I think I’m confused why the chosen distinction is something like “good/bad vs. agree/disagree” rather than “approve/disapprove vs. true/false.”
I do not have faith that people will use the agree/disagree voting for assessments of truth, which was the thing I personally wanted added to our voting system. Right now it feels like there are just two S1 monkey buttons and no nudge toward teasing out different strands of value.
Is there a difference between “agree/disagree” and “true/false”? They definitely parse in my mind as the same mental action (I mean, there are some very minor associative differences, but they do really point towards the same mental action for me).
I am pretty open to renaming the dimensions “approve/disapprove” vs. “true/false”. That’s pretty close to how I am referring to them in my mind. I think it’s also currently how almost everyone I’ve interviewed seems to interpret the current buttons, though you disagreeing is definitely evidence there is a broader distribution.
I think this would not be true if I had faith in the userbase, i.e. in a group composed entirely of Oli, Vaniver, Said, Logan, and Eliezer, I would trust the agreement/disagreement button.
I do not trust the mean, median, and modal LW users to reliably use “agree/disagree” to mean “true or mostly containing truth/false or mostly containing falsehood.”
So I don’t trust the aggregate of a lot of people using those buttons to be good signal rather than noise.
nods I think in that case we don’t disagree at all about the intention of the feature (the feature is intended to point people at true/false, so in as much as you picked up something else from the wording, seems good to clarify that). I do think we disagree about what the median user will think. I do actually think we should definitely say the words “true” and “false” in the hover-over (or maybe “correct” and “incorrect”, though I feel a bit confused about that one). Does putting the thing in the hover-over just resolve your crux?
(As a sort of sad but seems-good-to-share datapoint, on reading this comment it has 2 karma and −10 agrees, and I felt I had to explicitly undo my “woah, status punch!” reaction. On mousing over the agrees, it turns out it was only 1 vote, and that seemed to make it easier to undo the reaction; it was just one person, not the social winds.)
I think so? Depending on the exact wording or phrasing, but yeah: if it’s clear that the agreement or disagreement requested is an evaluation of truth/accuracy, then that resolves it.
There are some specific edge-cases that we hit on in another thread. In particular, I would like to somehow have a more principled distinction on whether pressing agree on sentences like “I believe X” means “I think you accurately report your beliefs” vs. “I would also report the same belief”. I think we almost always want to do the latter (since it’s more useful information), but “true” feels like it points a bit more toward the former. Maybe we can somehow massage that into the hover-over, or at least the FAQ.
Curious about your takes here. My sense is we are mostly on the same page on this distinction being important (and confusion between them seems like it could pretty easily cause a bunch of hurt).
The second button is an assessment of truth or falsehood, and in order to make that happen, we generally don’t click it one way or the other on somebody saying “I believe X.”
If I want to note that I would also report the same belief, I do a karma upvote and leave a comment.
Hmm, I think this would get rid of ~80% of the value for me, and also produce a lot of voting inconsistency, since it’s kind of author-specific how much they insert “I think X” vs. just saying “X”, and take the “I think” implicit.
I much prefer getting data on whether people agree with X in that case, and would really value that information.
Button overall. Like, I think I approximately never make a comment that doesn’t preface almost all of my (edit: not obviously correct) beliefs with “I think”, so this would cause no agree/disagree voting to happen on my comments.
I think this is why this button will be a very strong pressure away from LW, for me.
If the button claims to be about evaluating the truth or falsehood of the content of a comment, and also my comment has said a bunch of true stuff, and has a −17 on it or something, I will absolutely find this emotionally relevant and be Sad about it and want to spend much much less time on LW.
And if the button is not about the truth or falsehood of the content, and is just a signal of … how Other I am, versus how much I am Like the rest of the monkeys reading it, I expect to very frequently be receiving blunt You Are Not Like Us signals, all the time, and to have those signals permanently inscribed on all of my commentary (“look at what the guy that everybody disagrees with thinks!”) and to find this sad and alienating.
Like, I really cannot overstate the strength of the deterrent of the -n numbers on my comments on this post, alone. I’m keeping my hand on the hot stove because this feels important, but it does not feel good.
If this change sticks as it currently is, it will be really really difficult and painful for me to be on LW. Or, to be more specific: it’s already quite difficult and painful for me to be on LW, and I try very hard anyway/it takes up a disproportionate number of my spoons, and this will make that much worse.
I think that might just end up being fine/the cost that’s worth paying/the least bad option. Like, +10 good for thousands of users while −1000 for just Duncan is an obvious choice. But I wanted to be unequivocal about hating it, in its current state.
