_bite_the_newcomers]
Reason why is this neglected is that while if you are wronged on LW, you e.g.
write a blogpost, talk to people, etc. (and people notice, because it's you!),
if a random, bright, somewhat argumentative mathematician is "bitten" in her
first interaction with LW community, she just bounces, without leaving much
trace. The counterfactual damage is much less visible.
In contrast things like one very experienced, high status person writing
something wrong, harming other experienced, high status
4Duncan_Sabien3yI like this frame, and agree that there's a lot of value in comparing LW and WP.
I think having more defense for the bitten or bite-vulnerable is a big part of
my emotional motivation here, though it's because of empathy rather than game
theory (it seems justifiable under either).
I think it's helpful to have some sort of system to make sure that every comment gets read, but I think the ownership checkbox is potentially a bad way to do it. I'm mostly thinking of the incentives for moderators, here; it seems highly plausible that someone comes across a comment that feels off but they don't really know how to handle it; this means they don't want to say "yep, this one's mine" (because they don't want to handle it) but also not checking it is wrong somehow.
One of the things that I had considered when proposing the Sunshine Regiment was a 'report' button on every comment, available to all users, which was basically a "something about this should be handled by someone with more time and energy"--not necessarily "this post should get removed" but "oh man, I really want to see how Vaniver would respond to this comment," or something.
I also suspect there's something like Stack Exchange's edit queue that could be good, where several of the important pieces are 1) multiple eyes on any particular thing and 2) tracking how much people have eyes on things and 3) tracking when people's judgments disagree.
3zulupineapple3yWhy not just make a comment about it, as a reply to the offending post? E.g.
"There seem to be some problems in this thread that a third-party could help
with, but I'm not willing to do it, so don't reply to this". If that was an
established norm, then I think it's reasonably likely that such comment would be
noticed by someone (especially since we have "recent discussion" in the home
page).
Whether that would actually solve problems or just create more, is a separate
question. But trying things is generally a good idea.
8Vaniver3yIt is generally wise to solve social problems with tech, when possible. I also
think it is important here to have the someone who does the noticing be someone
who actually has the relevant skills, rather than thinking they have the
relevant skills (both because of overconfident bystanders and because of
underconfident skilled people, who won't feel licensed to point out such
problems unless handed a literal license to do so).
1zulupineapple3yYes, but giving people licenses is pretty easy. I'd be fine with you having one,
for example, though I guess I don't have the power to give it to you myself.
The problem is that tech takes time and effort to write, so writing tech to
solve problems that it may not actually solve is unwise. What I'm proposing is a
temporary prototype of some sort. If that worked out, then I agree, a proper
tech solution would be nice.
I think one of the main things I updated on is that a model I had of user discussions with users (that it's important to hug the most important part of the query, i.e. chasing and focusing on cruxes) is not applicable to mod actions (where it's important to catch each of the norm violations, instead of simply the most serious violation).
There are a handful of reasons for this, the most serious of which is the "that which is not punished is allowed"--Raemon made a comment on the Benquo's top comment that pointed out what Raemon saw as a defect in Benquo's presentation, and separated that from a criticism of Benquo's argument. But that meant that later, when I criticized the argument, Benquo responded with 'but Raemon wasn't criticizing my argument.' If I only point out what seems to me like the most serious error in a post, then there's a (reasonable!) argument that all the things that went unsaid weren't at the threshold of correction, and if I only point out the error that's easier to respond to, the same sort of thing goes through.
4Zvi3yIt's worth pointing out that in this case, Raemon explicitly said he wasn't
criticizing the argument itself, and he wrote a distinct comment to that effect.
I do think the general point is valid and a serious problem, though; if I
criticize X it can be read as giving a pass for Y even though it shouldn't be
(and the even worse parallel problem is when I give reason/argument X for my
point, and people then assume that means there's no additional good
reason/argument Y or Z, or otherwise don't actually add together evidence from
diverse angles, especially when they're relatively illegible).
Obvious brainstorm is, would it be sufficient to say 'The largest issue with' or
something like that, when you suspect there are additional problems? Or would
this create a worse burden; e.g. now if I don't say that, it is real evidence
that there are no additional problems, so now the burden to say anything goes
up, and oh no...
I really like when people put effort into providing alternatives-to/critiques-of how I work, so thanks for this.
