Summary of Core Feedback Collected by CEA in Spring/Summer 2019

by Ben_West 14d7th Nov 201912 min read30 comments

106


Introduction

The Centre for Effective Altruism (CEA) aims to grow and maintain the Effective Altruism (EA) movement. As part of that work, it is important for us to understand the needs, values, and concerns of members of the EA community.

CEA collects feedback from community members in a variety of ways (see “CEA’s Feedback Process” below). In the spring and summer of 2019, we reached out to about a dozen people who work in senior positions in EA-aligned organizations to solicit their feedback. We were particularly interested to get their take on execution, communication, and branding issues in EA. Despite this focus, the interviews were open-ended and tended to cover the areas each person felt was important.

This document is a summary of their feedback. The feedback is presented “as is,” without any endorsement by CEA. This feedback represents a small (albeit influential) portion of the EA community, and should be considered in context with other sources of feedback.

This post is the first in a series of upcoming posts where we aim to share summaries of the feedback we have received. The second is here.

CEA’s Feedback Process

CEA has, historically, been much better at collecting feedback than at publishing the results of what we collect. This post is part of our attempt to address that shortcoming and publish more feedback, but, until more can be published, we want to share more details about the types of feedback we collect.

As some examples of other sources of feedback CEA has collected this year:

  • We have received about 2,000 questions, comments and suggestions via Intercom (a chat widget on many of CEA’s websites) so far this year
  • We hosted a group leaders retreat (27 attendees), a community builders retreat (33 attendees), and had calls with organizers from 20 EA groups asking about what’s currently going on in their groups and how CEA can be helpful
  • Calls with 18 of our most prolific EA Forum users, to ask how the Forum can be made better.
  • A “medium-term events” survey, where we asked everyone who had attended an Individual Outreach retreat how the retreat impacted them 6-12 months later. (53 responses)
  • EA Global has an advisory board of ~25 people who are asked for opinions about content, conference size, format, etc., and we receive 200-400 responses to the EA Global survey from attendees each time.

The feedback summarized in this document sometimes agrees with other feedback we have received, and sometimes disagrees. This document generally presents feedback “as is” in an attempt to give an accurate summary of people’s responses, even if the feedback here disagreed with opinions we have gotten from other data sources.

Solutions Mentioned in this Document

In addition to examples of concerns respondents raised, this document contains efforts CEA has implemented which may address the concern. Efforts are ongoing, so these ideas are not intended to be final solutions, and we will continue to iterate as we gather more information about how things are working. Additionally, the solutions were not necessarily triggered by this feedback – many of these projects were started before the feedback round was run, were inspired by other feedback, etc.

Executive Summary

Things Which Are Going Well

  1. CEA’s Community Health and Events Projects. Respondents felt that the Community Health team does important work to keep the community safe, and there is a strong argument for a central entity like CEA to oversee community health. EA Global is the “flagship” event of the community, and smaller events run by CEA were also positively regarded.
  2. EA Community Members are Smart, Talented, and Thoughtful. Respondents frequently mentioned that the EA community is comprised of very smart, very promising community members who actually care about the world. They felt that EAs were generally focused on important questions, and genuinely trying to get the right answers.
  3. EA Community Members are Kind. Respondents also mentioned that EAs are “nice” or “kind” and have high integrity norms.

Things Which Could Be Improved

  1. Execution Issues. Respondents mentioned several times that CEA “overpromised and under delivered”. EA Grants is one example: CEA did not respond to applicants with grant decisions in the time we promised. Ensuring that we are able to make good on all of our promises has been a key focus of CEAs in 2019; for example, we did not have an open Grants round in 2019 to ensure that we were not stretched thin.
  2. Communication and Branding Issues. Respondents felt that some CEA projects, such as EA Global and the EA Forum, contain content which is not representative of what they feel is the best of EA. CEA has worked to make EA Global talks more representative, and to highlight the best content on the Forum.
  3. Community Welcomingness. Several respondents mentioned concerns with how welcoming the community is (or seems) to outsiders. Common concerns were a desire for more epistemic humility and welcomingness to outsiders. CEA has introduced several pilot projects, including the Guides program at EA Global, to begin to address these concerns in our own programs. For example, the Guides pilot project at EA Global matches first-time attendees with returning attendees, to create a warm welcome. 87% of attendees at EA Global SF 2019 stated that “EA Global is a place where I felt welcome.”
  4. Demographic Diversity. Respondents also mentioned concerns about demographic homogeneity in EA hampering the movement. CEA shares these concerns, as described in our statement on D&I. CEA has made some progress in this area. For example, when hiring CEA staff or recruiting EA Global speakers, we have processes in place to seek out excellent candidates from underrepresented groups. CEA staff and EA Global speakers are now approximately evenly gender balanced. However, we know we still have work to do here. We are piloting programs like Guides, Fellowships, and affinity group meetups at EA Global to build connections, solicit feedback, and further move the needle towards an EA that is more diverse.

