Katja_Grace

• Katja_Grace 13 Dec 2017 20:51 UTC

  5 points

  on: Four Or­ga­ni­za­tions EAs Should Fully Fund for 2018

  Do you have quan­ti­ta­tive views on the effec­tive­ness of donat­ing these or­ga­ni­za­tions, that could be com­pared to other ac­tions? (Or could you point me to any of the links go to some­thing like that?) Sorry if I missed them.

• Katja_Grace 27 Oct 2017 4:23 UTC

  7 points

  in reply to: Stefan_Schubert's comment on: Why & How to Make Progress on Diver­sity & In­clu­sion in EA

  It seems worth dis­t­in­guish­ing ‘effec­tive­ness’ in the sense of per­sonal com­pe­tence (as I guess is meant in the first case, e.g. ‘rea­son­ably sharp’) and ‘effec­tive­ness’ in the sense of try­ing to choose in­ter­ven­tions by cost-effec­tive­ness.

  Also re­mem­ber that se­lect­ing peo­ple to en­courage in par­tic­u­lar di­rec­tions is a sub­set of se­lect­ing in­ter­ven­tions. It may be that ‘E not A’ peo­ple are more likely to be helpful than ‘A not E’ peo­ple, but that chas­ing ei­ther group is less helpful than do­ing re­search on E that is helpful for whichever peo­ple already care about it. I think I have stronger feel­ings about E-im­prov­ing in­ter­ven­tions over­all be­ing good than about which peo­ple are more promis­ing al­lies.

• Katja_Grace 2 Jun 2015 4:23 UTC

  1 point

  in reply to: arrowind's comment on: Col­lec­tive Ac­tion and In­di­vi­d­ual Impact

  Yeah, and among com­mon in­tu­itions I think. But I thought EAs were mostly con­se­quen­tial­ists, so the in­tended role of obli­ga­tions is not ob­vi­ous to me.

• Katja_Grace 28 May 2015 16:22 UTC

  3 points

  on: Col­lec­tive Ac­tion and In­di­vi­d­ual Impact

  I’m cu­ri­ous about the im­plicit frame­work where some things are obli­ga­tory and some things are choices.

Im­pact Pur­chase: Round 2

• Katja_Grace 10 Apr 2015 15:30 UTC

  3 points

  in reply to: Evan_Gaensbauer's comment on: Im­pact pur­chase first round results

  We eval­u­ated all of the pro­jects other than the three I speci­fi­cally men­tioned not eval­u­at­ing. Sorry for not writ­ing up the other eval­u­a­tions—we just didn’t have time. We bought the ones that gave us the most im­pact per dol­lar, ac­cord­ing to our eval­u­a­tions (and based on the prices peo­ple wanted for their work). So we didn’t pur­chase Joao’s work this round be­cause we calcu­lated that it was some­what less cost-effec­tive than the things we did pur­chase, given the price. We may still pur­chase it in a later round.

Im­pact pur­chase first round results

The econ­omy of weirdness

When should an Effec­tive Altru­ist be veg­e­tar­ian?

• Katja_Grace 8 Nov 2014 10:14 UTC

  2 points

  in reply to: Evan_Gaensbauer's comment on: Why is effec­tive al­tru­ism new and ob­vi­ous?

  Chang­ing one’s val­ues does not more effec­tively pro­mote the val­ues one has ini­tially, so it seems one should be averse to it. I think the ex­pand­ing cir­cle case is more com­pli­cated—the ad­vo­cates of a wider cir­cle are try­ing to con­vince the oth­ers that those oth­ers are mis­taken about their own ex­ist­ing val­ues, and that by con­sis­tency they must care about some en­tities they think they don’t care about. This is why the phe­nomenon looks like an ex­pand­ing cir­cle—points just out­side a cir­cle look a lot like points just in­side it, so con­sis­tency pushes the cir­cle out­wards (this doesn’t ex­plain why the cir­cle ex­pands rather than con­tract­ing).

• Katja_Grace 7 Oct 2014 16:09 UTC

  1 point

  on: Sup­port­ive Scepticism

  It seems there are some com­mon states where this comes up, such as when one per­son is do­ing a thing which they think is good, given per­sonal con­straints which are hid­den to their con­ver­sa­tion part­ner, and wor­ries that they are harshly judged be­cause the con­straints are hid­den. Or where one per­son is try­ing out a thing, be­cause they think it might be very good, how­ever they don’t already think it is very good (ex­cept for VOI), and worry that oth­ers think they are ac­tu­ally ad­vo­cat­ing for some­thing sub­op­ti­mal. Or where one per­son doesn’t think what they are do­ing is likely to be op­ti­mal, but strug­gles to find some­thing ac­tu­ally bet­ter that they could fea­si­bly do.