It doesn’t mean I can always reply promptly, alas. Right now I’m in the process of taking all my finals (so I can get a visa and move to the Bay), and this will continue for a few weeks :(. Oli’s just coming back from the same, while Ray’s holding the fort.
It’s also regularly the case that doing detailed public write-ups of considerations around a decision/approach isn’t the right use of effort relative to just *building the product*, and that ap... (read more)
6habryka3yI am properly back since a few days ago, so I will probably write up my thoughts
on this sometime soon. I am quite excited about this post, and am looking
forward to the discussion around it.
This seems like it's pointing at a good thing. As a data point, the proposed responses to my comments would all have seemed friendly and helpful to me, and I'd have had an easy time engaging with the criticism.
His draft response to my VW comment would probably have motivated me to add a note to my initial comment deprecating that reference, or edit it out entirely (with a note clarifying that an edit had been made).
The asymmetry comment he pointed to as helpful (thanks, Vaniver!) actually would have been helpful, if SilentCal hadn't already taken the initiative (thanks, SilentCal!) and clarified what he took the term to mean.
I'm concerned that the described examples of holding individual comments to high epistemic standards don't seem to necessarily apply to top-level posts, or linked content- one reason I think this is bad is that it is hard to precisely critique something which is not in itself precise, or which contains metaphor, or which contains example-but-actually-pointing-at-a-class writing where the class can be construed in various different ways.
Critique of fuzzy intuitions and impressions and feelings often involves fuzzy intuitions and impressions and fe... (read more)
6Duncan_Sabien3yIf I've given the impression that I object to and want to restrict fuzziness
itself, then I'd like to clarify and reverse that impression ASAP. I think it's
entirely possible to do fuzzy intuitions and impressions and feelings in a
high-quality, high-epistemic way, and I think it's a critical part of the
overall ecosystem of ideas; as an example I'd point to ~the entire corpus of
work written by Anna Salamon.
I don't think there's a conflict between freedom of fuzziness and high
standards; the key is to flag things. If it's an intuition, note that it's an
intuition; if there's no formal argumentation behind it, note that; if you think
it would be resistant to update and also think that's correct, just be open
about that up front. That way, people who want to engage know what they're
dealing with, and know which lines of approach will be useful versus which will
be futile.
A sort of cheesy, empty example, since I'm trying to come up with it on the fly
and there's no actual topic at hand:
I'd upvote a comment like that in a heartbeat.
4Duncan_Sabien3yStrong agree. But I claim that I have seen myself and others attempt to do
exactly that in response to damaging falsehoods, and that the response has been
criticism of defensiveness.
Like, the point wasn't to salvage personal status; the point was the original
attack contained falsehoods, and this was dismissed as if the only activity
taking place was a status fight. There was a bucket error going on, a bucket
error that I think the ideal LW would have incentive slopes against, and
encourage people to grow out of being vulnerable to.
A quote that I believe I cited at the time, which feels relevant here, too:
... I believe that one of the foremost goals of LessWrong, at least from a
social norms standpoint, is to become the sort of place where Draco could say "I
think you should seriously consider the possibility that I didn't cast
Colloportus as strongly as I could have, and that therefore Hermione
counterspelling it isn't conclusive re: our relative magical strengths" and not
get laughed off stage. And I think that mods who care about that and are unified
in their commitment to it are a crucial ingredient.
(Edit: have updated the specific quote in the post to more clearly point at what
I'm actually advocating.)
4Zvi3yOne thing I noticed by writing a blog slash posting here, is that the rules for
posts and comments are importantly different. In comments you have discussion
norms in a way you don't for main posts, and for main posts you have an
on-the-record property that comments don't have. So when writing comments, one
needs to be very careful about certain types of norms of fairness and
politeness, and forms of argumentation, and be even more careful about
rhetorical flourish.
But in exchange there's an understanding that you're not held to your statements
and positions (outside of the comment section itself, at least) the way a post
would be. Thus, I draw a distinction that if I put something in a post it is
'fair game' to be quoted and held as my position in an outside context, whereas
a comment doesn't do that, it's on the record but it's exploratory, and in
several cases I've used it to say things in comments that I didn't feel
comfortable saying in full posts.
Right now I believe the way we handle issues with top-level posts is to not
promote them to front page if there are problems, and only curate them if
they're excellent, combined with lots of voting, which seems pretty good. When I
do things that are against LW norms (which I occasionally do since all my blog
posts are auto-posted here) I have no problem getting the message that I did
that (whether or not I already knew that I'd done that) through these systems.
