About three hours ago I posted a brief argument about why one of the US presidential candidates is strongly preferable to the other on five criteria of concern to EAs: AI governance, nuclear war prevention, climate change, pandemic preparedness and overall concern for people living outside the United States. I concluded by urging readers who were American citizens to vote accordingly, and encourage anyone whom they might potentially influence to do so. After about an hour the post had a karma of 23; shortly after that, it was removed from the front page and relegated to 'personal blogposts'. Not surprisingly, at that point it started to attract less attention.
Not having heard of 'personal blogposts', I checked the description, which suggested that they were appropriate for '[bullet point] topics that aren't closely related to EA; [bullet point] topics that are difficult to discuss rationally; [bullet point] topics of interest to a small fraction of the forum's readers (e.g. local events)'. Frankly, I can't see how my blogpost fit any of these descriptors. It focused on the candidates' positions on issues of core concern to EAs, and from the fact that it had a karma score of 23 after an hour, it was obviously of interest to the message board's readership. The remaining possibility--unless there were other unstated criteria--is that it was judged to be on a topic that was 'difficult to discuss rationally'.
If so, I think that's a troubling commentary on EA, or the moderator's conception of EA. My post was clearly partisan, but I don't think any reasonable observer would have called it a rant. This election will almost surely make more difference to most of the causes that EAs hold dear than any other event this year--perhaps any other event this decade. Shouldn't the EA community, if anybody, be able to discuss these issues in a reasonably rational manner? I'd be grateful for a response from the moderator justifying the decision to exclude them.
You can read our politics policy here.
This is weird to me. There are so many instances of posts on this forum having a “strong polarizing effect… [consuming] a lot of the community’s attention, and [leading] to emotionally charged arguments.” The several posts about Nonlinear last year strike me as a glaring example of this.
US presidential candidates’ positions on EA issues are more important to EA—and our ability to make progress on these issues—than niche interpersonal disputes affecting a handful of people. In short, it seems like posts about politics are ostensibly being held to a higher standard than other posts. I do not think this double standard is conducive to healthy discourse or better positions the EA community to achieve its goals.
I agree that this is inconsistent (looks like Ben's Nonlinear post is front page). But my conclusion is that community drama should also be made less visible except to those who opt in, not vice versa. The separate section for community posts was a decent start
No shade to the mods, but I'm just kind of bearish on mods' ability to fairly determine what issues are "difficult to discuss rationally," just because I think this is really hard and inevitably going to be subject to bias. (The lack of moderation around the Nonlinear posts, Manifest posts, Time article on sexual harassment, and so on makes me think this standard is hard to enforce consistently.) Accordingly, I would favor relying on community voting to determine what posts/comments are valuable and constructive, except in rare cases. (Obviously, this isn't a perfect solution either, but it at least moves away from the arbitrariness of the "difficult to discuss rationally" standard.)
This seems a question of what the policy is, not of judgement re how to apply it, in my opinion.
The three examples you gave obviously are in the category of "controversial community drama that will draw a lot of attention and strong feelings", and I trust the mod's ability to notice this. The question is whether the default policy is to make such things personal blog posts. I personally think this would be a good policy, and that anything in this category is difficult to discuss rationally. I do also consider the community pane a weaker form of low visibility, so there's something here already, but I would advocate for a stronger policy.
Another category is "anything about partisan US politics", which I don't think is that hard to identify, is clearly hard to discuss rationally, and in my opinion is reasonable to have a policy of lowering the visibility of.
I don't trust karma as a mechanism, because if the post is something that people have strong feelings about, and many of those feelings are positive (or at least, righteous anger style feelings), then posts often get high karma. Eg I think the Nonlinear posts got a ton of attention, in my opinion were quite unproductive and distracting, got very high karma, and if they had been less visible I think this would have been good
Upvoted, but I don't think one could develop and even-handedly enforce a rule on community-health disputes that didn't drive out content that (a) needed to be here, because it was very specifically related to this community or an adjacent one, and (b) called for action by this community. So I think those factors warrant treating community-health dispute content as frontpage content, even though it lets a lot of suboptimal content slip through.
