Centre for Effective Altruism (CEA): an overview of 2017 and our 2018 plans

This post is cross-posted from the Cen­tre for Effec­tive Altru­ism (CEA) blog. By post­ing to the EA Fo­rum we hope to in­crease visi­bil­ity on our work and give ev­ery­one the abil­ity to com­ment and ask ques­tions.

This has been a big year for the Cen­tre for Effec­tive Altru­ism (CEA). We launched EA Grants, which funded 21 pro­jects in the EA com­mu­nity; we cre­ated a new dona­tion plat­form called EA Funds to help peo­ple donate more effec­tively, and we ran three EA Global con­fer­ences to bring the com­mu­nity to­gether.

In this post, we share what CEA has been do­ing this year, and give you a taste of the things we will be work­ing on in 2018.

This post in­cludes:

  1. CEA’s mis­sion and vision

  2. High­lights from 2017

  3. A brief re­view of 2017 and plans for 2018 by team

  4. A non-ex­haus­tive list of our mis­takes and plans for im­prove­ment this year

  5. In­for­ma­tion on our cur­rent fund­ing situation

  6. An in­vi­ta­tion to join our sup­porter mailing list for monthly updates

For gen­eral in­quiries, please con­tact Kerry Vaughan, who is in charge of our In­di­vi­d­ual Outreach Team. If you would like to dis­cuss any parts of our plans in more depth, please reach out to the rele­vant team lead (first name [at] cen­tre­fore­ffec­tivealtru­ism.org)

CEA’s Mis­sion and Vision

CEA aims to solve the most im­por­tant prob­lems, re­gard­less of which species, lo­ca­tion, or gen­er­a­tion they af­fect. By do­ing this, we build to­wards our vi­sion of an op­ti­mal world.

Due to the scale of the po­ten­tial im­pact, our cur­rent best guess is that work to im­prove the long-term fu­ture is likely to be the best way to help oth­ers. How­ever, we think that there is a good chance that we are wrong. For this rea­son, we also want to con­tinue to de­vote re­sources to­wards find­ing bet­ter ways to ad­dress the world’s biggest prob­lems; as a part of this, we want to learn more about prob­lems we may not yet be pay­ing enough at­ten­tion to.

Im­prov­ing the world’s long-term tra­jec­tory will be very difficult. We be­lieve that long-scale change can­not be solved by in­di­vi­d­u­als. In­stead, it re­quires a com­mu­nity work­ing to­gether. We cur­rently think that such a com­mu­nity par­tic­u­larly needs peo­ple who are very en­gaged with these ideas, and who are able to do full-time re­search or policy work in the rele­vant ar­eas. As such, we fo­cus on at­tract­ing and sup­port­ing these highly-en­gaged and skil­led peo­ple.

This ex­plains our mis­sion:

Create a global com­mu­nity of peo­ple who have made helping oth­ers a core part of their lives, and who use ev­i­dence and sci­en­tific rea­son­ing to figure out how to do so as effec­tively as pos­si­ble.

CEA High­lights from 2017

The year started CEA go­ing through Y Com­bi­na­tor’s. startup ac­cel­er­a­tor pro­gram. Y Com­bi­na­tor is a startup in­cu­ba­tor that pro­vides seed fund­ing and ad­vice to star­tups. We were one of the few non-prof­its to be ac­cepted to their three-month pro­gram, which gave us ac­cess to one-on-one ad­vice from their founders. It was dur­ing this time that we built EA Funds, a plat­form that al­lows users to pool their money with like-minded donors so that fund man­agers can di­rect the money to the best giv­ing op­por­tu­ni­ties in differ­ent cause ar­eas. We of­ten talk about the EA com­mu­nity need­ing money, tal­ent, and ideas in or­der to suc­ceed. A time when we had ac­cess to some of the most suc­cess­ful en­trepreneurs seemed like the best time to build a product fo­cused on money.

While EA Funds was per­haps our high­est pro­file pro­ject, our EA Grants pro­gram in the sum­mer at­tracted over 700 ap­pli­cants. We wanted to find ways to sup­port the EA com­mu­nity in in­no­va­tive new pro­jects, and af­ter care­ful eval­u­a­tion, we de­cided to fund 21 pro­jects.

In Oc­to­ber, we set up CEA’s in­di­vi­d­ual out­reach pro­gram, which aims to help peo­ple get deeply in­volved with the effec­tive al­tru­ism com­mu­nity more quickly, through one-on-one men­tor­ing.

Most re­cently, we launched the Giv­ing What We Can pledge cam­paign. Our fo­cus has been get­ting cur­rent mem­bers to re­view where they donate and to en­courage peo­ple to think se­ri­ously about the ca­reer-long pledge.

In­ter­nally, we have con­soli­dated and built ca­pac­ity. Tara MacAu­lay, pre­vi­ously our COO, moved into the CEO role, which bet­ter re­flected the work she had been do­ing for some time. Will MacAskill moved from his CEO role to be­come pres­i­dent, and he will now fo­cus on aca­demic and pub­lic en­gage­ment roles. Around the same time, we con­soli­dated into five teams, with five team lead­ers:

1. Com­mu­nity (Larissa Hes­keth-Rowe),

2. Oper­a­tions (Miranda Dixon-Luinen­berg),

3.Tech (Sam Deere),

4. Re­search (Max Dal­ton),

5. In­di­vi­d­ual Outreach (Kerry Vaughan).

Over the course of the year, this man­age­ment ca­pac­ity helped us to grow the team, from 17 at the be­gin­ning of the year to 21 at the end of the year.

Below we give more de­tails of what each team has done in 2017 and their plans for 2018.

Re­search Team

The Re­search Team aims to com­mu­ni­cate new and im­por­tant ideas to the effec­tive al­tru­ism com­mu­nity. Or­ga­ni­za­tions, aca­demics, and in­de­pen­dent re­searchers within the effec­tive al­tru­ism com­mu­nity pro­duce valuable re­search, but such re­search can be difficult to find or ap­ply; we want to make that pro­cess eas­ier, so peo­ple can use and ex­plore the ideas we have.

There were sev­eral changes to re­search at CEA dur­ing 2017.

First, we dis­con­tinued the Philan­thropic Ad­vis­ing Team in Fe­bru­ary 2017. This pro­ject was ex­per­i­men­tal, and while the team had some suc­cess in pro­vid­ing ad­vice to philan­thropists, the re­turns were not com­pet­i­tive with other pro­jects, so we de­cided to end the pro­ject in or­der to be more fo­cused. Our policy work moved to a more nat­u­ral home at the Fu­ture of Hu­man­ity In­sti­tute (FHI), again to al­low us to fo­cus on other pro­jects.

As planned, the Global Pri­ori­ties In­sti­tute, a pro­ject that we in­cu­bated in 2016 and the first half of 2017, be­came a part of the Univer­sity of Oxford.

Fi­nally, the former Fun­da­men­tals Re­search Team, which had op­er­ated sep­a­rately from other parts of CEA, fully merged with the rest of the or­ga­ni­za­tion in May. This al­lows us to co­or­di­nate more eas­ily, which is es­pe­cially im­por­tant given that the team’s fo­cus has shifted to­ward com­mu­ni­cat­ing some of the key ideas in the com­mu­nity rather than con­duct­ing our own re­search. The team is now Max Dal­ton and Ste­fan Schu­bert, and it is ad­vised by Owen Cot­ton-Bar­ratt from FHI.

