I have also recently been thinking alot about “how should we want to deal with a scandal” but mostly in terms of how much time is being devoted to each of these scandals by a community who really advocates for using our minimal resources to do the most good. It makes me really disappointed.
<<i’m not sure the community even agrees internally on Bostrom, the Time article or Nonlinear>>
Forming a consensus additionally seems against the values of the EA community, particularly on quite complicated topics where there is alot of information asymmetry and people often update (as they should) based on new evidence, which is again an important part of the EA community to me at least. So I think I disagree and think it’s unrealistic to have a community as large as EA “to find agreement on these,” and I’m not sure how this would help.
But i fully agree it would be great if we had a better process or strategy of how to deal with scandals
“mostly in terms of how much time is being devoted to each of these scandals by a community who really advocates for using our minimal resources to do the most good. It makes me really disappointed.”
I think we talk about scandals too much without making progress but I’m not sure we spend too much time on them. Often it’s about trust. And communities need trust. If you are a not thick skinned person of colour, an outspoken autist, someone who runs an unconventional org, a normal person new to the job market, how these events are handled affects how much you can trust the community if these events happen to you.
It seems to me that most people want confidence that bad things won’t happen to them. If they don’t have that, they will probably leave. And that has it’s own, large, costs.
Yes sorry I think we are actually saying the same thing here, I meant your former statement not the later. I’m not saying we shouldn’t investigate things but the 300 plus comments on the 3-4 nonlinear posts doesn’t seem an optimal use of time and could probably be dealt with more efficiently, plus the thousands of people who have probably read the posts and comments is a lot of time! Maybe these things shouldn’t be handled in forum posts but in a different format.
I fully agree that these things have to be dealt with better my main concern about your point is over the consensus idea which I think is unrealistic in a community that tries to avoid group think and on topics (ftx aside) where there doesn’t seem to be a clear right or wrong.
This also seems right to me. I feel like there is a lot of unproductive conflict in the comments on these kinds of posts, which could be somehow prevented and would also be more productive if the conflict instead occurred between a smaller number of representative EA forum members, or something like that.
An very random idea in that direction that won’t work for many reasons, is some kind of “EA Forum jury” where you get randomly chosen to be one of the users in the comment section of a contentious post, and then you fight it out until you reach some kind of consensus, or at least the discussion dies down.
I do think the most standard way people have handled this in various contexts is to have panels, or courts, or boards or some kind of other system where some small subset of chosen representatives have the job of deciding on some tricky subject matter. I do kind of wish there was some kind of court system in EA that could do this.
One challenge with a “drama jury” is that the people who are most motivated to be heavy participants aren’t necessarily the people you want driving the discussion. But I guess that’s equally true in open posts. The solution in classical Athens was to pay jurors a wage; IIRC, many jurors were semiretired old men who had a lot of bandwidth.
Potentially, you’d have a few nonvoting neutrals in the mix to help facilitate discussion. It’s easier to be in a facilitating frame of mind when you are not simultaneously being asked to vote on a verdict.
What would happen if people kept posting it outside of the “courtroom”?
Not sure whether this is what you were implying, but I wasn’t thinking of private courts. My current guess is that it is important for courts to be at least observable, so that people can build trust in them (observable in the sense of how modern courts are observable, i.e. anyone can show up to the courtroom, but you might not be allowed to record it).
I think John meant that non-participants might keep commenting on the situation while the trial was in progress, and then after the trial. That might weaken some of the gains from having a trial in the first place (e.g., the hope that people will accept the verdict and move on to more productive things).
You could “sequester” the jury by making them promise not to read the non-courtroom threads until the jury had delivered a verdict. You could also have a norm that disputants would not comment in other threads while trial was ongoing. Not having the disputants in the non-courtroom thread would probably slow its velocity down considerably. You could even hide the courtroom thread from non-participants until the trial was over. That’s not a complete answer, but would probably help some.
The bottleneck feels more social than technological.
Also, I feel like someone else needs to do investigations for it to make sense for me to build the courtroom, since it does seem bad for one person to do both.
