Requesting community input on the upcoming EA Projects Platform


Ben­jamin Pence and I have been work­ing on Altru­, an ini­ti­a­tive to make more high EV (ex­pected value) pro­jects pos­si­ble. We origi­nally fo­cused on in­creas­ing fund­ing for high im­pact ven­tures, sparked by my ini­tial EA fo­rum post over the sum­mer. After talk­ing with Good Growth and other com­mu­nity mem­bers, we have broad­ened our scope to pro­ject suc­cess in gen­eral, of which pro­ject fund­ing is still a ma­jor com­po­nent.

In Oc­to­ber, we an­nounced the EA An­gel Group, a group that helps in­di­vi­d­ual EA fun­ders share grant op­por­tu­ni­ties and grant eval­u­a­tions with one an­other. The fun­ders gain the benefits of shar­ing in­for­ma­tion and op­por­tu­ni­ties within a small com­mu­nity. We are in­ter­ested in bring­ing these benefits to the EA com­mu­nity at large by cre­at­ing an on­line plat­form that re­solves ma­jor in­for­ma­tion and co­or­di­na­tion gaps in the high EV pro­ject space. Ben­jamin and I have cre­ated an op­er­a­tional pro­to­type of this plat­form and we are seek­ing feed­back on our de­sign choices.

Benefit 1: Discovery

It is very difficult for in­di­vi­d­ual EA fun­ders to dis­cover pro­jects to fund. With EA Grants, BERI Grants, the newly launched Effec­tive Altru­ism Foun­da­tion Fund, the four EA Funds, the EA An­gel Group, Let’s Fund, etc all seek­ing high im­pact ven­tures, we be­lieve a cen­tral lo­ca­tion for ap­pli­cants to ap­ply for fund­ing and fun­ders to dis­cover grant op­por­tu­ni­ties would be very helpful, par­tic­u­larly to EA com­mu­nity mem­bers and smaller grant­mak­ers.

The com­mu­nity has been ask­ing for this on an on­go­ing ba­sis: In the lat­est Long Term Fu­ture Fund grant dis­cus­sion in Novem­ber, the lat­est EA Grants ap­pli­ca­tion round an­nounce­ment in Septem­ber, and in a Face­book dis­cus­sion on the EA Face­book group ear­lier this year, com­mu­nity mem­bers ex­pressed sup­port for the idea of shar­ing fund­ing op­por­tu­ni­ties with the gen­eral pub­lic or with a sub­set of EAs.

Cur­rently, it is challeng­ing to re­quest and offer other pro­ject-re­lated op­por­tu­ni­ties such as vol­un­teer­ing, cofounder match­ing, and ad­vis­ing, as well as other op­por­tu­ni­ties like office space and web host­ing.

We en­vi­sion that the EA Pro­ject Plat­form will provide “op­por­tu­nity boards.” For ex­am­ple, in the case of vol­un­teer­ing, a pro­ject lead may dis­play vol­un­teer open­ings or in­di­vi­d­u­als in­ter­ested in vol­un­teer­ing can spec­ify the kinds of work they are will­ing to do such as graphic de­sign and trans­la­tion.

Re­gard­ing fund­ing, some EAs have pro­posed to us that all grant op­por­tu­ni­ties should be made pub­lic, while other EAs have sug­gested hid­ing some grant op­por­tu­ni­ties that could pose risks due to the in­for­ma­tion they in­clude or their po­ten­tially nega­tive out­comes. In or­der to ad­dress these con­cerns, our cur­rent in­tent is to make our en­tire plat­form visi­ble to EA com­mu­nity mem­bers only, and not pub­li­cly visi­ble or in­dex­able by search en­g­ines to en­courage greater in­for­ma­tion shar­ing and re­duce risks.

We will likely also im­pose a light re­view pro­cess (e.g. other EA fund­ing or­ga­ni­za­tions can de­lay or re­ject po­ten­tially harm­ful or con­cern­ing pro­jects) or qual­ity check (e.g. re­quiring at least 1 eval­u­a­tion) be­fore op­por­tu­ni­ties are pub­li­cly posted. While hid­ing po­ten­tially harm­ful grant op­por­tu­ni­ties may re­duce risks, post­ing the rest of the fund­ing op­por­tu­ni­ties with no way to share in­for­ma­tion re­gard­ing ex­pected im­pact could be quite in­effec­tive as well. That’s where the eval­u­a­tion com­po­nent of our plat­form comes in.