EDIT: “very strong pressure” as in, am currently right this minute trying to figure out where I will start posting essays in the future in the hypothetical where this change sticks, since probably-not-LW. =(
So someone can make a statement: “X”. X might be indexical or not. Indexical statements refer to the speaker, like “I think that probabilities are cool” or “I see a parrot.”. Non-indexical statements don’t, like “Probabilities track priors + evidence” or “There are parrots in the world”. The line is blurry: is “Probabilities are cool” implicitly indexical? Agree/disagree with X could be taken to mean, “It would be true if I said X, with the index pointing to me”, while true/untrue means, “X is the case”. If X is non-indexical, asserting agree/disagree is the same as asserting true/untrue. If X is indexical, they’re not the same; disagreeing with “I see a parrot” means “I (the disagree-er) don’t (myself) see a parrot”, while saying ” ‘I (the original speaker) see a parrot’ is untrue” means “No, you don’t see a parrot”.
Duncan, what would you think about a button that means agree/disagree in that sense, i.e., “I could also say this truthfully”? (As opposed to, it would be good for me to say this, or I would actually say this.) Is there a way to make that meaning clear? habryka, would that button get the value for you?
I like the sentence “I could also say this truthfully”, and I feel like it points towards the right generator that I have for what I would like “agree/disagree” to mean.
The tooltip of “Agree: Do you agree with the statements in this comment? Would the statements in this comment ring true if you said them yourself?” feels possibly good, though sure is a bit awkward and am not fully sure how reliably it would get the point across.
I’m pretty cynical about the ability to encourage any nuanced interpretation of such a simple input. Enough people will just use their first impression based on the icons and a quick reading of the labels that you will never be sure what the votes ACTUALLY mean, regardless of how clear your text guidance is.
I hope that people will just not use the agree/disagree voting for comments where it’s ambiguous what an entry would mean. If it doesn’t provide useful information about my reaction to the comment, why wouldn’t I just let my karma vote stand alone?
I do not trust the mean, median, and modal LW users to reliably use “agree/disagree” to mean “true or mostly containing truth/false or mostly containing falsehood.”
I’m confused by this, since to me it’s not even a question of trust, to me it seems like “agree/disagree” means “I think this is true/false”. In my head, to agree with a claim means that you think it’s true, and to disagree with a claim means that you think it’s false. (Of course, that also means that I’d be fine with changing the names.) Of course, “agree” does have some other meanings too (like “I agree to these terms of service”), but all of them seem clearly inapplicable to this context?
I strongly feel that penguins are the noblest of God’s creatures.
7:
I’m a utilitarian, and by a utilitarian calculus as I understand it, all pigeons should be replaced with penguins at once.
Would you say that, for each of these posts, “agree” means “I think this is true”? If so, what would it mean to “disagree” with any of these? They are (with one partial exception) simply reports of the commenter’s views. Does “disagree” mean “You are lying or mistaken about what you claim to believe”? If not, then it seems to me that “disagree” must (at least sometimes!) mean something different from, or at least something more subtle/nuanced than, merely “I think this is false”.
I think that you and similar people who are confused at my reaction (e.g. Oli, e.g. at least a little bit Rob) are basically … colorblind to something?
Like, I think that because it seems so obvious to you that agree/disagree is just about true/false that you’re not seeing how many many LWers would not and are not using it in that manner.
On a forum made up of just Kajs, Olis, and Robs, I would not have negative feelings about the way the second vote is used. But I think that its current agree/disagree label is much more ambiguous for people unlike yourselves, and so you’re not seeing why it needs to be more carefully specified (if we want distress like mine to be less in the mix).
It’s certainly possible that we’re colorblind to something, that’s why I was hoping for examples of what those alternative meanings could be so I could better understand what that something is. (And feel like I got them from Said’s response.)
Agree/Disagree are weird when evaluating your comment.
Agree with you asking the question (it’s the right question to ask) or disagree with your view?
I read Duncan’s comment as requesting that the labeling of the buttons be more explicit in some way, though I wasn’t sure if it was your way. (Also Duncan disagreeing with what they reflect).
I guess a different point is that, given what I understand to be the goals of LessWrong, I’m confused about valid reasons for liking something other than either:
This just seems true, irrespective of any of its other properties (e.g. whether it reduces the heat of a conversation or not)
This just seems like it moves the conversation in a better/more productive direction, irrespective of any of its other properties (e.g. whether it’s true or not)
Writing quality is a good one to mention; I suppose I have upvoted things purely on the grounds that I wanted to incentivize [more like this] for a comment that was clear and clearly effortful.
Is it good or bad? It’s impossible to decipher, I can’t tell. Is it true or false? No way to tell. (It doesn’t happen often, but it’s usually downvoted.)
ETA: I’m not sure at the moment what other aspects there are.
I think I’m confused why the chosen distinction is something like “good/bad vs. agree/disagree” rather than “approve/disapprove vs. true/false.”
