Effective Altruism Sweden plans for 2018

Effec­tive Altru­ism Swe­den will have a full-time em­ployee over the com­ing year (that’s me, Markus An­der­ljung). This doc­u­ment out­lines the plans of Effec­tive Altru­ism Swe­den and what we’ll do with that ex­tra la­bor over the com­ing year.

I think that the biggest op­por­tu­ni­ties in Swe­den that we ought to take ad­van­tage of are: po­ten­tial for com­mu­nity-build­ing, a vibrant startup com­mu­nity, the abil­ity to test work­ing on poli­tics and a fairly large x-risk com­mu­nity. Given this con­text, our cur­rent plans are to spend a sig­nifi­cant amount of re­sources on ca­pac­ity-build­ing efforts and on one larger pro­ject.

The ca­pac­ity-build­ing efforts will fo­cus on com­mu­nity-build­ing, pri­mar­ily us­ing meth­ods and rea­son­ing largely com­mon to most effec­tive al­tru­ism groups, with some ex­tra at­ten­tion paid to di­ver­sity-is­sues. Our rea­son­ing is laid out in more de­tail in Sec­tion 3.

The larger pro­ject we cur­rently plan on spend­ing a sig­nifi­cant amount of time on is the Rep­re­sen­ta­tion of Fu­ture Gen­er­a­tions. The idea is to im­ple­ment mechanisms in the Swedish poli­ti­cal sys­tem to en­sure that there is rep­re­sen­ta­tion of fu­ture gen­er­a­tions: that the in­ter­ests of fu­ture gen­er­a­tions are taken into ac­count in poli­ti­cal de­ci­sion-mak­ing. This work is mod­el­led on work car­ried out over the past year by CSER and FUSE in the UK, lead­ing to the for­ma­tion of the All-Party Par­li­a­men­tary Group on Fu­ture Gen­er­a­tions in Novem­ber. For more de­tails, see Sec­tion 4. The sec­tion also in­cludes de­scrip­tions of two pro­jects we de­cided not to pur­sue.

This doc­u­ment ex­plic­itly lays out the rea­son­ing be­hind our cur­rent plan, to al­low you to find mis­takes and help make the plan bet­ter. To aid this fur­ther, I try to make show where we think our un­cer­tain­ties are. I have also listed what I think are the biggest most cru­cial un­cer­tain­ties—our cru­cial con­sid­er­a­tions—are, in 3.3. and 4.3.

Thanks to Ste­fan Schu­bert, Carin Ism, Karolina Grundin, Ste­fan Ein­horn, Gabriella Overöd­der, To­bias Malm, Cather­ine Derieux, James Snow­den, Harri Besceli, Holly Mor­gan, Jonas Vol­lmer and Beth Barnes for use­ful dis­cus­sions on the plan so far. You can find a ver­sion of the doc­u­ment in­clud­ing foot­notes here.

Table of contents

1. Background

1.1. A brief (se­lec­tive) his­tory of effec­tive al­tru­ism in Sweden

1.2. Pur­pose of this document

1.3. What are the op­por­tu­ni­ties available to EA Swe­den?

2. Goals for the com­ing year

3. Ca­pac­ity-build­ing activities

3.1. Com­mu­nity-building

3.1.1. Get­ting peo­ple to pos­i­tively im­pact the world

3.1.2. Ex­am­ples of planned activities

3.1.3. Diver­sity and inclusiveness

3.1.4. Other activities

3.2. Or­ga­ni­za­tional capacity

3.3. Cru­cial con­sid­er­a­tions re­gard­ing ca­pac­ity-building

4. Direct work

4.1. Rep­re­sen­ta­tion of fu­ture generations

4.1.1. What’s the idea?

4.1.2. Rep­re­sen­ta­tion of fu­ture gen­er­a­tions in Sweden

4.1.3. Po­ten­tial benefits

4.1.4. Im­pact mechanism and failure modes

4.1.5. Do we have the re­sources to carry out this pro­ject?

4.1.6. Plan outline

4.2. Some pro­jects we have de­cided against for now

4.2.1. EA Fact Check­ers /​ The EA Re­pro­ducibil­ity Project

4.2.2. Aid effec­tive­ness-rank­ing of poli­ti­cal parties

4.3. Cru­cial con­sid­er­a­tions re­gard­ing di­rect work

1. Background

1.1. A brief (se­lec­tive) his­tory of effec­tive al­tru­ism in Sweden

A brief his­tory of the effec­tive al­tru­ism-move­ment in Swe­den may prove use­ful back­ground to the be­low. In 2015, peo­ple in Swe­den in­ter­ested in grow­ing the move­ment started meet­ing up. We did not set up Effec­tive Altru­ism Swe­den im­me­di­ately, but in­stead fo­cused on build­ing some mo­men­tum ini­tially. This was done by putting on some meet-ups, par­tic­u­larly in Stock­holm, but mainly in set­ting up three stu­dent groups at uni­ver­si­ties in Stock­holm: Stock­holm Univer­sity, Stock­holm School of Eco­nomics and the Royal In­sti­tute of Tech­nol­ogy.

In Septem­ber 2016, we felt that there was enough of a mo­men­tum and peo­ple in­ter­ested in effec­tive al­tru­ism to war­rant a na­tional or­ga­ni­za­tion and EA Swe­den was cre­ated. Dur­ing 2017, EAS had a com­mit­tee of five peo­ple. The main fo­cus was set­ting up the in­fras­truc­ture for the or­ga­ni­za­tion, putting on com­mu­nity-build­ing events and out­reach.

