People who have behaved like this should publicly apologise as close to the time as possible.
Agreevote to agree. I don’t necessarily endorse this but I think it’s valuable for us to know what we think on stuff like this—I bet we disagree a surprising amount.
I actually endorse the idea of polls on this but don’t want to make one.
Why?
I’m in several text and real life conversations with women right now and none of them are commenting here because we’re sad and annoyed and frustrated. So they’re not voting.
The person returning to the community role needs to accept that there may be longer-term restrictions on, or added oversight of, their conduct. That may involve a level of restriction and oversight that would seem onerous or heavy-handed if applied to someone without a history of inappropriate conduct.
Some of you should write other polls in this post, so we can figure out what we think. I’m not gonna write them, but I suggest women in EA might want to figure out what they think, because again, I sense a broad set of views and many people who don’t want to comment, but might agree/disagree.
I struggle with seeing polls like this as being a valid source of information that change my beliefs because we have no way of knowing who is actually voting and whether or not they are trolls.
I think you’re right to be cautious, although most of the other sources of information I could think of have some significant flaws of some sort as well. So it’s probably best to look for consensus from several different information flows that don’t have too many overlapping weaknesses.
Also they really exploit the forum’s karma system. Should we strong vote for strong agree? If so, mine would be worth about 3x the average forum user’s. Even if I only ‘weak’ vote, it automatically gives +/-2 to the karma score now.
Even if sock puppet accounts do not currently influence votes, the more the EA community will use such polls to decide on things the more likely such influence will become.
I think we currently see too little community sensemaking, not too much. And if sockpuppets could be used on polls they could be used everywhere. Why not complain that the whole voting system is wrong?
I mean, trolling requires EAs (the only people who are looking at a page at the bottom of the home page) creating additional accounts. I think it’s unlikely they would do that.
Good point. Maybe the forum could introduce up- and downvoting restrictions until a new account has reached a few milestones, like being active for more than a month and adding comments with overall positive upvotes. This is how I remember StackOverflow handling this.
We have at least some basic tools to detect account-creation and mass-voting. I don’t think the voting system is perfect, but at least on LW I would notice if someone was creating multiple accounts to vote, as I have noticed multiple times in the past.
A very valid concern. Thankfully, if I’m not mistaken, the moderators block troll votes (70% confidence) and have done so already in the past (50% confidence).
Here are some reasons to downvote this poll and all the other polls (I have not yet done this). Some people feel very frightened and stressed etc, and for them, it is not yet the time to engage with these polls.
In addition, it feels wrong that justice is ruled by a weird anonymous voting procedure which does not include a bunch of deliberation (and that does not include more people that are specialised in this problem) . EA is not a democratic organisation (for good or bad reasons) - ultimately, this is not how EVF makes decisions. Suppose, someone did X (X is clearly morally wrong), a community does a vote, and more than 50% think it is acceptable, that does not mean the behaviour is acceptable (or even if 80% agree).
The EVF board hired an external group to look into this issue (see their announcement post)- I don’t understand why you don’t first trust them (you can still disagree if they conclude).
Perhaps, though by that argument maybe we shouldn’t have long and complex arguments either. We are already engaging in discussion while upset and stressed. What I would like is for us at least to be able to see some of the landscape.
I don’t think this suggests justice, I think it allows us to see what each other think.
It’s not that I don’t trust the external group—it’s that I am interested in what we all think and feel.
Some personal thoughts about the the polls generally—apologies in advance if this comes across confrontational, as I know you are well-intentioned! Also to be clear, I enjoyed and appreciated what you did with the cluster poll, as they provided new information RE: clusters—I’m referring mainly to comment polls like this.[1]
I understand your desire and goals RE: polling, but I personally find many of the polls to be of unclear value, and it often clogs up the comment sections (even if they are downvoted to the bottom, as this is, I often stay updated by looking at “recent discussion” on the front page).
I hear that you are interested in what we all think and feel—I am too! But I don’t know if the polls actually answer this, or at least I find it difficult to meaningfully update on them.
Perhaps, though by that argument maybe we shouldn’t have long and complex arguments either.
