weeatquince

• weeatquince 13 Apr 2019 13:33 UTC

  8 points

  on: An­nounc­ing EA Hub 2.0

  This is amaz­ing. Great work for ev­ery­one who in­putted. Was think­ing that a pos­si­ble fu­ture fea­tures (al­though per­haps not a pri­or­ity) would be in­te­gra­tion to the EA funds dona­tion track­ing and maybe LinkedIn pro­file data.

• weeatquince 9 Apr 2019 14:01 UTC

  9 points

  on: Can my film­mak­ing/​song­writ­ing skills be used more effec­tively in EA?

  Your videos are great.

  I am sure there is space for con­tent cre­ators to be hav­ing a pow­er­ful im­pact on the world. Not en­tirely sure how but I did want to flag that the Long Term Fu­ture EA Fund has just given a $39,000 grant to a video pro­ducer: https://​​fo­rum.effec­tivealtru­ism.org/​​posts/​​CJJDw­gyqT4gXktq6g/​​long-term-fu­ture-fund-april-2019-grant-de­ci­sions .

  Maybe get in touch or have a look into what was suc­cess­ful there (I get the im­pres­sion that they found an im­por­tant ar­eas where there was oth­er­wise a lack of good video con­tent).

• weeatquince 21 Mar 2019 18:36 UTC

  3 points

  on: Should You Offer to Send Copies of Do­ing Good Bet­ter to Your Friends on So­cial Me­dia?

  Awe­some post.

  Sugges­tion: I have found in per­son feed­back to use­ful alongside sur­veys. Suggest mak­ing a bit of effort to talk to peo­ple in per­son, es­pe­cially if it is friends you see any­way, and in­clud­ing this data into a fi­nal im­pact es­ti­mate.

• weeatquince 21 Mar 2019 11:43 UTC

  10 points

  in reply to: Flemming's comment on: In­tro­duc­ing GPI’s new re­search agenda

  There are maybe 100+ as im­por­tant other steps to policy. In rough chronolog­i­cal or­der I started list­ing some of them be­low (I got bored part way through and stopped at what looks like 40 points).

  I have aimed to have all of these is­sues at a roughly similar or­der of mag­ni­tude of im­por­tance. The scale of these is­sues will de­pend on coun­try to coun­try and the tractabil­ity of try­ing to change these is­sues will vary with time and de­pend on in­di­vi­d­ual to in­di­vi­d­ual.

  Over­all I would say that vot­ing re­form is not ob­vi­ously more or less im­por­tant than the other 100+ things that could be on this list (al­though I guess it is of­ten likely to some­where in the top 50% of is­sues). There is a lot more un­cer­tainty about what the best vot­ing mechanisms look like than many of the other is­sues on the list. It is also an is­sue that may be hard to change com­pared to some of the oth­ers.

  Either way vot­ing re­form is a tiny part of an in­cred­ibly long pro­cess, a pro­cess with some huge ar­eas for im­prove­ments in other parts.

  SETTING BOUNDARIES

  • con­sti­tu­tion and hu­man rights and set­ting re­mits of poli­ti­cal pow­ers to change fun­da­men­tal struc­tures of country

  • de­volu­tion and set­ting re­mits of cen­tral poli­ti­cal pow­ers verses lo­cal poli­ti­cal bodies

  • term limits

  CHOOSING POLITICIANS

  • elec­toral com­mis­sion body set­ting or ad­just­ing bor­ders of vot­ing ar­eas /​ constituencies

  • ini­tial policy re­search by po­ten­tial can­di­dates (of­ten with very limited re­sources)

  • man­i­festo writ­ing (this is hugely im­por­tant to set the agenda and hard to change)

  • pub­lic /​ par­ties choos­ing can­di­dates (of­ten a lot of in­ter­nal party squab­bling be­hind the scenes)

  • cam­paign fundrais­ing (maybe un­due in­fluences)

  • cam­paign­ing and in­for­ma­tion spread­ing (maybe is­sues with false in­for­ma­tion)

  • tack­ling voter ap­a­thy /​ engagement

  • Vot­ing mechanism

  • coal­i­tion form­ing (of­ten very un­trans­par­ent)

  • gov­ern­ment/​leader as­signs topic ar­eas to ministers /​ se­niors (very poli­ti­cal, ev­i­dence that un­der­stand­ing a topic is in­versely pro­por­tional to how long a minister will work on that topic)

  CIVIL SERVICE STAFFING

  • hiring staff into gov­ern­ment (hiring pro­cesses, lack of ex­per­tise, di­ver­sity is­sues)

  • how staff in gov­ern­ment are man­aged (val­ues, team build­ing, re­wards, pro­gres­sion, di­ver­sity)