A se­lec­tion of ac­tivi­ties over the year Oct 2016 - Oct 2017:

  • Or­ga­nized a talk by Will MacAskill with ap­prox. 180 at­ten­dees and or­ga­nized ac­com­pa­ny­ing in­ter­views in Swedish media

  • Car­ried out six ex­ter­nal lec­tures/​talks, e.g. at Almedalsveckan, a big event in Swedish poli­tics, and the largest con­fer­ence for char­i­ties in Sweden

  • Or­ga­niz­ing reg­u­lar mee­tups and workshops

  • The stu­dent groups car­ried out a num­ber of events, in­clud­ing read­ing groups and ca­reer workshops

  • An EA co-liv­ing space was set up: Neurora

1.2. Pur­pose of this doc­u­ment

The pur­pose of this doc­u­ment is to ex­plic­itly out­line the plans of Effec­tive Altru­ism Swe­den (EAS) dur­ing 2018 in or­der to clar­ify our plans in­ter­nally and to en­able other to provide use­ful feed­back. I also aim to make ex­plicit as many cru­cial con­sid­er­a­tions as pos­si­ble. Th­ese are un­cer­tain­ties and ques­tions where the an­swers will be par­tic­u­larly im­pact­ful on the work we do. There are few state­ments in this doc­u­ments that we’re sure of, and so feed­back and crit­i­cism is greatly ap­pre­ci­ated.

1.3. What are the op­por­tu­ni­ties available to EA Swe­den?

In or­der to de­ter­mine what we ought to fo­cus on over the com­ing year, it’s use­ful to think about what the main op­por­tu­ni­ties for an EA or­ga­ni­za­tion to do good in Swe­den are. In short, we think that the (not mu­tu­ally ex­clu­sive) op­por­tu­ni­ties and re­lated re­sources are:

  • Com­mu­nity build­ing: This seems like a good op­por­tu­nity re­gard­less of where your EA group is lo­cated. There may be rea­sons to think that Swe­den is par­tic­u­larly good to do com­mu­nity-build­ing in, but I would not take these par­tic­u­larly se­ri­ously: Anec­do­tally, there is a fairly large num­ber of EAs from Swe­den rel­a­tive to the coun­try’s size, de­spite there hav­ing been lit­tle or­ga­niz­ing over the past few years. Within the x-risk com­mu­nity some of the most in­fluen­tial figures are Swedish, in­clud­ing An­ders Sand­berg, Nick Bostrom and Max Teg­mark are all Swedish. The num­ber of Giv­ing What We Can pledgers might be an­other mea­sure of how promis­ing com­mu­nity-build­ing in Swe­den is. In terms of pledgers per cap­ita, Swe­den ranks 7th (5th if out­liers with small pop­u­la­tions and one pledger are re­moved).

  • Startup com­mu­nity: Rel­a­tive to its size, Swe­den has a fairly large startup com­mu­nity, be­ing the place where e.g. Spo­tify, Ma­jong (cre­ators of Minecraft), King, Klarna and iZet­tle started. To provide an in­di­ca­tion re­gard­ing the size, $1.4 billion was in­vested into Stock­holm-based star­tups in 2016. In ad­di­tion there is an op­por­tu­nity in the high-im­pact startup space: 2017 saw the cre­ation of Norrsken Foun­da­tion in Stock­holm. The or­ga­ni­za­tion has an en­dow­ment of at least $50 mil­lion, “a strong be­lief in Effec­tive Altru­ism” and aims to cre­ate high-im­pact star­tups. Another in­di­ca­tion of this be­ing a promis­ing op­por­tu­nity is that Founders Pledge re­ceived an Open Philan­thropy Pro­ject grant to ex­pand to Ber­lin, Paris and Stock­holm, Swe­den.

  • Poli­tics: Peo­ple within the effec­tive al­tru­ism move­ment have typ­i­cally been timid about get­ting in­volved with poli­ti­cal is­sues, es­pe­cially with an ex­plicit as­so­ci­a­tion with effec­tive al­tru­ism. How­ever, there is a lot of po­ten­tial to do good if effec­tive al­tru­ism can be brought into the poli­ti­cal sphere in the right way. Given this con­text, there are ad­van­tages to try­ing out poli­ti­cal cam­paign­ing in a small coun­try: Firstly, small coun­tries, in­clud­ing Swe­den, have a tighter con­nec­tion be­tween the poli­ti­cal pro­cess and the peo­ple. This means that a poli­ti­cal cam­paign has a higher chance of suc­cess al­low­ing shorter feed­back loops. Se­condly, the risks are smaller. If a poli­ti­cal cam­paign ac­ci­den­tally causes rep­u­ta­tional dam­age to effec­tive al­tru­ism (e.g. the ideas come to be seen as weird), dam­age may be con­tained to the Swedish-speak­ing world as op­posed to the English-speak­ing world. Thirdly, it is eas­ier to pro­mote the im­ple­men­ta­tion of a policy if it has already been im­ple­mented some­where else. If it is less costly to im­ple­ment new policy in smaller coun­tries, then it is ad­van­ta­geous to let smaller coun­tries be this “some­where else”.

  • X-risk: Globally speak­ing, Swe­den is likely a small con­trib­u­tor to global catas­trophic risks. How­ever, there are rea­sons to be­lieve the coun­try might be a use­ful place to do work in the area. As men­tioned above, there are sub­stan­tial num­bers of x-risk re­searchers with a con­nec­tion to Swe­den, in­clud­ing Nick Bostrom, Max Teg­mark, An­ders Sand­berg and Olle Häg­gström. There are also a num­ber of Swedish or­ga­ni­za­tions and in­sti­tu­tions that provide ca­pac­ity: the In­sti­tute for Fu­ture Stud­ies and the Global Challenges Foun­da­tion.

2. Goals for the com­ing year

The ac­tivi­ties that EAS car­ries out can lead to a num­ber of benefits. Broadly defined, these seem to fit into the cat­e­gories be­low:

  • Ca­pac­ity-build­ing: Ca­pac­ity-build­ing is about lay­ing the ground­work for fu­ture im­pact. It can roughly be di­vided into com­mu­nity-build­ing and build­ing or­ga­ni­za­tional ca­pac­ity. Com­mu­nity-build­ing is about fos­ter­ing a com­mu­nity of peo­ple with a com­mit­ment to and ca­pa­ble of im­prov­ing the world effec­tively, who also have an un­der­stand­ing of how to do so. A cru­cial part of this com­mu­nity-build­ing will be to make sure that we are a di­verse and in­clu­sive com­mu­nity. Or­ga­ni­za­tional ca­pac­ity is about the or­ga­ni­za­tion, and not just the peo­ple within it, be­ing able to do good. This in­cludes the or­ga­ni­za­tion be­ing good at ac­ti­vat­ing peo­ple within the com­mu­nity, hav­ing full-time staff, hav­ing pro­cesses to sup­port its work etc.