There is a potential discussion to be had someday about whether upvotes and downvotes themselves can be considered to be "in error" given the specific milieu of LW, or whether upvotes and downvotes are sacred à la free speech. Looking at votes on this and other threads, I have frequently had the sense that people were Doing It (Objectively) Wrong, not fundamentally as in wrong-opinions, but culturally as in participation-in-this-culture-means-precommitting-to-supporting-X-and-othering-Y.
I'm aware that there is a strongly-felt libertarian ar... (read more)
5Raemon3yThis question and its subcomponents are type erroring for me or something.
I think people can downvote/upvote in ways that are not game theoretically
optimal given their values.
I think criticizing people's upvotes/downvotes probably tends to lead to bad
outcomes. (Citation needed, but this is what I currently believe)
The notion that you shouldn't criticize people's votes because it leads to bad
outcomes feels white to me rather than red (which may have actually been what
you meant but my impression was you meant Red to be the color of "vote however
you want")
-2Duncan_Sabien3yAgreed. But as long as I'm personally forming lists of things that LW culture
should get right that most internet culture fails at abysmally ... I don't know
whether "users can criticize other users' votes without the conversation being
derailed into status fights" should be on that list or not.
There are certainly comments where the fact of their very high or very low karma
scores seems to me to be meaningful evidence re: whether it makes sense to be
hopeful about the LessWrong project in general.
Am gathering a little informal data on this on FB now; will report back later.
(EDIT: 70 votes, 91% in favor of the position that votes themselves can be
considered to be in error; no data collected on how to operationalize that in
the social space.)
2Duncan_Sabien3yHere's a comment that I think should be defended, on the ideal LW (which maybe
it would've anyway in the natural course of affairs; it's only been up for a few
hours and it's only been downvoted by a few people). It's a bit snide/aggressive
in its framing, but it raises a completely valid point, and if the snideness is
a problem it can be objected to separately from the content.
In response, I'd say yeah—I clearly overreached in my phrasing. What I actually
meant to convey was something like, different from the majority of "typical"
places on the internet. I wanted to convey the sense that LW ought to be
something like 98th percentile on this axis, but I was incorrect to imply that
the ideal LW would be at the absolute far right tail of all websites and
discussion fora, and didn't actually mean that in the first place.
FYI, AFAICT I am the only person who downvoted it (bringing it from 3 to -3). As long as we're having the meta-est of conversations, I wanted to talk about why.
The single most common criticism I hear from good writers on LW, is that the comments feel nitpicky in a way that isn't actually helpful, and that you have to defend your points against a hostile-seeming crowd rather than collaboratively building something.
This comment seemed to be doing two things in that direction: a) it was bit aggressive as you noted, b) the point it was making just... didn't seem very relevant. Yes, there are places on the internet aspiring to similar things as LW. But a reasonable read on your statement was "most of the internet isn't trying to do this much at all, LW is different." While some humility is nice, I really don't understand your point better now that you've rephrased it to the 98th percentile thing.
So the main impact of this sort of comment is both to spend time on things that don't matter much, and increase the latent hostility of the thread, and I think people don't appreciate enough how bad that is for discourse. Both of that seem like something to me that is better silently downvoted than engaged with.
1Duncan_Sabien3yOh, yeah, there's maybe another crux between us—while I buy the argument that a
policy of silently downvoting this class of comment may be more sustainable in
the long run, I don't share your instincts re: demon threads (to the point that
I'm relatively dismayed that that label for them is starting to stick, because
it implies hopelessness or helplessness or external-locus-of-control that I
don't feel). My approach to threads like those is maybe best summed up by "Let
the Light win, and if trouble comes of it ... let the Light win again."
EDIT: Also, I'm pretty sure other up- and downvotes are in the mix by now,
because I've seen it at a lot of other scores in the interim. I think -4 and 0
and -6.
-3zulupineapple3yIf you see a comment that is technically correct but nitpicky and unhelpful, you
could reply "this is technically correct, but nitpicky and unhelpful".
Downvoting correct statements just looks bad.
I think there is a more charitable reading of TAG's comment. Not only are there
places in the internet aspiring to find the truth, there are, in fact, very few
places that are not aspiring to find it. The point isn't that there are more
places like LW. The point is that "truth seeking" isn't the distinguishing
characteristic of LW.