  • how staff in gov­ern­ment are trained (feed­back mechanisms, train­ing)

  ACCOUNTABILITY

  • split­ting out ar­eas where poli­ti­cal lead­er­ship is needed and ar­eas where tech­no­cratic lead­er­ship is needed

  • de­sign­ing clear mechanisms of ac­countabil­ity to top­ics so that poli­ti­ci­ans and civil ser­vants are aware of what their re­spon­si­bil­ities are and can be held to ac­count for their ac­tions (this is su­per im­por­tant)

  • en­sur­ing poli­ti­cal rep­re­sen­ta­tion so each in­di­vi­d­ual has di­rect ac­cess to a poli­ti­cian who is ac­countable for their con­cerns

  • putting in place sys­tems that al­low changes to the sys­tem if an ac­countabil­ity mechanisms is not work­ing

  • en­sur­ing ac­countabil­ity for un­known un­known is­sues that may arise

  • how poor perfor­mance of poli­ti­cal and civil staff is ad­dressed (poor perfor­mance pro­ce­dures, whistle­blow­ing)

  • how cor­rup­tion is rooted out and ad­dressed (yes there is cor­rup­tion in de­vel­oped coun­tries)

  • mechanisms to al­low par­ties /​ pop­u­la­tions to kick out bad lead­ers if needed

  • En­sur­ing mechanisms for cross party di­alogue and that par­ti­san-ism of poli­tics does not lead to dis­tor­tions of truth

  AGENDA SETTING AND INITIAL RESEARCH

  • car­ry­ing out re­search to un­der­stand what the policy prob­lems are (of­ten un­clear how to do this)

  • un­der­stand­ing what the pop­u­la­tion wants (pub­lic of­ten ig­nored, need good pro­ce­dures for in­for­ma­tion gath­er­ing, pub­lic con­sul­ta­tion, etc)

  POLICY DEVELOPMENT

  • Devel­op­ment of policy op­tions to ad­dress problems

  • Mechanisms for Cost Benefit Anal­y­sis and Im­pact Assess­ments to de­cide best policy options

  • ac­cess to ex­per­tise ad­vice and best prac­tice (lack of com­mu­ni­ca­tion be­tween academia and policy)

  • mea­sur­ing im­pact of a policy pro­posal once in place (en­sur­ing that mechanisms to mea­sure im­pact are ini­ti­ated at the very start of the policy im­ple­men­ta­tion)

  • ac­tu­ally us­ing in­for­ma­tion on

  • how poli­ti­ci­ans are al­lowed to change their mind given new ev­i­dence (up­dat­ing is of­ten seen as weak­ness)

  • mechanisms to en­sure is­sues that are not poli­ti­cally im­me­di­ately nec­es­sary are tack­led (lack of long term think­ing)

  LEGISLATIVE PROCESS

  POLICY IMPLEMENTATION

  POLICY REVIEW

  PUBLIC COMMUNICATION

  RISK MANAGEMENT

  GENERAL

  • flex­i­bil­ity to deal with shocks of ev­ery step of the above pro­cess (of­ten lack­ing)

  • trans­parency of ev­ery step of the above pro­cess (of­ten lack­ing)

• weeatquince 6 Mar 2019 8:59 UTC

  2 points

  on: Cli­mate Change Is, In Gen­eral, Not An Ex­is­ten­tial Risk

  Another thing to con­sider is that, given cli­mate mod­el­ing is so im­pre­cise and reg­u­larly flawed, that our mod­els are wrong and the risk is sig­nifi­cantly differ­ent than pre­dicted.

  (Similar to some of Toby’s stuff on the Large Hadron Col­lider risks: http://​​blog.prac­ti­calethics.ox.ac.uk/​​2008/​​04/​​these-are-not-the-prob­a­bil­ities-you-are-look­ing-for/​​)

  This could go both ways.

• weeatquince 6 Mar 2019 8:41 UTC

  11 points

  on: In­tro­duc­ing GPI’s new re­search agenda

  This is re­ally re­ally im­pres­sive. An amaz­ing col­lec­tion of re­ally im­por­tant ques­tions.
  
  POSITIVES. I like the fact that you in­tend to re­search:
  * In­sti­tu­tional ac­tors (2.8). Sig­nifi­cant changes to the world are likely to come through in­sti­tu­tional ac­tors and the EA com­mu­nity has largely ig­nored them to date. The ex­ist­ing re­search has fo­cused so much on the benefits of marginal dona­tions (or marginal re­search) that our views on cause pri­ori­ti­sa­tion can­not be eas­ily ap­plied to states. As some­one into EA in the busi­ness of in­fluenc­ing states this is a re­ally prob­le­matic over­sight of the com­mu­nity to date, that we should be look­ing to fix as soon as pos­si­ble.
  * De­ci­sion-the­o­retic is­sues (2.1)
  * The use of dis­count rates. This is prac­ti­cally use­ful for de­ci­sion mak­ers.
  