I’d much rather just see more discussion on the actual question being polled, because then people provide their reasoning. In this case, even if those who aren’t on the forum or aren’t in a position to comment or engage, I can see if I agree with the reasoning myself, and update more strongly on reasoning I find more compelling, etc. On the other hand, if all I see is “7 agreement karma 22 votes”, that doesn’t give me a lot of information other than something like “ok seems like people voted both ways on this”.
I also worry that it may even be counterproductive and give the wrong impression of what the community’s views are.
One recent example that comes to mind:
Apparently, the majority opinion is that what Owen did was not “wrong and very serious”.
If this is actually representative of the EA community’s opinion, then many people I know who have been dismayed / distressed about recent events will be doing a fair amount of updating about how this community sees sexual harassment, abuse of power, navigating professional / personal boundaries, etc.[2]
It’s also a little draining in some sense—I feel like I want to ignore most of these polls, but I also want to signal to people in this community who are more personally affected by these that there are others who care about the things they care about. This isn’t your fault though, just raising how these are real considerations when I’m reading your polls, and imagining how my friends who might see these polls may react. And on topics where those most deeply affected have strong reasons to not engage with the poll, this is going to be a larger issue.
I also agree with earlier comments about the potential for trolls (contra your earlier claim, there is recent evidence to support this), and the negative incentives that this may contribute to if these polls end up being action guiding (or even just perceived to be action guiding) in some way.
This has like 6 different claims. Are people agreeing with all 6? the general vibe? just three? Is the 19 voters to 35 karma indicative of 19 people saying weakly yes, or a bimodal distribution from people who feel very strongly?
I can also already see the quote in the next hit piece: “In an internal poll, a majority of EAs believe that [OCB’s actions] were not wrong/very serious”.
Thank you for raising this. I appreciate pushback.
I don’t think this issue is being brigaded. If it were we’d see a load of chat on twitter. I think that explains too much. I just would bet it isn’t the case. Some people are making new burners but I’d guess that’s it.
My top level comment is in positive karma. Just as you don’t like the polls, many people do. In fact I’ve received 2-3 messages to that end. What you and those who agree with you are doing is expressing a preference. Others have different preferences. And the evidence suggests that your preference is not the overwhelming majority.
I would like to be better and more gracious at doing this. But I think that it provides information on balance. Again, as the karma shows, people agree.
I like polls sure, and martyrdom, but also feel scared and tired and like the truth is getting ignored because it’s impolitic. The truth seems to be that people have more complex views on this than is comfortable, in many directions.
I am deeply confused by the idea that we want to make comments and look at the voting (which we do, people regularly comment on it) but that trying to establish simple ideas is somehow a different class of thing.
I suggest that many people don’t like what these polls show and so instead think “they must be inaccurate or inappropriate”. I don’t think so (70%), and I’m pretty well calibrated. In time, I think just as criticisms of me saying FTX would go really badly were false, so will the idea that EAs hold simple views on the topic of this case. I think people would do well to be curious—what do people in EA believe? What do women believe and want? What norms will create a community that does the most good? Again, I don’t know, but I don’t think that anyone else does either
On the what “Owen did was not “wrong and very serious” ” poll, I think that is a deeply unpopular view to express under one’s own name, so I’m not surprised people haven’t said it. But I think it’s a thing people believe—I’ve talked to some. I think it’s easy to be repulsed by this, to theorize that these people don’t care about the accusers welfare or are closing ranks around Cotton Barrett. I don’t know! Which is why I ask. It seems clear to me that this discussion is a scary and exhausting place, not just for people who ignore the polls but for those who are voting in ways that surprise you.
I think of this community as a brain. And when my brain is scared and hurt it comes to quick conclusions “Owen is really bad” “casual sex should not happen in EA” or “Owen was misunderstood” “We should be able to do what we want”. But this often isn’t useful. Instead I want to understand what happened, what makes people able to keep doing good work and what good norms should be. And allowing the community to raise it’s prefereces is the first part of that. Many people seemingly thought that SBF was reckless and broke promises. I wish we’d had some comment polls about that.
Bruce, I am glad you raised this and thanks to any who agree with him. I like you and I like your desire to create a safe environment for discourse. I know this feels weird and I think it could be done better. I desire that we have some sense of what people believe and I feel that strongly. I feel sad and frustrated and curious and like making polls and crying and many other things. I wish you all well.