I honestly believe that attacking people's points is a good way to learn
something. I don't know what you mean by "collaboratively building something",
I'd appreciate examples where that has happened in the past. I suspect that
you're overestimating how valuable or persistent this "something" is.
I don't think you've provided strong arguments that it actually is bad for
discourse. Yes, demon threads don't usually go anywhere, but regular threads
don't usually go anywhere either. And people can actually learn from demon
threads, even if they're not willing to admit it right away. I certainly have.
Not only are there places in the internet aspiring to find the truth, there are, in fact, very few places that are not aspiring to find it.
?? If you look at the Alexa top 50 sites for the US, how many of them are about aspiring to find the truth? I count between 3 and 4 (Google, Wikipedia, and Bing for sure, Wikia maybe).
4Duncan_Sabien3yYou forgot amazon.com and walmart.com, which have the tightest instrumental
rationality feedback loops of them all.
</banter>
0zulupineapple3yI think you're confusing "aspiring to find truth" with "finding truth". Your
crackpot uncle who writes facebook posts about how Trump eats babies isn't doing
it because he loves lies and hates truth, he does it because he has poor
epistemic hygiene.
So in this view almost every discussion forum and almost every newspaper is
doing their best to find the truth, even if they have some other goals as well.
Also, of course, I'm only counting places that deal with anything like
propositions at all, and excluding things like jokes, memes, porn, shopping,
etc, which is a large fraction of the internet.
Meta-meta point: the fact that you used your own post as an example, rather than a post where you weren't involved, made a very large difference in how I engaged with this Moderating LessWrong article. It prevented me from finishing it. Specifically, when I got to the point where you quoted Benquo, I felt honor-bound to stop reading and go check the source material, which turned out to be voluminous and ambiguous enough that I ran out of time before I came back. This is because, long ago, I made it a bright-line rule to never read someone quoting someone they disagree with in a potentially-hostile manner without checking the original context. That wouldn'tve felt necessary if the examples were drawn from discussions where you weren't involved.
(Meta, in case it's relevant to anyone: it's felt to me like LW has been deluged in the last few weeks by people saying things that seem very clearly wrong to me (this is not at all specific to this post / discussion), but in such a way that it would take a lot of effort on my part to explain clearly what seems wrong to me in most of the cases. I'm not willing to spend the time necessary to address every such thing or even most of them, but I also don't have a clear sense of how to prioritize, so I approximately haven't been commenting at all as a result because it just feels exhausting. I'm making an exception for this post more or less on a whim.)
There's a lot I like about this post, and I agree that a lot of benquo's comments were problematic. That said, when I put on my circling hat, it seems pretty clear to me that benquo was speaking from a place of being triggered, and I would have confidently predicted that the conversation would not have improved unless this was addressed in some way. I have some sense of how I would address this in person but less of a sense of how to reasonably address it on LW.
There's something in Duncan... (read more)
Endorsed as well, although I think I might have a disagreement re: how reasonable it is for the site to expect people to conform to certain very concrete and grokkable standards even while triggered. I would claim, for instance, that even in the Dragon Army posts, where I was definitely triggered, you can't find any examples of egregious or outright violations of the principles laid out in this essay, and that fewer than 15% of my comments would contain even moderate, marginal violations of them.
I note that this is a prediction that sticks its neck out, if anybody wants to try. I have PDFs of both threads, including all comments, that I share with anyone who requests them.
I am curious about the trivial inconvenience, here; why not just share the PDFs with a link, instead of requiring people to ask you for them first?
Highly endorse this. And this is fact might be the entirety of my crux of disagreement.
(expressing wariness about delving into the details here. I am not willing to delve into details that focus on the recent Benquo thread until I've had a chance to talk to Ben in more detail. Interested in diving into details re: past discussions I've had with Duncan, but would probably prefer to do that in a non-public setting because the nature-of-the-thing makes it harder)
[edit: I think I have cruxes separate from this, but they might be similar/entangled]
There has to be some way to add the hypothesis that someone is triggered into the conversation. Like, sure, maybe the given example doesn't cut it, and maybe it's hard/tricky/subtle/we won't get it on the first five tries/no one solution will fit all situations. And maybe you're pointing at something like, this isn't really a hypothesis, but is an assertion clothed so as to be sneakily defensible.
But people do get triggered, and LessWrong has got to be the kind of place where that, itself, can be taken as object—if not by the person who's currently in the middle of a triggered state, then at least by the people around them.