I do not have faith that people will use the agree/disagree voting for assessments of truth, which was the thing I personally wanted added to our voting system. Right now it feels like there are just two S1 monkey buttons and no nudge toward teasing out different strands of value.
Is there a difference between “agree/disagree” and “true/false”? They definitely parse in my mind as the same mental action (I mean, there are some very minor associative differences, but they do really point towards the same mental action for me).
I am pretty open to renaming the dimensions “approve/disapprove” vs. “true/false”. That’s pretty close to how I am referring to them in my mind. I think it’s also currently how almost everyone I’ve interviewed seems to interpret the current buttons, though you disagreeing is definitely evidence there is a broader distribution.
I do not trust the mean, median, and modal LW users to reliably use “agree/disagree” to mean “true or mostly containing truth/false or mostly containing falsehood.”
So I don’t trust the aggregate of a lot of people using those buttons to be good signal rather than noise.
nods I think in that case we don’t disagree at all about the intention of the feature (the feature is intended to point people at true/false, so in as much as you picked up something else from the wording, seems good to clarify that). I do think we disagree about what the median user will think. I do actually think we should definitely say the words “true” and “false” in the hover-over (or maybe “correct” and “incorrect”, though I feel a bit confused about that one). Does putting the thing in the hover-over just resolve your crux?
(As a sort of sad but seems-good-to-share datapoint, on reading this comment it has 2 karma and −10 agrees, and I felt I had to explicitly undo my “woah, status punch!” reaction. On mousing over the agrees, it turns out it was only 1 vote, and that seemed to make it easier to undo the reaction; it was just one person, not the social winds.)
(I feel like I should be clear that it wasn’t me.)
I think so? Depending on the exact wording or phrasing, but yeah: if it’s clear that the agreement or disagreement requested is an evaluation of truth/accuracy, then that resolves it.
There are some specific edge-cases that we hit on in another thread. In particular, I would like to somehow have a more principled distinction on whether pressing agree on sentences like “I believe X” means “I think you accurately report your beliefs” vs. “I would also report the same belief”. I think we almost always want to do the latter (since it’s more useful information), but “true” feels like it points a bit more toward the former. Maybe we can somehow massage that into the hover-over, or at least the FAQ.
Curious about your takes here. My sense is we are mostly on the same page on this distinction being important (and confusion between them seems like it could pretty easily cause a bunch of hurt).
I think that the correct norm is:
The second button is an assessment of truth or falsehood, and in order to make that happen, we generally don’t click it one way or the other on somebody saying “I believe X.”
If I want to note that I would also report the same belief, I do a karma upvote and leave a comment.
Hmm, I think this would get rid of ~80% of the value for me, and also produce a lot of voting inconsistency, since it’s kind of author-specific how much they insert “I think X” vs. just saying “X”, and take the “I think” implicit.
I much prefer getting data on whether people agree with X in that case, and would really value that information.
80% of the value in those cases, or of the button overall? ’Cos if the latter, it seems like that’s our real disagreement.
Button overall. Like, I think I approximately never make a comment that doesn’t preface almost all of my (edit: not obviously correct) beliefs with “I think”, so this would cause no agree/disagree voting to happen on my comments.
I think this is why this button will be a very strong pressure away from LW, for me.
If the button claims to be about evaluating the truth or falsehood of the content of a comment, and also my comment has said a bunch of true stuff, and has a −17 on it or something, I will absolutely find this emotionally relevant and be Sad about it and want to spend much much less time on LW.
And if the button is not about the truth or falsehood of the content, and is just a signal of … how Other I am, versus how much I am Like the rest of the monkeys reading it, I expect to very frequently be receiving blunt You Are Not Like Us signals, all the time, and to have those signals permanently inscribed on all of my commentary (“look at what the guy that everybody disagrees with thinks!”) and to find this sad and alienating.
Like, I really cannot overstate the strength of the deterrent of the -n numbers on my comments on this post, alone. I’m keeping my hand on the hot stove because this feels important, but it does not feel good.
If this change sticks as it currently is, it will be really really difficult and painful for me to be on LW. Or, to be more specific: it’s already quite difficult and painful for me to be on LW, and I try very hard anyway/it takes up a disproportionate number of my spoons, and this will make that much worse.
I think that might just end up being fine/the cost that’s worth paying/the least bad option. Like, +10 good for thousands of users while −1000 for just Duncan is an obvious choice. But I wanted to be unequivocal about hating it, in its current state.