  • Ex­plo­ra­tion value: Swe­den is a small part of the global EA move­ment. If the work EAS does can help oth­ers do good bet­ter in the fu­ture, that’s very valuable. This can be by learn­ing how to bet­ter build a com­mu­nity or carry out a spe­cific kind of pro­ject.

  • Direct good: The work we do can also have di­rect benefits to the world. Ex­am­ples of this would be through af­fect­ing policy, helping an or­ga­ni­za­tion work more effec­tively or through dona­tions be­cause of our com­mu­nity-build­ing work.

In light of the above, we will fo­cus on two types of ac­tivi­ties: ca­pac­ity-build­ing ac­tivi­ties (sec­tion 3) and di­rect work (sec­tion 4). The ca­pac­ity-build­ing ac­tivi­ties are those that pri­mar­ily fo­cus on ca­pac­ity-build­ing, but they may nonethe­less lead to some di­rect good. The di­rect work, similarly, pri­mar­ily aims at do­ing di­rect good, but also pro­vides ex­plo­ra­tion value and the build­ing of ca­pac­ity.

We have also defined mea­surables to track our im­pact. The most im­por­tant mea­surables are data con­cern­ing the fun­nel of 80k, in ad­di­tion to dona­tions and pledges through Giv­ing What We Can/​EA Funds.

3. Ca­pac­ity-build­ing activities

Below is a brief sum­mary of our cur­rent plans re­lated to ca­pac­ity-build­ing. As men­tioned pre­vi­ously I cat­e­gorise these in terms of ac­tivi­ties chiefly aimed at com­mu­nity-build­ing and those aimed at build­ing or­ga­ni­za­tional ca­pac­ity.

3.1. Com­mu­nity-building

3.1.1. Get­ting peo­ple to pos­i­tively im­pact the world

The ul­ti­mate goal of com­mu­nity-build­ing within effec­tive al­tru­ism is not to grow effec­tive al­tru­ism: it is to pos­i­tively im­pact the world. We want ca­pa­ble peo­ple with an un­der­stand­ing of how to im­prove the world as much as pos­si­ble to spend a sig­nifi­cant amount of their re­sources on im­prov­ing the world. This is a big ask. And so for most peo­ple, it is a step­wise pro­cess from hear­ing about effec­tive al­tru­ism for the first time to ded­i­cat­ing a sig­nifi­cant amount of your time and money to im­prov­ing the world effec­tively. This step-wise na­ture needs to be taken into ac­count in com­mu­nity-build­ing. Depend­ing on where they are, peo­ple will have differ­ent needs and differ­ent ac­tivi­ties will be needed to cater to these differ­ent needs.

A use­ful anal­ogy to think­ing about this is a sales fun­nel (see be­low):

Illus­tra­tion 1: A sales fun­nel for effec­tive al­tru­ism.

Over the past year, EAS has been shift­ing our efforts more to­wards the lat­ter parts of the sales fun­nel. Over the com­ing year, we will con­tinue to put a lot of our efforts in that area. I think this is a good idea since I tend to find that aware­ness and knowl­edge of effec­tive al­tru­ism is not enough for peo­ple to take sig­nifi­cant ac­tion on the ideas. For most peo­ple, in my ex­pe­rience, a so­cial con­text and nudges to­wards ac­tion are needed. My sense is that this is the case even for peo­ple who find that they agreed with a lot of the core effec­tive al­tru­ism prin­ci­ples right away. How­ever, I ac­knowl­edge that given the im­por­tance of the ques­tion, this rea­son­ing is likely not enough. There­fore, we are con­sid­er­ing putting more en­ergy into an­swer­ing the ques­tion of how best to get peo­ple from be­ing aware of the ideas to act­ing on them.

3.1.2. Ex­am­ples of planned activities

Most of our events and ac­tivi­ties will be ex­plic­itly put in one of the lev­els of the sales fun­nel. That way, we can adapt our events to the rele­vant tar­get au­di­ence. For ex­am­ple, a Level 3-event can benefit from be­ing in a some­what pri­vate space, e.g. some­one’s home. How­ever, if you have only just now heard about effec­tive al­tru­ism, that might feel too per­sonal. Below, I’ll list some of the ac­tivi­ties that we have planned for the differ­ent lev­els:

Level 1: Awareness

There are sev­eral ways in which we plan to make more peo­ple aware of effec­tive al­tru­ism and EAS. The first is word-of-mouth or per­sonal con­tacts, which was the most com­mon way for peo­ple to find out about EA in the 2017 Effec­tive Altru­ism Sur­vey (15.5% of re­spon­dents). I find that a lot of peo­ple within the EA move­ment find it difficult to talk about effec­tive al­tru­ism with friends and fam­ily. There­fore, we have already and are plan­ning to keep run­ning events about how to do so.

In ad­di­tion to the above, we will do more con­ven­tional out­reach, pri­mar­ily lec­tures, talks and one-on-one meet­ings for in­di­vi­d­u­als we think par­tic­u­larly re­cep­tive and im­pact­ful. We will also spend some re­sources on me­dia out­reach, but pri­mar­ily in re­la­tion to our di­rect work.

Level 2: Knowledge

Our main ac­tivi­ties on this level will be a few re­cur­ring ac­tivi­ties, listed be­low:

  • Socials

  • In­tro­duc­tion to effec­tive al­tru­ism (2-4 hour work­shop)

  • Giv­ing games

  • Read­ing groups

Level 3: Action

Ac­tivi­ties un­der this level aim to get peo­ple to take ac­tion on their knowl­edge of effec­tive al­tru­ism. In short, these ac­tions can be donat­ing money, chang­ing your ca­reer or do­ing di­rect work. By di­rect work, I mean for ex­am­ple helping or­ga­nize a lo­cal group, but also run­ning spe­cific pro­jects such as those dis­cussed in sec­tion 4.