I don’t agree with this at all. In fact, I’d say precisely the opposite: Less Wrong has got to be exactly the kind of place where “someone is triggered” should not be added into the conversation—neither by the “triggered” person, nor by others.
My emotional state is my own business. If we are having a conversation on Less Wrong, and I do something which violates some norm, by all means confront me about that violation. I will not use being “triggered” as an excuse, a justification, or even a thing that you have any obligation at all to consider; in return, you will not psychoanalyze me and ask things like whether I am “triggered”. That is—or, rather, that absolutely should be—the social contract.
On Less Wrong (or any similar forum), the interface that we implement is “person who does not, in the context of a conversation/debate/discussion/argument, have any
... (read more)Maybe this "social contract" is a fine thing for LessWrong to uphold.
But rationalists should not uphold it, in all places, at all times. In fact, in places where active truth-seeking occurs, this contract should be deliberately (consensually) dropped.
Double Cruxing often involves showing each other the implementation details. I open up my compartments and show you their inner workings. This means sharing emotional states and reactions. My cruxes are here, here, and here, and they're not straightforward, System-2-based propositions. They are fuzzy and emotionally-laden expectations, movies in my head, urges in my body, visceral taste reactions.
... (read more)Then chalk up another reason to disdain Double Cruxing.
This (“go to the depths of truth-seeking together”) is certainly an attitude that I would not like to see become prevalent on Less Wrong.
(Noting disagreement with your position here, and willingness to expand on that at some other point in time when there are not 5 discussions going on on LW that I feel like I want to participate in more. Rough model outline is something like: "I think close friendships and other environments with high barriers to entry can indeed strongly benefit from people modeling each other's implementation details. Most of the time when environments with a lower standard of selection or trust try to do this, it ends badly, though it's not obvious to me that they always have to go badly, or whether there are hypothetical procedures or norms that allow this conversation to go well even in lower-trust environments, though I haven't yet seen such a set of norms robustly in action.")
Something that I guess I've never quite gotten is, in your view Said, what is Less Wrong for? In 20 years if everything on LW went exactly the way you think is ideal, what are the good things that would have happened along the way, and how would we know that we made the right call?
(I have my own answers to this, which I think I've explained before but if I haven't done a clear enough job, I can try to spell them out)
That’s a good question, but a tricky one to answer directly. I hope you’ll forgive me if (à la Turing) I substitute a different one, which I think sheds light on the one you asked:
In the time Less Wrong has existed, what has come out of it, what has happened as a result, that is good and positive; and, contrariwise, what has happened that is unfortunate or undesirable?
Here’s my list, which I expect does not match anyone else’s in every detail, but the broad outlines of which seem to me to be clearly reasonable. These lists are in no particular order, and include great and small things alike:
Pros
Just want to note that I think you may be underestimating the extent to which these things on your Cons list have contributed to these things on your Pros list.
For example:
If you found out some of those cons (or some close version of them) were necessary in order to achieve those pros, would anything shift for you?
For instance, if you see people acting to work on/improve/increase the cons... would you see those people as acting badly/negatively if you knew it was the only realistic way to achieve the pros?
(This is just in the hypothetical world where this is true. I do not know if it is.)
Like, what if we just live in a "tragic world" where you can't achieve things like your pros list without... basically feeding people's desire for community and connection? And what if people's desire for connection often ends up taking the form of wanting to live/work/interact together? Would anything shift for you?
(If my hypothetical does nothing, then could you come up with a hypothetical that does?)
Confused and curious about why you put Kocherga in positives and all other rationalist social/community/meatspace things in negatives. I don't think the difference between the two is that large. (I'm a Bay Area rationalist-type person who has been to a couple of things at Kocherga)
FWIW, I appreciated Said giving a response that was a succinct but comprehensive answer – I think further details might make sense as a top-level post but would probably take this thread in too many different directions. I think there's something useful for people with really different worldviews being able to do a quick exchange of the high level stuff without immediately diving into the weeds.
Well, empirically, we also say the stuff about being triggered. I’m saying that we shouldn’t say either sort of thing.
To be clear on my view (as a mod), it is fine for you to ask for evidence (note that habryka did as well, earlier), and also fine for Duncan to disengage. I suspect that the world where he disengages is better than the one where he responds, primarily because it seems to me like handling things in a de-escalatory way often requires not settling smaller issues until more fundamental ones are addressed.