Benefit 2: Evaluation

Even more prob­le­matic than a lack of aware­ness re­gard­ing pro­jects could be a lack of in­for­ma­tion about a pro­ject’s ex­pected im­pact. Without aware­ness of ex­pected im­pact, peo­ple could in­ad­ver­tently ex­e­cute low EV or even nega­tive EV pro­jects. Fun­ders and vol­un­teers could in­ad­ver­tently back in­effec­tive or harm­ful pro­jects, which is es­pe­cially rele­vant for an open plat­form.

We will sup­port pro­ject eval­u­a­tions on the plat­form. All users or a cer­tain sub­set of users will be able to cre­ate an eval­u­a­tion and we will strongly em­pha­size com­mu­nity norms around in­di­cat­ing the rigor that went into the eval­u­a­tion (e.g. time spent) as well as the depth of rele­vant knowl­edge. We be­lieve that pub­li­cly shar­ing all eval­u­a­tion in­for­ma­tion makes it eas­ier for po­ten­tial fun­ders and con­trib­u­tors to get a full sense of all of the rele­vant con­sid­er­a­tions around a pro­ject. How­ever, a pub­lic eval­u­a­tion sys­tem may en­counter var­i­ous pit­falls.

As pro­jects progress and re­ceive feed­back, they will likely change in plan or per­son­nel; there­fore, the plat­form will likely strongly cou­ple eval­u­a­tions of a pro­ject with a stage in its de­vel­op­ment or a cer­tain point in time to en­courage pro­ject idea iter­a­tion, re­duce bi­as­ing fu­ture eval­u­a­tors, and avoid cer­tain prob­lems (e.g. a self-fulfilling prophecy where a pro­ject fails be­cause it did not at­tract fund­ing be­cause an eval­u­a­tion pre­dicted it would fail).

Fur­ther­more, it may be valuable to in­clude other mea­sures like anony­mous pro­ject pro­pos­als to be eval­u­ated, anony­mous eval­u­a­tions, and the pos­si­bil­ity of limit­ing eval­u­a­tion visi­bil­ity to re­duce the nega­tive effects that gain­ing pro­ject feed­back at too early of a stage or gain­ing feed­back from the wrong eval­u­a­tors may have on pro­ject suc­cess.

Benefit 3: Coordination

Ear­lier this week, some­one pitched a pro­ject idea to me with­out know­ing that it was pre­vi­ously an ac­tual EA pro­ject with vol­un­teers (a pro­ject that I had pre­vi­ously con­tributed to, in fact). There is cur­rently no easy way for com­mu­nity mem­bers to dis­cover ac­tive pro­jects, past pro­jects, and pro­ject ideas that are in the com­mu­nity. This has un­de­sir­able effects such as du­pli­cated efforts, missed col­lab­o­ra­tion op­por­tu­ni­ties, and wasted time as peo­ple pur­sue pro­ject ideas that have already been tried un­suc­cess­fully in the past.

Our pro­ject plat­form will re­volve around pro­jects. It will be easy to find pro­jects that match spe­cific crite­ria and all their re­lated fund­ing and vol­un­teer­ing op­por­tu­ni­ties. While we will not re­move con­tent from the plat­form, users will be able to filter by cre­ated and last mod­ified meta­data to en­sure that they are see­ing con­tent that is still rele­vant.

We be­lieve it may be valuable to sep­a­rate ac­tive pro­jects and pro­ject ideas. Hav­ing a list­ing of ideas may pos­i­tively in­fluence the de­sign of pro­jects that are be­ing ac­tively ex­e­cuted as well as help EAs and EA or­ga­ni­za­tions eas­ily judge com­mu­nity sen­ti­ment around cer­tain pro­pos­als, es­pe­cially if ideas can be pro­posed anony­mously. Good Growth men­tioned to us that sup­port­ing pro­ject in­cu­ba­tion with plat­form el­e­ments like ex­pert-gen­er­ated ideas, pre-funded pro­ject pro­pos­als, and pairing pro­ject pro­pos­als with teams that have the re­quired back­ground and ex­per­tise could be highly valuable.