Un­der this level, we plan to carry out some of the fol­low­ing ac­tivi­ties:

  • Ca­reer workshops

  • Work­shops on other ac­tion­able top­ics, e.g. how to talk to your friends about effec­tive al­tru­ism and how to make your­self live in ac­cor­dance with your values

  • One-on-one con­ver­sa­tions with par­tic­u­larly high-po­ten­tial members

  • Pro­vid­ing easy ways for peo­ple to get in­volved in di­rect work

3.1.3. Diver­sity and inclusiveness

Lack of di­ver­sity is a sig­nifi­cant risk to the effec­tive al­tru­ism move­ment. In short, a lack of di­ver­sity means that we are los­ing out on a lot of tal­ent. Fur­ther­more, there are rea­sons to think that a com­mu­nity not be­ing di­verse is self-re­in­forc­ing: peo­ple tend to know and be more com­fortable get­ting to know peo­ple who are similar to them­selves. As an ex­am­ple of how triv­ial this type of effect can be: when I lived in the UK, I was far more likely to befriend some­one if they were Swedish. And so, if a group were to be­come pre­dom­i­nantly stu­dents, they would tend to bring in other stu­dents and non-stu­dents would feel less at home in the set­ting. Be­cause of the po­ten­tial loss in value and the self-re­in­forc­ing na­ture of ho­mo­gene­ity in a group, it is im­por­tant to ex­plic­itly put effort into pro­mot­ing di­ver­sity and in­clu­sive­ness and it is im­por­tant to do that at an early stage to avoid cal­ci­fi­ca­tion of a cer­tain de­mo­graphic.

We have put work into this is­sue over the past year and cur­rently the EA Swe­den board is 46 fe­male. How­ever, the ma­jor­ity of our mem­bers are still white men. Given this, more work in this area is war­ranted. We hope to be pi­o­neers within the EA com­mu­nity in terms of di­ver­sity.

Below is a short list of the types of ac­tions that we will and have taken in the past to im­prove our di­ver­sity and in­clu­sive­ness:

  • Put ex­tra effort in the “Aware­ness”-stage on groups that are cur­rently un­der­rep­re­sented and would make valuable con­tri­bu­tions to the com­mu­nity. This can in­clude putting more en­ergy into con­tact­ing those or­ga­ni­za­tions whose mem­bers may im­prove our di­ver­sity and in­clu­sive­ness.

  • Pro­ject the com­mu­nity we want to be. That is, make sure that the peo­ple who are visi­ble within the com­mu­nity are di­verse in terms of back­ground, eth­nic­ity and gen­der. This can be done when post­ing pic­tures on­line, but also when choos­ing who to put on a stage. Of course, this needs to be done with a lot of sen­si­tivity and tact.

  • Ex­plic­itly dis­cuss the im­por­tance of di­ver­sity/​in­clu­sive­ness and ways to im­prove on it. As an ex­am­ple, we ran a work­shop on this is­sue a few months be­fore set­ting up EA Swe­den.

  • Show­ing ev­ery­one that they are wel­come. If, at some event, we sus­pect that some­one might not feel as wel­come as ev­ery­one else, we will put ex­tra effort into chang­ing that.

3.1.4. Other activities

  • Or­ga­niz­ing EAGx Scan­d­i­navia: We have been in dis­cus­sions with EA Nor­way about putting on an EAGx Scan­d­i­navia and are cur­rently fairly likely to go through with the pro­ject. The thought is that this pro­vides a use­ful ways to in­volve peo­ple in an EA pro­ject and to fur­ther grow the com­mu­nity.

  • Seed­ing new groups: So far, EA Swe­den has largely been syn­ony­mous with EA Stock­holm. This is some­thing we will put some en­ergy into chang­ing over the com­ing year, pri­mar­ily by con­nect­ing in­di­vi­d­u­als who are ex­cited about start­ing lo­cal groups with each other and giv­ing them nudges and sup­port in set­ting up.

3.2. Or­ga­ni­za­tional capacity

Or­ga­ni­za­tional ca­pac­ity is about the or­ga­ni­za­tion, and not just the peo­ple within it, be­ing able to do good. This in­cludes the or­ga­ni­za­tion be­ing good at com­mu­nity-build­ing and con­duct­ing di­rect work. Or­ga­ni­za­tional ca­pac­ity can be con­tributed to by hav­ing a valuable brand, hav­ing the right cul­ture, hav­ing sys­tems and pro­cesses for com­mu­nity-build­ing, hav­ing full-time staff and hav­ing the ca­pac­ity to sup­port mem­bers in do­ing di­rect work.

A big part of the plan for the year is to grow the or­ga­ni­za­tional ca­pac­ity of EAS. This will be done through a num­ber of ac­tivi­ties, rang­ing from:

  • Set­ting/​im­prov­ing on up our digi­tal in­fras­truc­ture: web­site, track­ing of mem­bers, col­lect­ing data from events etc.

  • Set­ting up pro­cesses and poli­cies re­lated to hav­ing staff, in­clud­ing payroll

  • Learn­ing more sys­tem­at­i­cally how we can best carry out our com­mu­nity-building

  • Pro­vid­ing mem­bers with ways to get in­volved in di­rect work

  • In­creas­ing the num­ber of effec­tive al­tru­ism lead­ers, ca­pa­ble of lead­ing lo­cal groups and lec­tur­ing on effec­tive altruism

The main out­stand­ing ques­tion re­lated to or­ga­ni­za­tional ca­pac­ity whether we should fo­cus on grow­ing the num­ber of staff. Cur­rently, the plan is to put effort into con­sid­er­ing this ques­tion later dur­ing the spring. How­ever, out­side per­spec­tives on this ques­tion would be very valuable. Cru­cial ques­tions re­lated to this in­clude:

  • What would be the value of an­other EAF-, CEA- or RC-style or­ga­ni­za­tion?

  • If such an or­ga­ni­za­tion would be valuable, what ought the role of EAS in the global EA ecosys­tem be?

  • If the or­ga­ni­za­tion ought to ex­pand, should only effec­tive al­tru­ist donors be con­sid­ered, or also ad­di­tional sources of fund­ing?

