Reasons for and against posting on the EA Forum

By MichaelA🔸 @ 2021-05-23T11:29 (+32)

I think many people should be writing for or posting on the EA Forum more! And when giving career advice or talking to people with interesting ideas, one of the most common things I say is “Maybe you should start posting?”

But of course, not everyone should be posting to the Forum, and not all the time.^[1] So how can people decide?

This post discusses reasons for and against writing for / posting on the Forum, which readers can consider in light of their specific situation in order to make decisions that are right for them.

Feel free to skim or jump around this post; each section should make sense by itself.

This post doesn’t necessarily represent the views of my employers.

Summary

1. Reasons that sometimes push in favour of writing for or posting on the Forum include that doing so could help a person:
   1. Test their fit for various things (more)
   2. Improve their knowledge, “models”, & skills (more)
   3. Help other people improve the world (more)
   4. Credibly signal good things about themselves (more)
   5. Build their network (more)
   6. Gain enjoyment, satisfaction, etc. (more)
2. Reasons that sometimes push against writing for or posting on the Forum include the possibilities that doing so could:^[2]
   1. Have substantial opportunity cost (i.e., take up time and energy the writer could otherwise spend on other things) (more)
   2. Not really improve - or actually worsen - other people’s knowledge, “models”, skills, and priorities (more)
   3. Create information hazards (more)
   4. Harm the writer’s reputation, make it harder to build a network, or similar (more)
   5. Lead to unpleasant experiences after posting (more)
   6. Be something about which the writer feels fear, anxiety, or aversion (more)

I’ve put those reasons in roughly descending order by how much I think they should influence people’s decisions, in a typical case. But:

• I think that, in practice, people are often influenced more than they should be by some of these reasons - perhaps particularly some of the reasons against posting.
• Whether a given reason should influence someone’s decision will vary substantially between different people and situations.
• I encourage you to consider whether you specifically are predisposed to focus overly strongly on some of those reasons, and to consider talking this over with someone who doesn’t live in your brain.
  • See also Should You Reverse Any Advice You Hear?
  • See also Impostor syndrome

The scope of this post

• I have in mind decisions about both:
  • Whether to write something specifically for the Forum
  • Whether to adapt/cross-post for the Forum something that one has already written or will write anyway (e.g., a thesis, a paper, a post for your personal blog)^[3]
• In either case, the posts could take the form of regular posts, question posts, or shortform posts
• See How to use the Forum
• Somewhat similar points would also apply to decisions about whether to comment on the Forum
• This post isn’t focused on venues publication venues other than the EA Forum
  • E.g., academic journals, arxiv, news media, LessWrong, blog hosts like Medium
  • But of course, many people should be writing for those venues instead or as well!
    • I’ve chosen to focus on the Forum here simply because I think some people should focus on the Forum sometimes, and I happen to know more about the Forum than about other publication venues
  • That said, I do think modified versions of many of these points would apply to decisions about writing for (or sharing writings on) those other venues as well
• This post doesn’t explicitly compare the value of posting on the Forum to the value of other specific actions
  • But I hope what I write here will help you do that yourself
• This post isn’t focused on things like how to write for the Forum, pick topics to write about, get feedback, etc.
  • For that, see Notes on EA-related research, writing, testing fit, learning, and the Forum

1. Reasons that sometimes push in favour of posting

I’ll use italics for statements that are more like guesswork or based on my own experiences.

(Many of these reasons relate to EA’s ability to scalably use labour, including our ability to improve the EA-aligned research pipeline and deal with vetting constraints and organisation capacity bottlenecks.)

1.1 Test your fit for various things

• The things you’re testing your fit for could include:
  • Writing or research in general
  • Writing or research of an “EA style” or for EA audiences
  • Writing or research on specific topics
• You could learn more about your fit for these things via:
  • The process of writing the posts or of adapting other writings for the EA Forum
  • Comments you receive on drafts
    • E.g., if you send a draft to specific people for review
  • Comments you receive on the writings once posted
  • People giving you feedback in other ways
    • E.g., if you include a link for anonymous feedback
    • E.g., if you run a survey about the quality and impact of your writings