Re­search Team Ac­tivity in 2017

For the first half of the year, the Fun­da­men­tals Re­search Team was fo­cused on pro­duc­ing new cause pri­ori­ti­za­tion re­search. This in­cluded work to clar­ify im­por­tant con­cepts like diminish­ing re­turns and cause neu­tral­ity. It also in­cluded dis­cus­sion of com­mu­nity norms and why we should be es­pe­cially wary of hard-to-re­verse de­ci­sions.

In the sec­ond half of the year, we fo­cused more on com­mu­ni­cat­ing ex­ist­ing re­search and ideas in effec­tive al­tru­ism. We no­ticed that there were many ideas that were un­pub­lished or scat­tered across a va­ri­ety of per­sonal blogs. We wanted to make these ideas more ac­cessible to peo­ple who want to more deeply en­gage with effec­tive al­tru­ism. We pub­lished a se­ries of cause pro­files on the long-term fu­ture, an­i­mal welfare, global health, and effec­tive al­tru­ism com­mu­nity build­ing. We also rewrote the_In­tro­duc­tion to Effec­tive Altru­ism_, and we tran­scribed and col­lated some of the best re­cent re­search in effec­tive al­tru­ism on our re­sources page. The re­sources page is in­tended to be a beta ver­sion of our 2018 pro­ject, which is a se­ries of re­search ar­ti­cles cov­er­ing some of the key ideas in effec­tive al­tru­ism. We hope this se­ries will quickly in­tro­duce key con­cepts to those who are new to effec­tive al­tru­ism.

In ad­di­tion, we hosted sev­eral re­search fel­lows over the sum­mer, sup­port­ing them to pro­duce origi­nal re­search and train­ing them in rele­vant skills. Most of this re­search was posted on the EAfo­rum (e.g., Danae Ar­royos-Calvera on DALYs, Alex Barry and Denise Melchin on the Causal Net­works Model, and Emily Tench on Com­mu­nity Norms).

Im­pact Review

It is gen­er­ally difficult to eval­u­ate the im­pact of re­search, since many of the effects are in­di­rect.

Over­all, while our more ex­plo­ra­tory work ear­lier in the year seemed some­what use­ful, we found that we were not able to re­li­ably gen­er­ate im­por­tant new con­sid­er­a­tions. This was why are now fo­cus­ing more on com­mu­ni­cat­ing ex­ist­ing ideas that are not widely shared. We are par­tic­u­larly ex­cited about changes we made to the con­tent of www.effec­tivealtru­ism.org. We think it now offers a good sum­mary of cur­rent think­ing in effec­tive al­tru­ism; the web­site should be clearer for peo­ple new to the com­mu­nity and a good refer­ence for more es­tab­lished com­mu­nity mem­bers.

We also think that run­ning the re­search fel­low­ship was worth­while. Not only did the sum­mer fel­lows pro­duce use­ful re­search, but they de­vel­oped skills and un­der­stand­ing that will al­low them to have a greater im­pact in the fu­ture.

Plans for 2018

In 2018, we hope to make re­search in effec­tive al­tru­ism even more ac­cessible.

Our key pro­ject will be a se­ries of ar­ti­cles, to be pub­lished on Effec­tiveAltru­ism.org, which will give an overview of cur­rent think­ing in effec­tive al­tru­ism and our ap­proach to cause pri­ori­ti­za­tion. Although there are many good ideas in the com­mu­nity, many are un­pub­lished or pub­lished in an ob­scure place. We be­lieve that we can provide a lot of value by mak­ing these ideas more ac­cessible, as well as by cre­at­ing a com­mon refer­ence work for the com­mu­nity.

We will also con­tinue to im­prove other as­pects of the con­tent and de­sign of Effec­tiveAltru­ism.org. We will fur­ther sup­port the in­tel­lec­tual com­mu­nity work­ing on effec­tive al­tru­ism. This will in­clude work­ing with the tech team on the de­sign of a new Effec­tive Altru­ism Fo­rum, and sup­port­ing more com­mu­ni­ca­tion and col­lab­o­ra­tion be­tween pro­fes­sional EA re­searchers.

In­di­vi­d­ual Outreach Team

The goal of the In­di­vi­d­ual Outreach Team is (1) to iden­tify peo­ple within the effec­tive al­tru­ism com­mu­nity that we ex­pect will make big con­tri­bu­tions to im­por­tant pro­jects and (2) to help them to have a greater im­pact.

This is a new team, and we are ex­per­i­ment­ing with this con­cept be­cause of two con­sid­er­a­tions:

  1. The heavy-tailed dis­tri­bu­tion the­sis: It seems plau­si­ble that the dis­tri­bu­tion of im­pact is “heavy-tailed” in that a small num­ber of peo­ple might provide a sig­nifi­cant amount of the value that the com­mu­nity cre­ates.

  2. Self-sort­ing: Peo­ple tend to in­ter­act with oth­ers who they per­ceive are similar to them­selves.

If both of these claims are true, then the way in which we have fo­cused our com­mu­nity build­ing efforts may be miss­ing some peo­ple. At pre­sent, en­gage­ment with EA usu­ally in­volves get­ting new peo­ple in­volved with the EA com­mu­nity through lo­cal groups, events like EA Global, and on­line dis­cus­sion fo­rums. How­ever, it seems plau­si­ble that the next Peter Singer or Nick Bostrom will be seek­ing a very spe­cific peer group, and thus may not get in­volved with the com­mu­nity via lo­cal groups or effec­tive al­tru­ism con­fer­ences alone.

The in­di­vi­d­ual out­reach team be­lieves that it is im­por­tant to iden­tify and con­nect the peo­ple with the largest po­ten­tial for im­pact, even if these peo­ple are not in­ter­act­ing with the EA com­mu­nity through our stan­dard fo­rums. That is why we’re putting more em­pha­sis on de­vel­op­ing our abil­ity to make in­di­vi­d­ual con­nec­tions be­tween peo­ple who we can help and who may be able to move the nee­dle on im­por­tant prob­lems.

In­di­vi­d­ual Outreach Team Ac­tivity in 2017

The In­di­vi­d­ual Outreach Team was formed in Septem­ber 2017, so the team is rel­a­tively new. How­ever, here are some of the ac­tivi­ties that we have en­gaged in so far.

Meet­ings at EA Global: London

We re­viewed ap­pli­ca­tions to EA Global and or­ga­nized short, one-to-one meet­ings with around one-third of all EA Global at­ten­dees dur­ing the con­fer­ence.

Dur­ing these meet­ings, we looked for ways that we could de­liver value to at­ten­dees by helping them im­prove their plans, get a bet­ter sense of the land­scape of EA pro­jects, and meet oth­ers work­ing on similar things.

Post-EAG: Lon­don retreat

We held a small re­treat for around 30 highly-en­gaged com­mu­nity mem­bers af­ter EA Global Lon­don. We aimed to help at­ten­dees think through their ca­reers and to dis­cuss with one an­other.

