Running an AMA on the EA Forum
AMAs (âAsk Me Anythingâ) are some of my favorite posts on the Forum, and Iâd love to see more of them.
I sometimes ask specific people if theyâd like to run AMAs, but you donât need an invitation to do this! Anyone can run an AMA if they want to.
How do AMAs work?
(Optional) You can reach out to me if you have questions or concerns.
You write a post explaining who you are, and what sorts of things youâd like to answer questions about.
You tell people when you plan to answer questions, and whether youâll still answer more questions that pop up after that.
You hit âpublishâ, and then share the post as you see fit (the more places you tell people youâre running an AMA, the more questions youâll get).
Two great examples of AMA posts: Ajeya Cotra and Owen Cotton-Barratt (you can also check the tag for lots of other examples)
Other notes
You donât have to commit to answering every question. Feel free to set a narrow window for when you plan to respond, and to skip questions as you see fit.
You donât have to share your name to run an AMA. However, if you want to run one anonymously, a moderator may reach out to ask for evidence about your experience (this is standard practice on Reddit, where the AMA format first became popular).
We frequently pin AMAs and share them on social media, but this may depend on what else is pinned/âbeing shared when your AMA goes up â I canât guarantee it for any specific AMA.
What makes a good AMA?
Some good reasons you could run an AMA (though Iâm sure there are others I havenât thought of):
Your job has some relation to EAâs goals and you want to share your experiences
Note that this doesnât just mean âworking at an organization with an explicitly EA missionâ. It could also mean:
Working at a nonprofit in an EA-linked area (e.g. the Gates Foundation)
Working at a for-profit in an EA-linked area (e.g. Impossible Foods)
Working in a position where youâve learned things about running an organization well (e.g. COO at a fast-growing startup)
Working in a position where you know a lot about some area that is important to EAâs mission (e.g. political lobbyist)
You know a lot about something EA-related and want to share your knowledge
Example #1: A PhD student in economics who recently worked on an RCT in Kenya and can answer questions about âon-the-groundâ development work
Example #2: An avid reader whoâs been exploring nuclear policy for fun and has a strong grasp of that literature
Example #3: Michael Aird gave this a try, and I really appreciate him for it!
You have some other experience you think people on the Forum would benefit from hearing about
Example #1: You grew up in an area with a lot of extreme poverty and have firsthand experience with the types of people GiveWellâs charities aim to help
Example #2: Your parents own a farm that raises animals for large meat producers, and you have firsthand experience with that industry (and with how some farmers react to animal welfare efforts)
Example #3: Youâve run a bunch of AI-box experiments and have thoughts on what they tell us about AI alignment (or maybe you think the whole thing was a waste of time, which would also be interesting to hear about)
If you have a very narrow range of things you want to talk about, a regular post might be better than an AMA (e.g. a post summarizing a research paper you wrote, rather than an AMA about the paper). But thereâs nothing wrong with trying the AMA option.
Any questions?
If you think you might want to run an AMA, but have questions or concerns, weâd be happy to talk!
Reach out to the Forum team at forum@centreforeffectivealtruism.org, weâll be in touch soon.
- AMA: 10 years of EarnÂing To Give by (10 Dec 2024 22:12 UTC; 116 points)
- AMA: EarnÂing to Give by (7 Nov 2023 16:20 UTC; 92 points)
- AMA: EarnÂing to Give by (LessWrong; 7 Nov 2023 16:20 UTC; 53 points)
- AMA: Ian David Moss, stratÂegy conÂsulÂtant to founÂdaÂtions and other institutions by (2 Mar 2021 16:55 UTC; 40 points)
I like answering questions when Iâm asked them, but I suspect that there wonât be (m)any questions asked to me if I did an AMA. Should I do one anyway?
Maybe a group AMA thread would be a good idea, where everyone can post a comment that they are happy to answer questions?
Another idea is to set up conditional AMAs, e.g. âI will commit to doing an AMA if at least n people commit to asking questions.â This has the benefit of giving each AMA its own time (without competing for attention with other AMAs) while trying to minimize the chance of time waste and embarrassment.
This is a nice idea!
One slight downside of a group AMA thread is that itâs not quite as easily indexable as smaller AMA threads would be (harder to get good tags on it, harder for people to know itâs a thing theyâre interested in based on the title). But I still think it would be a clear net positive if this were to exist.
I could post one of these, but I think it would also be good to have it come from a more grassroots source than the head moderatorâDenise, would you be down to post something, assuming that you would be willing to answer questions on some topic? (I can also do it, just wanted to give you the first shot.)
I had similar concerns about our Operations AMA recently. It wasnât wildly popular, but we got 7 questions and I still felt like it was a good use of my time. Several people in the group said they really enjoyed it and would be interested in doing another one, and I liked it enough that Iâm planning to do another AMA for one of my other projects as well.
Iâll also mention that itâs a (relatively) low-effort way to create content (and get karma, if you care). I often feel like I should post to the Forum more but either donât feel like I have anything worth posting, or donât have the time to write anything out, but the nice thing about AMAs is that you donât have to come up with a novel topic that fits neatly into a typical EA Forum post, and the standard for quality as far as formatting/âorganization/âetc. is lower.
The only downsides of posting that I see is time spent on creating the post (I estimate we collectively spent about an hour on this, though I think you could do a less detailed one in 15 minutes), and I suppose the possible embarrassment of not getting asked any questions, but I think this is unlikely (I donât think itâs ever happened on the Forum), and you can always delete the post if youâre really concerned about that.
FWIW I think youâd be well-suited to do an AMA :)
Yeah, my view is that it takes little time to set up an AMA, and can take very little time if you go for a barebones approach, such that the âmay as well just run an AMA and seeâ bar is fairly low.
Personally, I had relatively little engagement with my AMA, which was basically what I expected (as I was focusing on a somewhat niche topic that Iâm not a proper expert on). But I still think setting it up made sense ex ante, since it took me little time and there was some chance it wouldnât been more useful.
(Also, focusing on Jamie Harrisâs case specifically, I share Marisaâs view that an AMA by Jamie would get a decent amount of engagement.)
Agree with Marisa that youâd be well suited to do an AMA
You should!
Iâll note that an AMA doesnât have to be âtime-limitedâ: you can set it up, and then link to it in your Forum bio/âelsewhere so that people know they can always ask questions.
For example, if you left a detailed response on someoneâs post/âcomment, you could add: âIf you have other questions about this area that arenât exactly on this topic, you can email me or put them on my AMA thread.â Or something like that.
An alternative to an AMA might be an open discussion thread on a given topic, perhaps with specific people committing to be active in the discussions (could be experts, but not necessarily)
Iâve thought of this as an alternative in cases where the person thinks that there are likely many people with more experience than themselves, but where they can still generate useful answers and insights.