3.3. Cru­cial con­sid­er­a­tions re­gard­ing ca­pac­ity-building

Below are ques­tions where we cur­rently are both very un­cer­tain and where the an­swer seems to be par­tic­u­larly im­por­tant to how our work should be car­ried out:

  • What is the best way to get more peo­ple to act upon the prin­ci­ples of effec­tive al­tru­ism? On one ex­treme, the best strat­egy is sim­ply ex­pos­ing a lot of peo­ple to a short pitch and on the other is sim­ply fo­cus­ing on the lat­ter parts of the sales fun­nel, let­ting aware­ness of EA mainly come through on­line chan­nels and word-of-mouth. EAS cur­rent an­swer is that a com­bi­na­tion of the two is needed.

  • Should we at­tempt to grow our num­ber of staff?

  • What are the best meth­ods for us to in­volve our mem­bers in di­rect work?

  • Should we put on EAGx Scan­d­i­navia? We think the most cru­cial ques­tions to an­swer are: How valuable is it? Is the com­mu­nity large enough for a con­fer­ence?

4. Direct work

In ad­di­tion to fo­cus­ing on com­mu­nity-build­ing, we will also fo­cus on di­rect work, partly due to its im­por­tance to com­mu­nity-build­ing.

The cur­rent plan is to fo­cus our efforts on a pro­ject re­lated to x-risk, both due to the im­por­tance of the cause, the re­sources available in car­ry­ing out such a pro­ject (see sec­tion 1.3) and that the pro­posed pro­ject seems promis­ing. Below, the planned pro­ject and pro­jects that were con­sid­ered but de­cided against, are out­lined.

4.1. Rep­re­sen­ta­tion of fu­ture generations

4.1.1. What’s the idea?

Fu­ture gen­er­a­tions sel­dom make it to the vot­ing sta­tion. Be­cause of this, their in­ter­ests are poorly rep­re­sented in cur­rent poli­ti­cal sys­tems. By fu­ture gen­er­a­tions here, I mean peo­ple other be­ings who have yet to come into ex­is­tence. The aim of the pro­ject (“RFG” be­low) is to bet­ter rep­re­sent fu­ture gen­er­a­tions in the Swedish poli­ti­cal sys­tem. The hope is that this rep­re­sen­ta­tion can af­fect policy which in turn de­creases e.g. x-risks. The idea repli­cates work that has been done in the pri­mar­ily UK over the past year by among oth­ers FUSE and CSER, which for ex­am­ple lead to the for­ma­tion of the All-Party Par­li­a­men­tary Group on Fu­ture Gen­er­a­tions.

The cur­rent strat­egy within work on x-risk has fo­cused mainly on re­search­ing the risks and rais­ing aware­ness. How­ever, in or­der for this re­search to de­crease x-risks, there needs to be a way for the re­search to af­fect be­havi­our of rele­vant ac­tors. So far, this has mainly been done by rais­ing aware­ness among those who are likely to de­velop the po­ten­tially harm­ful tech­nol­ogy—i.e. re­searchers and com­pa­nies. Another ap­proach is to do this by af­fect­ing policy. The aim of this pro­ject is to cre­ate av­enues for x-risk re­search to do so.

There are many mechanism that could rep­re­sent fu­ture gen­er­a­tions. A forth­com­ing pa­per, writ­ten by mem­bers of CSER and FUSE, pro­vides a good in­tro­duc­tion to the topic, in­clud­ing differ­ent ways in which coun­tries have rep­re­sented fu­ture gen­er­a­tions and the suc­cesses of these var­i­ous mechanisms. The pa­per ends with con­crete sug­ges­tions con­cern­ing how fu­ture gen­er­a­tions ought to re­ceive rep­re­sen­ta­tion in a UK con­text. In short, the mechanisms can range from ex­plic­itly grant­ing rights to fu­ture gen­er­a­tions, re­quiring that fu­ture gen­er­a­tions be taken into ac­count in the leg­is­la­tive pro­cess, hav­ing a body with pow­ers to veto leg­is­la­tion or groups where par­li­a­men­tar­i­ans can dis­cuss is­sues con­nected to fu­ture gen­er­a­tions. One con­clu­sion re­gard­ing the differ­ent at­tempts to rep­re­sent fu­ture gen­er­a­tions drawn in the pa­per is that soft power-mechanisms are prefer­able. Mechanisms where a body e.g. re­ceives the power to veto leg­is­la­tion do not last very long. We are cur­rently not sure what modes of rep­re­sen­ta­tion are suit­able for the Swedish con­text. Work­ing on this ques­tion will be one of the first tasks of the pro­ject.

4.1.2. Rep­re­sen­ta­tion of fu­ture gen­er­a­tions in Sweden

There has been at­tempts to do some­thing similar in Swedish gov­ern­ment over the past few decades. Below is a short sum­mary of some shal­low re­search into the topic.

Most of the pre­vi­ous at­tempts seem to have con­cerned cre­at­ing a body that pro­vides the gov­ern­ment with strate­gic think­ing, helps co­or­di­nate be­tween de­part­ments in the gov­ern­ment and fo­cuses on ques­tions such as the en­vi­ron­ment, the fu­ture of work and global co­op­er­a­tion. From Oc­to­ber 2014 to May 2016, Kristina Pers­son served as Minister for Strate­gic Devel­op­ment and presided over the Coun­cil on Strate­gic and Fu­ture Is­sues. The Coun­cil pub­lished three re­ports in June 2016 on The Work of the Fu­ture, The En­vi­ron­men­tal Tran­si­tion and Swedish Com­pet­i­tive­ness, in ad­di­tion to Global Co­op­er­a­tion. Both the Coun­cil and the Minister for Strate­gic Devel­op­ment were dis­banded in May 2016, dur­ing a larger reshuffle of the cab­i­net.

Cur­rently, the in­sti­tu­tions or mechanisms that most re­sem­ble a rep­re­sen­ta­tion of fu­ture gen­er­a­tions are the Fu­ture Com­mis­sion—which re­placed the Coun­cil on Strate­gic and Fu­ture Is­sues and is part of the gov­ern­ment—and the Agenda 2030-Del­e­ga­tion. I have so far been un­suc­cess­ful in find­ing any in­for­ma­tion as to what the Fu­ture Com­mis­sion has done since its start in June 2016, while the fo­cus of the Agenda 2030-del­e­ga­tion has been to de­ter­mine how the Sus­tain­able Devel­op­ment Goals should be im­ple­mented in Swedish poli­tics.

