The EA forum (and LessWrong) are both structured primarily as a newsfeed of posts sorted by date. This caters well to immediate engagement, but is much worse for building up a repository of knowledge which is accessible and relevant over a long time period. LessWrong 2.0 has (to some extent) managed to avoid this problem by having a) curated content, so that people don't have to look at literally everything which is posted, and b) sequences which store great posts in a format that makes them easily accessible a long time afterwards. The EA forum has neither. This makes it rather frustrating to try to use it to build on existing intellectual progress, as I recently found out while reviewing forum posts on career advice. Why don't we have any mechanisms for ensuring good content lasts, and what can be done about this? (Even just a blanket 'curate everything above x karma' strategy would help, while requiring very little moderator effort. EDIT: I actually no longer believe this last part, I think the key thing is collating material from across the internet.)
35
35
Reactions
1 Answers sorted by
Update from CEA:
We've been planning to implement "collections of recommended reading" (name not yet finalized) since the Forum was launched (see "core series of posts" here).
As Habryka mentioned in his comment, we're bottlenecked on developer time, but the most important factor behind our pushing back the timeline was Max's move from being the head of CEA's Content team to becoming our interim Executive Director.
Max had been leading the charge to make "collections" happen, and his move dropped our content team from two people to one (myself). For the first few months of this year, I've been prioritizing existing content projects over new features that are content-intensive.
Still, we think that a "collection" feature would be very valuable, and we do intend for it to exist on the Forum at some point (though I can't yet commit to a specific release date). I have an initial "introduction to EA" collection in the works, and I look forward to seeing what other people build when the feature goes live.
Meanwhile, if anyone is champing at the bit to create EA reading lists, you could publish your list in a post and ask for community feedback. That would give you a head start on making a collection when that becomes an option!
I remember from conversations with the EA Forum team that adding this is indeed planned, but because CEA is currently quite bottlenecked on developer-hours, they wanted to minimize the code-footprint of the EA Forum and adapt its functionality iteratively. I expect that these features will be ported over eventually, though I am not sure about the timelines on that.
I did spend a day or two collating some potential curated sequences for the forum.
- I still have a complete chronological list of all public posts between Eliezer and Holden (&friends) on the subject of Friendly AI, which I should publish at some point
- I spent a while reading through people's work like Nick Bostrom and Brian Tomasik (I didn't realise how much amazing stuff Tomasik had written)
- I found a bunch of old EA blogs by people like Paul Christiano, Carl Shulman, and Sam Bankman-Fried that would be good to collate the best pieces from
- I constructed a mini versions of things like the sequences, the codex, and Owen Cotton-Barratt's excellent intro to EA (prospecting for gold) as ideas for curated sequences on the Forum.
I think it would be good from a long-term community norms standpoint to know that great writing will be curated and read widely.
Alas, CEA did not seem to have the time to work through any sequences (seemed like there was a lot of worries about what signals the sequences would send, and working through the worries was very slow going). At some point if this ever gets going again, it would be good to have a discussion pointing to any good old posts that should be included.
It's very cool that you took the time to do so. I agree that preserving and showcasing great content is important in the long term, and am sad that this hasn't come to anything yet. Of course the EA forum is still quite new, but my intuition is that collating a broadly acceptable set of sequences (which can always be revised later) is the sort of thing that would take only one or two intern-weeks.
My intuition, having seen proposals from people both inside and outside of CEA, is that this collation will almost certainly take longer than a week or two:
- A higher standard than "broadly acceptable" seems important, since whatever posts are chosen will be seen as having CEA's endorsement (assuming CEA is the one doing the collation). A few critics can contribute a lot of negative public feedback, and even a single unfortunate line in a curated post may cause problems later.
- I also think there's a lot of value to publishing a really good collection the first time around:
- Making frequent revisions to a "curated" collection of posts makes them look a lot less curated, and removes comments from the public eye that authors may have worked on assuming they'd stick around.
- It's also not great if Post A is chosen for curation despite Post B being a much stronger take on the same subject; assembling a collection of posts that are roughly the best posts on their respective topics takes a lot of experience with EA content and consultation with other experienced people (no one has read everything, and even people who've read almost everything may differ in which pieces they consider "best").
That said, the task is doable, and I'm consulting with other CEA staff who work on the Forum to draft a top-level answer about our plans for this feature.
