Denise_Melchin

• Denise_Melchin 5 Jun 2019 13:04 UTC

  32 points

  on: What’s the best struc­ture for op­ti­mal al­lo­ca­tion of EA cap­i­tal?

  I find some of the state­ments in your post a bit jar­ring, and this is not the first time I feel like this when read­ing your writ­ing. The founders of Good Ven­tures are multi-billion­aires who have been in­fluenced by EA ideas which stem to some ex­tent from the EA com­mu­nity. This is ex­cel­lent. But the EA com­mu­nity does not have own­er­ship over this money. Your writ­ing makes it sound like it does, which I find pre­sump­tu­ous and off-putting.

  For the fu­ture, I would recom­mend that you should try to bet­ter un­der­stand the re­la­tion­ships be­tween differ­ent in­di­vi­d­u­als and in­sti­tu­tions that are as­so­ci­ated with the Effec­tive Altru­ism com­mu­nity be­fore ask­ing ques­tions like this one.

  (writ­ing in per­sonal ca­pac­ity here, not as a mod)

• Denise_Melchin 8 May 2019 11:05 UTC

  6 points

  in reply to: Larks's comment on: Is pre­vent­ing child abuse a plau­si­ble Cause X?

  I think it’s worth not­ing what the re­port says about fam­ily struc­tures that fall in nei­ther cat­e­gory—both mar­ried and un­mar­ried par­ents where one par­ent isn’t biolog­i­cally re­lated to the child but still takes on a parental role as well as sin­gle par­ents with­out a part­ner fall some­where in be­tween mar­ried biolog­i­cal par­ents and sin­gle par­ents with part­ner in terms of child abuse rates.

  (I was a bit con­fused and thought ‘sin­gle par­ents with part­ner’ in­cluded cases in which the part­ner takes on parental re­spon­si­bil­ity so the high rate seemed off to me.)

• Denise_Melchin 7 May 2019 16:49 UTC

  16 points

  in reply to: Robert_Wiblin's comment on: Why isn’t GV psychedelics grant­mak­ing housed un­der Open Phil?

  But ask­ing pri­vately only gives one per­son the an­swer, in­stead of many. I’m a bit sur­prised by your re­sponse—I had ex­pected that the group who knows the an­swer usu­ally has bet­ter things to do than an­swer ran­dom emails, while there are a lot of in­di­vi­d­u­als who prob­a­bly have knowl­edge like this whose time isn’t as valuable.

• Denise_Melchin 1 May 2019 14:15 UTC

  11 points

  on: Will splashy philan­thropy cause the biose­cu­rity field to fo­cus on the wrong risks?

  This seems like a rea­son­able piece to me, lay­ing out the ba­sic ground­work for more fu­ture scrutiny on philan­thropy’s im­pact on the biose­cu­rity field, but not more than that. (‘Estab­lish­ing com­mon knowl­edge’ seems like a good sum­mary to me.)

  A large in­flux of money can sig­nifi­cantly change a field, and gen­er­ally speak­ing, it is much harder for sud­den big changes to im­prove the state of af­fairs than to make them worse. That said, sud­den large changes, even if net pos­i­tive over­all, will of­ten have some nega­tive side effects, and I would ex­pect more money for a ‘do good-ing’ field to lead to more good over­all.

  Some­thing that might be in­ter­est­ing to see would be a sur­vey of top peo­ple in the biose­cu­rity field how this has changed their field and whether they view this change as pos­i­tive. Gen­er­ally speak­ing, I would ex­pect them to have a much bet­ter grasp of em­piri­cal pri­ori­ti­sa­tion ques­tions in biose­cu­rity than a few peo­ple at a large foun­da­tion, no mat­ter how care­ful they are and how much work they put in. The more work large foun­da­tions put into be­ing in touch with peo­ple in the field, the less con­cerned one needs to be I think.

  Similar crit­i­cisms also ex­ist in other fields, e.g. about the Gates foun­da­tion drown­ing out pri­mary health care work by fo­cus­ing on vac­ci­na­tions and spe­cific dis­eases and in­ad­ver­tently caus­ing some harm this way. I have not in­ves­ti­gated the mer­its of this crit­i­cism, but it seems like a worth­while thing to do.