Ideas could im­prove the eval­u­a­tion po­ten­tial of the plat­form. Pro­ject leads look­ing to get a quick check on com­mu­nity sen­ti­ment be­fore start­ing a pro­ject can post an anony­mous idea for feed­back, and af­ter a pro­ject gets off the ground, it can re­quest in-depth eval­u­a­tions by reach­ing out to com­mu­nity mem­bers and have eval­u­a­tors post a sum­mary of their eval­u­a­tion of the pro­ject on the plat­form.

Thanks for read­ing! Any com­ments, ques­tions, and eval­u­a­tions are ap­pre­ci­ated. The fol­low­ing para­graphs are my com­ments on re­lated com­mu­nity ini­ti­a­tives.

It may be the case that an even broader plat­form that ap­plies to the en­tire EA com­mu­nity, en­com­pass­ing all com­mu­nity mem­bers and EA or­ga­ni­za­tions, would be even bet­ter for in­creas­ing the im­pact of the effec­tive al­tru­ism move­ment. Such a sys­tem could make it easy for EAs to share in­for­ma­tion and col­lab­o­rate with other EAs in rele­vant ar­eas, and for the com­mu­nity to see all of the ini­ti­a­tives hap­pen­ing across EA and col­lec­tively share eval­u­a­tions of EA ini­ti­a­tives and provide fund­ing and sup­port. In my mind, such a plat­form would likely in­clude a dis­tinc­tion be­tween or­ga­ni­za­tions, pro­jects, lo­cal groups, ideas, and other en­tities, as well as provide strong so­cial net­work­ing and col­lab­o­ra­tion fea­tures.

Any in­sights on whether a broader sys­tem would be valuable would be ap­pre­ci­ated. I be­lieve this is the ul­ti­mate vi­sion Re­think Char­ity’s EA Hub team is work­ing to­wards. After speak­ing with the EA Hub, my as­sess­ment is that our MVP (min­i­mum vi­able product) con­cept and their MVP con­cept are quite differ­ent but may con­verge in the fu­ture; their MVP em­pha­sizes com­mu­nity pro­files and ge­olo­ca­tion, whereas our MVP em­pha­sizes pro­jects, pro­ject eval­u­a­tions, and pro­ject sup­port op­por­tu­ni­ties. I hope our plat­form de­sign and launch will provide valuable in­for­ma­tion for fu­ture at­tempts to im­prove com­mu­nity co­or­di­na­tion with soft­ware and we can in­te­grate with or port over all in­for­ma­tion to fu­ture com­mu­nity co­or­di­na­tion soft­ware ini­ti­a­tives that gain trac­tion.

Re­gard­ing more limited ap­pli­ca­tions of our pro­ject plat­form idea, EA Tech Ini­ti­a­tives and the EA Work Club are the only cur­rently ac­tive pro­jects I am aware of. The .im­pact Hack­pad is the only past pro­ject I am aware of.

EA Tech Ini­ti­a­tives is suc­cess­ful, but it only per­tains to tech­nol­ogy pro­jects, and all of its pro­ject list­ings and op­por­tu­ni­ties are in a Google Spread­sheet. Our plat­form ap­pears to help with the di­rec­tory func­tion­al­ity of the Google Spread­sheet while en­abling the EA Tech Ini­ti­a­tives com­mu­nity to con­tinue col­lab­o­rat­ing across var­i­ous mediums like Re­think Char­ity’s Slack team.

The EA Work Club fo­cuses on work op­por­tu­ni­ties and differs from our in­tended plat­form in many ways. I view our eval­u­a­tion sys­tem as an iter­a­tion of its up­vote sys­tem, and our time-based fil­ter­ing in­stead of au­to­matic pro­ject list­ing re­moval is likely a bet­ter im­ple­men­ta­tion that en­sures users see rele­vant con­tent while pre­serv­ing data that could still be use­ful.

The .im­pact Hack­pad val­i­dated that gath­er­ing in­for­ma­tion about ac­tive pro­jects is pos­si­ble and use­ful. I be­lieve the abil­ity for our plat­form to sup­port fil­ter­ing hun­dreds or thou­sands of pro­jects and op­por­tu­ni­ties based on meta­data like cause area, cre­ation date, and stage of pro­ject will help it scale and al­low users to bet­ter dis­cover in­for­ma­tion.