1.2 Improve your knowledge, models, and skills

• This includes knowledge and models relevant to the domains you’re writing about
  • By “improving your models”, I mean something like “building reasonable inside-view models, even if it’s still very unclear how accurate they are”, or “more fully thinking through and crystallising your beliefs about something that seems decision-relevant, but on which there’s still a lot of room for reasonable disagreement”
    • One could also call this “improving your opinions”
    • E.g., I think no one can currently know with high confidence the answer to various crucial questions about optimal timing of work and donations, but writing that linked post allowed me to come to my own tentative opinions, which has informed some donation and career decisions in a way that seems more likely to have been for the better than for the worse.
• This also includes reasoning skills, research skills, and writing and communication skills
  • These could be generic or specific to the topics you wrote about or to writing/research that’s “EA style” or “for an EA audience”^[4]
• You could improve your knowledge, models, and skills via:
  • The process of writing the posts or of adapting other writings for the EA Forum
  • Comments you receive on drafts
    • E.g., if you send it to specific people for review
  • Comments you receive on the writings once posted
  • People giving you feedback in other ways
    • E.g., if you include a link for anonymous feedback
    • E.g., if you run a survey about the quality and impact of your writings
  • I learned less from comments on published posts than I might’ve guessed.
  • But I learned a lot during the process of writing posts and from a survey I ran on my writings.
  • And in retrospect, early on, I probably should’ve more often asked people who knew more about topics I’d written drafts about to review those drafts; I imagine that would’ve helped me learn a lot.

1.3 Help other people improve the world (via your posts themselves)

• Your posts could improve other people’s knowledge, models, skills, and priorities
• This could of course be broken down into various ways in which your posts could improve people’s knowledge, models, skills, and priorities, and various ways in which that could translate into impact (e.g., improving decisions about research, careers, funding, or policies)
  • And then one could also consider how much of each of those benefits one could expect depending on various features of a post (e.g., its topic, the time invested to write it, the writing style, the author’s relevant prior knowledge, etc.)
    • But I won’t do that here.
    • For related things, see Why you (yes, you) should post on the EA Forum, Readings and notes on how to do high-impact research, and/or Impact assessment.
• I think some people overestimate the difficulty of having a nontrivial amount of impact via Forum poss
  • The EA community is relatively small, it’s relatively new, and its members are on average relatively young. I think that, partly because of this, there are still so many:
    • Bodies of knowledge that very few EAs know about, and that no one has summarised and drawn implications from for an EA audience
    • Bodies of knowledge generated by professional EA researchers which haven’t been summarised for a broader audience
    • Collections or databases that “obviously” should be made but haven’t yet
    • “Obvious” and important considerations or project ideas that have yet to be highlighted
    • Etc.
  • See also Suggestion: EAs should post more summaries and collections
  • So I think there’s a good chance that you are already well-placed to write something that would be useful to many EAs
    • *This is of course especially likely if you have an area of deep expertise that’s relatively rare in the EA community  *
  • Here are a couple examples of things I’ve made/written that really anyone could’ve done but no one had, and that I think have been pretty useful:
    • Database of existential risk estimates
    • A central directory for open research questions
    • (I also think that many of my “research insights” have been relatively obvious, and that in some cases other people had already thought similar things, but they just hadn’t been written up clearly yet)
• Regarding the impact of my own work as of approximately August 2020, I have some data and reflections here and some further thoughts here
• (I called this section “Help other people improve the world (via your posts themselves)” because posting could also make you more able to in other ways help other people improve the world, e.g. by helping you test your fit for things.)

1.4 Credibly signal good things about you

See also EA hiring, EA is vetting-constrained, and this comment by Rob Wiblin.