We thought the re­treat would be suc­cess­ful if one of the at­ten­dees made a ma­jor up­date to their plans. In fact, we saw four ma­jor plan changes, as well as a num­ber of less sig­nifi­cant plan changes.

EA Grants

The In­di­vi­d­ual Outreach Team has taken on EA Grants, which was pre­vi­ously es­tab­lished by mem­bers of the Re­search Team. EA Grants aims to provide fund­ing for high-im­pact pro­jects, es­pe­cially those that may not be funded by or­ga­ni­za­tions like the Open Philan­thropy Pro­ject (Open Phil). We al­lo­cated £370,000 to 21 pro­jects in our first round of Grants. More de­tails can be found here.

Im­pact Review

Our main im­pact this quar­ter came from the post-EAG re­treat. Im­me­di­ately af­ter the re­treat, four at­ten­dees made sig­nifi­cant changes to their plans. (Three peo­ple moved from earn­ing to give to di­rect work, and one changed their re­search tra­jec­tory.) In ad­di­tion, a num­ber of other at­ten­dees en­gaged in more mod­est plan im­prove­ments.

It is worth not­ing that not all of these changes were en­tirely due to the re­treat. Some might have hap­pened any­way, but hap­pened sooner due to the re­treat, and some changes might turn out to be less valuable than we an­ti­ci­pate. Nev­er­the­less, the level of changes in­di­vi­d­u­als made vastly ex­ceeded our ex­pec­ta­tions.

Since this is a new pro­ject, one of our key goals is to learn more. Meet­ing with hun­dreds of peo­ple at EA Global helped us to get rapid feed­back on how we could be more helpful. The re­sults of the Novem­ber re­treat in­di­cated that re­treats may be a use­ful mechanism for al­low­ing peo­ple to de­velop their plans quickly.

It is difficult to as­sess the full im­pact of the grants we dis­tributed at this stage. How­ever, there ap­pear to be some early wins:

  • The new LessWrong web­site is be­ing well used (see strat­egy doc­u­ment here).

  • David Denken­berger, one of our grant re­cip­i­ents, has been pub­lish­ing work we funded to the EA Fo­rum, and he has also set up a new or­ga­ni­za­tion work­ing on re­lated is­sues.

  • Some of the peo­ple we funded for ma­chine learn­ing re­search are pro­duc­ing pub­li­ca­tions, and gained op­por­tu­ni­ties, based on the work that we funded.

Plans for 2018

Grants: We are plan­ning to run EA Grants through­out 2018, with an an­ti­ci­pated bud­get of around £2m. There are some changes from the last round.

First, we plan to ac­cept ap­pli­ca­tions year-round with quick re­views and re­sponses for ur­gent ap­pli­ca­tions and quar­terly re­views for less ur­gent ap­pli­ca­tions.

Se­cond, we plan to move the eval­u­a­tion pro­cesses even fur­ther in the di­rec­tion of mostly eval­u­at­ing the mer­its of the ap­pli­cants them­selves rather than their spe­cific plans. This is be­cause:

  1. We ex­pect most plans to be rel­a­tively spec­u­la­tive and there­fore sub­ject to change;

  2. We are time and re­source-con­strained in how con­tin­u­ously we can mon­i­tor pro­jects, so we need to make sure we have high con­fi­dence in grantees; and

  3. We do not think we can de­velop ex­per­tise in all pos­si­ble pro­jects, but we can de­velop ex­per­tise in eval­u­at­ing the ap­pli­cants.

Fi­nally, we plan to move fur­ther in the di­rec­tion of a hits-based giv­ing ap­proach, us­ing EAGrants to place bets on risky, un­usual, or con­tro­ver­sial pro­jects that seem plau­si­bly very valuable in ex­pec­ta­tion.

Re­treats: We plan to work with the Com­mu­nity Team to run more re­treats through­out the year, with a tar­get of run­ning ap­prox­i­mately one per quar­ter. We also plan to ex­per­i­ment more with differ­ent for­mats and ac­tivi­ties dur­ing these events.

Men­tor­ing: We plan to men­tor promis­ing com­mu­nity mem­bers on a weekly or bi-weekly ba­sis as a way of gain­ing more in-depth feed­back on how we can help peo­ple ac­com­plish their goals more quickly.

Com­mu­nity team

CEA be­lieves that the EA com­mu­nity could be an im­por­tant way to in­fluence the world’s long-term tra­jec­tory for the bet­ter. We be­lieve that a tightly co­or­di­nated group of peo­ple, work­ing to­gether, can have much more of an im­pact than each in­di­vi­d­ual work­ing alone. This means that — be­yond the money, tal­ent, and ideas that we of­ten dis­cuss as be­ing nec­es­sary for suc­cess — we also need to be able to co­or­di­nate as a com­mu­nity.

The Com­mu­nity Team at CEA works to en­courage that co­or­di­na­tion. We fa­cil­i­tate some of the spaces where the com­mu­nity comes to­gether, both on­line and in-per­son (some­times in part­ner­ship with other or­ga­ni­za­tions in the com­mu­nity). We also work to im­prove co­op­er­a­tion by helping to shape com­mu­nity norms and cul­ture. We ad­vise lo­cal groups, hold events, sup­port the Giv­ing What We Can com­mu­nity, mod­er­ate on­line dis­cus­sions and me­di­ate dis­putes.

As such, we have a lot of cross-over with the other teams at CEA. Many of the peo­ple that our In­di­vi­d­ual Outreach Team looks to men­tor come from lo­cal groups, and our events are a great way for in­di­vi­d­u­als to get one-on-one ad­vice.

Although the Com­mu­nity Team was cre­ated in the sum­mer of 2017, many of our pro­jects have been run­ning in some form for much longer.

Below is a sum­mary of the Com­mu­nity Team’s ac­com­plish­ments this year and plans for the fu­ture.


Events in 2017

This year, we ran three EA Global con­fer­ences, with over 1,600 at­ten­dees in to­tal. Our con­fer­ence in Bos­ton fo­cused on the fron­tiers of re­search in EA, our San Fran­cisco event was fo­cused on the EA com­mu­nity, and in Lon­don, we ex­per­i­mented with fo­cus­ing more on ex­ist­ing com­mu­nity mem­bers and less on in­tro­duc­tory con­tent. The unify­ing theme of the three con­fer­ences was “do­ing good to­gether.”

Alongside these three con­fer­ences, we also ran two smaller ex­ter­nal events (one with the In­di­vi­d­ual Outreach Team and one with lead­ers of EA or­ga­ni­za­tions) and three in­ter­nal offsite events for CEA staff. We also sup­ported lo­cal com­mu­ni­ties in run­ning four EAGx events: EAGxMadi­son, EAGxPhilly, EAGxAus­tralia and EAGxBer­lin. In to­tal, there were ap­prox­i­mately 500 at­ten­dees.

You can find de­tails of many of the larger ex­ter­nal events that we have run on our web­site, and videos of many of the talks are available on our YouTube chan­nel.

