SiebeRozendal

• SiebeRozendal 27 Feb 2019 14:40 UTC

  1 point

  in reply to: kierangreig's comment on: Ro­dents farmed for pet snake food

  I thinks there’s a de­cent chance that if one were to dive deeper into the farm­ing of in­ver­te­brates then this could lead to dis­cov­er­ies of tens of billions of ad­di­tional farmed an­i­mals the move­ment largely cur­rently ne­glects.

  I agree.

  Say this is the case, what would be the im­pli­ca­tions? It does seem that more gen­eral anti-speciecism efforts be­come com­par­a­tively more effec­tive the more an­i­mal suffer­ing is widely dis­persed.

• SiebeRozendal 17 Jan 2019 14:50 UTC

  2 points

  on: Ac­countabil­ity bud­dies: a pro­posed system

  Hey Holly! Since I read this post I found an ac­countabil­ity part­ner a bit later. The sys­tem works well and we’ve built it out ac­cord­ing to our own needs, but I like that you’ve kept it sim­ple. A lot of sys­tems and life­hacks are very spe­cific which makes them a pain to im­ple­ment.

  So thank you and good job! :)

  IF I’d make an ad­di­tion to your ad­vice, it would prob­a­bly be: plan a brief eval­u­a­tion af­ter x (3? 6?) months to take a big­ger pic­ture:

  • Has the sys­tem ac­tu­ally helped you ac­com­plish things, rather than just be­ing pro­duc­tive?

  • How do you like the role of each other? Would you like them to be more strict? Would you like them to fo­cus on spe­cific ques­tions or pit­falls? (E.g. I asked my buddy to guard my from tak­ing on too many re­spon­si­bil­ities—when I plan a new pro­ject she asks me “ok, so what are you go­ing to drop to make place for it?“)

• SiebeRozendal 20 Dec 2018 22:31 UTC

  9 points

  on: Women’s Em­pow­er­ment: Founders Pledge re­port and recommendations

  Thanks for post­ing this. Some com­ments and ques­tions:

  I echo Habryka’s re­luc­tance about cause-par­tial re­search, and I would have ap­pre­ci­ated if you’d shared a lit­tle more con­text for EA’s, given that this is the EA Fo­rum. (for ex­am­ple, why Founders Pledge de­cided to re­search this, how it’s rele­vant from an EA per­spec­tive and how it’s not, which as­sump­tions it makes)

  Some spe­cific ques­tions:

  1. Why haven’t you com­pared char­i­ties on a com­pa­rably met­ric, such as the DALY, or Life-Satis­fac­tion Points? Is this be­cause the re­port as­sumes what is (in­trin­si­cally or ro­bustly) im­por­tant is em­pow­er­ment, and that is hard to mea­sure?

  2. Why do you recom­mend both Village En­ter­prise and Band­han’s ‘Tar­get­ing the Hard­core Poor’ pro­gramme? The lat­ter ap­pears 3.5 times as cost-effec­tive (in terms of in­creas­ing pur­chas­ing power), and the or­gani­sa­tions ap­pear to have similar strength, room for more fund­ing, and wider benefits.

  3. Do you have a pub­li­cly available cost-effec­tive­ness model for the 15 char­i­ties con­sid­ered? By how much did these char­i­ties differ on com­pa­rable mea­sures?

• SiebeRozendal 18 Dec 2018 13:45 UTC

  2 points

  in reply to: Jonas Vollmer's comment on: Launch­ing the EAF Fund

  We always point out that the fund is fo­cused on re­duc­ing suffer­ing in the long-term fu­ture.

  Also, why should they donate to that other fund in­stead? E.g., the Long-Term Fu­ture Fund is also im­por­tantly mo­ti­vated by “as­tro­nom­i­cal waste” type con­sid­er­a­tions which those donors don’t un­der­stand ei­ther, and might not agree with.

  Yes, I’m not say­ing you’re mis­lead­ing your donors, nor that they are less in­formed than donors of other funds. Just that there are many rea­sons peo­ple are donat­ing to a par­tic­u­lar fund, and I think prop­erly nam­ing a fund is a step in the right di­rec­tion.