• The things posting could help you signal include specific areas of knowledge, skills, and interest
• This could help with things like:
  • Getting a role at an EA organisation
  • Getting funding for a project
  • Getting collaborators for a project you want to work on
  • Being reached out to to collaborate with or give input to other people
    • E.g., I’m often reached out to for career advice, input on draft posts/reports, input on project plans, etc., partly because I’ve written posts about relevant topics and thereby demonstrated relevant knowledge, skills, and interest
      • I tend to enjoy giving this input, and I think doing so is often (but not always) more impactful than what I’d accomplish with the same time otherwise
  • Reducing the costs of vetting
    • E.g., allowing you to share a post you’ve written instead of doing a work test, saving you time and saving the organisation you’re applying to money (if they’d have compensated you for your time spent on the work test)
      • But I think this has only happened for me once, and I’ve applied for ~20 roles at EA orgs. (I’ve often provided Forum posts as writing samples, but only once instead of a work test.)
    • E.g., making it easier for hirers to find good candidates without doing an open hiring round, thereby saving time for both the hirers and the people who would ultimately be rejected in an open hiring round
      • My tentative independent impression is that it’s plausible that it’s usually or always best to do an open hiring round anyway. But I could definitely be wrong.
• Posting on the EA Forum would presumably mostly help with credibly signalling to - or aiding in vetting by - people/organisations closely associated with the EA community, rather than other people.
  • This is because:
    • The hirer, funder, etc. is far more likely to have read your Forum post if they’re closely associated with the EA community
    • You’re more likely to share a Forum post to hirers, funders, etc. if they’re closely associated with the EA community
    • A key reason Forum posts might be useful as credible signals is that Forum posts may demonstrate knowledge, skills, reasoning styles, and communication styles that are particularly relevant to EA topics or particularly valued in EA
      • (If not for that fact, it would often be about as good or better to share some other writing sample, such as a university assignment.)
  • But, of course, probably everyone should apply for at least some roles, funding, etc. from people/orgs who aren’t closely associated with the EA community, and many people should primarily focus on roles, funding, etc. from such orgs.
    • This reduces, but does not eliminate, this particularly benefit of posting on the Forum.
    • See also working at EA vs non-EA orgs
• Posting on the EA Forum would presumably mostly help with credibly signalling fit for - or aiding in vetting for - research-related roles or funding.
• I think posting to the Forum might have improved/accelerated my own career trajectory.
  • In particular, there are multiple times when people who had read one or more of my Forum posts were involved in selection processes where (a) I was asked to apply (including to non-public hiring rounds), (b) I got fairly close to receiving an offer, or (c) I received an offer.
  • But it’s hard to know what the counterfactual was. E.g.:
    • Perhaps those people would’ve learned about my previous work even if I hadn’t posted it on the Forum
    • I would have applied to the public hiring rounds even if I wasn’t reached out to
    • I think the hirers’ decisions were based on work tests more than on Forum posts

1.5 Build your network

• There are multiple ways in which the process of posting could help you build your network:
  • Sending a draft to a relevant person for review, or sharing it in a more open space like the Effective Altruism Editing and Review, could:
    • make people aware of you and your knowledge, skills, and interests
    • establish initial contact between you and them
    • give you more of a sense of how useful or enjoyable further interactions with them might be
  • Posting on the Forum will make more people aware of you and of your knowledge, skills, and interests
    • This could lead to them reaching out to you then or later (e.g., at a conference)
  • People commenting on your posts could:
    • establish initial contact between you and them (at least if you reply)
    • give you more of a sense of how useful or enjoyable further interactions with them might be
• As with credibly signalling good things about yourself / aiding with vetting (see the previous section):
  • Building your network could help with things like:
    • Getting a role at an EA organisation
    • Getting funding for a project
    • Getting collaborators for a project you want to work on
    • Being reached out to to collaborate with or give input to other people
    • Reducing the costs of vetting
  • Posting on the EA Forum would presumably mostly help with credibly signalling to, or aiding in vetting by, people/organisations closely associated with the EA community, rather than other people
  • Posting on the EA Forum would presumably mostly help with credibly signalling fit for, or aiding in vetting for, research-related roles or funding
  • Posting on the Forum definitely helped me build my network, which in turn definitely led to me being reached out to to collaborate with or give input to other people, and may have accelerated or improved my career trajectory.
    • See the previous section for discussion of these points.

1.6 Enjoyment, satisfaction, etc.

AKA acquire additional utils

• For example, you could get enjoyment or satisfaction from the writing process, upvotes and positive/interesting comments on your posts, and good interactions that come from building your network (see the previous section).

2. Reasons that sometimes push against posting

As above, I’ll use italics for statements that are more like guesswork or based on my own experiences.

2.1 Opportunity cost (i.e., you’d have to spend time and energy that you could otherwise spend on other things)

• I think this is the most common reason for it being a good idea to not write a post for, or share writings to, the Forum
• But it is often possible to write / share a post relatively quickly
  • This goes especially for sharing things one already has written anyway or already would write anyway
    • E.g., I think many people in the EA community have written or will in any case write potentially EA-relevant things, for example for university assignments or their personal blogs. I think that more of these things should be posted to the Forum.^[5]
  • That said, there’s of course usually a tradeoff between achieving more of the benefits of posting and reducing the time and energy cost
  • For specific tips on relatively low-cost ways to make useful posts, see Notes on EA-related research, writing, testing fit, learning, and the Forum

2.2 Posting might not really improve - or might actually worsen - other people’s knowledge, models, skills, and priorities