In sum­mary, un­less the Fu­ture Com­mis­sion is in fact ac­tive and will last a sig­nifi­cant amount of time, there seems to be a need for a mechanism to rep­re­sent fu­ture gen­er­a­tions. Fur­ther­more, given a cur­sory glance at the pre­vi­ous at­tempts sug­gests that most of them have suffered from be­com­ing poli­ti­cized. Most of the pre­vi­ous bod­ies fo­cus­ing on the fu­ture seem to have been set up by gov­ern­ments rather than be­ing in­de­pen­dent. This needs to be taken into ac­count in de­ter­min­ing what type of mechanism for the rep­re­sen­ta­tion of fu­ture gen­er­a­tions should be ad­vo­cated for.

4.1.3. Po­ten­tial benefits

The po­ten­tial benefits of this pro­ject would be:

  • Direct im­pact: If the pro­ject is suc­cess­ful, it has the po­ten­tial to im­prove the lot of fu­ture gen­er­a­tions. One could be skep­ti­cal of the im­pact here see­ing as Swe­den is not a par­tic­u­larly large coun­try. How­ever, the smaller size likely means that there is a higher chance of suc­cess. In ad­di­tion, policy in one coun­try af­fects the policy of other coun­tries. It is of­ten eas­ier to im­ple­ment a new policy if it has already been done in an­other coun­try. This sug­gests that a good strat­egy is to first im­ple­ment new poli­cies in small coun­tries and lev­er­age that in other coun­tries.

  • Com­mu­nity-build­ing: The pro­ject would be a good op­por­tu­nity to heav­ily in­volve a small group of in­di­vi­d­u­als, mak­ing them more knowl­edge­able and more ca­pa­ble of car­ry­ing out im­por­tant work in the fu­ture. In ad­di­tion to this, the pro­ject would be use­ful to other as­pects of com­mu­nity-build­ing, both in be­ing able to reach out to a wider au­di­ence, and be able to show that effec­tive al­tru­ism is more than a group of peo­ple dis­cussing in­ter­est­ing ideas.

  • Ex­plo­ra­tion value: Po­ten­tially the largest im­pact from car­ry­ing out this pro­ject is learn­ing about how to do pro­jects of this kind in the fu­ture. That is, both pro­jects con­cern­ing bring­ing re­search con­cern­ing x-risk to bear on policy and poli­ti­cal pro­cesses, but also the ideas of effec­tive al­tru­ism more gen­er­ally. Many have wor­ried that bring­ing effec­tive al­tru­ism to bear on poli­tics poses a risk of effec­tive al­tru­ism be­ing per­ceived as par­ti­san. How­ever, Swe­den can act as a kind of test­ing ground here. If effec­tive al­tru­ism does be­come per­ceived as par­ti­san and that turns out to be costly, it would be far less costly than if that was the case in a larger or an English-speak­ing coun­try. This is not to say that the risk of be­ing per­ceived as par­ti­san should not be miti­gated, but merely that a nega­tive out­come would be less costly.

4.1.4. Im­pact mechanism and failure modes

A use­ful way to iden­tify how the pro­ject might fail and find­ing ways to miti­gate those risks is to think about the im­pact mechanism of the pro­ject, iden­tify ways for it to go wrong and how to miti­gate those risks.

In or­der for the pro­ject to have a di­rect pos­i­tive im­pact the fol­low­ing needs to oc­cur: (i) some mechanism of rep­re­sen­ta­tion of fu­ture gen­er­a­tions needs to be im­ple­mented and sus­tained, (ii) that mechanism must af­fect policy in some way and (iii) it af­fects policy pos­i­tively. The failure modes for this kind of pro­ject are the in­verse of the steps in the im­pact mechanism. There are there­fore three failure modes:

  1. No mechanism for the rep­re­sen­ta­tion of fu­ture gen­er­a­tions is im­ple­mented or al­ter­na­tively, the mechanism does not last a long time

  2. The im­ple­mented mechanism does not af­fect policy

  3. The im­ple­mented mechanism af­fects policy negatively

For a graph­i­cal rep­re­sen­ta­tion, see illus­tra­tion 2 be­low.

Illus­tra­tion 2. Im­pact mechanism and failure modes of the project

Failure mode 1 re­ceives the most at­ten­tion from the forth­com­ing pa­per out of CSER and FUSE. They fo­cus on a num­ber of fac­tors that might lead to this failure mode. In par­tic­u­lar, they ar­gue that modes of rep­re­sen­ta­tion that use in­for­mal power are prefer­able, mainly be­cause they last longer. Ex­am­ples of for­mal pow­ers in­clude be­ing able to veto leg­is­la­tion. When these types of mechanisms have been im­ple­mented, they have of­ten been re­moved within a short amount of time, pre­sum­ably be­cause the in­sti­tu­tion eas­ily be­comes poli­ti­cized. Ex­em­plify­ing this, the Is­raeli par­li­a­ment es­tab­lished a Com­mis­sion for Fu­ture Gen­er­a­tions in 2001, with scope rang­ing across 12 policy ar­eas, the abil­ity to ini­ti­ate bills and the abil­ity to veto leg­is­la­tion that did not com­ply with the in­ter­ests of fu­ture gen­er­a­tions. The body was given a five-year man­date which was not re­newed, offi­cially due to bud­getary rea­sons, but there are in­di­ca­tions that par­li­a­men­tar­i­ans felt the body had too much au­thor­ity.

In ad­di­tion to the above, the Swedish con­text needs to be taken into ac­count. There is a Swedish par­li­a­men­tary elec­tion com­ing up in Septem­ber. This means that the me­dia and poli­ti­ci­ans will have their at­ten­tion fixed on more near-term poli­ti­cal is­sues: taxes, im­mi­gra­tion, school­ing etc. The elec­tion also means that many par­li­a­men­tar­i­ans will be chang­ing jobs in Septem­ber. There­fore, re­sources be­ing put into try­ing to im­ple­ment mechanisms for the rep­re­sen­ta­tion of fu­ture gen­er­a­tions be­fore Septem­ber risk be­ing wasted. Ad­di­tion­ally, pre­vi­ous at­tempts at es­tab­lish­ing mechanisms for the rep­re­sen­ta­tion of fu­ture gen­er­a­tions ought to be used as case stud­ies to learn from.