  I wouldn’t call it a co­or­di­na­tion prob­lem in the game-the­o­retic sense

  I see it as co­or­di­na­tion be­tween differ­ent fund man­agers, where each wants to max­i­mize the amount of funds for their own fund. As such, there are some in­cen­tives to not max­i­mally in­form one’s donors if other fund­ing pos­si­bil­ities, or the best ar­gu­ments against donat­ing to the fund they are fundrais­ing for.

  I’m not say­ing that this type of self­ish be­hav­ior is very pre­sent in the EA com­mu­nity—I’ve heard that it is quite the op­po­site. But I do think that the situ­a­tion is not yet op­ti­mal: cur­rent al­lo­ca­tion of re­sources is not based largely on care­ful weigh­ing the rele­vant ev­i­dence and ar­gu­ments. I also think we can move closer to this op­ti­mal al­lo­ca­tion.

  Any­way, I didn’t mean to make this into a large de­bate :) I’m glad the fund ex­ists and I’d be happy if the name changes to some­thing more dis­t­in­guish­able!

• SiebeRozendal 17 Dec 2018 13:37 UTC

  2 points

  in reply to: Jonas Vollmer's comment on: Launch­ing the EAF Fund

  I still per­ceive it as sub­op­ti­mal, al­though I un­der­stand you don’t like any of the po­ten­tial names.

  We se­ri­ously con­sid­ered “S-risk Fund” but ul­ti­mately de­cided against be­cause it seems harder to fundraise for from peo­ple who are less fa­mil­iar with ad­vanced EA con­cepts (e.g., poker pros in­ter­ested in im­prov­ing the long-term fu­ture).

  I think this touches on a se­ri­ous worry with fundrais­ing from peo­ple un­fa­mil­iar with EA con­cepts: why should they donate to your fund rather than an­other EA fund if they don’t un­der­stand the ba­sic goal you are aiming for with your fund?

  I can imag­ine that peo­ple donate to the EAF fund for so­cial rea­sons (e.g. you hap­pen to be well-con­nected to poker play­ers) more than in­tel­lec­tual rea­sons (i.e. fun­ders donate be­cause they pri­ori­tize s-risk re­duc­tion). If that is the case, I’d find it prob­le­matic that the fund is not clearly named: it makes it less likely that peo­ple donate to par­tic­u­lar funds for the right rea­sons.

  Of course, this is part of a larger co­or­di­na­tion prob­lem in which all kinds of non-in­tel­lec­tual rea­sons are driv­ing dona­tion de­ci­sions. I am not sure what the ideal solu­tion is*, but I wanted to flag this is­sue.

  *Per­haps it should be a best prac­tice for EA fundraisers to always recom­mend fun­ders to all go through a (to be cre­ated) dona­tion de­ci­sion tool that takes them through some of the rele­vant ques­tions. That tool would be a bit like the flowchart from the Global Pri­ori­ties Pro­ject, but more user-friendly.

• SiebeRozendal 4 Dec 2018 7:54 UTC

  3 points

  on: Launch­ing the EAF Fund

  Great to see this set up!

  Small note on the name: is not at all clear for peo­ple un­fa­mil­iar with EAF what the fund is about, and a name change for the fund would prob­a­bly get you more at­ten­tion. Some­thing like “suffer­ing pre­ven­tion fund”, “suffer­ing re­duc­tion fund”, “long-term suffer­ing re­duc­tion fund”, “suffer­ing risk fund” would all be sig­nifi­cantly clearer (even though they feel in­ad­e­quate to de­scribe the fund’s goals).

• SiebeRozendal 20 Nov 2018 11:49 UTC

  4 points

  in reply to: Halstead's comment on: William MacAskill mis­rep­re­sents much of the ev­i­dence un­der­ly­ing his key ar­gu­ments in “Do­ing Good Bet­ter”

  Okay, points taken. I should have been much more care­ful given the strength of the ac­cu­sa­tion, and the ac­cu­sa­tion that DGB was writ­ten “in bad faith” seems (far) too strong.