  OMISSIONS. I did how­ever note a few things that I would have ex­pected to be in­cluded, to not be men­tioned in this re­search agenda in par­tic­u­lar there was no dis­cus­sion on
  * Use­ful mod­els for think­ing about and talk­ing about cause pri­ori­ti­sa­tion. In par­tic­u­lar the scale ne­glect­ed­ness and tractabil­ity frame­work is of­ten used and of­ten crit­i­cised. What other mod­els can or should be used by the EA com­mu­nity.
  * So­cial change. Within sec­tion 1 there is some dis­cus­sion of broad verses nar­row fu­ture fo­cused in­ter­ven­tions, and so I would have ex­pected a similar dis­cus­sion in sec­tion 2 on so­cial change in­ter­ven­tions verses tar­geted in­ter­ven­tions in gen­eral. This was not men­tioned.
  * (which risks to the fu­ture are most con­cern­ing. Although I as­sume this is be­cause those top­ics are be­ing cov­ered by oth­ers such as FHI.)
  
  CONCERN
  Like I said above I think the ques­tions within 2.8 are re­ally im­por­ta­tion for EA to fo­cus on. I hope that the fact it is low on the list does not mean it is not pri­oro­tised.
  I also note that there is a sub-ques­tion in 2.8 on “what is the best fea­si­ble vot­ing sys­tem”. I think this is­sue comes up too much and is of­ten a dis­trac­tion. It feels like a minor sub part of the ques­tion on “what is the op­ti­mal in­sti­tu­tion de­sign” which peo­ple grav­i­tate too be­cause it is the most visi­ble part of many poli­ti­cal sys­tems, but is re­ally un­likely to be thing on the mar­gin that most needs im­prov­ing.

  I hope that helps, Sam

• weeatquince 6 Mar 2019 8:09 UTC

  28 points

  on: After one year of ap­ply­ing for EA jobs: It is re­ally, re­ally hard to get hired by an EA organisation

  CEA run the EA com­mu­nity fund to provide fi­nan­cial sup­port EA com­mu­nity group lead­ers.

  The key met­ric that CEA for eval­u­at­ing the suc­cess of the groups they fund is the num­ber of peo­ple from each lo­cal group who reach in­ter­view stage for high im­pact jobs, which largely means jobs within EA or­gani­sa­tions. Bonus points available if they get the job.

  This in­for­ma­tion feels like a rele­vant piece of the puz­zle for any­one think­ing through these is­sues. It could be (that in hind­sight) CEA push­ing chap­ter or­ganisers to push peo­ple to fo­cus on jobs in EA or­gani­sa­tions in many ways might not be the best strat­egy.

• weeatquince 13 Jan 2019 23:43 UTC

  3 points

  on: Tac­ti­cal mod­els to im­prove in­sti­tu­tional de­ci­sion-making

  I found this ar­ti­cle un­clear about what you were talk­ing about when you say “im­prov­ing in­sti­tu­tional de­ci­sion mak­ing” (in policy). I think we can break this down into two very differ­ent things.

  A: Im­prov­ing im­prov­ing the de­ci­sion mak­ing pro­cesses and sys­tems of ac­countabil­ity that policy in­sti­tu­tions use to make de­ci­sions so that these in­sti­tu­tions will more gen­er­ally be bet­ter de­ci­sion mak­ers. (This is what I have always meant by and un­der­stood by the term “im­prov­ing in­sti­tu­tional de­ci­sion mak­ing”, and what JEss talks about in her post you link to)

  B: Hav­ing in­fluence in a spe­cific situ­a­tion on the policy mak­ing pro­cess. (This is ba­si­cally what peo­ple tend to call “lob­by­ing” or some­times “cam­paign­ing”.)

  I felt that the DFID story and the three mod­els were all fo­cused on B: lob­by­ing. The mod­els were use­ful for think­ing about how to do B well (as­sum­ing you know bet­ter than the policy mak­ers what policy should be made). The­o­ret­i­cal ad­vice on lob­by­ing is a nice thing to have* if you are in the field (so thank you for writ­ing them up, I may give them some thought in my up­com­ing work). And if you are try­ing to change A it would be use­ful to un­der­stand how to do B.

  The mod­els were very use­ful for ad­vis­ing on how to do A: im­prov­ing how in­sti­tu­tions work gen­er­ally. And A is where I would say the value lies.