Failure mode 2 seems like a sub­stan­tial risk: if the mechanism uti­lizes soft power, it seems that it would be less ca­pa­ble of af­fect­ing policy. In ad­di­tion, if the x-risk re­search field is not ma­ture enough to make spe­cific policy recom­men­da­tions, there is a risk that the mechanism has no policy to dis­cuss or put for­ward.

Failure mode 3 is surely some­thing to be avoided as it not only poses risks hav­ing di­rect nega­tive effects, but it might also lead to se­vere rep­u­ta­tional costs. This failure mode could po­ten­tially arise in one of two ways: the recom­men­da­tions of x-risk re­searchers turn out to be harm­ful or the mechanism is co-opted. Both of these risks hinge on the ques­tion of whether the x-risk re­search field is ma­ture enough to come up with use­ful recom­men­da­tions.

To avoid these failure modes the pro­ject ought to:

  • Con­cen­trate re­sources into pub­lic ad­vo­cacy af­ter the elec­tion or a sig­nifi­cant amount of time be­fore the election

  • In­clude high-pro­file aca­demics who can lend cred­i­bil­ity to the idea

  • En­sure that there is a sus­tain­able con­nec­tion be­tween the mechanism and the x-risk re­search com­mu­nity. This has for ex­am­ple been done in re­gards to the APPG by CSER pro­vid­ing the sec­re­tariat for the group

  • Learn from peo­ple with sig­nifi­cant ex­pe­rience of cam­paign­ing and the poli­ti­cal pro­cess in Sweden

  • Learn from pre­vi­ous mechanisms to rep­re­sent fu­ture gen­er­a­tions in Swedish politics

4.1.5. Do we have the re­sources to carry out this pro­ject?

This pro­ject will re­quire a lot of ca­pa­bil­ities, of which we think we have some. Below is a list of needed re­sources and our ac­cess to these re­sources:

  • Enough re­search within the x-risk space to war­rant the mechanism: A mechanism for the rep­re­sen­ta­tion of fu­ture gen­er­a­tions can only provide di­rect value if we know what policy it ought to pro­mote or speak out against. To fur­ther in­ves­ti­gate this ques­tion we need to dis­cuss the is­sue with more re­searchers in the x-risk field.

  • Re­sources to carry out the pro­ject: To carry out the pro­ject, we will need the right work­force. Between me work­ing full-time and in­ter­ested vol­un­teers, we feel con­fi­dent we will be able to carry it out (al­though with a high chance of failure).

  • An un­der­stand­ing of and a rele­vant net­work within Swedish poli­tics: Car­ry­ing out the pro­ject suc­cess­fully re­quires an un­der­stand­ing of how the Swedish poli­ti­cal sys­tem works and a re­lated net­work. We likely have ac­cess to this net­work.

  • A net­work of x-risk re­searchers and aca­demics: We have ac­cess to a net­work of x-risk re­searchers through effec­tive al­tru­ism.

4.1.6. Plan outline

The out­line of the plan is as fol­lows: Be­fore the elec­tion (Septem­ber 2018) the ground­works for a cam­paign are cre­ated. After the elec­tions, a pub­lic cam­paign will be rol­led out. The na­ture of this cam­paign is yet to be de­ter­mined.

The ac­tivi­ties to be car­ried out as part of set­ting the ground­works in­clude:

  • Gather feed­back on the pro­ject idea from rele­vant ex­perts (within the EA com­mu­nity, but mainly those with ex­per­tise in Swedish poli­tics)

  • Assess in­ter­est in the topic from Swedish politicians

  • Put to­gether a work­ing group of high-im­pact in­di­vi­d­u­als who can work on and sup­port the project

  • Con­duct an anal­y­sis of the his­tory of rep­re­sen­ta­tion of fu­ture gen­er­a­tions in Swedish politics

  • Develop a strat­egy for the rep­re­sen­ta­tion of fu­ture gen­er­a­tions in a Swedish con­text. This in­cludes an­swer­ing ques­tions such as: What mechanisms for the rep­re­sen­ta­tion of fu­ture gen­er­a­tions are suit­able? Should these mechanisms be im­ple­mented in a spe­cific or­der? What are some spe­cific poli­cies that these mechanisms could ex­am­ine?

  • Develop a short-list of the is­sues and poli­cies that we would like a mechanism to rep­re­sent fu­ture gen­er­a­tions to address

  • Build in­ter­est in the is­sue, by putting on a num­ber of workshops

  • Develop a con­crete plan of ac­tion for af­ter the election

  • Pre­pare for launch of cam­paign post-election

4.2. Some pro­jects we have de­cided against for now

A lot of pro­ject ideas have been con­sid­ered. Below is a short ex­pla­na­tion of each of them. If you’re keen, they’re yours for the tak­ing!

4.2.1. EA Fact Check­ers /​ The EA Re­pro­ducibil­ity Project

To do good in the world, it is im­por­tant to have ac­cu­rate be­liefs about the world: facts are im­por­tant. Be­cause of this it is par­tic­u­larly im­por­tant to up­hold high epistemic stan­dards. One way in which this could be done would be to have an in­de­pen­dent or­ga­ni­za­tion that checks on claims made within the effec­tive al­tru­ism com­mu­nity.

One way to im­ple­ment this idea—the ver­sion I call EA Fact Check­ers—would be to fo­cus in on facts or memes that are com­monly cited within the effec­tive al­tru­ism com­mu­nity. Ex­am­ples of such claims are: “donat­ing to char­ity makes you hap­pier”, “some char­i­ties are 1000x more effec­tive than oth­ers” and the graph from Ord (2013) show­ing the differ­ence in cost-effec­tive­ness be­tween differ­ent in­ter­ven­tions. Most of these claims are likely not acted upon di­rectly, and so fact-check­ing them would not have a di­rect pos­i­tive effect on the world. It would likely not lead to money be­ing al­lo­cated more effec­tively. How­ever, there likely would be pos­i­tive effects in up­hold­ing epistemic stan­dards within the com­mu­nity.