I think you may have a point on "positions on EA issues" narrowly defined -- but that is going to be a tough boundary to enforce. Once someone moves to the implied conclusion of "vote for X," then commenters will understandably feel that all the reasons not to vote for X are fair commentary whether or not they involve "positions on EA issues." [ETA: I say narrowly defined because content about how so-and-so is a fascist, or mentally unstable, or what have you is not exactly in short supply. I have little reason to believe that anyone is going to change their minds about such things from reading discussions on the Forum.]
There's also a cost to having a bunch of partisan political content -- the vast majority of which would swing in one direction for the US -- showing up when people come to EA's flagship public square. We have to work with whoever wins, and tying ourselves to one team or the other more than has already happened poses some considerable costs. There is much, much less broader risk on community-health disputes like Nonlinear (one can simply choose not to read them).
Yeah, just to be clear, I am not arguing that the "topics that are difficult to discuss rationally" standard should be applied to posts about community events, but instead that there shouldn't be a carveout for political issues specifically. I don't think political issues are harder to discuss rationally or less important.
Could you provide examples of political discussions on the EA Forum that appear to have negatively impacted the forum’s environment or impaired its ability to achieve its objectives? While I find this plausible, I’d also expect the EA Forum to be one of the most conducive spaces online for constructive political discourse.
My understanding is that the forum’s primary goal is to support discussions relevant to effective altruism and facilitate the coordination of related projects. Given that politics is highly relevant to these aims, I believe there should be a strong(er) justification for any restrictions on political topics.
I think the repeated guilt-by-association posts pointing out that someone in EA associated with someone who has some right wing views are pretty negative.
Which posts? (you don't need to list them, just briefly describe them so I can find them myself)
While My experience at the controversial Manifest 2024 (and several related posts) was (were) not explicitly about policies or politicians, I think it's largely the underlying political themes that made it so heated.
Manifest was advertised on Forum and the controversial speakers were IIRC largely advertised and invited guests. Some of the talks were at least adjacent to the objected-to views.
That seems a significantly tighter connection than "someone in EA associated with someone who has some right wing views."
Thanks! Yeah, I thought maybe this was what Larks was referring to. Putting to one side the question of whether that was a valuable discussion or not, I wouldn't put that in the same category as OP's post. The Manifest discussion was about whether an organisation such as Manifest should give a platform to people with views some people consider racist, OP's post is an analysis of the policy platform of a leading candidate in what is arguably the world's most important election. I wouldn't describe the former discussion as 'political' in the same way that I would describe the OP's post. But perhaps others see it differently?
As far as I remember, the political discussions have been quite civilized on the EA Forum. But I think this is because of the policies and culture the EA Forum has. If political discussions were a lot more frequent, the culture and discussion styles could get worse. For example, it might attract EA-adjacent people or even outsiders to fight their political battles on the EA Forum. Maybe this can be solved by hiring additional moderators though.
Also, politics can get a lot of attention that would be better spend elsewhere. For example this post about Trump generated 60 comments, and I am not sure if it was worth it.
So you think so far it's mostly been OK? If that's the case, and if it's plausible that high-quality discussions about politics would be valuable, shouldn't we lean towards loosening the policy and seeing what happens?
Best case, good discussions happen and the forum does a better job of meeting its objective. Worst case, bad discussion happens, but then it should be pretty simple to tighten the policy up and no lasting harm would be done.
Not sure what to make of it, but one of 80k hrs top recommendations is Government and policy - here it seems like several top careers could need to consider what party to work for. I might agree that discussions about who to vote for might not be high priority (although I think Rob Wiblin made a really good point that in "swing districts" voting might be really high value in expectation). That said, there might be a trade-off for many people, perhaps even EA as a whole between whether to try to make those issues that are still not partisan (like AI perhaps) stay non-partisan and using our resources to galvanize a political faction around issues that are partisan.