• Denise_Melchin 20 Apr 2019 6:14 UTC

  17 points

  on: Thoughts on 80,000 Hours’ re­search that might help with job-search frustrations

  80,000 Hours thinks earn­ing to give is the best op­tion for a sub­stan­tial num­ber of peo­ple—those for whom it’s their com­par­a­tive ad­van­tage. They are keen, how­ever, to make sure that peo­ple fully con­sider di­rect work op­tions, in­stead of de­fault­ing to earn­ing to give be­cause they’ve heard it is the best way to do good with one’s ca­reer.

  If I re­mem­ber cor­rectly, 80,000 Hours has stated that they think 15% of peo­ple in the EA Com­mu­nity should be pur­su­ing earn­ing to give. Have they re­vised this opinion or am I re­mem­ber­ing it in­cor­rectly?

  If not, your de­scrip­tion seems a bit mis­lead­ing to me. Sub­stan­tial num­ber sounds like a sig­nifi­cantly higher frac­tion of peo­ple to me, per­haps some­thing like 40% in­stead of 15%.

• Denise_Melchin 1 Apr 2019 12:34 UTC

  7 points

  in reply to: AviNorowitz's comment on: EA Fo­rum Prize: Win­ners for Fe­bru­ary 2019

  As one of the peo­ple who voted, I was also sur­prised and dis­ap­pointed by this. But differ­ent vot­ers ap­plied differ­ent stan­dards on what kind of con­tent they wish to sup­port.

• Denise_Melchin 21 Mar 2019 18:22 UTC

  12 points

  in reply to: Jan_Kulveit's comment on: EA Pro­jects eval­u­a­tion platform

  (I still feel like I don’t re­ally un­der­stand where you’re com­ing from.)

  I am con­cerned that your model of how idea pro­pos­als get eval­u­ated (and then plau­si­bly funded) is a bit off. From the origi­nal post:

  hard to eval­u­ate which pro­ject ideas are ex­cel­lent , which are prob­a­bly good, and which are too risky for their es­ti­mated re­turn.

  You are miss­ing one ma­jor cat­e­gory here: pro­jects which are sim­ply bad be­cause they do have ap­prox­i­mately zero im­pact, but aren’t par­tic­u­larly risky. I think this cat­e­gory is the largest of the the four.

  Which pro­jects have a chance of work­ing and which don’t is of­ten pretty clear to peo­ple who have ex­pe­rience eval­u­at­ing pro­jects quite quickly (which is why Oli sug­gested 15min for the ini­tial in­ves­ti­ga­tion above). It sounds to me a bit like your model of ideas which get pro­posed is that most of them are pretty valuable. I don’t think this is the case.

  When fun­ders give gen­eral opinions on what should or should not get started or how you value or not value things, again, I think you are at greater risk of hav­ing too much of an in­fluence on the com­mu­nity. I do not be­lieve the knowl­edge of the fun­ders is strictly bet­ter than the knowl­edge of grant ap­pli­cants.

  I am con­fused by this. Knowl­edge of what?

  The role of fun­ders/​eval­u­a­tors is to eval­u­ate pro­jects (and maybe pro­pose some for oth­ers to do). To do this well they need to have a good men­tal map of what kind of pro­jects have worked or not worked in the past, what good and bad signs are, ideally from an ex­plicit feed­back loop from fund­ing pro­jects and then see­ing how the pro­jects turn out. The role of grant ap­pli­cants is to come up with some ideas they could ex­e­cute. Do you dis­agree with this?

• Denise_Melchin 21 Mar 2019 15:10 UTC

  10 points

  in reply to: Jan_Kulveit's comment on: EA Pro­jects eval­u­a­tion platform

  I think it much harder to give open feed­back if it is closely tied with fund­ing. Feed­back from fun­ders can eas­ily have too much in­fluence on peo­ple, and should be very care­ful and nu­anced, as it comes from some po­si­tion of power. I would ex­pect adding fi­nan­cial in­cen­tives can eas­ily be detri­men­tal for the pro­cess. (For self-refer­en­tial ex­am­ple, just look on this dis­cus­sion: do you think the fact that Oli dis­likes my pro­posal and sug­gest LTF can back some­thing differ­ent with $20k will not cre­ate at least some un­con­scious in­cen­tives?)