• This is the flipside of “Help other people improve the world (via your posts themselves)”; see that section for some relevant thoughts
• This problem could occur if your post:
  • says things most readers would already have thought of
  • says things that aren’t really relevant to any important decisions
  • is unclear
  • says things that are false or misleading
  • distracts people from - or makes it harder to find - more useful or important things
  • (This is not necessarily an exhaustive list)
• These things of course can happen
  • I also think some people are probably insufficiently concerned about these things
    • See also
  • But I think it’s more common to be overly concerned about these things
    • And I think the fact you’re reading this sentence is some evidence that you’re the kind of person who’d tend to worry too much rather than too little
      • See also Should You Reverse Any Advice You Hear?
    • One of the main reasons I say this is because I think there are various mechanisms that mean there may be very little harm even if you do say things that are (relatively) obvious, unimportant, unclear, incorrect, etc.
      • Mechanisms you can use include:
        • providing an epistemic status
        • providing a clear summary
        • trying to maintain reasoning transparency
        • getting feedback before posting (see here for tips on getting feedback)
      • Meanwhile, mechanisms that don’t rely on you using them include:
        • The karma system (on average, the less useful the post, the fewer upvotes it’ll get, and so the faster it’ll slip off the front page)
        • Comments (people can highlight reasons why the post is flawed or may not be very useful)
        • Tags (tags on your post can help people decide whether it’s relevant to them, and the fact people can use tags to find other posts means it’s less bad to “clutter up the front page”)
        • The search bar (which again reduces how much bad “cluttering up the front page” is)

2.3 Information hazards

• An information hazard is a “risk that arises from the dissemination or the potential dissemination of (true) information that may cause harm or enable some agent to cause harm” (Bostrom, 2011).
  • One common - but commonly overlooked - type of information hazard is an attention hazard. This refers to cases where the “mere drawing of attention to some particularly potent or relevant ideas or data increases risk, even when these ideas or data are already ‘known’” (Bostrom, 2011).
• I think that information hazards are often a good reason to post on the Forum
  • And there are several things I’ve written but not posted for that reason
  • And there are a few cases where I’ve felt that something that was posted perhaps shouldn’t have been posted, for information hazard reasons
• That said, I also think some people are more worried about information hazards in a given specific instance than they really should be
• This is a complicated topic, and I won’t provide specific guidance here. If someone thinks something they want to post might pose information hazards, I recommend that they read some writings on information hazards and the unilateralist’s curse and what to do about those things.

2.4 Posting might harm your reputation, make it harder to build a network, or similar

• This is the flipside of “Credibly signal good things about you” and “Build your network”; see those sections for some relevant thoughts
• It’s worth quickly noting that, in some cases, a harm to an aspect of your reputation - or it being harder for you to build a network in some area - might be good for the world and perhaps even for you
  • This is if the harm and difficulty just appropriately reflects you indeed not being a good fit for something
  • But there could of course be other cases where the harm and difficulty are not appropriately calibrated, and where they’re bad for you and for the world
• I think that this risk is worth at least a little bit of consideration, at least for people who (a) might want to get jobs at explicitly EA organisations and (b) might by default fire off many quickly written posts
  • E.g., people very much like me
  • I also have some indications that this downside has occurred to a small extent for me
    • However, I think that, overall, posting the way I have is more likely to have benefitted than harmed my career
    • And I’m a relatively extreme case (e.g., I think I might’ve made the most non-link-post Forum posts of anyone last year)

2.5 Posting might lead to unpleasant experiences

• This is the flipside of “Enjoyment, satisfaction, etc.”
• These negative experience could also perhaps reduce your engagement with and retention in EA, and increase your value drift

2.6 Fear, anxiety, or aversion about posting

• I think that this is an entirely understandable reason for people to not write / share posts on the Forum
  • *And it seems to unfortunately be one of the most common and strongest reasons influencing people *
  • And I sympathise with people who feel this way
  • But I think this fear, anxiety, or aversion is unnecessarily holding people back, and that it would be good to - at a community or individual level - try to find ways to mitigate or push through it
    • I’m hoping that this post, my other post Notes on EA-related research, writing, testing fit, learning, and the Forum, and some of the links in that post can help
      • E.g., maybe the options listed in that post for getting feedback on drafts could help people get “over the hurdle” for their first few posts

See also

• Aaron Gertler: Why you (yes, you) should post on the EA Forum
• Notes on EA-related research, writing, testing fit, learning, and the Forum
• How you can make an impact on the EA Forum
• Goals we might have when taking actions to improve the EA-aligned research pipeline
• Potential benefits and downsides of making and/or sharing a research agenda [I’ll be posting this soon]

Acknowledgements

I’m grateful to Neil Dullaghan, Aaron Gertler, Peter Hurford, David Moss, and Saulius Šimčikas for comments on an earlier draft of this post. This does not imply that these people endorse all aspects of this post.