I think the polls here were too soon—among other things, it’s not clear that we want to be quantifying hot takes as opposed to more settled views a week or so after breaking news.
I also suggest the Forum software is just not set up for polling. I think a pol.is could be potentially useful on this topic to identify clusters, or a traditional survey site to allow capturing somewhat more nuanced responses. The existence of strong votes is particularly ill suited to polling where the number of each vote type isn’t disclosed.
Trolls are a concern, although that’s true of both voting and general Forum activity/upvotes. I think one concern with relying on Forum comments to gauge community sentiment is that we’re largely seeing a few dozen people who are either particularly passionate about the particular question at hand or are quite frequent posters. So that’s unlikely to be representative of the broader community.
Alternative possibility: Assuming that immediate removal was not warranted (but not taking a position on that either way), Owen should have been given a stern warning that (1) any future incidents would warrant removal, and (2) for any future reports, the reporter would get the benefit of the doubt.
(Agreevote to agree/disagree: I’m not committing to a position, mainly pointing out that there are options between removal and doing very little.)
The idea was that a final-chance warning would hopefully deter future incidents.
I didn’t specify either way what the consequences would be for future reports of prior-to-warning events (especially if pre-EVF) . That wouldn’t have been necessary to resolve upfront because one cannot deter past events.
Poll:
People who have behaved like this should publicly apologise as close to the time as possible.
Agreevote to agree. I don’t necessarily endorse this but I think it’s valuable for us to know what we think on stuff like this—I bet we disagree a surprising amount.
I actually endorse the idea of polls on this but don’t want to make one. Why? I’m in several text and real life conversations with women right now and none of them are commenting here because we’re sad and annoyed and frustrated. So they’re not voting.
Poll:
There should be a way for people who have behaved like this to hold community roles afterwards if they can meet some standard.
Agreevote to agree. I don’t necessarily endorse this
Poll, contingent on the above being true:
The person returning to the community role needs to accept that there may be longer-term restrictions on, or added oversight of, their conduct. That may involve a level of restriction and oversight that would seem onerous or heavy-handed if applied to someone without a history of inappropriate conduct.
Some of you should write other polls in this post, so we can figure out what we think. I’m not gonna write them, but I suggest women in EA might want to figure out what they think, because again, I sense a broad set of views and many people who don’t want to comment, but might agree/disagree.
I struggle with seeing polls like this as being a valid source of information that change my beliefs because we have no way of knowing who is actually voting and whether or not they are trolls.
I think you’re right to be cautious, although most of the other sources of information I could think of have some significant flaws of some sort as well. So it’s probably best to look for consensus from several different information flows that don’t have too many overlapping weaknesses.
Also they really exploit the forum’s karma system. Should we strong vote for strong agree? If so, mine would be worth about 3x the average forum user’s. Even if I only ‘weak’ vote, it automatically gives +/-2 to the karma score now.
I think they are better than nothing. I would like specific voting objects but we make do with what we have.
Let’s be real, many people upvoting here don’t actually think there are trolls, but the polls make you uneasy, right?
perhaps because you feel they are an oversimplification? or because you feel they are unrepresentative?
Even if sock puppet accounts do not currently influence votes, the more the EA community will use such polls to decide on things the more likely such influence will become.
I think we currently see too little community sensemaking, not too much. And if sockpuppets could be used on polls they could be used everywhere. Why not complain that the whole voting system is wrong?
I agree that such concerns are not limited to the polls on this page. (I’ve also wrote a comment related to such concerns here.)
I mean, trolling requires EAs (the only people who are looking at a page at the bottom of the home page) creating additional accounts. I think it’s unlikely they would do that.
I think the amount of press EA has gotten lately makes it more likely for there to be trolls but I could be wrong.
Good point. Maybe the forum could introduce up- and downvoting restrictions until a new account has reached a few milestones, like being active for more than a month and adding comments with overall positive upvotes. This is how I remember StackOverflow handling this.
I don’t think we have any evidence this is happening.