  I think the main point is just on how easy the ar­ti­cle was to read. I found the ar­ti­cle it­self was very con­fus­ing as to if you were talk­ing about A or B at many points.

  *Also in gen­eral I think the field of lob­by­ing is as one might say “more of an art than a sci­ence” and al­though a the­o­ret­i­cal un­der­stand­ing of how it works is nice it is not su­per use­ful comapred to ex­pe­rience in the field in the spe­cific coun­try that you are in.

• weeatquince 13 Jan 2019 23:15 UTC

  3 points

  on: Cli­mate Change Is, In Gen­eral, Not An Ex­is­ten­tial Risk

  I would be cu­ri­ous about any views or re­search you may have done into geo­eng­ineer­ing risk?

  My un­der­stand­ing is that cli­mate change is not it­self an ex­is­ten­tial risk but that it may lead to other risks (such war which as Peter Hur­ford men­tions). One other risk is geo­eng­ineer­ing where hu­man­ity starts think­ing it can con­trol plane­tary tem­per­a­tures and makes a mis­take (or the tech­nol­ogy is used mal­i­ciously) and that pre­sents a risk.

• weeatquince 13 Jan 2019 23:08 UTC

  6 points

  on: EAs Should In­vest All Year, then Give only on Giv­ing Tuesday

  Just to flag that the case for this is much much weaker out­side the USA.

  The match­ing limits for dona­tions out­side the US is much lower and you may also lose your tax benefits of donat­ing.

  See: https://​​docs.google.com/​​doc­u­ment/​​d/​​1hCCfv-1DI4FD5I5Pw5E3Ov4O46uzpIdI0QqRrWzvRsI/​​edit

• weeatquince 22 Sep 2018 7:20 UTC

  6 points

  in reply to: weeatquince's comment on: CEA on com­mu­nity build­ing, rep­re­sen­ta­tive­ness, and the EA Summit

  Hi Kerry, Thank you for the call. I wrote up a short sum­mary of what we dis­cussed. It is a while since we talked so not perfect. Please cor­rect any­thing I have mis­re­mem­bered.

  ~

  1.

  ~ ~ Set­ting the scene ~ ~

  • CEA should cham­pion cause pri­ori­ti­sa­tion. We want peo­ple who are will­ing to pick a new cause based on ev­i­dence and re­search and a com­mu­nity that con­tinues to work out how to do the most good. (We both agreed this.)

  • There is a differ­ence be­tween “cause im­par­tial­ity”, as defined above, and “ac­tual im­par­tial­ity”, not hav­ing a view on what causes are most im­por­tant. (There was some con­fu­sion but we got through it)

  • There is a differ­ence be­tween long-ter­mism as a method­ol­ogy where one con­sid­ers the long run fu­ture im­pacts which CEA should 100% pro­mote and long-ter­mism as a con­clu­sion that the most im­por­tant thing to fo­cus on right now is shap­ing the long term fu­ture of hu­man­ity. (I as­serted this, not sure you ex­pressed a view.)

  • A ra­tio­nal EA de­ci­sion maker could go through a pro­cess of cause pri­ori­ti­sa­tion and very le­gi­t­i­mately reach differ­ent con­clu­sions as to what causes are most im­por­tant. They may have differ­ent skills to ap­ply or differ­ent ethics (and we are far away from solv­ing ethics if such a thing is pos­si­ble). (I as­serted this, not sure you ex­pressed a view.)

  ~

  2.

  ~ ~ Create space, build trust, ex­press a view, do not be perfect ~ ~

  • The EA com­mu­nity needs to cre­ate the right kind of space so that peo­ple can reach their own de­ci­sion about what causes are most im­por­tant. This can be a phys­i­cal space (a lo­cal com­mu­nity) or an on­line space. Peo­ple should feel em­pow­ered to make their own de­ci­sions about causes. This means that they will be more adept at cause pri­ori­ti­sa­tion, more likely to be­lieve the con­clu­sions reached and more likely to come to the cor­rect an­swer for them­selves, and EA is more likely to come to a cor­rect an­swers over­all. To do this they need good tools and re­sources and to feel that the space they are in is neu­tral. This needs trust...

  • Creat­ing that space re­quires trust. Peo­ple need to trust the tools that are guid­ing and ad­vis­ing them. If peo­ple feel they be­ing sub­tly pushed in a di­rec­tion they will re­ject the re­sources and tools be­ing offered. Any sign of a break­down of trust be­tween peo­ple read­ing CEA’s re­sources and CEA should be taken very se­ri­ously.

  • Creat­ing that space does not mean you can­not also ex­press a view. You just want to dis­t­in­guish when you are do­ing this. You can cre­ate cause pri­ori­ti­sa­tion re­sources and tools that are truly neu­tral but still have a sep­a­rate sec­tion on what an­swers do CEA staff reach or what is CEA’s an­swer.