Im­pact Review

Our EA Global con­fer­ences fo­cused on ex­ist­ing mem­bers of the EA com­mu­nity and helping them im­prove their plans, their com­mit­ment, and their un­der­stand­ing of EA. In our sur­veys, we, there­fore, asked about the goals of at­ten­dees, their en­gage­ment, what they learned, and whether their plans changed. Un­for­tu­nately, in efforts to gather bet­ter data, we changed our sur­vey ques­tions be­tween events, which made com­par­i­sons harder. Of re­spon­dents from our Bos­ton event sur­vey, 92% learned some­thing new, and 14% of re­spon­dents said our San Fran­cisco con­fer­ence would lead them to make sig­nifi­cant plan changes. From Lon­don, 28% of sur­vey re­spon­dents ex­pect to make ma­jor plan changes, in­clud­ing chang­ing di­rec­tion within a field (25%) or com­pletely chang­ing cause ar­eas (3.1%).

Beyond the benefit to in­di­vi­d­u­als, there are also com­mu­nity-wide benefits to in­creas­ing co­op­er­a­tion. The longer-term effects are hard to pin­point, but we were pleased with how well this year’s theme of do­ing good to­gether seemed to go at the con­fer­ences.

We are cur­rently con­duct­ing a more in-depth anal­y­sis of our EA Global Lon­don data, in­clud­ing con­duct­ing in­ter­views with some at­ten­dees who ei­ther had large plan changes or no plan changes from at­tend­ing.

As men­tioned in the In­di­vi­d­ual Outreach Team sec­tion above, our smaller events seem to have helped in­di­vi­d­u­als make sig­nifi­cant plan changes and helped or­ga­ni­za­tions co­or­di­nate.

Plans for 2018

We plan to ex­per­i­ment with fur­ther small events this year. Th­ese events will help with the In­di­vi­d­ual Outreach Team’s work and provide us with faster feed­back loops for learn­ing about the best event for­mats than our larger events. Th­ese events in­clude a small AI strat­egy re­treat in Jan­uary and a lo­cal group lead­ers re­treat.

Given the higher num­bers of ap­pli­cants and at­ten­dees for this year’s San Fran­cisco and Lon­don EA Global events than for the Bos­ton event, we are con­sid­er­ing hav­ing two 2018 EA Global con­fer­ences: one in Lon­don and one in the San Fran­cisco Bay Area in the US.

Giv­ing What We Can

Giv­ing What We Can Ac­tivi­ties in 2017

Giv­ing What We Can is a com­mu­nity of peo­ple who have pledged to donate 10% of their in­come over the course of their ca­reers to the most im­pact­ful or­ga­ni­za­tions that they can find. This year was our first full year of run­ning Giv­ing What We Can as a pro­ject within CEA, rather than it run­ning as a sep­a­rate or­ga­ni­za­tion. This tran­si­tion saw the con­soli­da­tion of the Giv­ing What We Can Trust into CEA and a new pres­i­dent of the pro­ject. In June, Ju­lia Wise (a mem­ber of the Com­mu­nity Team at CEA) took the role of pres­i­dent of Giv­ing What We Can.

This year, we par­tic­u­larly fo­cused on em­pha­siz­ing the se­ri­ous­ness of the com­mit­ment when tak­ing the Pledge, en­courag­ing more peo­ple to use Try Giv­ing, as a way to make short-term com­mit­ments to giv­ing be­fore you take the life­time pledge. We have also been work­ing to im­prove the effec­tive­ness of dona­tions through the cre­ation of EA Funds, a plat­form to make donat­ing to effec­tive causes eas­ier. This is now the home of the Giv­ing What We Can pledge form and will soon house My Giv­ing, our dona­tion-track­ing plat­form. All of this will make it eas­ier for mem­bers to donate and to record whether or not they have fulfilled their pledge. Many mem­bers are already donat­ing through the EA Funds and their dona­tion his­tory will au­to­mat­i­cally be added to their My Giv­ing record. We ex­pect the new com­bined sys­tem to provide more com­plete and re­li­able in­for­ma­tion about the com­mu­nity’s pat­tern of dona­tions and pledge fol­low-through than we have had in the past.

We also fo­cused on im­prov­ing the com­mu­nity’s un­der­stand­ing of the pledge, with a fo­rum post clar­ify­ing com­mon mis­con­cep­tions and a talk on the pledge at EA Global Bos­ton.

We cel­e­brated reach­ing 2,500 mem­bers with a re­cep­tion in San Fran­cisco and 3,000 mem­bers with a re­cep­tion in Lon­don.

Im­pact Review

At the time of writ­ing, we have had 848 new mem­bers join in 2017 (a 35% in­crease on the 2430 mem­bers at the be­gin­ning of the year). Dur­ing the same pe­riod last year, we had 850 new mem­bers, which was a 58% in­crease on the ini­tial 1460 that year. This slow-down in the rate of growth re­flects our change from em­pha­siz­ing re­cruit­ment of new mem­bers to em­pha­siz­ing the Pledge as a se­ri­ous life­time com­mit­ment to be thor­oughly con­sid­ered.

Plans for 2018

Our main pri­ori­ties for 2018 in­clude get­ting the new plat­form be­hind Giv­ing What We Can and EA Funds run­ning so that we can get bet­ter data on mem­ber dona­tions. We will also up­date the Giv­ing What We Can web­site to more ac­cu­rately re­flect the range of cause ar­eas our mem­bers care about and to re­flect more cur­rent in­for­ma­tion about past and pro­ject giv­ing by mem­bers.

Lo­cal Groups

Lo­cal Groups Ac­tivity in 2017

This year has seen a shift in fo­cus for the lo­cal group sup­port pro­vided by CEA. Par­tic­u­larly this aca­demic year, we have been ded­i­cat­ing more time to giv­ing more in-depth sup­port to the most es­tab­lished groups rather than more ba­sic, blan­ket sup­port to all groups. In part, we were able to do this be­cause of the lo­cal group sup­port offered by Re­think Char­ity and EAF.

We re­al­ize that the EA com­mu­nity has grown a lot, but his­tor­i­cally, we have put more en­ergy into that growth than into sup­port­ing the peo­ple we already have in the com­mu­nity to deepen their en­gage­ment. Our fo­cus on a smaller num­ber of groups is in part to rec­tify that.

It seems that the most en­gaged EA Groups may provide many times more value than an av­er­age group, so we have sharp­ened our fo­cus on the most en­gaged groups for that rea­son.

We have pro­vided in­tern­ships at the CEA Oxford office for group lead­ers with a par­tic­u­lar fo­cus on one-to-one sup­port as they de­velop pro­jects for their groups. With sup­port from vol­un­teers, we provide on­go­ing video call sup­port for 35 groups as part of a men­tor­ing pro­gram run in con­junc­tion with EA Cam­bridge. We referred 25 peo­ple to 80,000 Hours coach­ing this aca­demic year and given $40,000 in fund­ing to EA Groups. If you run a lo­cal group, we strongly en­courage you to ap­ply for fund­ing via the EA Groups page.

Im­pact review

We have not yet con­ducted a sys­tem­atic re­view of EA Groups sup­port for 2017. This is partly due to our chang­ing pri­ori­ties in 2017 mean­ing chang­ing ways of mea­sur­ing suc­cess.