Another way to im­ple­ment this idea—EA Re­pro­ducibil­ity Pro­ject—would be to fo­cus in on facts and claims on which peo­ple base de­ci­sions: e.g. recom­men­da­tions on char­ity dona­tions. The epistemic stan­dards put on re­search within the effec­tive al­tru­ism com­mu­nity is likely high. How­ever, my im­pres­sion is that the ma­jor­ity of check­ing on claims is done within EA or­ga­ni­za­tions them­selves. Check­ing from out­side of or­ga­ni­za­tions is pri­mar­ily in­for­mal: ask­ing peo­ple one knows for feed­back, pub­li­cly and trans­par­ently com­mu­ni­cat­ing recom­men­da­tions through e.g. the EA Fo­rum. What does not ex­ist is a sys­tem­atic way in which recom­men­da­tions are ex­ter­nally vet­ted. How­ever, this might be needed as the EA move­ment grows larger.

An ini­tial sketch of how to carry out this EA Re­pro­ducibil­ity Pro­ject is as fol­lows: Ini­tially, one would fo­cus on the sim­plest, most me­chan­i­cal way of vet­ting recom­men­da­tions, where there is lit­tle room for in­ter­pre­ta­tion. This would likely be by ran­domly choos­ing a recom­men­da­tion made by an EA or­ga­ni­za­tion and check­ing whether its quan­ti­ta­tive mod­els in­clude mis­takes or the stud­ies cited sup­ported the claims made. If the first few iter­a­tions proved suc­cess­ful, the pro­ject could ex­pand into ei­ther fo­cus­ing on a more of the re­search within the EA com­mu­nity (e.g. vet­ting cita­tions in blog post from EA or­ga­ni­za­tions) or into car­ry­ing out more sub­stan­tive re­pro­duc­tions (e.g. by look­ing at whether the re­search left out im­por­tant stud­ies or made du­bi­ous ar­gu­ments). Fur­ther down the line, this in­de­pen­dent or­ga­ni­za­tion or pro­ject could be a part of set­ting com­mon epistemic stan­dards or set­ting up sys­tems of effec­tive al­tru­ism peer-re­view.

We de­cided against work­ing on this pro­ject since we do not see our­selves as hav­ing a com­pet­i­tive ad­van­tage in car­ry­ing it out. How­ever, I think it could be quite valuable, both in up­hold­ing epistemic stan­dards and in be­com­ing a part of the effec­tive al­tru­ism re­searcher pipeline.

4.2.2. Aid effec­tive­ness-rank­ing of poli­ti­cal parties

A sim­ple way to po­ten­tially af­fect Swedish aid policy would be to con­duct a rank­ing of poli­ti­cal par­ties in terms of the qual­ity of their aid poli­cies

Similar pro­jects have been run in the past. The Effec­tive Altru­ism Foun­da­tion did a similar rank­ing in con­nec­tion with the Swiss elec­tion, but fo­cus­ing on the poli­tics of spe­cific can­di­dates, not par­ties, and look­ing at all policy, not just aid policy. EA NTNU in Nor­way did a similar pro­ject dur­ing the re­cent Nor­we­gian elec­tions, which prompted the Nor­we­gian Aid Minister to dis­cuss the sug­ges­tions at one of their events. In both of these cases, all of the work was done by the lo­cal group it­self. This is likely good for e.g. skill-build­ing. How­ever, to af­fect policy, there likely needs to be a greater in­volve­ment of aca­demics with cred­i­bil­ity, both to make sure the con­clu­sions are more likely to be cor­rect and to lend the pro­ject cred­i­bil­ity.

Another no­table similar pro­ject, and which this type of pro­ject would likely take a lot of in­put from, is the Cen­tre for Global Devel­op­ment’s Com­mit­ment to Devel­op­ment In­dex, a rank­ing of coun­tries in terms of their con­tri­bu­tion to global de­vel­op­ment.

The po­ten­tial benefits of car­ry­ing out such a pro­ject would be:

  • Ex­plo­ra­tion value in learn­ing about how effec­tive al­tru­ism can be ap­plied to the poli­ti­cal sphere and po­ten­tially pro­vid­ing a blueprint for car­ry­ing out the same pro­ject in other countries

  • Direct im­pact through po­ten­tially af­fect­ing policy. Pre­sum­ably, the biggest ma­jor­ity of the po­ten­tial im­pact would not be by sway­ing vot­ers to vote for par­ties with bet­ter poli­cies, but rather through get­ting poli­ti­ci­ans to pay at­ten­tion to the is­sue. This would be ac­com­plished both through them be­liev­ing that the is­sue will sway vot­ers and by sim­ply the pro­ject ex­pos­ing them to con­crete policy pro­pos­als.

  • Ca­pac­ity-build­ing, es­pe­cially by pro­vid­ing a good op­por­tu­nity to en­gage high-pro­file aca­demics about the ideas of effec­tive altruism

4.3. Cru­cial con­sid­er­a­tions re­gard­ing di­rect work

Below are ques­tions where we cur­rently are both very un­cer­tain and where the an­swer seems to be par­tic­u­larly im­por­tant to how or if the pro­ject should be car­ried out:

  • Re­gard­ing the rep­re­sen­ta­tion of fu­ture gen­er­a­tions pro­ject (RFG), is there re­search within the x-risk field that could be im­ple­mented through the pro­posed mechanism?

  • Re­gard­ing RFG, should this kind of cam­paign for the rep­re­sen­ta­tion of fu­ture gen­er­a­tions be ex­plic­itly linked to effec­tive al­tru­ism?

  • Re­gard­ing RFG, how can the mis­takes of pre­vi­ous similar at­tempts in Swe­den be avoided?

  • Re­gard­ing RFG, should the roll-out of the pro­ject wait un­til af­ter the elec­tion? Should e.g. a dec­la­ra­tion with sig­na­tures be made dur­ing the elec­tion cam­paign?

  • Given the above, ought we to fo­cus on RFG?