  • Perfec­tion is not re­quired as long as there is trust and the sys­tem is not break­ing down.

  • For ex­am­ple: pro­vid­ing policy ad­vice I gave the ex­am­ple of writ­ing ad­vice to a Gov Minister on a con­tro­ver­sial poli­ti­cal is­sue, as a civil ser­vant. The first ~85% of this imag­i­nary ad­vice has an im­par­tial sum­mary of the back­ground and the prob­lem and then a se­ries of sug­gested ac­tions with eval­u­a­tions of their im­pact. The fi­nal ~15% has a recom­mended ac­tion based on the civil ser­vant’s view of the mat­ter. The im­por­tant thing here is that there gen­er­ally is trust be­tween the Minister and the Depart­ment that ad­vice will be neu­tral, and that in this case the Minister trusts that the sec­tion/​space set­ting out the back­ground and pos­si­ble ac­tions is neu­tral enough for them to make a good de­ci­sion. It doesn’t need to be perfect, in fact the Minister will be aware that there is likely some amount of bias, but as long as there is suffi­cient trust that does not mat­ter. And there is a recom­men­da­tion which the Minister can choose to fol­low or not. In many cases the Minister will fol­low the recom­men­da­tion.

  ~

  3.

  ~ ~ How this goes wrong ~ ~

  • Imag­ine some­one who has iden­ti­fied cause X which is su­per im­por­tant comes across the EA com­mu­nity. You do not want the com­mu­nity to ei­ther be so fo­cused on one cause that this per­son is ei­ther put off or is per­suaded that the cur­rent EA cause is more im­por­tant and for­gets about cause X

  • I men­tioned some of the things that dam­age trust (see the foot of my pre­vi­ous com­ment).

  • You men­tioned you had seen signs of trib­al­ism in the EA com­mu­nity.

  ~

  4.

  ~ ~ Con­clu­sion ~ ~

  • You said that you saw more value in CEA cre­at­ing a space that was “ac­tual im­par­tial” as op­posed to “cause im­par­tial” than you had done pre­vi­ously.

  ~

  5.

  ~ ~ Ad­den­dum: Some thoughts on ev­i­dence ~ ~

  Not dis­cussed but I have some ex­tra thoughts on ev­i­dence.

  There are two ar­eas of my life where much of what I have learned points to­wards the views above be­ing true.

  • Coach­ing. In coach­ing you need to make sure the coachee feels like you are there to help them not in any way with you own agenda (that is differ­ent from theirs).

  • Policy. In policy mak­ing you need trust and neu­tral­ity be­tween Minister and civil ser­vant.

  There is value in fol­low­ing per­ceived wis­dom on a topic. That said I have been look­ing out for any strong ev­i­dence that these things are true (eg. that coach­ing goes badly if they think you are sub­tly bi­ased one way or an­other) and I have yet to find any­thing par­tic­u­larly per­sua­sive. (Coun­ter­point: I know one friend who knows their ther­a­pist is overly-bias to­wards push­ing them to have ad­di­tional ses­sions but this does not put them off at­tend­ing or mean they find it less use­ful). Per­haps this de­serves fur­ther study.

  Also worth bear­ing in mind there maybe dis­similar­i­ties be­tween what CEA does and the fields of coach­ing and policy.

  Also worth flag­ging that the ex­am­ple of policy ad­vice given above is some­what ar­tifi­cial, some policy ad­vice (es­pe­cially where con­tro­ver­sial) is like that but much of it is just: “please ap­prove ac­tion x”

  In con­clu­sion my views on this are based on very lit­tle ev­i­dence and a lot of gut feel­ing. My in­tu­itions on this are strongly guided by my time do­ing coach­ing and do­ing policy ad­vice.

• weeatquince 16 Sep 2018 1:08 UTC

  13 points

  in reply to: jpaddison's comment on: Ad­di­tional plans for the new EA Forum

  Fea­ture idea: If you co-write an ar­ti­cle with some­one be­ing able to post as co-au­thors.

Self-care ses­sions for EA groups

• weeatquince 26 Aug 2018 12:56 UTC

  3 points

  in reply to: Kerry_Vaughan's comment on: CEA on com­mu­nity build­ing, rep­re­sen­ta­tive­ness, and the EA Summit

  Hi Kerry, Some more thoughts prior to hav­ing a chat.

  -

  Is longter­mism a cause?

  Yes and no. The term is used in mul­ti­ple ways.

  A: Con­sid­er­a­tion of the long-term fu­ture.