At the start of 2017, GWWC pledges were the core met­ric for EA groups, but as men­tioned above, we changed our ap­proach, em­pha­siz­ing the Pledge as a se­ri­ous life­time com­mit­ment. It there­fore no longer felt ap­pro­pri­ate for this to be the main met­ric for lo­cal groups.

We spent the sum­mer plan­ning new pro­jects, and this aca­demic year, 80,000 Hours coach­ing refer­rals is now be­ing used by EA groups sup­port as the main met­ric. The num­bers on this went well, but fur­ther ac­tivi­ties to gen­er­ate coach­ing refer­rals were post­poned, as 80,000 Hours had less coach­ing ca­pac­ity dur­ing their re­view pe­riod. Of our 25 refer­rals, 15 peo­ple ei­ther have already or will re­ceive coach­ing. Of those, one is now spend­ing a month work­ing at FHI, one is do­ing con­tract work for CEA, and an­other will be at­tend­ing a fol­low-up event. It is too early to tell the long-term im­pli­ca­tions of this, but we have helped some lo­cal group mem­bers gain new ex­pe­rience and im­prove their plans.

We hope to gain ad­di­tional in­for­ma­tion from the Re­think Char­ity LEAN im­pact eval­u­a­tion. Some of their ini­tial find­ings are sum­ma­rized here. This re­port in­cludes in­for­ma­tion from an EA­groups sur­vey that Re­think Char­ity, EAF and CEA col­lab­o­rated on.

Plans for 2018

We plan to con­tinue our fo­cus on the larger EA groups by run­ning a re­treat for group lead­ers and a se­ries of in­tern­ships. The aim of these will be to bring group lead­ers up to speed on our think­ing and to give groups time to trial pro­ject ideas with ded­i­cated sup­port from us. We are cur­rently look­ing into pro­vid­ing fund­ing to sup­port some groups to pro­fes­sion­al­ize with full-time, paid lo­cal group or­ga­niz­ers.

We are close to pub­lish­ing an EA com­mu­nity build­ing guide with our most up-to-date think­ing on how lo­cal groups can best help their mem­bers have an im­pact.

Com­mu­nity Health

Com­mu­nity Health Ac­tivi­ties in 2017

Helping the EA com­mu­nity thrive is a key part of the Com­mu­nity Team’s work. We try to im­prove on­line dis­course, provide re­sources for han­dling com­mon is­sues in lo­cal groups, and re­duce risks to the EA com­mu­nity. This in­cludes a num­ber of ac­tivi­ties such as:

  • Hav­ing Ju­lia Wise serve as a point per­son to col­late in­for­ma­tion from around the com­mu­nity about prob­lems that arise, such as peo­ple act­ing badly to­ward oth­ers in the com­mu­nity. A point per­son who ad­dresses such prob­lems re­duces the risk of sev­eral com­mu­nity mem­bers in­de­pen­dently ex­pe­rienc­ing a prob­lem but not think­ing that their in­di­vi­d­ual ex­pe­rience is worth act­ing on, or not be­ing in a good po­si­tion to act on it.

  • En­courag­ing more ac­tive mod­er­a­tion of EA Face­book groups by their re­spec­tive mod­er­a­tors to re­duce di­vi­sive “flame­war” style dis­cus­sions and to steer to­ward civil, pro­duc­tive dis­course.

  • Manag­ing the EA Fo­rum, which CEA took over re­spon­si­bil­ity for this year.

  • Pro­vid­ing re­sources to lo­cal EA groups, such as a train­ing on han­dling protests at speaker events and a guide on host­ing jour­nal­ists at lo­cal events.

We have also been work­ing on proac­tive ap­proaches to com­mu­nity health, such as cre­at­ing the EA Guid­ing Prin­ci­ples, to which many EA or­ga­ni­za­tions have added their sup­port, in or­der to help the effec­tive al­tru­ism com­mu­nity stay true to its best el­e­ments. We’ve tried to shape com­mu­nity norms through con­tent at EA Global about self-care, di­ver­sity, and mak­ing lo­cal groups more wel­com­ing. In De­cem­ber, we have been work­ing with the Re­search Team to do a re­view of the po­ten­tial risks to the EA com­mu­nity and ways to miti­gate them.

Im­pact review

Suc­cess in this work gen­er­ally looks like mem­bers of the EA com­mu­nity not notic­ing prob­lems that have been averted, so the im­pact is hard to see. How­ever, we think that, given the value of the EA com­mu­nity, re­duc­ing risks to the com­mu­nity is im­por­tant. We have not pre­vi­ously con­ducted a sys­tem­atic re­view, so our re­cent work with the Re­search Team in­volves iden­ti­fy­ing ar­eas we can track to un­der­stand our progress in the fu­ture.

Plans for 2018

Our plans for 2018 are to con­tinue re­spond­ing to prob­lems that arise in the com­mu­nity, while ex­plor­ing ways we can be more proac­tive in pre­vent­ing these prob­lems. This par­tic­u­larly in­cludes our work with the Re­search Team to iden­tify the biggest risks to the EA Com­mu­nity and pos­si­ble steps to re­duce such risks.

If you would like to dis­cuss any of the Com­mu­nity Team’s work, please con­tact Larissa Hes­keth-Rowe (larissa@cen­tre­fore­ffec­tivealtru­ism.org), as feed­back is always wel­come.

Oper­a­tions Team

The Oper­a­tions Team at CEA sup­ports the effec­tive­ness of other teams.

Like the other teams at CEA, the op­er­a­tions team as a group of peo­ple with their own man­ager, met­rics, and team mem­bers, is rel­a­tively new as our op­er­a­tions were pre­vi­ously man­aged part-time by CEA staff with other roles. All four mem­bers of the cur­rent Oper­a­tions Team joined CEA in 2017. Hav­ing a full team has al­lowed other staff to fully con­cen­trate full-time on CEA’s other pro­jects (such as event man­age­ment) rather than ad­di­tion­ally hav­ing to do op­er­a­tions work.

The Oper­a­tions Team man­ages all of CEA’s fi­nan­cial and le­gal needs and lends in­valuable sup­port to some of CEA’s largest pro­jects, such as lo­gis­tics at EA Global con­fer­ences, set­ting the fi­nan­cial and le­gal frame­work for EA Grants and man­ag­ing the hiring and re­ten­tion of CEA’s staff.

Oper­a­tions Team Ac­tivi­ties in 2017

This year we:

  • Hired and on-boarded 9 new staff.

  • Set up an en­tirely new office in Berkeley.

  • Made ren­o­va­tions to the Oxford office to help our staff be more pro­duc­tive, in­clud­ing stand­ing desks, noise-can­cel­ling par­ti­tions, faster Wi-Fi, and daylight lamps.

  • Pro­vided lo­gis­ti­cal sup­port for EA Global con­fer­ences, Lead­ers Fo­rum, and team re­treats.

  • Dealt with bud­gets, con­tracts, grants, and pay­ments nec­es­sary for the on­go­ing func­tion of CEA. This in­cluded man­ag­ing ap­prox­i­mately ten times the vol­ume of dona­tions as last year.

  • En­abled the func­tion­ing of both EA Funds and EA Grants by in­ves­ti­gat­ing le­gal risks and re­quire­ments, track­ing dona­tion in­for­ma­tion, work­ing with lawyers to write up con­tracts, cor­re­spond­ing with donors, and pay­ing out grants to re­cip­i­ents.