Things Which Are Going Well

CEA’s Projects

Community Health

Respondents were uniformly positive about CEA’s Community Health project. Julia Wise has been the contact person for the community for several years, and Sky Mayhew joined in January 2019.

Some respondents compared the EA community to other communities they are a part of and felt that EA takes the safety and welfare of its constituents more seriously. Respondents noted that having someone whose full-time job it is to help the community be healthy makes a serious impact on the community’s welfare.

Several respondents had personally worked with CEA on previous community health cases, and felt that they were handled responsibly, professionally, and with a commitment to confidentiality and other best practices.

Events

Respondents appreciated CEA’s events, particularly EA Global. They felt that having an overall conference for the community was helpful and lends EA credibility.

Respondents mentioned that EA Global:

  • has smoothly running logistics,
  • is lower cost than most conferences,
  • has a welcoming atmosphere,
  • and is thoughtful about the admissions process

EA Community Members are Smart, Talented, and Thoughtful

Community Members are High Achievers

Respondents felt that the EA community is made up of extremely smart and talented people. EA has a large early-career demographic, and respondents felt that those people generally had promising careers ahead of them.

Respondents described community members as generally analytical and thoughtful. Community members have internalized ideas around how important it is to be careful, to be open to changing one’s mind, be driven by one’s best understanding of the world, and be accurate. Some respondents mentioned that EA leaders have especially strong epistemic rigor.

Respondents also stated that, because the EA community is doing such important work, EA has a very compelling pitch for a certain subset of people, and has been successful in attracting them.

Community Members Genuinely Care

Several respondents felt that one of the most remarkable aspects of the EA community is that people were not only asking important questions but also genuinely trying to get the right answers and take actions to make the world better based on those answers. These questions revolve around a genuine desire to help others and make the world a better place. EA community members are willing to make profound commitments to improve the world.

EA Community Members are Kind

Several respondents remarked that EA community members are “nice” or “kind”. EA community members combine analytical understanding with compassion and motivation to be good.

Some respondents felt that the EA community is better at talking things over and disagreeing than other communities they are part of.

Overall, respondents felt that the EA community serves as a “magnet” for people who both care about the world and have an analytic and thoughtful approach.

Things Which Could Be Improved

Execution Issues

Grants

Execution problems in EA Grants were notable to several respondents. Our current Grants staff member has a post forthcoming with a more in-depth review, and I will update this post with a link once that is published. Edit: that post is now published here.

Funds

Respondents mentioned two broad concerns about EA Funds:

  1. Funds did not adequately communicate with community members and donors.
  2. Funds was targeted to meet the needs of a small set of donors, but was advertised to the entire EA community.

The first concern is related to one listed on our mistakes page:

Our shortcoming: During the first 21 months of the EA Funds, some funds made regular disbursements of donations but others disbursed more rarely than was expected. Community members requested more transparency about how much money was in each fund and when it would likely be disbursed. In April 2018 we stated in a comment on the EA Forum that we expected to publish a post with more information in the next few weeks, but we did not publish more information until August 2018 (after an additional post from a community member expressing concerns).

Steps taken: In October 2018, we announced new management teams for most of the funds, and a new three-times-yearly regular grantmaking schedule. We implemented a dashboard for each fund, where the public can see how much money is in each fund.

In addition to the new management teams, schedule and dashboard, we have implemented a system for proactively emailing donors who opt in when distributions are made, and several funds teams have posted regular grant reports on the Forum, including the Long Term Future Fund April recommendation, which became one of the most commented Forum posts of all time.

On the second concern: our initial proposal for what the Funds might grant to included things like “Providing a funding stream for more unusual, risky, or time-sensitive projects, particularly where the Open Philanthropy Project might have brand-risk concerns.” Many donors may not want their donations going towards “unusual, risky, or time-sensitive projects”, and respondents were concerned that the Funds were advertised to too broad a set of donors, including those for whom the Funds may not have been a good fit.

CEA has several processes which address this: (1) we've updated the Fund pages to more clearly explain why someone may choose not to donate to a fund (e.g. the Long-Term Future Fund), (2) we have published past grants, and (3) we don't currently undertake paid advertising, email marketing, or direct outreach to people outside our networks. Note that we do link to EA Funds from community-focused websites such as EffectiveAltruism.org and have not prioritized removing links to the Funds from those pages, which may have been a mistake on our part (our staff does not have a consistent view on whether those links should be there).[1]

EAGx

Respondents reported concerns that CEA promised support to EAGx organizers that was not delivered. For example, we took a long time to respond to questions from organizers, and did not provide them with content we had promised.