  I guess I have a ten­dency to sup­port efforts that challenge com­mon be­liefs that might not be held for the right rea­sons (in this case “DGB is a rigourously writ­ten book, and a good in­tro­duc­tion to effec­tive al­tru­ism”). This seemed to out­weigh the costs of crit­i­cism, likely be­cause my in­tu­ition of­ten un­der­es­ti­mates the costs of crit­i­cism. How­ever, the OP challenged a much stronger com­mon be­lief (“Will MacAskill is not an un­trust­wor­thy per­son”) and I should have bet­ter dis­t­in­guished those (both in my mind and in writ­ing).

  When I was writ­ing it, I was very doubt­ful about whether I was phras­ing it cor­rectly, and I don’t think I suc­ceeded. I think my in­ten­tion for “writ­ten in bad faith” was meant less strongly, but a bit more than ‘writ­ten with less at­ten­tion to de­tail than it could have been’: i.e. that less at­ten­tion was given to de­tails that wouldn’t pan out in favour of EA. More along the lines of this:

  “sloppy, and per­haps with a sub­con­scious finger on the scale tilt­ing the er­rors to be favourable to the the­sis of the book” rather than de­ceit, mal­ice, or other ‘bad faith’.

  I also have a lower cre­dence in this now. I should add that my use of “con­vinc­ing” was also too strong a term, as it might be in­ter­preted as >95% cre­dence, in­stead of the >60% cre­dence I ob­served at the time of writ­ing.

• SiebeRozendal 17 Nov 2018 14:19 UTC

  0 points

  on: William MacAskill mis­rep­re­sents much of the ev­i­dence un­der­ly­ing his key ar­gu­ments in “Do­ing Good Bet­ter”

  [EDIT: this was not a very care­ful com­ment and mul­ti­ple claims were stated more strongly than I be­lieved them, as well that my be­liefs might have been not so well-sup­ported]

  I ad­mire the amount of effort that has gone into this post and its level of rigor. I think it’s very im­por­tant for an epistem­i­cally healthy move­ment that high-sta­tus peo­ple can be crit­i­cised suc­cess­fully.

  I think your premises do not fully sup­port the con­clu­sion that MacAskill is com­pletely un­trust­wor­thy. How­ever, I agree that the book mis­rep­re­sents sources struc­turally, and this is a con­vinc­ing sign it is writ­ten in bad faith.

  I hope that MacAskill has already re­al­ized the book was not up to the stan­dards he now pro­motes. Writ­ing an in­tro­duc­tion to effec­tive al­tru­ism was and re­mains a very difficult task, and at the time there was still a mind­set of “push EA even if it’s at the cost of some epistemic hon­esty”. I think the com­mu­nity has been mov­ing away from this mind­set since, and this post is a good ad­di­tion to that.

  We need a bet­ter in­tro­duc­tory book. (Also be­cause it’s out­dated.)

• SiebeRozendal 13 Nov 2018 10:41 UTC

  2 points

  on: Which World Gets Saved

  But we might op­pose efforts like the Nu­clear Threat Ini­ti­a­tive, which dis­pro­por­tionately save vi­o­lent-psy­chol­ogy wor­lds.

  Does it? It seems a lot of risk comes from ac­ci­den­tal catas­tro­phes rather than in­ten­tional ones. Ac­ci­den­tal catas­tro­phes to me don’t seem like proof of the fu­ture be­ing vi­o­lent.

  Also, I think that we should treat our efforts to re­duce the risk of in­ten­tional catas­tro­phe or in­flicted suffer­ing as ev­i­dence. Why wouldn’t the fact that we choose to re­duce the im­pact of mal­i­cious ac­tors be proof that mal­i­cious ac­tors’ im­pact will be cur­tailed by other ac­tors in the fu­ture?