  Per­haps it would be helpful if there was a thing you could say in these rec­og­nized cir­cum­stances to let your con­ver­sa­tion part­ner know that you know that what you are do­ing doesn’t look op­ti­mal, and you are already aware of the situ­a­tion.

• Katja_Grace 2 Oct 2014 7:17 UTC

  2 points

  in reply to: atucker's comment on: Why is effec­tive al­tru­ism new and ob­vi­ous?

  the other is that the par­tic­u­lar style in which the EA com­mu­nity pur­sues that idea (look­ing for in­ter­ven­tions with ro­bust aca­demic ev­i­dence of effi­cacy, and then sup­port­ing or­ga­ni­za­tions im­ple­ment­ing those in­ter­ven­tions that ac­countably have a high amount of in­ter­ven­tion per marginal dol­lar) is novel, but mostly be­cause the cul­tural back­ground for it seem­ing pos­si­ble as an op­tion at all is new.

  The kinds of ev­i­dence available for some EA in­ter­ven­tions, e.g. ex­is­ten­tial risk ones, doesn’t seem differ­ent in kind to the ev­i­dence prob­a­bly available ear­lier in his­tory. Even in the best cases, EAs of­ten have to lean on a com­bi­na­tion of more rigor­ous ev­i­dence and some not very rigor­ous or ev­i­denced guesses about how in­di­rect effects work out etc. So if the more rigor­ous ev­i­dence available were sub­stan­tially less rigor­ous than it is, I think I would ex­pect things to look pretty much the same, with us just hav­ing lower stan­dards—e.g. only be­ing will­ing to trust cer­tain peo­ple’s re­ports of how in­ter­ven­tions were go­ing. So I’m not con­vinced that some re­cently at­tained level of good ev­i­dence has much to do with the over­all phe­nomenon of EA.

• Katja_Grace 2 Oct 2014 5:35 UTC

  0 points

  in reply to: Ben_Kuhn's comment on: Why is effec­tive al­tru­ism new and ob­vi­ous?

  I agree that those I men­tioned are prob­a­bly not the only con­tentious claims, and that the one you men­tion in par­tic­u­lar is prob­a­bly an­other.

• Katja_Grace 2 Oct 2014 5:34 UTC

  2 points

  in reply to: Peter_Hurford's comment on: Why is effec­tive al­tru­ism new and ob­vi­ous?

  In­ter­est­ing sug­ges­tions.

  I’d ex­pect the in­ter­net to make many minor­ity causes and in­ter­ests more suc­cess­ful by let­ting their rare sup­port­ers get to­gether, and I think it has had this effect. How­ever that doesn’t seem to ex­plain why they are minor­ity causes to be­gin with.

  Do you mean that be­fore com­puter pro­gram­ming the philo­soph­i­cally minded just didn’t have lu­cra­tive pro­fes­sions?

  Have we re­cently passed some thresh­old in high qual­ity ev­i­dence for what works in aid? I’d ex­pect in fu­ture we think of 2014 level of ev­i­dence as low, and still say we only re­cently got good ev­i­dence.

Why is effec­tive al­tru­ism new and ob­vi­ous?

• Katja_Grace 16 Sep 2014 3:17 UTC

  1 point

  on: Su­per­in­tel­li­gence read­ing group

  This has be­gun: http://​​less­wrong.com/​​r/​​dis­cus­sion/​​lw/​​ku6/​​su­per­in­tel­li­gence_read­ing_group_sec­tion_1_past/​​

• Katja_Grace 10 Sep 2014 9:02 UTC

  0 points

  on: High Im­pact Science

  The last four para­graphs of this post are re­peated from ear­lier, and ap­pear to be cut­ting out some of the origi­nal post.

• Katja_Grace 10 Sep 2014 8:50 UTC

  14 points

  on: Cosmopolitanism

  Good point. I have won­dered be­fore whether some­thing like anti-cos­mopoli­tanism some­times ac­counts for dis­agree­ment that is at­tributed to anti-effec­tive­ness or anti-al­tru­ism. Cos­mopoli­tanism seems like a much more plau­si­ble thing for a hu­man to dis­like than effec­tive­ness, es­pe­cially given the pop­u­lar­ity of effec­tive­ness in other ar­eas of life.

Su­per­in­tel­li­gence read­ing group

Ap­ples and or­anges? Some ini­tial thoughts on com­par­ing di­verse benefits