We have at least some basic tools to detect account-creation and mass-voting. I don’t think the voting system is perfect, but at least on LW I would notice if someone was creating multiple accounts to vote, as I have noticed multiple times in the past.
A very valid concern. Thankfully, if I’m not mistaken, the moderators block troll votes (70% confidence) and have done so already in the past (50% confidence).
Poll:
What Owen did was (given the best of our knowledge) wrong and very serious.
Agreevote to agree. I don’t necessarily endorse this
I think people are accidentally down-voting instead of disagree-voting, which makes the comment hidden.
The up/down vote is on the left, agree/disagree is on the right.
No, I think it’s deliberate.
The agree-votes have pretty directly proven you correct.
Poll:
Owen should have been removed from the board as soon as this was known, regardless of his accuser’s opinion.
Agreevote to agree. I don’t necessarily endorse this
Here are some reasons to downvote this poll and all the other polls (I have not yet done this). Some people feel very frightened and stressed etc, and for them, it is not yet the time to engage with these polls.
In addition, it feels wrong that justice is ruled by a weird anonymous voting procedure which does not include a bunch of deliberation (and that does not include more people that are specialised in this problem) . EA is not a democratic organisation (for good or bad reasons) - ultimately, this is not how EVF makes decisions. Suppose, someone did X (X is clearly morally wrong), a community does a vote, and more than 50% think it is acceptable, that does not mean the behaviour is acceptable (or even if 80% agree).
The EVF board hired an external group to look into this issue (see their announcement post)- I don’t understand why you don’t first trust them (you can still disagree if they conclude).
Thanks for saying why you disagree.
Perhaps, though by that argument maybe we shouldn’t have long and complex arguments either. We are already engaging in discussion while upset and stressed. What I would like is for us at least to be able to see some of the landscape.
I don’t think this suggests justice, I think it allows us to see what each other think.
It’s not that I don’t trust the external group—it’s that I am interested in what we all think and feel.
Some personal thoughts about the the polls generally—apologies in advance if this comes across confrontational, as I know you are well-intentioned! Also to be clear, I enjoyed and appreciated what you did with the cluster poll, as they provided new information RE: clusters—I’m referring mainly to comment polls like this.[1]
I understand your desire and goals RE: polling, but I personally find many of the polls to be of unclear value, and it often clogs up the comment sections (even if they are downvoted to the bottom, as this is, I often stay updated by looking at “recent discussion” on the front page).
I hear that you are interested in what we all think and feel—I am too! But I don’t know if the polls actually answer this, or at least I find it difficult to meaningfully update on them.
I’d much rather just see more discussion on the actual question being polled, because then people provide their reasoning. In this case, even if those who aren’t on the forum or aren’t in a position to comment or engage, I can see if I agree with the reasoning myself, and update more strongly on reasoning I find more compelling, etc. On the other hand, if all I see is “7 agreement karma 22 votes”, that doesn’t give me a lot of information other than something like “ok seems like people voted both ways on this”.
I also worry that it may even be counterproductive and give the wrong impression of what the community’s views are.
One recent example that comes to mind:
Apparently, the majority opinion is that what Owen did was not “wrong and very serious”.
If this is actually representative of the EA community’s opinion, then many people I know who have been dismayed / distressed about recent events will be doing a fair amount of updating about how this community sees sexual harassment, abuse of power, navigating professional / personal boundaries, etc.[2]
It’s also a little draining in some sense—I feel like I want to ignore most of these polls, but I also want to signal to people in this community who are more personally affected by these that there are others who care about the things they care about. This isn’t your fault though, just raising how these are real considerations when I’m reading your polls, and imagining how my friends who might see these polls may react. And on topics where those most deeply affected have strong reasons to not engage with the poll, this is going to be a larger issue.
I also agree with earlier comments about the potential for trolls (contra your earlier claim, there is recent evidence to support this), and the negative incentives that this may contribute to if these polls end up being action guiding (or even just perceived to be action guiding) in some way.
This has like 6 different claims. Are people agreeing with all 6? the general vibe? just three? Is the 19 voters to 35 karma indicative of 19 people saying weakly yes, or a bimodal distribution from people who feel very strongly?
I can also already see the quote in the next hit piece: “In an internal poll, a majority of EAs believe that [OCB’s actions] were not wrong/very serious”.