• SiebeRozendal 13 Nov 2018 10:37 UTC

  2 points

  in reply to: aarongertler's comment on: Which World Gets Saved

  I’ve been read­ing Phil Tor­res’s book on ex­is­ten­tial risks and I agree with him to the ex­tent that peo­ple have been too dis­mis­sive about the amount of om­ni­ci­dal agents or their ca­pa­bil­ity to de­stroy the world. I think his re­ac­tion to Pinker would be that the level of com­pe­tence needed to cre­ate dis­rup­tion is de­creas­ing be­cause of tech­nolog­i­cal de­vel­op­ment. There­fore, his­tor­i­cal prece­dent is not a great guide. See: Who would de­stroy the world? Om­ni­ci­dal agents and re­lated phe­nom­ena

  Ab­stract:

  The ca­pac­ity for small groups and even sin­gle in­di­vi­d­u­als to wreak un­prece­dented havoc on civ­i­liza­tion is grow­ing as a re­sult of dual-use emerg­ing tech­nolo­gies. This means that schol­ars should be in­creas­ingly con­cerned about in­di­vi­d­u­als who ex­press om­ni­ci­dal, mass geno­ci­dal, anti-civ­i­liza­tional, or apoc­a­lyp­tic be­liefs/​de­sires. The pre­sent ar­ti­cle offers a com­pre­hen­sive and sys­tem­atic sur­vey of ac­tual in­di­vi­d­u­als who have har­bored a death wish for hu­man­ity or de­struc­tion wish for civ­i­liza­tion. This pa­per thus pro­vides a strong foun­da­tion for fu­ture re­search on “agen­tial risks” and re­lated is­sues. It could also serve as a helpful re­source for coun­tert­er­ror­ism ex­perts and global risk schol­ars who wish to bet­ter un­der­stand our evolv­ing threat en­vi­ron­ment.

• SiebeRozendal 13 Nov 2018 10:24 UTC

  3 points

  on: W-Risk and the Tech­nolog­i­cal Wavefront (Nell Wat­son)

  It’s an in­ter­est­ing ques­tion to ask how likely it is to re­cover from civ­i­liza­tional col­lapse, and talk­ing about ‘step­ping down in com­plex­ity’ might be use­ful. I’ve pre­vi­ously only seen it dis­cussed as whether we lose agri­cul­ture, sci­ence, or in­dus­try (see e.g. Baum et al., 2018). It seems the au­thor is im­plic­itly refer­ring to the En­ergy-Com­plex­ity Spiral by Joseph Tain­ter, a fas­ci­nat­ing con­cept:

  The com­mon view of his­tory as­sumes that com­plex­ity and re­source con­sump­tion have emerged through in­no­va­tion fa­cil­i­tated by sur­plus en­ergy. This view leads to the sup­po­si­tion that com­plex­ity and con­sump­tion are vol­un­tary, and that we can there­fore achieve a sus­tain­able fu­ture through con­ser­va­tion. Such an as­sump­tion is sub­stan­tially in­cor­rect. His­tory sug­gests that com­plex­ity most com­monly in­creases to solve prob­lems, and com­pels in­crease in re­source use. This pro­cess is illus­trated by the his­tory of the Ro­man Em­pire and its col­lapse. Prob­lems are in­evitable, re­quiring in­creas­ing com­plex­ity, and con­ser­va­tion is there­fore in­suffi­cient to pro­duce sus­tain­abil­ity.

  It seems most x-risk schol­ars be­lieve the prob­a­bil­ity of re­cov­ery is re­ally high (>90%) as long as some­thing like the sci­en­tific method is pre­served (last few peo­ple prob­lem). I think this is likely to be cor­rect, and that the failed re­cov­er­ies are ei­ther by ex­tinc­tion (70% of failed re­cov­er­ies) or by loss of the sci­en­tific method (30% of failed re­cov­er­ies). Per­ma­nent loss of tech­nol­ogy seems un­likely to me, as tech­nolog­i­cal de­vel­op­ment offers many ad­van­tages and is ob­served in most cul­tures.