  • Suc­cess­fully com­pleted an au­dit for CEA UK.

  • Ac­quired H1B cap ex­emp­tion, mak­ing it eas­ier for CEA to hire peo­ple in the fu­ture.

  • Ac­quired visas for many staff so they were able to move to the US.

Im­pact assessment

Our op­er­a­tions work is vi­tal for the func­tion­ing of CEA, and this year has al­lowed us to scale up re­cruit­ment, fi­nances and office space. There are still some ar­eas where we can im­prove, how­ever, in­clud­ing:

  • In­creas­ing the timeli­ness with which we deal with donor and other enquiries

  • Up­dat­ing our ac­count­ing so that it is trans­par­ent out­side of CEA and faster to au­dit.

Plans for 2018

Our main pri­or­ity in 2018 is to build ca­pac­ity so that we can con­tinue to scale both in terms of staff and in dona­tions. This means build­ing more ro­bust fi­nan­cial, le­gal and HR pro­cesses that suit the or­ga­ni­za­tion in its cur­rent, larger form (e.g., bet­ter com­mu­ni­ca­tion pro­to­cols with ac­coun­tants, bet­ter ex­pens­ing pro­ce­dures for em­ploy­ees, and bet­ter-doc­u­mented pro­cesses for op­er­a­tional pro­ce­dures.)

Over­all, 2017 was largely a year of cre­ation for the Oper­a­tions Team it­self and 2018 will be fo­cused on im­prov­ing what has been built to meet the needs of the or­ga­ni­za­tion’s size and scope.

Tech Team

The Tech Team pro­vides on­line in­fras­truc­ture and tech­ni­cal ad­vice to other teams, main­tain­ing and build­ing soft­ware to help the EA com­mu­nity be more effec­tive and to in­crease CEA’s op­er­a­tional effi­ciency. EA Funds is now also man­aged by the Tech Team.

The team scaled up sig­nifi­cantly this year, from one em­ployee to four, adding two de­vel­op­ers and a product man­ager. This should al­low us to sig­nifi­cantly re­duce de­vel­op­ment times and provide more ca­pac­ity to work on more com­plex pro­jects.

Tech Team Ac­tivi­ties in 2017

Our par­ti­ci­pa­tion in Y Com­bi­na­tor al­lowed us to build use­ful net­works and build skills. Dur­ing Y Com­bi­na­tor, we de­vel­oped the idea of EA Funds. Our work cul­mi­nated in the pro­ject’s re­lease in March 2017.

EA Funds has re­ceived over $2m in dona­tions to its philan­thropic funds to date, and of this amount, re­granted around $1.1m. In ad­di­tion to the do­main-ex­pert-man­aged philan­thropic funds, EA Funds has also been serv­ing as a cen­tral­ized dona­tion gate­way for donors to give to EA-al­igned or­ga­ni­za­tions. Donors can eas­ily set a preferred al­lo­ca­tion to any com­bi­na­tion of the Funds and to any of the non-profit or­ga­ni­za­tions cur­rently sup­ported on the plat­form. (This in­cludes all GiveWell top char­i­ties, most stand­out char­i­ties, and sev­eral EA“meta” or­ga­ni­za­tions). So far, a fur­ther ap­prox­i­mately $600,000 has been donated through the plat­form in this man­ner.

EA Funds is part of Effec­tiveAltru­ism.org, which we are build­ing out as our flag­ship on­line plat­form. The plat­form has ex­panded to in­clude the Giv­ing What We Can Pledge, and tools for eas­ier dis­cov­ery and man­age­ment of lo­cal groups. Th­ese prod­ucts are still in early stages of user test­ing, but they have proven that the ar­chi­tec­ture of the web app can be used for a broad range of pur­poses. This will even­tu­ally provide mem­bers of the EA com­mu­nity with a sin­gle lo­gin for ac­cess­ing a range of core on­line ser­vices (Funds dona­tions, Giv­ing What We Can or Try Giv­ing pledge, and My Giv­ing record, lo­cal group mem­ber­ships, and even­tu­ally EA Fo­rum and EA Global tick­et­ing), all in one place.

One of our key achieve­ments this year was to dras­ti­cally in­crease our ca­pac­ity for next year. The key bot­tle­neck prior to the hiring round was hav­ing all as­pects of CEA’s on­line in­fras­truc­ture man­aged by a sin­gle in­di­vi­d­ual. This hiring round al­lowed us to pro­fes­sion­al­ize, and the greater di­vi­sion of re­spon­si­bil­ities means that in­di­vi­d­ual prod­ucts re­ceive much more ded­i­cated at­ten­tion.

Im­pact Review

Our most ob­vi­ous suc­cess this year is EA Funds, which grew from an idea in Jan­uary to a widely used plat­form by the end of the year. We pro­cessed 10,000 dona­tions from al­most 2,500 in­di­vi­d­ual donors, to­tal­ing ~$2.6m. We see EA Funds as a key piece of com­mu­nity in­fras­truc­ture, as well as a well-tested spring­board to launch new pro­jects that provide value to the com­mu­nity.

While we con­sider EA Funds to be a suc­cess­ful pro­ject, there re­mains con­sid­er­able room for im­prove­ment. We should have pri­ori­tized build­ing sys­tems that provide more reg­u­lar in­sight into the amount of money in each fund. This would have benefited both the fund man­agers (who do not cur­rently have an easy way to check how much is in their re­spec­tive funds at any given time) and would in­crease the com­mu­nity’s trust in EA Funds by pro­vid­ing greater trans­parency. In par­tic­u­lar, we did not pub­lish grant pay­out re­ports on the web­site as quickly as we should have. We are up­dat­ing our pro­cesses to ad­dress these is­sues, and we have pri­ori­tized cre­at­ing a unified dash­board where donors can find more in­for­ma­tion about the cur­rent bal­ance of each fund, with work ex­pected to com­mence in Q1 2018.

Pri­ori­ti­za­tion has been difficult. There is a ten­sion be­tween:

  • Busi­ness as usual (e.g., GWWC Pledge up­grade, main­tain­ing EA Funds)

  • Ca­pac­ity build­ing (e.g., pro­jects for the Oper­a­tions team)

  • Im­proved com­mu­nity in­fras­truc­ture (e.g., new EA Fo­rum, event tick­et­ing)

It seems im­por­tant to day to day work done, even though it of­ten seems lower pri­or­ity than im­prov­ing in­fras­truc­ture. We would like to spend more time in 2018 clar­ify­ing how to make these trade-offs so that we fo­cus on the de­vel­op­ment work that adds the most value.

Plans for 2018

The Tech Team will con­tinue to sup­port all other CEA teams in achiev­ing their goals, listen closely to com­mu­nity and user feed­back, and de­velop pro­jects that help EAs be­come more effec­tive.

We are broadly fol­low­ing the ‘ag­ile de­vel­op­ment’ model, which in­volves (1) seek­ing reg­u­lar in­put from other teams on what to pri­ori­tize and (2) build­ing tight feed­back loops so that we can test hy­pothe­ses and make course cor­rec­tions. The be­low are our best guesses at pri­or­ity pro­jects for 2018 but are sub­ject to change and repri­ori­ti­za­tion as we get in­put from col­leagues and users.