CEA has scaled back the number of EAGx’s from 11 in 2016 to 3 in 2019, and increased the number of staff working on events, to ensure that we can provide appropriate levels of support. Our mistakes page has more detail.

Groups

Respondents pointed out that most of our groups, including some of our strongest groups at key universities, have never received a visit from CEA staff and that the amount of support we provide to groups is limited. CEA has expanded our groups support, including making the newsletter monthly, and providing Community Building Grants. More information can be found in our 2019 update. More information about mistakes can be found on our mistakes page.

General

A general theme running through all of these examples is that CEA promised more than we delivered. In 2019, we have made a concerted effort to be careful with our commitments and only agree to things we are confident we can deliver. This is reflected in internal processes, such as a commitments project in Asana where we record and regularly track progress on any commitments we have made, as well as external humility in scaling back the number of programs and promises we make.

Communication and Branding Issues

EA Global

Respondents mentioned two primary concerns with past EA Global events:

  1. CEA attempting to “push” ideas like longtermism on attendees without transparent reasoning.
  2. Demographic diversity at EA Global is not representative of what respondents want.

Dylan Matthews’ article about the EA Global SF 2015 conference is one notable example of the first concern: because AI Safety was simply presented as a crucial problem area for EA without background information and reasoning about why AI safety is so crucial, attendees (including Matthews himself) were turned off. Respondents also mentioned that some posters and themes at previous EAG’s felt insincere or paternalistic. Respondents generally felt that recent EAG’s were significantly better than earlier ones.

Our EA Global advisory board was mixed on the question of whether to have public themes, with the majority saying they “neither agree nor disagree” with our having public themes. In 2019 EA Global did not have public “themes”, and our content was chosen in line with our view on representativeness. At EA Global SF 2019, for example: 9 talks were on AI, 7 on animal welfare, 7 on global health and development, 4 on horizon scanning and 9 on meta. We discussed the theme of “stewardship” with each speaker as a way to encourage them to think about welcomingness and humility in engaging with outside stakeholders.

Diversity is discussed below but respondents noted that, since EA Global is the “flagship” event of EA, demographic skews at the event are especially impactful.

EA Forum

Respondents felt that the quality of content on the Forum is mixed, and that the posts are not always representative of EA thinking. For example: there have been 12 posts on psychedelics on the forum in 2019, and only 4 on malaria, despite malaria being a much more mainstream cause within EA. Respondents were particularly concerned that newcomers may get an inaccurate picture of what the community values.

CEA has recently introduced the “community favorites” section to better highlight high-quality posts and is adapting the Effective Altruism Handbook for the Forum, to lead new readers through key ideas in EA. Our monthly Forum prize highlights what the moderators feel are the best posts and comments in that month. We also proactively reach out to high-quality contributors to encourage them to post content, and make it easy for people and organizations to crosspost to the EA forum.

EA Community

Welcomingness

Respondents reported that while EAs are generally kind, the EA community can be unwelcoming to people who are unfamiliar with EA. This seems particularly true of social events in the Berkeley area. Common concerns were people not making an effort to reach out to newcomers and get them involved in conversations, and people doing a “naïve consequentialist” evaluation of everyone they meet (an extreme version of this is something like: “I only want to be friends with people who are working on 80,000 Hours priority paths”).

There is a curious tension between respondents reporting the EA community as exceptionally nice while still reporting it as being sometimes unwelcoming. Respondents suggested a few possible explanations:

  1. Prioritization is an inherent part of EA, but it’s challenging to do in a way that is not offputting. Given this, EAs are welcoming, but that may still be unwelcoming to certain people.
  2. Similarly, EAs tend to use things like quantitative models in regular speech, in a way which could be perceived as “cold” to outsiders.
  3. Finally, respondents agreed that there was wide variation among EAs in welcomingness, and that this can explain people’s different experiences. Several respondents said something to the effect of “most of the people I would want to introduce a promising young person to are involved in EA, but many of the people involved in EA are not people I would want to introduce a promising young person to.” The EA Facebook group was frequently cited as an example source of offputting conversations, while people in leadership roles in EA organizations were cited as being unusually welcoming/friendly.