I have also recently been thinking alot about “how should we want to deal with a scandal” but mostly in terms of how much time is being devoted to each of these scandals by a community who really advocates for using our minimal resources to do the most good. It makes me really disappointed.
<<i’m not sure the community even agrees internally on Bostrom, the Time article or Nonlinear>>
Forming a consensus additionally seems against the values of the EA community, particularly on quite complicated topics where there is alot of information asymmetry and people often update (as they should) based on new evidence, which is again an important part of the EA community to me at least. So I think I disagree and think it’s unrealistic to have a community as large as EA “to find agreement on these,” and I’m not sure how this would help.
But i fully agree it would be great if we had a better process or strategy of how to deal with scandals
I think we talk about scandals too much without making progress but I’m not sure we spend too much time on them. Often it’s about trust. And communities need trust. If you are a not thick skinned person of colour, an outspoken autist, someone who runs an unconventional org, a normal person new to the job market, how these events are handled affects how much you can trust the community if these events happen to you.
It seems to me that most people want confidence that bad things won’t happen to them. If they don’t have that, they will probably leave. And that has it’s own, large, costs.
Yes sorry I think we are actually saying the same thing here, I meant your former statement not the later. I’m not saying we shouldn’t investigate things but the 300 plus comments on the 3-4 nonlinear posts doesn’t seem an optimal use of time and could probably be dealt with more efficiently, plus the thousands of people who have probably read the posts and comments is a lot of time! Maybe these things shouldn’t be handled in forum posts but in a different format.
I fully agree that these things have to be dealt with better my main concern about your point is over the consensus idea which I think is unrealistic in a community that tries to avoid group think and on topics (ftx aside) where there doesn’t seem to be a clear right or wrong.
This also seems right to me. I feel like there is a lot of unproductive conflict in the comments on these kinds of posts, which could be somehow prevented and would also be more productive if the conflict instead occurred between a smaller number of representative EA forum members, or something like that.
An very random idea in that direction that won’t work for many reasons, is some kind of “EA Forum jury” where you get randomly chosen to be one of the users in the comment section of a contentious post, and then you fight it out until you reach some kind of consensus, or at least the discussion dies down.
I do think the most standard way people have handled this in various contexts is to have panels, or courts, or boards or some kind of other system where some small subset of chosen representatives have the job of deciding on some tricky subject matter. I do kind of wish there was some kind of court system in EA that could do this.
One challenge with a “drama jury” is that the people who are most motivated to be heavy participants aren’t necessarily the people you want driving the discussion. But I guess that’s equally true in open posts. The solution in classical Athens was to pay jurors a wage; IIRC, many jurors were semiretired old men who had a lot of bandwidth.
Potentially, you’d have a few nonvoting neutrals in the mix to help facilitate discussion. It’s easier to be in a facilitating frame of mind when you are not simultaneously being asked to vote on a verdict.
I’ve had similar thoughts. I think the biggest questions are:
- Who would organise them?
- What powers would they have?
- What would happen if people kept posting it outside of the “courtroom”?
Not sure whether this is what you were implying, but I wasn’t thinking of private courts. My current guess is that it is important for courts to be at least observable, so that people can build trust in them (observable in the sense of how modern courts are observable, i.e. anyone can show up to the courtroom, but you might not be allowed to record it).
I think John meant that non-participants might keep commenting on the situation while the trial was in progress, and then after the trial. That might weaken some of the gains from having a trial in the first place (e.g., the hope that people will accept the verdict and move on to more productive things).
Ah, thanks, that makes sense
You could “sequester” the jury by making them promise not to read the non-courtroom threads until the jury had delivered a verdict. You could also have a norm that disputants would not comment in other threads while trial was ongoing. Not having the disputants in the non-courtroom thread would probably slow its velocity down considerably. You could even hide the courtroom thread from non-participants until the trial was over. That’s not a complete answer, but would probably help some.
Do you think you’ll build and test this on LessWrong? Feels doable.
The bottleneck feels more social than technological.
Also, I feel like someone else needs to do investigations for it to make sense for me to build the courtroom, since it does seem bad for one person to do both.