  It is a core part of cause pri­ori­ti­sa­tion to avoid availa­bil­ity bi­ases: to con­sider the plights of those we can­not so eas­ily be aware of, such as an­i­mals, peo­ple in other coun­tries and peo­ple in the fu­ture. As such, in my view, it is im­per­a­tive that CEA and EA com­mu­nity lead­ers pro­mote this.

  B: The long-term cause area.

  Some peo­ple will con­clude that the op­ti­mal use of their limited re­sources should be putting them to­wards shap­ing the far fu­ture. But not ev­ery­one, even af­ter full ra­tio­nal con­sid­er­a­tion, will reach this view. Nor should we ex­pect such una­n­im­ity of con­clu­sions. As such, in my view, CEA and EA com­mu­nity lead­ers can recom­mend peo­ple to con­sider this causes area, but should not tell peo­ple this is the an­swer.

  -

  Thread­ing the needle

  I agree with the 6 points you make here.

  (Although in­ter­est­ingly I per­son­ally do not have ev­i­dence that “area alle­giance is op­er­at­ing as a kind of tribal sig­nal in the move­ment cur­rently”)

  -

  CEA and cause-impartiality

  I think CEA should be care­ful about how to ex­press a view. Do­ing this in wrong way could make it look like CEA is not cause im­par­tial or not rep­re­sen­ta­tive.

  My view is to give recom­men­da­tions and tools but not an­swers. This is similar to how we would not ex­pect 80K to have a view on what the best job is (as it de­pends on an in­di­vi­d­ual and their skills and needs) but we would ex­pect 80K to have recom­men­da­tions and to have ad­vice on how to choose.

  I think this ap­proach is also use­ful be­cause:

  • Peo­ple are more likely to trust de­ci­sions they reach through their own think­ing rather than con­clu­sions they are pushed to­wards.

  • It han­dles the fact that ev­ery­one is differ­ent. The ad­vice or rea­son­ing that works for one per­son may well not make sense for some­one else.

  I think (as Khor­ton says) it is perfectly rea­son­able for an or­gani­sa­tion to not have a con­clu­sion.

  -

  (One other thought I had was on ex­am­ples of ac­tions I would be con­cerned about CEA or an­other move­ment build­ing or­gani­sa­tions tak­ing would be: Ex­press­ing cer­tainty about a area (in in­ter­nal policy or ex­ter­nally), bas­ing im­pact mea­sure­ment solely on a sin­gle cause area, hiring staff for cause-gen­eral roles based on their views of what causes is most im­por­tant, at­tempt­ing to push as many peo­ple as pos­si­ble to a spe­cific cause area, etc)

• weeatquince 16 Aug 2018 7:27 UTC

  1 point

  in reply to: RandomEA's comment on: CEA on com­mu­nity build­ing, rep­re­sen­ta­tive­ness, and the EA Summit

  Yes thanks. Edited.

• weeatquince 15 Aug 2018 23:50 UTC

  27 points

  in reply to: jpaddison's comment on: CEA on com­mu­nity build­ing, rep­re­sen­ta­tive­ness, and the EA Summit

  We would like to hear sug­ges­tions from fo­rum users about what else they might like to see from CEA in this area.

  Here is my two cents. I hope it is con­struc­tive:

  ---

  1.

  The policy is ex­cel­lent but the challenge lies in im­ple­men­ta­tion.

  Firstly I want to say that this post is fan­tas­tic. I think you have got the policy cor­rect: that CEA should be cause-im­par­tial, but not cause-ag­nos­tic and CEA’s work should be cause-gen­eral.

  How­ever I do not think it looks, from the out­side, like CEA is fol­low­ing this policy. Some ex­am­ples:

  • EA Lon­don staff had con­cerns that they would need to be more fo­cused on the far fu­ture in or­der to re­ceive fund­ing from CEA.

  • You ex­plic­itly say on your web­site: “We put most of our cre­dence in a wor­ld­view that says what hap­pens in the long-term fu­ture is most of what mat­ters. We are there­fore more op­ti­mistic about oth­ers who roughly share this wor­ld­view.“[1]

  • The ex­am­ple you give of the new EA handbook

  • There is a close as­so­ci­a­tion with 80000 Hours who are ex­plic­itly fo­cus­ing much of their effort on the far fu­ture.

  Th­ese are all quite sub­tle things, but col­lec­tively they give an im­pres­sion that CEA is not cause im­par­tial (that it is x-risk fo­cused). Of course this is a difficult thing to get cor­rect. It is difficult to draw the line be­tween say­ing: ‘our staff mem­bers be­lieve cause___ is im­por­tant’ (a use­ful fac­toid that should definitely be said), whilst also putting across a strong front of cause im­par­tial­ity.

  ---

  2.