I do note some unpleasantness here around the question of who gets "the last word" before things are handled a different way, where any call to change methods while a particular person is "up" is like that person attempting to score a point, and I frown on people making attempts to score points if they expect the type of conversation to change shortly.
As a last point, the word "indefinitely" stuck out to me because of the combination with "temporary" earlier, and I note that the party who is more interested in repeatedly doing the 'disengage until facilitated conversation' move is also opening themselves up to sniping in this way.
There's a set of moderation-challenges that the post doesn't delve into, which are the ones I struggle most with – I don't have a clear model of what it'd mean to solve these, whereas the challenges pointed to in the OP seem comprehensible, just hard. I'm interested in thoughts on this.
1. Difficulty with moderating just-over-the-line comments, or non-legibly-over-the-line comments
The most common pattern I run into, where I'm not sure what to do, is patterns of comments from a given user that are either just barely over the line, or where each given comment is under the line, but so close to a line that repetition of it adds up to serious damage – making LW either not fun, or not safe feeling.
The two underlying generators I'm pointing at here seem to be:
- Not actually approaching a discussion collaboratively. Sometimes this is legit bad faith, and sometimes it's just... not enough willingness to do interpretive labor, or a vague underlying sense that you're not trying to be on the same team.
- Not being up-to-speed enough to contribute to a discussion. This feels harshest to take action on, even when I'm putting on my maximally friendly h
... (read more)What I used to do on the #lesswrong IRC was put every time I see someone make a comment like this into a journal, and then once I find myself really annoyed with them I open the journal to help establish the pattern. I'd also look at peoples individual chat history to see if there's a consistent pattern of them doing the thing routinely, or if it's a thing they just sometimes happen to do.
I definitely agree this is one of the hardest challenges of moderation, and I pretty much always see folks fail it. IMO, it's actually more important than dealing with the egregious violations, since those are usually fairly legible and just require having a spine.
My most important advice would be don't ignore it. Do not just shrug it off and say "well nothing I can do, it's not like I can tell someone off for being annoying". You most certainly can and should for many kinds of 'annoying'. The alternative is that the vigor of a space slowly gets sucked out by not-quite-bad-actors.
Yeah, these are two of the things that have been turning me off from trying to keep up with comments the most. I don't really have any ideas short of incredibly aggressive moderation under a much higher bar for comments and users than has been set so far.
This article gave me a bunch of food for thought. I don't think it addresses my main cruxes re: previous disagreements I've had with Duncan, but it definitely gave me some new ideas and new vantage points to view old ones.
(Note 1: I won't be commenting on Duncan's comments on Benquo's comments because I'm still in the process with chatting with Benquo about it. I have a number of relevant disagreements with both Ben and Duncan, hope to resolve those disagreements at some point, but meanwhile don't have the bandwidth I'd require to engage with both of them at once)
Some thoughts so far:
1. Hierarchy of Goals
The hierarchy of "purposes of LessWrong" that Duncan describes is roughly the same one I'd describe. A concern or difference in framing I have here is that several the of the stages reinforce each other in a cyclical fashion.
I'm not quite sure you can cleanly prioritize truth over truthseeking culture.
If our culture isn't outputting useful accumulation of knowledge, then it's failing at our core mission. Definitely. But in the situations where truthseeking-culture vs truth seem to be in conflict, I think it... (read more)
Hm. I am in favor of high standards of discourse but I am remarkably resistant to Duncan imposing his high standards of discourse because he has remarkably different social/discourse intuitions from me.
To me it seems there is a certain tension between In Defense of Punch Bug and this post.
As I understand it, while "In defense of Punch Bang" in some parts argues people should not spend huge amount of attention on basically random noise, this calls for very high attention to detail on the part of moderators, like
"be attentive enough to be the one to catch the slipped-in hidden assumption in the middle of the giant paragraph, and to point out the uncharitable summary even when it’s carefully and politely phrased, and to follow the subthread all the way down to its twentieth reply".
This sounds surprisingly similar as a call why people should be diligently watching for microaggressions: so I would just point to the In Defense of Punch Bang for a reasonable counterargument.
I think it's helpful to have some sort of system to make sure that every comment gets read, but I think the ownership checkbox is potentially a bad way to do it. I'm mostly thinking of the incentives for moderators, here; it seems highly plausible that someone comes across a comment that feels off but they don't really know how to handle it; this means they don't want to say "yep, this one's mine" (because they don't want to handle it) but also not checking it is wrong somehow.