EDIT: “very strong pressure” as in, am currently right this minute trying to figure out where I will start posting essays in the future in the hypothetical where this change sticks, since probably-not-LW. =(
So someone can make a statement: “X”. X might be indexical or not. Indexical statements refer to the speaker, like “I think that probabilities are cool” or “I see a parrot.”. Non-indexical statements don’t, like “Probabilities track priors + evidence” or “There are parrots in the world”. The line is blurry: is “Probabilities are cool” implicitly indexical? Agree/disagree with X could be taken to mean, “It would be true if I said X, with the index pointing to me”, while true/untrue means, “X is the case”. If X is non-indexical, asserting agree/disagree is the same as asserting true/untrue. If X is indexical, they’re not the same; disagreeing with “I see a parrot” means “I (the disagree-er) don’t (myself) see a parrot”, while saying ” ‘I (the original speaker) see a parrot’ is untrue” means “No, you don’t see a parrot”.
Duncan, what would you think about a button that means agree/disagree in that sense, i.e., “I could also say this truthfully”? (As opposed to, it would be good for me to say this, or I would actually say this.) Is there a way to make that meaning clear? habryka, would that button get the value for you?
I like the sentence “I could also say this truthfully”, and I feel like it points towards the right generator that I have for what I would like “agree/disagree” to mean.
The tooltip of “Agree: Do you agree with the statements in this comment? Would the statements in this comment ring true if you said them yourself?” feels possibly good, though sure is a bit awkward and am not fully sure how reliably it would get the point across.
I’m pretty cynical about the ability to encourage any nuanced interpretation of such a simple input. Enough people will just use their first impression based on the icons and a quick reading of the labels that you will never be sure what the votes ACTUALLY mean, regardless of how clear your text guidance is.
I hope that people will just not use the agree/disagree voting for comments where it’s ambiguous what an entry would mean. If it doesn’t provide useful information about my reaction to the comment, why wouldn’t I just let my karma vote stand alone?
I find the solution of “I could also say this truthfully” to be pretty clever and my gut sense is that it would resolve the distress.
I’m confused by this, since to me it’s not even a question of trust, to me it seems like “agree/disagree” means “I think this is true/false”. In my head, to agree with a claim means that you think it’s true, and to disagree with a claim means that you think it’s false. (Of course, that also means that I’d be fine with changing the names.) Of course, “agree” does have some other meanings too (like “I agree to these terms of service”), but all of them seem clearly inapplicable to this context?
Consider the following hypothetical posts:
1:
2:
3:
4:
5:
6:
7:
Would you say that, for each of these posts, “agree” means “I think this is true”? If so, what would it mean to “disagree” with any of these? They are (with one partial exception) simply reports of the commenter’s views. Does “disagree” mean “You are lying or mistaken about what you claim to believe”? If not, then it seems to me that “disagree” must (at least sometimes!) mean something different from, or at least something more subtle/nuanced than, merely “I think this is false”.
These were useful examples, thanks.
I think that you and similar people who are confused at my reaction (e.g. Oli, e.g. at least a little bit Rob) are basically … colorblind to something?
Like, I think that because it seems so obvious to you that agree/disagree is just about true/false that you’re not seeing how many many LWers would not and are not using it in that manner.
On a forum made up of just Kajs, Olis, and Robs, I would not have negative feelings about the way the second vote is used. But I think that its current agree/disagree label is much more ambiguous for people unlike yourselves, and so you’re not seeing why it needs to be more carefully specified (if we want distress like mine to be less in the mix).
It’s certainly possible that we’re colorblind to something, that’s why I was hoping for examples of what those alternative meanings could be so I could better understand what that something is. (And feel like I got them from Said’s response.)
Agree/Disagree are weird when evaluating your comment.
Agree with you asking the question (it’s the right question to ask) or disagree with your view?
I read Duncan’s comment as requesting that the labeling of the buttons be more explicit in some way, though I wasn’t sure if it was your way. (Also Duncan disagreeing with what they reflect).
Upvote (Like**)
Quality*
Agreement (Truth)
Veracity
Not present***: Value?Judgement? (Good/bad)
Good/Bad
**This is in ()s because it’s the word that shows up in bold when hovering over a button.
*How well something is written?
***That is a harsher bold than I was going for.
I guess a different point is that, given what I understand to be the goals of LessWrong, I’m confused about valid reasons for liking something other than either:
This just seems true, irrespective of any of its other properties (e.g. whether it reduces the heat of a conversation or not)
This just seems like it moves the conversation in a better/more productive direction, irrespective of any of its other properties (e.g. whether it’s true or not)
Writing quality is a good one to mention; I suppose I have upvoted things purely on the grounds that I wanted to incentivize [more like this] for a comment that was clear and clearly effortful.
Yeah. When something is very unclear, it’s like
Is it good or bad? It’s impossible to decipher, I can’t tell. Is it true or false? No way to tell. (It doesn’t happen often, but it’s usually downvoted.)
ETA: I’m not sure at the moment what other aspects there are.