  I’m a bit con­fused here. I think I dis­agree with you, but maybe I am not un­der­stand­ing you cor­rectly.

  I con­sider hav­ing peo­ple giv­ing feed­back to have ‘skin in the game’ to be im­por­tant for the ac­cu­racy of the feed­back. Most peo­ple don’t en­joy dis­cour­ag­ing oth­ers they have so­cial ties with. Often re­view­ers with­out suffi­cient skin in the game might be tempted to not be as openly nega­tive about pro­pos­als as they should be.

  Fun­ders in­stead can give you a strong sig­nal—a sig­nal which is un­for­tu­nately some­what bi­nary and lacks nu­ance. But some­one be­ing will­ing to fund some­thing or not is a much stronger sig­nal for the value of a pro­posal than com­ments from friends on a GoogleDoc. This is es­pe­cially true if peo­ple propos­ing ideas don’t take into ac­count how hard it is to dis­cour­age peo­ple and don’t in­ter­pret feed­back in that light.

• Denise_Melchin 19 Mar 2019 15:42 UTC

  18 points

  on: A guide to im­prov­ing your odds at get­ting a job in EA

  EA jobs, un­like many other jobs, do not com­pare very well to other kinds of work ex­pe­rience,

  I’m pretty scep­ti­cal of this claim (not just made here, but also made in many other posts). I think this might be true for some roles like the Re­search An­a­lyst po­si­tions at the Open Philan­thropy Pro­ject which com­bine aca­demic re­search with grant­mak­ing which is un­usual in the wider job mar­ket.

  But I don’t see why e.g. op­er­a­tions at an av­er­age EA or­gani­sa­tion would not com­pare well to other kinds of work ex­pe­rience in op­er­a­tions. I’m happy to hear coun­ter­ar­gu­ments to this.

  The un­der­ly­ing crux here might be that I’m gen­er­ally wary of any claims of ‘EA ex­cep­tion­al­ism’.

• Denise_Melchin 19 Mar 2019 15:37 UTC

  24 points

  on: A guide to im­prov­ing your odds at get­ting a job in EA

  This list seems roughly rea­son­able. What most stands out to me is that your sug­ges­tions are ex­tremely time con­sum­ing, es­pe­cially in ag­gre­gate. The hours ap­pli­cants to jobs at EA or­gani­sa­tions spend on timed work tests and hon­ing their CVs pale in com­par­i­son.

  I also think your sug­ges­tions are ap­pli­ca­ble to some other fields which might be of in­ter­est to peo­ple who are try­ing to have a high im­pact. It is not un­usual for de­sir­able roles in e.g. in­ter­na­tional de­vel­op­ment to re­quire hun­dreds to thou­sands of hours of in­vest­ment.

  How­ever, if peo­ple are in­vest­ing those thou­sands of hours into learn­ing about EA, they will not spend them in­vest­ing in in­ter­na­tional de­vel­op­ment or nu­clear se­cu­rity.

  While peo­ple fol­low­ing your sug­ges­tions might benefit in­di­vi­d­u­ally, as a move­ment we and the world might be worse off.

• Denise_Melchin 13 Mar 2019 13:33 UTC

  54 points

  in reply to: Ben_Kuhn's comment on: EA is vet­ting-constrained

  (Fund­ing man­ager of the EA Meta Fund here)

  We have run an ap­pli­ca­tion round for our last dis­tri­bu­tion for the first time. I con­ducted the very ini­tial in­ves­ti­ga­tion which I com­mu­ni­cated to the com­mit­tee. Pre­vi­ous grantees came all through our per­sonal net­work.