CEA hopes people will experience the EA community as exceptionally considerate. We are continually implementing changes to hold ourselves accountable to this internally. We think individuals and leaders can make a difference in their own communities and organizations in this regard, and hope that our efforts encourage others to do the same. As some examples:

  1. On the Forum, Aaron posts friendly and engaged comments on posts from first-time authors, both to improve those authors’ experience as well as publicly display the kind of norms we want to see.
  2. EA Global SF 2019 piloted two programs: the Guides program, which connects first-time attendees with more experienced attendees, and the Fellowship program, which connects people interested in a career path with others further along in that career path. 87% of attendees at EAG “agreed” or “strongly agreed” that “EA Global is a place where I felt welcome.” Both pilots were successful and will be scaled up for EA Global London.
  3. Some CEA staff send a personal message to a newcomer once per week thanking them for being involved.

Epistemic Humility

Respondents felt that EAs sometimes come across as overconfident or arrogant, and that the community can be too inward-looking. They felt EAs should be patient and open to people with different views who are seriously and thoughtfully trying to engage in analysis of hard problems. We should be mindful of our relative inexperience about many things, and grateful for busy people who take the time to try to contribute their ideas.

Many of the projects referred to in the “welcomingness” section are also intended to help with humility, as CEA wants our programs to reflect our own commitments to the social norms we want to see, including both welcomingness and humility.

EA Global brings in thought leaders from outside the EA community to speak on issues relevant to our work, such as Philip Tetlock and Bonnie Jenkins.

Diversity

Respondents reported that the relative demographic homogeneity of the EA community is a concern. Respondents were aware that the EA community appears to be more young, white, irreligious, male, and from Europe and North America than the world as a whole, and generally they agreed with CEA’s stance on diversity and inclusion. They noted that underrepresentation from certain groups can contribute to EAs from some groups not feeling as welcome as other community members, and reduces the chances of achieving our shared goal for a better future.

Respondents also mentioned a desire for more diversity of intellectual backgrounds, skills, experience, and interests. Some respondents were concerned that EA selects too heavily for certain types of people (e.g. those interested in quantitative, totalizing moral frameworks) at the expense of selecting out people with other relevant skills and experience. Some respondents felt homogeneity in these areas is related to or causes demographic homogeneity (e.g. people with more experience tend to be older).

As mentioned above, this is a concern CEA shares. We do not yet know all the ways that CEA might contribute to addressing this concern, but we are very interested in exploring this question and continuing to support diversity and inclusion efforts in our organization and programs.

We think doing good work on diversity and inclusion should be about a lot more than mere numbers of people; it should focus on the experiences of individuals and groups. That said, we also think it can be helpful to have common knowledge about current data. As of this writing, CEA staff is 47% female. 49% of speakers at EA Global London 2019 were female (compared to 29% of respondents to the 2018 EA Survey who identified as female). 26% of speakers at EA Global London 2019 were people of color (compared to 22% of EA Survey 2018 respondents who identified as nonwhite). For EA Global London (Oct 2019), we have added demographic questions to the registration form and post-event survey, in the hopes that we can better compare attendee data to EA Survey data in the future.

Actually Doing EA

Respondents pointed out that EA is hard: digging through sources and creating detailed models is laborious, time-consuming work. They were concerned that too few EAs are doing this work; for example, one person reported that ALLFED’s cost-effectiveness model was the only cost-effectiveness model they could recall having been posted on the Forum in the past year, and it received relatively few comments.

The past several months have seen a significant uptick in the amount of detailed analysis on the Forum. Luisa Rodriguez’s work on nuclear war, Leopold Aschenbrenner’s work on existential risk and economic growth, and Saulius Šimčikas’ work on animal welfare commitments are a few examples.

Conclusion

The EA community has accomplished a lot, but we have a long way to go. CEA greatly appreciates respondents taking the time to provide feedback, and you taking the time to read this.

We have two requests:

  1. First, if you have any feedback for CEA as a whole, please feel free to email me (ben.west@centreforeffectivealtruism.org) or fill out this form. We are also happy to connect you to specific project owners.
  2. Second, we encourage you to be exemplars of the welcoming, humble, and intellectually rigorous community our respondents felt was important for success, and to keep encouraging us at CEA to do the same.

Credits

(Note: Unlike most of my posts on the forum, this is written in my official capacity at CEA.) Almost everyone at CEA contributed to writing this document, but I would like to especially thank Sky Mayhew, Julia Wise, and Amy Labenz for substantive edits. I would also like to thank the people who provided feedback for their careful thoughts and being generous with their time.


  1. A previous version of this article said that CEA "does not currently proactively advertise EA funds"; it has been updated to clarify that we do link to it from some websites, but do not do other sorts of advertisement. ↩︎

106