  Sugges­tion: CEA should ac­tively cham­pion cause impartiality

  If you gen­uinely want to be cause im­par­tial I think most of the solu­tions to this are around be­ing su­per vigilant about how CEA comes across. Eg:

  • Have a clear in­ter­nal style guide that sets out to staff good and bad ways to talk about causes

  • Have ‘cause im­par­tial­ity’ as a staff value

  • If you do an ac­tion that does not look cause im­par­tial (say EA Grants mostly grants money to far fu­ture causes) then just ac­knowl­edge this and say that you have noted it and ex­plain why it hap­pened.

  • Public posts like this one set­ting out what CEA believes

  • If you want to do lots of “pre­scrip­tive” ac­tions split them off into a sub pro­ject or a sep­a­rate in­sti­tu­tion.

  • Ap­ply the above retroac­tively (re­move lines from your web­site that make it look like you are only fu­ture fo­cused)

  Beyond that, if you re­ally want to cham­pion cause im­par­tial­ity you may also con­sider ex­tra things like:

  • More fo­cus on cause pri­ori­ti­sa­tion re­search.

  • Hiring peo­ple who value cause im­par­tial­ity /​ cause pri­ori­ti­sa­tion re­search /​ com­mu­nity build­ing, above peo­ple who have strong views on what causes are im­por­tant.

  ---

  3.

  Be­ing rep­re­sen­ta­tive is about mak­ing peo­ple feel listened too.

  Your sec­tion on rep­re­sen­ta­tives feels like you are try­ing to pin down a way of find­ing an ex­act num­ber so you can say we have this many ar­ti­cles on topic x and this many on topic y and so on. I am not sure this is quite the cor­rect fram­ing.

  Things like the EA hand­book should (as a lower bound) have enough of a di­ver­sity of causes men­tioned that the broader EA com­mu­nity does not feel mis­rep­re­sented but (as an up­per bound) not so much that CEA staff [2] feel like it is mis­rep­re­sent­ing them. Any­thing within this range seems fine to me. (Eg. with the EA hand­book both groups should feel com­fortable hand­ing this book to a friend.) Although I do feel a bit like I have just typed ‘just do the thing that makes ev­ery­one happy’ which is eas­ier said than done.

  I also think that “rep­re­sen­ta­tive­ness” is not quite the right is­sue any way. The im­por­tant thing is that peo­ple in the EA com­mu­nity feel listened too and feel like what CEA is do­ing rep­re­sents them. The % of con­tent on differ­ent top­ics is only part of that. The other parts of the solu­tion are:

  • Com­ing across like you listen: see the afore­men­tioned points on cham­pi­oning cause im­par­tial­ity. Also ex­press­ing un­cer­tainty, men­tion­ing that there are op­pos­ing views, giv­ing two sides to a de­bate, etc.

  • Listen­ing—ie. con­sult­ing pub­li­cly (or with trusted par­ties) wher­ever pos­si­ble.

  If any­thing get­ting these two things cor­rect is more im­por­tant than get­ting the ex­act per­centage of your work to be rep­re­sen­ta­tive.

  ---

  Sam :-)

  ---

  [1] https://​​www.cen­tre­fore­ffec­tivealtru­ism.org/​​a-three-fac­tor-model-of-com­mu­nity-building

  [2] Un­less you have rea­son to think that there is a sys­tem­atic bias in staff, eg if you ac­tively hired peo­ple be­cause of the cause they cared about.

• weeatquince 8 Aug 2018 23:04 UTC

  0 points

  in reply to: MarekDuda's comment on: EA Funds—An up­date from CEA

  YAY <3

• weeatquince 8 Aug 2018 11:17 UTC

  6 points

  in reply to: jpaddison's comment on: EA Funds—An up­date from CEA

  Marek, well done on all of your hard work on this.

  Separate from the man­aged funds. I re­ally like the work that CEA is do­ing to help money be moved around the world to other EA char­i­ties. I would love to see more or­gani­sa­tions on the list of places that dona­tions can be made through the EA Funds plat­form. Eg, REG or An­i­mal Char­ity Eval­u­a­tors or Re­think Char­ity. Is this in the works?

  https://​​app.effec­tivealtru­ism.org/​​dona­tions/​​new/​​organizations

• weeatquince 5 Aug 2018 21:57 UTC

  32 points

  in reply to: Geoff_Anders's comment on: Lev­er­age Re­search: re­view­ing the ba­sic facts

  count­ing our re­search as 0 value, and us­ing the move­ment build­ing im­pact es­ti­mates from LEAN, we come out well on EV com­pared to an av­er­age char­ity … I will let read­ers make their own calculations

  Hi Ge­off. I gave this a lit­tle thought and I am not sure it works. In fact it looks quite plau­si­ble that some­one’s EV (ex­pected value) calcu­la­tion on Lev­er­age might ac­tu­ally come out as nega­tive (ie. Lev­er­age would be caus­ing harm to the world).