  Things we learnt dur­ing our ap­pli­ca­tion round:

  i) We got sig­nifi­cantly fewer ap­pli­ca­tions than we ex­pected and would have been able to spend more time vet­ting pro­jects. This was not a bot­tle­neck. After some in­ves­ti­ga­tion through per­sonal out­reach I have the im­pres­sion there are not many pro­jects be­ing started in the Meta space (this is differ­ent for other fund­ing spaces).

  ii) We were able to fund a de­cent frac­tion of the ap­pli­ca­tions we re­ceived (25%?). For about half of the ap­pli­ca­tions I was rea­son­ably con­fi­dent that they did not meet the bar so I did not in­ves­ti­gate fur­ther. The re­main­ing quar­ter felt bor­der­line to me, I of­ten still in­ves­ti­gated but the re­sults con­firmed my ini­tial im­pres­sion.

  My cur­rent im­pres­sion for the Meta space is that we are not vet­ting con­strained, but more men­tor­ing/​pro-ac­tive out­reach con­strained. One thing we want to do in the fu­ture is to run a re­quest for pro­pos­als pro­cess.

• Denise_Melchin 13 Mar 2019 8:37 UTC

  11 points

  in reply to: RyanCarey's comment on: SHOW: A frame­work for shap­ing your tal­ent for di­rect work

  This isn’t re­ally com­par­ing like with like how­ever—in one case you’re do­ing cold out­reach and in oth­ers there are es­tab­lished ap­pli­ca­tion pro­cesses. It might make more sense to com­pare the de­mand for re­searcher po­si­tions with e.g. the Toby Ord’s Re­search As­sis­tant po­si­tion.

  But if your point is that peo­ple should be more will­ing to do cold out­reach for re­search as­sis­tant po­si­tions like you did, that seems fair.

• Denise_Melchin 28 Feb 2019 14:02 UTC

  11 points

  in reply to: agdfoster's comment on: After one year of ap­ply­ing for EA jobs: It is re­ally, re­ally hard to get hired by an EA organisation

  many can­di­dates treated the pro­cess like a 2 way ap­pli­ca­tion the whole way through the pro­cess. This three off my in­tu­itions and nor­mally I would have dropped all can­di­dates who weren’t sig­nal­ling they were speci­fi­cally very ex­cited about my role. First call ex­cluded.

  I won­der whether this is just a re­sult of peo­ple on both sides of the ap­pli­ca­tion pro­cess know­ing each other in a so­cial con­text.

  If the can­di­date knows they will in­ter­act with peo­ple mak­ing the hiring de­ci­sion in the fu­ture, they might not want them to feel bad about re­ject­ing them. The peo­ple mak­ing the hiring de­ci­sion might ar­guably feel less bad about not hiring some­one if the can­di­date wasn’t that ex­cited. Lack of ex­cite­ment also al­lows the can­di­date to save face if they get re­jected, which also only mat­ters be­cause the can­di­date and the per­son mak­ing the hiring de­ci­sion might in­ter­act so­cially in the fu­ture.

• Denise_Melchin 26 Feb 2019 14:48 UTC

  53 points

  in reply to: FlorentBerthet's comment on: After one year of ap­ply­ing for EA jobs: It is re­ally, re­ally hard to get hired by an EA organisation

  I don’t re­ally agree with your sec­ond and third point. See­ing this prob­lem and re­spond­ing by try­ing to cre­ate more ‘cap­i­tal let­ter EA jobs’ strikes me as con­tin­u­ing to pur­sue a failing strat­egy.

  What (in my opinion) the EA Com­mu­nity needs is to get away from this idea of chan­nel­ling all com­mit­ted peo­ple to a few or­gani­sa­tions—the com­mu­nity is grow­ing faster* than the or­gani­sa­tions, and those num­bers are un­likely to add up in the mid term.

  Com­mit­ting all our peo­ple to a few or­gani­sa­tions se­ri­ously limits our im­pact in the long run. There are plenty of op­por­tu­ni­ties to have a large im­pact out there—we just need to ap­pre­ci­ate them and pur­sue them. One thing I would like to see is stronger pro­fes­sion-spe­cific net­works in EA.