• SiebeRozendal 13 Nov 2018 10:01 UTC

  6 points

  on: Cross-post: Think twice be­fore talk­ing about ‘tal­ent gaps’ – clar­ify­ing nine mis­con­cep­tions, by 80,000 Hours.

  To help com­mu­ni­cate this, it may be helpful if or­gani­sa­tions pub­lished typ­i­cal ra­tios of ap­pli­cants to hires to let peo­ple plan ac­cord­ingly.

  This would be helpful in­deed! Although I sus­pect or­ga­ni­za­tions would be re­luc­tant to do this as it might put off ap­pli­ca­tions. On the other hand, it might turn off ap­pli­ca­tions that would not be good any­way and save a lot of work. This data would also be use­ful for job seek­ers to de­cide whether they should ap­ply mostly within EA, or fo­cus efforts on im­pact­ful op­por­tu­ni­ties out­side EA (like me). If there are 1,000 ap­pli­cants for 50 po­si­tions yearly, I should not bet too much on get­ting into an EA org. I think this should es­pe­cially en­courage ap­pli­cants who don’t have the right skills or knowl­edge yet to ap­ply out­side EA to skill up, in­stead of hang­ing around in EA limbo.

  I think it is worth think­ing more about how the ap­pli­ca­tion pro­cesses of EA orgs can cre­ate more value (for the com­mu­nity). I think it is an op­por­tu­nity for the or­ga­ni­za­tion to help the wider com­mu­nity: re­view­ing hun­dreds of ap­pli­cants is a unique data set. One con­crete ac­tion is pub­lish­ing or send­ing gen­eral feed­back about what differ­en­ti­ated top ap­pli­cants from non-top ap­pli­cants, which skills/​back­grounds were in wide sup­ply and which were rare and use­ful. Th­ese can also be referred to for later hiring pro­cesses for similar roles. There’s more dis­cus­sion about this topic in this Face­book post.

• SiebeRozendal 13 Nov 2018 9:36 UTC

  4 points

  in reply to: aarongertler's comment on: Keep­ing Ab­solutes in Mind

  Some­thing along those lines. Thanks for in­ter­pret­ing! :)

  What I was get­ting at was mostly that praise/​recog­ni­tion should be a smooth func­tion, such that things not branded as EA still get recog­ni­tion if they’re only 1/​10th effec­tive, in­stead of the cur­rent situ­a­tion (as I per­ceive it) that when some­thing is not max­i­mally effec­tive it’s not rec­og­nized at all. I no­tice in my­self that I find it harder to as­sess and rec­og­nize the im­pact of non-EA branded pro­jects.

  I ex­pect this is partly be­cause I don’t have ac­cess/​don’t un­der­stand the rea­son­ing so can’t as­sess the ex­pected value, but partly be­cause I’m nor­mally lean­ing on sta­tus for EA-branded pro­jects. For ex­am­ple, if Will MacAskill is start­ing a new pro­ject I will pre­dict it’s go­ing to be quite effec­tive with­out know­ing any­thing about the pro­ject, while I’d be skep­ti­cal about an un­known EA.

• SiebeRozendal 7 Nov 2018 9:43 UTC

  3 points

  on: Keep­ing Ab­solutes in Mind

  Ab­solutely agree! :) I think this also ex­tends to “non-EA” causes and pro­jects that do good: sure, they’re not most effec­tive, but they’re still im­prov­ing or sav­ing lives and that’s praise­wor­thy.

  Re­lat­edly, I think it’s hard to be mo­ti­vated by sub­jec­tive ex­pected value, even if that’s what most peo­ple think we should max­i­mize. When some­thing turns out to not be suc­cess­ful al­though it was re­ally high ex­pected value (so not the re­sult of bad anal­y­sis), the ac­tion should be praised. I’m afraid that the rank­ing of ac­tions by ex­pected value di­verges sig­nifi­cantly from the rank­ing by ex­pected recog­ni­tion (from one­self and oth­ers) and I think this should be some­what wor­ry­ing.