EA Funds

  • A re­port­ing solu­tion for the Fund Man­agers and EA Funds users, im­prov­ing the trans­parency of each fund’s tak­ings and payouts

  • Donor lot­tery func­tion­al­ity (beta to be re­leased mid-De­cem­ber, fur­ther tweaks ex­pected in sub­se­quent runs of the lot­tery)

  • Po­ten­tial ex­pan­sion of EA Funds on offer and in­ves­ti­ga­tion of differ­ent mod­els for run­ning and us­ing funds

  • Au­toma­tion of pay­roll giving

  • In­clu­sion of PayPal as a sup­ported pay­ment gateway

Effec­tivealtru­ism.org web app

  • Bring­ing the cur­rent con­tent on Effec­tiveAltru­ism.org into the web app, thus con­soli­dat­ing al­most all our prod­ucts un­der one login

  • Var­i­ous ex­pe­rience and de­sign im­prove­ments across the web app

In­ter­nal effi­ciency and mon­i­tor­ing software

  • An im­proved ad­min por­tal for Effec­tiveAltru­ism.org for use by the op­er­a­tions team to stream­line the day-to-day ad­minis­tra­tion of dona­tions, re­grants and cus­tomer support

  • A dash­board for the op­er­a­tions team fo­cused on their key metrics

  • Com­mu­nity team Cus­tomer Re­la­tion­ship Man­age­ment (CRM) solu­tion, ei­ther built in-house or an off-the-shelf solu­tion in­te­grated with our sys­tems. The goal is to have a solu­tion that al­lows us to pre­sent rele­vant data from all sources in one place.

  • Fo­cus on im­prov­ing our in­ter­nal an­a­lyt­ics processes

Offi­cial roll-out of the EA Groups platform

  • We have user test­ing planned for early Jan­uary and in­tend to use feed­back re­ceived to fi­nal­ize the plat­form. We ex­pect a wide re­lease later that month.

Fi­nal­iz­ing the mi­gra­tion of Giv­ing What We Can site func­tion­al­ity to effec­tivealtru­ism.org

  • The Pledge has already been mi­grated to effec­tivealtru­ism.org.

  • Mi­gra­tion of the My Giv­ing dash­board. This will au­to­mat­i­cally im­port dona­tions through EAFunds. It will al­low users to re­port and mon­i­tor their in­comes, dona­tions, and Pledge ad­her­ence. This is due late 2017/​early 2018.

  • Mi­gra­tion of ex­ist­ing My Giv­ing users to the new system

EA Forum

  • Fi­nal­ize the han­dover of the cur­rent EA Fo­rum’s code­base from Trike Apps (cur­rent main­tain­ers)

  • In­ves­ti­gate op­tions for build­ing a new com­mu­nity dis­cus­sion plat­form and ex­e­cute on the plan that comes out of this process

Develop in-house event man­age­ment software

  • Cen­tral tick­et­ing sys­tem for EA Global/​EAGx events, im­prov­ing user ex­pe­rience, re­duc­ing ac­count­ing time, and re­duc­ing re­li­ance on of­ten-in­ad­e­quate third-party event man­age­ment soft­ware.

CEA’s Mistakes

We have of course made mis­takes. While some of these are cov­ered in our sec­tions re­view­ing im­pact, we felt it was im­por­tant to clearly note our short­com­ings here, too. A few of the more-sig­nifi­cant ways we can im­prove which we have iden­ti­fied that cut across mul­ti­ple pro­jects are as fol­lows:

  • We as­pire to high stan­dards of trans­parency, so the com­mu­nity and our donors know what we’re do­ing and so other ac­tors in effec­tive al­tru­ism are able to make in­formed de­ci­sions about how to in­ter­act with CEA and the ser­vices we provide. How­ever, we have not always lived up to this stan­dard. In some cases, we pri­ori­tized mov­ing on to new pro­jects be­fore suffi­ciently com­mu­ni­cat­ing our cur­rent think­ing re­gard­ing ex­ist­ing ones. This is one of the mo­ti­va­tions for writ­ing this post and for en­courag­ing in­ter­ested com­mu­nity mem­bers to sign up to our sup­port­ers mailing list, but we still need to do more in this area.

  • We also could have put more em­pha­sis on en­sur­ing that when staff move be­tween pro­jects, they clearly hand over their re­port­ing re­quire­ments.

  • In some cases, we also had trou­ble com­mu­ni­cat­ing strat­egy or plans for re­struc­tur­ing in­ter­nally. This caused stress and re­duced pro­duc­tivity for some staff mem­bers.

  • In a few cases, we were poor at com­mu­ni­cat­ing hiring de­ci­sions to ap­pli­cants.

Below is a non-ex­haus­tive list of short­com­ings, or­ga­nized by pro­ject team:

Re­search Team

Towards the be­gin­ning of the year, we failed to have a suffi­ciently fo­cused re­search agenda. This was part of the mo­ti­va­tion for the in­te­gra­tion of the re­search team with the rest of CEA mid-year.

In the sec­ond half of the year, we failed to make enough progress on pro­duc­ing origi­nal con­tent, partly be­cause we were split­ting our time be­tween this, strat­egy work, sup­port­ing EA Grants, and col­lat­ing con­tent for Effec­tiveAltru­ism.org. We plan to be more fo­cused on origi­nal con­tent in the com­ing year.

In some cases, we should have spent more time plan­ning pro­jects for sum­mer re­search fel­lows, and we should have en­couraged sum­mer fel­lows to share their ideas and col­lab­o­rate with each other more than we did this year. We will be care­fully as­sess­ing whether and how to run any re­search fel­low­ships in the fu­ture.

In­di­vi­d­ual Outreach Team

EA Grants

Note that we have in­cluded this sec­tion un­der the In­di­vi­d­ual Outreach Team be­cause they now run EA Grants, but this was not the case for most of this year.

Our com­mu­ni­ca­tion around EA Grants was con­fus­ing. We ini­tially an­nounced the pro­cess with lit­tle ad­ver­tise­ment. Then, we ad­ver­tised it on the EA Newslet­ter, but only shortly be­fore the ap­pli­ca­tion dead­line, and ex­tended the dead­line by two days.

We un­der­es­ti­mated the num­ber of ap­pli­ca­tions we would re­ceive, which gave us less time per can­di­date in the ini­tial eval­u­a­tion than we would have liked. It also caused de­lays, which we did not ad­e­quately com­mu­ni­cate to ap­pli­cants. We should have been less am­bi­tious in set­ting our ini­tial dead­lines for re­ply­ing, and we should have com­mu­ni­cated all changes in our timetable im­me­di­ately and in writ­ing to all ap­pli­cants.

Our ad­ver­tise­ment did not make suffi­ciently clear that we might not be able to fund ed­u­ca­tional ex­penses through CEA. For­tu­nately, the Open Philan­thropy Pro­ject was re­cep­tive to con­sid­er­ing some of the aca­demic ap­pli­cants.