  It’s catas­trophic that new and long-term EAs now con­sider their main EA ac­tivity to be to ap­ply for the same few jobs in­stead of try­ing to in­crease their dona­tions or in­vest­ing in non-‘cap­i­tal let­ter EA’ promis­ing ca­reers.

  But this is hardly sur­pris­ing given past mes­sag­ing. The only rea­son EA or­gani­sa­tions can get away with hav­ing very ex­pen­sive hiring rounds for the ap­pli­cants is be­cause there are a lot of strongly com­mit­ted peo­ple out there will­ing to take on that cost. Or­gani­sa­tions can­not get away with this in most of the for-profit sec­tor.

  *Though this might be slow­ing down some­what, per­haps be­cause of this ‘be­ing an EA is ap­ply­ing un­suc­cess­fully for the same few jobs’ phe­nom­ena.

• Denise_Melchin 26 Feb 2019 12:35 UTC

  48 points

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

  I re­ally ap­pre­ci­ate you writ­ing this. You are not the first per­son to con­sider do­ing so and I ap­plaud you for ac­tu­ally do­ing it.

• Denise_Melchin 8 Feb 2019 15:16 UTC

  15 points

  on: EA grants available to in­di­vi­d­u­als (cross­post from LessWrong)

  Hi Jame­son,

  I’m a fund man­ager for the EA Meta Fund. Your as­sess­ment in your post is in­cor­rect—we are also open to in­di­vi­d­ual grant ap­pli­ca­tions, though ap­pli­ca­tions for the Fe­bru­ary dis­tri­bu­tion have now closed. I’d ex­pect them to open again in a cou­ple of months.

  I’m cu­ri­ous how you got the im­pres­sion that we aren’t open to ap­pli­ca­tions. It’s im­por­tant to us that we are able to reach all in­ter­ested in­di­vi­d­u­als so any in­sight into where we may have failed to com­mu­ni­cate that is use­ful to us.

EA Meta Fund: we are open to applications

• Denise_Melchin 23 Dec 2018 11:30 UTC

  14 points

  on: Should donor lot­tery win­ners write re­ports?

  My main worry about donor lot­tery re­ports is some­what differ­ent. Usu­ally, peo­ple seem to as­sign some ex­tra cred­i­bil­ity to a donor’s rea­son­ing if the dona­tion/​s is/​are large. This seems rea­son­able to me, since donors who donate large sums of­ten have a lot more ex­pe­rience with mak­ing dona­tion de­ci­sions. But donor lot­tery win­ners have much less ex­per­tise than the av­er­age per­son who makes large dona­tions (and only just as much as those long-term large donors had when they made a large dona­tion for the first time).

  In sum my con­cern is that peo­ple will trust donor lot­tery win­ner’s eval­u­a­tions of dona­tion tar­gets more than they should.

• Denise_Melchin 20 Dec 2018 13:47 UTC

  9 points

  in reply to: alexherwix's comment on: EA Meta Fund AMA: 20th Dec 2018

  Hello Alex,

  We are in­ter­ested in fund­ing new pro­jects (see also Alex Foster’s re­sponse above).

  I am also con­cerned about the difficulty of promis­ing new pro­jects to be dis­cov­ered. Per­son­ally, I am happy to in­vest some time into eval­u­at­ing new pro­jects. This is why we have a grant con­sid­er­a­tion form you can fill out to be con­sid­ered for re­ceiv­ing a grant. That said, we are time ca­pac­ity con­strained and would not be able to han­dle 100 ap­pli­ca­tions per month in our cur­rent setup.

  I have per­son­ally con­sid­ered putting out pro­pos­als like you are sug­gest­ing, but am con­cerned about the time in­vest­ment. First I would like to see how much in­ter­est we can gather in differ­ent ways.

• Denise_Melchin 20 Nov 2018 21:59 UTC

  2 points

  in reply to: Ben Pace's comment on: Take­aways from EAF’s Hiring Round

  To be clear, I meant ask­ing for a refer­ence be­fore an offer is ac­tu­ally made, at the stage when offers are be­ing de­cided (so that ap­pli­cants who don’t re­ceive offers one way or the other don’t ‘use up’ their refer­ences).