  Com­ing back to the post, I also think the drop in recog­ni­tion is too large when ab­solute value re­al­ized is not max­i­mal. I’m cu­ri­ous to figure out what is the op­ti­mal recog­ni­tion func­tion is (square root of ex­pected value?), but I think that’s a bit be­sides the point of this post!

• SiebeRozendal 24 Oct 2018 12:54 UTC

  0 points

  on: EA Con­cepts: Share Im­pres­sions Be­fore Credences

  It would be use­ful to define what you mean by im­pres­sions and cre­dences. Espe­cially “cre­dences” is not a word com­monly used out­side of EA or fields re­lated to de­ci­sion the­ory.

• SiebeRozendal 24 Oct 2018 12:46 UTC

  2 points

  in reply to: jpaddison's comment on: On Be­com­ing World-Class

  Hey JP, I don’t mean to sound con­de­scend­ing or polic­ing, but I find this com­ment to be full of (un­nec­es­sary) jar­gon which would be very hard to read for new­com­ers. What would help:

  • Writ­ing out ab­bre­vi­a­tions (“EV” --> “ex­pected value”)

  • Writ­ing out the rea­son­ing in­stead of a con­cept (“the value of the op­por­tu­nity cost” --> “the benefits of pur­su­ing a ‘tra­di­tional’ EA-al­igned ca­reer)

• SiebeRozendal 14 Oct 2018 18:14 UTC

  4 points

  in reply to: Denise_Melchin's comment on: Many EA orgs say they place a lot of fi­nan­cial value on their pre­vi­ous hire. What does that mean, if any­thing? And why aren’t they hiring faster?

  I just want to note that not ev­ery re­jected ap­pli­ca­tion has been burnt value for me and most have ac­tu­ally been pos­i­tive, es­pe­cially in terms of things learned. In the ones I got far it has re­sulted in more rather than less mo­ti­va­tion. In the case I had to do work-re­lated tasks (write re­search pro­posal, or ex­e­cute a sam­ple of typ­i­cal re­search) I learned a lot.

  On the other hand, in­creas­ing the ap­pli­cants:hired-ra­tio would mostly in­crease the pro­por­tion of peo­ple not get­ting far in the ap­pli­ca­tion pro­cess which is where least of the value pos­i­tive fac­tors are and most of the nega­tive.

• SiebeRozendal 13 Oct 2018 10:03 UTC

  13 points

  in reply to: jpaddison's comment on: Many EA orgs say they place a lot of fi­nan­cial value on their pre­vi­ous hire. What does that mean, if any­thing? And why aren’t they hiring faster?

  The nar­ra­tive that “EA is tal­ent con­strained” has had, as men­tioned, some nega­tive and un­in­tended con­se­quences. One more I’d like to add is that on this nar­ra­tive the ad­vised ac­tion “go work for a larger or­ga­ni­za­tion, like the gov­ern­ment” might feel for many peo­ple as failure or “not liv­ing up to their goals.” I’m afraid this leads to more value drift—peo­ple’s val­ues sid­ing away from EA-al­igned val­ues—be­cause they feel they’re not “liv­ing EA” any­way.

• SiebeRozendal 13 Oct 2018 9:57 UTC

  6 points

  in reply to: jpaddison's comment on: Many EA orgs say they place a lot of fi­nan­cial value on their pre­vi­ous hire. What does that mean, if any­thing? And why aren’t they hiring faster?

  A first thought I have is “if the cur­rent or­ga­ni­za­tions can­not grow fast enough to use all the available tal­ent, why not in­crease the growth rate of the move­ment?” That is, should we start more (paid) pro­jects?

  I think there is a case to be made that we should: there seem to be many small pro­jects with ca­pa­ble peo­ple seek­ing fund­ing. They are not as effec­tive per em­ployee as some of the core or­ga­ni­za­tions, but the core or­ga­ni­za­tions don’t need the money. I don’t hold this opinion strongly, and I’m cu­ri­ous what other peo­ple think.

• SiebeRozendal 8 Sep 2018 16:33 UTC

  2 points

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

  Roughly when will the new fo­rum be launched, if I may ask?