In my opinion, organizations may do best to avoid officially endorsing anything other than the most central content that they produce in order to reduce these PR headaches, regarding both
- what's said in the endorsed articles and
- which articles were or weren't chosen to begin with (the debate over the EA Handbook comes to mind).
As an alternative, maybe individual people could create their own non-CEA-endorsed lists of recommended content, and these could be made available somewhere. Having many such lists would allow for diversity based on interests and values. (For example, "The best global poverty articles", "The best career articles", "The best articles for suffering-focused altruists", etc.)
I haven't yet finished thinking about how the EA Forum Team should go about doing this, given their particular relationship to the site's members, but here's a few thoughts.
I think, for a platform to be able to incentivise long-term intellectual progress in a community, it's important that there are individuals trusted on the platform to promote the best content to a place on the site that is both lasting and clearly more important than other content, like I and others have done on the AI Alignment Forum and LessWrong. Otherwise the site devolves into a news site, with a culture that depends on who turns up that particular month.
I do think the previous incarnation of the EA Forum was much more of a news site, where the most activity occurred when people turned up to debate the latest controversy posted there, and that the majority of posts and discussion on the new Forum are much more interested in discussion of the principles and practice of EA, rather than conflict in the community.
(Note that, while it is not the only or biggest difference, LessWrong and Hacker News both have the same sorting algorithm on their posts list, yet LW has the best content shown above the recent content, and thus is more clearly a site that rewards the best content over the most recent content.)
It's okay to later build slower and more deliberative processes for figuring out what gets promoted (although you must move much more quickly than the present day academic journal system, and with more feedback between researchers and evaluators). I think the Forum's monthly prize system is a good way to incentivise good content, but it crucially doesn't ensure that the rewarded content will continue to be read by newcomers 5 years after it was written. (Added: And similarlycurrent new EAs on the Forum are not reading the best EA content of the past 10 years, just the most recent content.)
I agree it's good for members of the community to be able to curate content themselves. Right now anyone can build a sequence on LessWrong, then the LW team moves some of them up into a curated section which later get highlighted on the front page (see the library page, which will become more prominent on the site after our new frontpage rework). I can imagine this being an automatic process based on voting, but I have an intuition that it's good for humans to be in the loop. One reason is that when humans make decisions, you can ask why, but when 50 people vote, it's hard to interrogate that system as to the reason behind its decision, and improve its reasoning the next time.
(Thanks for your comment Brian, and please don't feel any obligation to respond. I just noticed that I didn't intuitively agree with the thrust of your suggestion, and wanted to offer some models pointing in a different direction.)
It crucially doesn't ensure that the rewarded content will continue to be read by newcomers 5 years after it was written... New EAs on the Forum are not reading the best EA content of the past 10 years, just the most recent content.
This sentence deserves a strong upvote all by itself, it is exactly the key issue. There is so much good stuff out there, I've read pretty widely on EA topics but continue to find excellent material that I've never seen before, scattered across a range of blogs. Gathering that together seems vital as the movement gets older and it gets harder and harder to actually find and read everything.
I can imagine this being an automatic process based on voting, but I have an intuition that it's good for humans to be in the loop. One reason is that when humans make decisions, you can ask why, but when 50 people vote, it's hard to interrogate that system as to the reason behind its decision, and improve its reasoning the next time.
I think that's true when there are moderators who are able to spend a lot of time and effort thinking about what to curate, like you do for Less Wrong. But right now it seems like the EA forum staff are very time-constrained, and in addition are worried about endorsing things. So in addition to the value of decentralising the work involved, there's an additional benefit of voting in that it's easier for CEA to disclaim endorsement.
Given that, I don't have a strong opinion about whether it's better for community members to be able to propose and vote on sequences, or whether it's better for CEA to take a strong stance that they're going to curate sequences with interesting content without necessarily endorsing it, and ensure that there's enough staff time available to do that. The former currently seems more plausible (although I have no inside knowledge about what CEA are planning).
The thing I would like not to happen is for the EA forum to remain a news site because CEA is too worried about endorsing the wrong things to put up the really good content that already exists, or sets such a high bar for doing so that in practice you get only a couple of sequences. EA is a question, not a set of fixed endorsed beliefs, and I think the ability to move fast and engage with a variety of material is the lifeblood of an intellectual community.