Quick thoughts:
Thank you for raising this. I appreciate pushback.
I don’t think this issue is being brigaded. If it were we’d see a load of chat on twitter. I think that explains too much. I just would bet it isn’t the case. Some people are making new burners but I’d guess that’s it.
My top level comment is in positive karma. Just as you don’t like the polls, many people do. In fact I’ve received 2-3 messages to that end. What you and those who agree with you are doing is expressing a preference. Others have different preferences. And the evidence suggests that your preference is not the overwhelming majority.
I would like to be better and more gracious at doing this. But I think that it provides information on balance. Again, as the karma shows, people agree.
I like polls sure, and martyrdom, but also feel scared and tired and like the truth is getting ignored because it’s impolitic. The truth seems to be that people have more complex views on this than is comfortable, in many directions.
I am deeply confused by the idea that we want to make comments and look at the voting (which we do, people regularly comment on it) but that trying to establish simple ideas is somehow a different class of thing.
I suggest that many people don’t like what these polls show and so instead think “they must be inaccurate or inappropriate”. I don’t think so (70%), and I’m pretty well calibrated. In time, I think just as criticisms of me saying FTX would go really badly were false, so will the idea that EAs hold simple views on the topic of this case. I think people would do well to be curious—what do people in EA believe? What do women believe and want? What norms will create a community that does the most good? Again, I don’t know, but I don’t think that anyone else does either
On the what “Owen did was not “wrong and very serious” ” poll, I think that is a deeply unpopular view to express under one’s own name, so I’m not surprised people haven’t said it. But I think it’s a thing people believe—I’ve talked to some. I think it’s easy to be repulsed by this, to theorize that these people don’t care about the accusers welfare or are closing ranks around Cotton Barrett. I don’t know! Which is why I ask. It seems clear to me that this discussion is a scary and exhausting place, not just for people who ignore the polls but for those who are voting in ways that surprise you.
I think of this community as a brain. And when my brain is scared and hurt it comes to quick conclusions “Owen is really bad” “casual sex should not happen in EA” or “Owen was misunderstood” “We should be able to do what we want”. But this often isn’t useful. Instead I want to understand what happened, what makes people able to keep doing good work and what good norms should be. And allowing the community to raise it’s prefereces is the first part of that. Many people seemingly thought that SBF was reckless and broke promises. I wish we’d had some comment polls about that.
Bruce, I am glad you raised this and thanks to any who agree with him. I like you and I like your desire to create a safe environment for discourse. I know this feels weird and I think it could be done better. I desire that we have some sense of what people believe and I feel that strongly. I feel sad and frustrated and curious and like making polls and crying and many other things. I wish you all well.
I think the polls here were too soon—among other things, it’s not clear that we want to be quantifying hot takes as opposed to more settled views a week or so after breaking news.
I also suggest the Forum software is just not set up for polling. I think a pol.is could be potentially useful on this topic to identify clusters, or a traditional survey site to allow capturing somewhat more nuanced responses. The existence of strong votes is particularly ill suited to polling where the number of each vote type isn’t disclosed.
Trolls are a concern, although that’s true of both voting and general Forum activity/upvotes. I think one concern with relying on Forum comments to gauge community sentiment is that we’re largely seeing a few dozen people who are either particularly passionate about the particular question at hand or are quite frequent posters. So that’s unlikely to be representative of the broader community.
Alternative possibility: Assuming that immediate removal was not warranted (but not taking a position on that either way), Owen should have been given a stern warning that (1) any future incidents would warrant removal, and (2) for any future reports, the reporter would get the benefit of the doubt.
(Agreevote to agree/disagree: I’m not committing to a position, mainly pointing out that there are options between removal and doing very little.)
Sure but in this case, what is the issue with what happened? I guess maybe there were other smaller behaviours afterwards?
Note that the future reports did come and therefore he should’ve been removed before this point in time.
The idea was that a final-chance warning would hopefully deter future incidents.
I didn’t specify either way what the consequences would be for future reports of prior-to-warning events (especially if pre-EVF) . That wouldn’t have been necessary to resolve upfront because one cannot deter past events.