  This is be­cause:

  • Most EA orgs calcu­late their coun­ter­fac­tual ex­pected value by tak­ing into ac­count what the peo­ple in that or­gani­sa­tion would be do­ing oth­er­wise if they were not in that or­gani­sa­tion and then deduct this from their im­pact. (I be­lieve at least 80K, Char­ity Science and EA Lon­don do this)

  • Given Lev­er­age’s ten­dency to hire am­bi­tious al­tru­is­tic peo­ple and to look for peo­ple at EA events it is plau­si­ble that a sig­nifi­cant pro­por­tion of Lev­er­age staff might well have ended up at other EA or­gani­sa­tions.

  • There is a tal­ent gap at other EA or­gani­sa­tions (see 80K on this)

  • Lev­er­age does spend some time on move­ment build­ing but I es­ti­mate that this is a tiny pro­por­tion of the time, >5%, best guess 3%, (based on hav­ing talked to peo­ple at lev­er­age and also based on look­ing at your achieve­ments to date com­pared it to the ap­par­ent 100 per­son-years figure)

  • There­fore if the amount of staff who could be ex­pected to have found jobs in other EA or­gani­sa­tions is thought to be above 3% (which seems rea­son­able) then Lev­er­age is ac­tu­ally dis­plac­ing EAs from pro­duc­tive ac­tion so the to­tal EV of Lev­er­age is negative

  Of course this is all as­sum­ing the value of your re­search is 0. This is the as­sump­tion you set out in your post. Ob­vi­ously in prac­tice I don’t think the value of your re­search is 0 and as such I think it is pos­si­ble that the to­tal EV of Lev­er­age is pos­i­tive*. I think more trans­parency would help here. Given that al­most no re­search is available I do think it would be rea­son­able for some­one who is not at Lev­er­age to give your re­search an EV of close to 0 and there­fore con­clude that Lev­er­age is caus­ing harm.

  I hope this helps and maybe ex­plains why Lev­er­age gets a bad rep. I am ex­cited to see a more trans­parency and a new ap­proach to pub­lic en­gage­ment. Keep on fight­ing for a bet­ter world!

  *sen­tence ed­ited to bet­ter match views

• weeatquince 4 Aug 2018 7:34 UTC

  14 points

  in reply to: jpaddison's comment on: Prob­lems with EA rep­re­sen­ta­tive­ness and how to solve it

  Hi Joey, thank you for writ­ing this.

  I think call­ing this a prob­lem of rep­re­sen­ta­tion is ac­tu­ally un­der­stat­ing the prob­lem here.

  EA has (at least to me) always been a com­mu­nity that in­spires en­courages and sup­ports peo­ple to use all the in­for­ma­tion and tools available to them (in­clud­ing their in­di­vi­d­ual pri­ors in­tu­itions and sense of moral­ity) to reach a con­clu­sion about what causes and ac­tions are most im­por­tant for them to take to make a bet­ter world (and of course to then take those ac­tions).

  Even if 90% of ex­pe­rienced EAs /​ EA com­mu­nity lead­ers cur­rently con­verge on the same con­clu­sion as to where value lies, I would worry that a strong fo­cus on that is­sue would be detri­men­tal. We’d be at risk of los­ing the em­pha­sis on cause pri­ori­ti­sa­tion—ar­guably most use­ful in­sight that EA has pro­vided to the world.

  • We’d risk los­ing an abil­ity to sup­port peo­ple though cause pri­ori­ti­sa­tion (coach­ing, EA or oth­er­wise, should not pre-empt the an­swers or have ul­te­rior mo­tives)

  • we risk cre­at­ing a com­mu­nity that is less able to switch to fo­cus on the most im­por­tant thing

  • we risk stifling use­ful debate

  • we risk cre­at­ing a com­mu­nity that does not benefits from col­lab­o­ra­tion by peo­ple work­ing in differ­ent areas

  • etc

  (Note: Prob­a­bly worth adding that if 90% of ex­pe­rienced EAs /​ EA com­mu­nity lead­ers con­verged on the same con­clu­sion on causes my in­tu­itions would sug­gest that this is lik­ley to be ev­i­dence of founder effects /​ group-think as much as it is ev­i­dence for that cause. I ex­pect this is be­cause I see a huge di­ver­sity in peo­ple’s val­ues and think­ing and a difficulty in reach­ing strong con­clu­sions in ethics and cause pri­ori­ti­sa­tion)