A de­lay in im­ple­ment­ing some of the re­cur­ring pay­ment pro­cess­ing logic in EA Funds meant that users who cre­ated re­cur­ring pay­ments be­fore May did not have their sub­scrip­tions pro­cessed. The is­sue has since been fixed and re­cur­ring pay­ments have been work­ing nor­mally since mid-May. We did not charge make-up pay­ments, which meant sev­eral thou­sand dol­lars worth of pay­ments that should have been pro­cessed were not. We in­formed donors as soon as we were aware of the is­sue. How­ever, a more ro­bust pri­ori­ti­za­tion pro­cess could prob­a­bly have avoided the is­sue.

A bug in our mes­sage queue sys­tem meant that some pay­ment in­struc­tions were pro­cessed twice. Due to poor timing (an au­dit, fol­lowed by a team re­treat), the bug was not dis­cov­ered for sev­eral days, lead­ing to around 20 donors be­ing charged for their dona­tions twice. As soon as the fault was dis­cov­ered, we no­tified donors and re­funded their pay­ments. We now pe­ri­od­i­cally in­spect the mes­sage queue to en­sure mes­sages are be­ing de­liv­ered cor­rectly, and we have im­ple­mented an ad­di­tional layer of dedu­pli­ca­tion logic.

A failure to perform server main­te­nance on Giv­ing What We Can’s server caused it to in­ter­mit­tently stop re­spond­ing to net­work con­nec­tions. This caused many peo­ple frus­tra­tion as they tried to log in or take the Pledge. Due to on­go­ing is­sues with this sys­tem, and pri­ori­tiz­ing other pro­jects, we did not iden­tify the cause of the fault as fast as we should have. This func­tion­al­ity has since been mi­grated to the Effec­tiveAltru­ism.org, and the server will be de­com­mis­sioned soon.

We failed to keep the EA Funds web­site up to date, mean­ing that many users were un­sure how their money was be­ing used. With the ar­rival of Marek Duda as product man­ager, we are now ad­dress­ing this. We are plan­ning to pub­lish a fur­ther post, by the end of 2017, with a deeper dive on the EA Funds plat­form.

Our hiring pro­cess took longer than we an­ti­ci­pated be­cause we had to de­velop a pro­cess for tech­ni­cal hiring rounds. We think that we have now learned how to run such a round in the fu­ture.

Com­mu­nity Team


Our ad­ver­tis­ing around EA Global events, es­pe­cially the Lon­don con­fer­ence, was con­fus­ing. We made the de­ci­sion to shift to­wards more ad­vanced con­tent aimed at ex­ist­ing mem­bers of the com­mu­nity mid­way through the five-month ap­pli­ca­tion pe­riod. This led to con­fu­sion about the in­tended au­di­ence for the con­fer­ence and newer mem­bers of the EA com­mu­nity who were pre­vi­ously en­couraged to at­tend our events felt shut out. This is a not good way of wel­com­ing peo­ple to our com­mu­nity. In the fu­ture, we will make sure that our ad­ver­tis­ing, ad­mis­sions, and con­tent for events is more con­sis­tent and com­mu­ni­cated fur­ther in ad­vance.

We did not provide enough sup­port to or­ga­niz­ers of EAGx con­fer­ences. We hope that in­creased ca­pac­ity on the team, through hiring a new events spe­cial­ist and hand­ing op­er­a­tions work to the Oper­a­tions Team, will help with this.

Lo­cal groups

Com­mu­ni­ca­tion with EA Group Or­ga­niz­ers should have been more fre­quent and more re­li­able. For ex­am­ple, many stu­dent group or­ga­niz­ers dur­ing EAG Lon­don re­ported hav­ing an un­clear un­der­stand­ing of the goals of EA groups and CEAs’ think­ing in gen­eral. The EA Com­mu­nity build­ing guide is in­tended to ad­dress this, but it is yet to be pub­lished. Speed of com­mu­ni­ca­tion should have been pri­ori­tized over depth here. There have also been oc­ca­sions where group lead­ers have been left wait­ing be­cause of CEA. For ex­am­ple, the launch of the EA Groups plat­form has been de­layed mul­ti­ple times while we have been build­ing the ca­pac­ity in our Tech Team.

Our work on lo­cal groups was at times in­suffi­ciently fo­cused. In some cases, we tried sev­eral ap­proaches, but not long enough to prop­erly as­sess whether they had suc­ceeded or not. For ex­am­ple, we be­gan a beta ver­sion of an EA Con­ver­sa­tions plat­form to fa­cil­i­tate con­ver­sa­tions be­tween EAs but dis­con­tinued work on it de­spite ini­tial suc­cess, largely be­cause of com­pet­ing time de­mands. We have been us­ing a quar­terly goal set­ting and re­view pro­cess to try and im­prove this.

Giv­ing What We Can

We failed to keep some of the con­tent on this web­site up-to-date, as some of the figures we use have changed. Similarly, while the Pledge is cause-neu­tral and no longer fo­cused solely on the de­vel­op­ing world our web­site doesn’t fully re­flect that. We plan to ad­dress these is­sues in 2018.

At times, par­tic­u­larly ear­lier in the year, we fo­cused too much on pro­mot­ing the main Pledge, even though this might not be the right op­tion for some peo­ple. For this rea­son, we have shifted more em­pha­sis to Try Giv­ing.

Com­mu­nity health

In re­spond­ing to ex­ist­ing prob­lems, we have not pri­ori­tized pre­vent­ing or miti­gat­ing other prob­lems as much as we could have. We have com­mis­sioned the Re­search Team to pro­duce a re­port on proac­tive things we can do to mea­sure and ad­dress com­mu­nity health, and we in­tend to be more proac­tive on this in 2018.


As dis­cussed above, oc­ca­sion­ally, the team was slow to re­spond to donor and other en­quiries. We have been build­ing our team’s knowl­edge, ca­pac­ity and pro­cesses to try and im­prove this. For ex­am­ple, ear­lier in the year ac­cess to donor in­for­ma­tion to re­spond to en­quiries was re­stricted to a few staff mem­bers, slow­ing our re­sponse times.

We still need to im­prove our ac­count­ing, so that it is more trans­par­ent to peo­ple out­side CEA, and to re­duce the time that au­dits take up.


Our cur­rent fund­ing situ­a­tion is se­cure, with ap­prox­i­mately two years of run­way. How­ever, we are plan­ning to scale up some ac­tivi­ties in 2018, in­clud­ing ex­panded grant­ing through EA Grants and via lo­cal groups, as well as a larger pro­gram of events. Given our growth, it is also likely we will need to make new hires dur­ing the year. While much of this ex­pen­di­ture will be cov­ered by larger donors, we are also fundrais­ing to make sure we have di­ver­sity in our sources of in­come (es­pe­cially since some of our fund­ing agree­ments are con­tin­gent on us hav­ing mul­ti­ple back­ers). If you would like to sup­port CEA, please donate us­ing this link.


We hope that this post has given you some in­sight into our work this year and our plans for 2018. As our mis­sion is to sup­port the EA com­mu­nity in do­ing the most good we can, we want to keep you in the loop and hear your feed­back.

If you would like to re­ceive our monthly sup­port­ers emails for more reg­u­lar up­dates on our work, please sign up here. If you would like to dis­cuss any of our plans in depth, please con­tact the team lead listed at the be­gin­ning of this post or com­ment here.