I also think there's a lot of value to publishing a really good collection the first time around
The EA handbook already exists, so this could be the basis for the first sequence basically immediately. Also EA concepts.
More generally, I think I disagree with the broad framing you're using, which feels like "we're going to get the definitive collection of essays on each topic, which we endorse". But even if CEA manages to put together a few such sequences, I predict that this will stagnate once people aren't working on it as hard. By contrast, a more scalable type of sequence could be something like: ask Brian Tomasik, Paul Christiano, Scott Alexander, and other prolific writers, to assemble a reading list of the top 5-10 essays they've written relating to EA (as well as allowing community members to propose lists of essays related to a given theme). It seems quite likely that at least some of those points have been made better elsewhere, and also that many of them are controversial topics within EA, but people should be aware of this sort of thing, and right now there's no good mechanism for that happening except vague word of mouth or spending lots of time scrolling through blogs.
I'm sorry if my framing was misleading: When this feature goes live on the Forum, other users will be able to use it freely. CEA still wants to have its own "collections" be as close to "definitive" as we can reasonably get, with occasional updates/added material.
Meanwhile, until the feature goes live, I'm considering ways to more reliably expose Forum visitors to collections of introductory material that already exist, like the material compiled on EA.org. Maybe a pinned post, or maybe a page that shows by default to non-logged-in users; that's still in the works.
Isn't all the code required for curation already implemented for Less Wrong? I guess adding functionality is rarely easy, but in this case I would have assumed that it was more work to remove it than to keep it.
The EA forum is a forum with just two categories so it functions more like a news board, which means most conversation ends up happening in bespoke Facebook groups for various causes, careers and interest groups. Facebook has a similar problem of losing good posts and conversations, which might have been stickied in a traditional forum with multiple categories and sub-categories.
There may be advantages to having these discussions happen on Facebook as people might be more likely to check, although it does make it hard to find out where the conversations are happening.
Agreed that subforums are a good idea, but the way they're done on facebook seems particularly bad for creating common knowledge, because (as you point out) they're so scattered. Also the advantage of people checking facebook more is countered, for me, by the disadvantage of facebook being a massive time sink, so that I don't want to encourage myself or others to go on it when I don't have to. So it would be ideal if the solution could be a modification or improvement to the EA forum - especially given that the code for curation already exists!
Wasn't it announced at launch, that this would be implemented at some point?
Here's the post I believe Yannick was thinking of. (Find the phrase "core series of posts".)
This is still something we plan to do in the future; I'm consulting with other CEA staff who work on the Forum to draft a top-level answer to Richard's question.
Related thought: I haven't had a chance to look closely at this yet, but it appears to be the nth proposal for something like this I've seen on the EA forum. This might also represent a failure of the "newsfeed of posts sorted by date" model.
What if there was a box to check for "this is a proposal", and then once that box was checked, we got "bump" style forum mechanics where every time someone left a comment, the post went back to the top of the front page? That way a proposal could stay on the front page long enough for details to get hashed out & eventually the proposal might actually get implemented.
Proposals could also sort comments by "new" by default, so you could write a comment summarizing all of the discussion so far & suggesting a way to synthesize it, and that comment would be the first comment to get displayed. Perhaps there could also be something to nudge the user towards reading most or all of the comments in a proposal thread before leaving a comment of their own?
(Are there other categories of posts that would benefit from additional time on the front page? BTW, another advantage of the bumping mechanic is it makes it easier to have an influence on the discussion even if you only check the forum occasionally. Which is likely going to be true for important people who have a lot of other responsibilities.)
Another situation where it can be valuable for a post to spend more time on the frontpage: This essay argues it's important to have 4 layers of intellectual conversation. The number 4 seems arbitrary to me, but I agree with the overall point that back-and-forth is valuable and necessary. But if a post falls off the frontpage partway through that back-and-forth, people are less motivated to continue the back-and-forth because the audience is smaller.
Another situation where it can be valuable for a post to spend more time on the frontpage: This essay argues it's important to have 4 layers of intellectual conversation. The number 4 seems arbitrary to me, but I agree with the overall point that back-and-forth is valuable and necessary. But if a post falls off the frontpage partway through that back-and-forth, people are less motivated to continue the back-and-forth because the audience is smaller.