One of the things that I had considered when proposing the Sunshine Regiment was a 'report' button on every comment, available to all users, which was basically a "something about this should be handled by someone with more time and energy"--not necessarily "this post should get removed" but "oh man, I really want to see how Vaniver would respond to this comment," or something.
I also suspect there's something like Stack Exchange's edit queue that could be good, where several of the important pieces are 1) multiple eyes on any particular thing and 2) tracking how much people have eyes on things and 3) tracking when people's judgments disagree.
I think one of the main things I updated on is that a model I had of user discussions with users (that it's important to hug the most important part of the query, i.e. chasing and focusing on cruxes) is not applicable to mod actions (where it's important to catch each of the norm violations, instead of simply the most serious violation).
There are a handful of reasons for this, the most serious of which is the "that which is not punished is allowed"--Raemon made a comment on the Benquo's top comment that pointed out what Raemon saw as a defect in Benquo's presentation, and separated that from a criticism of Benquo's argument. But that meant that later, when I criticized the argument, Benquo responded with 'but Raemon wasn't criticizing my argument.' If I only point out what seems to me like the most serious error in a post, then there's a (reasonable!) argument that all the things that went unsaid weren't at the threshold of correction, and if I only point out the error that's easier to respond to, the same sort of thing goes through.
I really like when people put effort into providing alternatives-to/critiques-of how I work, so thanks for this.
It doesn’t mean I can always reply promptly, alas. Right now I’m in the process of taking all my finals (so I can get a visa and move to the Bay), and this will continue for a few weeks :(. Oli’s just coming back from the same, while Ray’s holding the fort.
It’s also regularly the case that doing detailed public write-ups of considerations around a decision/approach isn’t the right use of effort relative to just *building the product*, and that ap... (read more)
Moved to meta.
This seems like it's pointing at a good thing. As a data point, the proposed responses to my comments would all have seemed friendly and helpful to me, and I'd have had an easy time engaging with the criticism.
His draft response to my VW comment would probably have motivated me to add a note to my initial comment deprecating that reference, or edit it out entirely (with a note clarifying that an edit had been made).
The asymmetry comment he pointed to as helpful (thanks, Vaniver!) actually would have been helpful, if SilentCal hadn't already taken the initiative (thanks, SilentCal!) and clarified what he took the term to mean.
I'm concerned that the described examples of holding individual comments to high epistemic standards don't seem to necessarily apply to top-level posts, or linked content- one reason I think this is bad is that it is hard to precisely critique something which is not in itself precise, or which contains metaphor, or which contains example-but-actually-pointing-at-a-class writing where the class can be construed in various different ways.
Critique of fuzzy intuitions and impressions and feelings often involves fuzzy intuitions and impressions and fe... (read more)
There is a potential discussion to be had someday about whether upvotes and downvotes themselves can be considered to be "in error" given the specific milieu of LW, or whether upvotes and downvotes are sacred à la free speech. Looking at votes on this and other threads, I have frequently had the sense that people were Doing It (Objectively) Wrong, not fundamentally as in wrong-opinions, but culturally as in participation-in-this-culture-means-precommitting-to-supporting-X-and-othering-Y.
I'm aware that there is a strongly-felt libertarian ar... (read more)
Because all those science, logic, maths and philosophy forums are just full of lies.
FYI, AFAICT I am the only person who downvoted it (bringing it from 3 to -3). As long as we're having the meta-est of conversations, I wanted to talk about why.
The single most common criticism I hear from good writers on LW, is that the comments feel nitpicky in a way that isn't actually helpful, and that you have to defend your points against a hostile-seeming crowd rather than collaboratively building something.
This comment seemed to be doing two things in that direction: a) it was bit aggressive as you noted, b) the point it was making just... didn't seem very relevant. Yes, there are places on the internet aspiring to similar things as LW. But a reasonable read on your statement was "most of the internet isn't trying to do this much at all, LW is different." While some humility is nice, I really don't understand your point better now that you've rephrased it to the 98th percentile thing.
So the main impact of this sort of comment is both to spend time on things that don't matter much, and increase the latent hostility of the thread, and I think people don't appreciate enough how bad that is for discourse. Both of that seem like something to me that is better silently downvoted than engaged with.
?? If you look at the Alexa top 50 sites for the US, how many of them are about aspiring to find the truth? I count between 3 and 4 (Google, Wikipedia, and Bing for sure, Wikia maybe).