Bluefalcon

• Bluefalcon 21 Mar 2019 17:58 UTC

  1 point

  on: Can­di­date Scor­ing Sys­tem, Se­cond Release

  De­laney’s hit­ting 2% on Pre­dic­tit for the first time AFAIK. Did your quasi-en­dorse­ment move the mar­kets?

• Bluefalcon 14 Feb 2019 4:36 UTC

  5 points

  in reply to: Aidan O'Gara's comment on: A sys­tem for scor­ing poli­ti­cal can­di­dates. RFC (re­quest for com­ments) on method­ol­ogy and positions

  This is awe­some and I’ve been want­ing some­thing like it but am too lazy to cre­ate it my­self. So I’m re­ally glad kbog did.

  I vote for con­tin­u­ing to in­clude weight­ings for e.g. can­di­date health. The in­ter­est­ing ques­tion is who is ac­tu­ally likely to do the most good, not who be­lieves the best things. So to model that well you need to cap­ture any per­sonal fac­tors that sig­nifi­cantly af­fect their prob­a­bil­ity of car­ry­ing out their agenda.

  I think AI safety and biorisk de­serve some weight­ing here even if can­di­dates aren’t ad­dress­ing them di­rectly. You could use proxy is­sues that the can­di­dates are more likely to have records on and that rele­vant ex­perts have a con­sen­sus are helpful or un­helpful (e.g. ac­tions likely to lead to an arms race with China). And then ad­just for un­cer­tainty by giv­ing them a some­what lower weight than you would give a di­rect vote on some­thing like cre­at­ing an un­friendly AI.

• Bluefalcon 13 Jan 2019 4:46 UTC

  1 point

  in reply to: Denkenberger's comment on: The Global Pri­ori­ties of the Copen­hagen Consensus

  agreed

• Bluefalcon 11 Jan 2019 6:55 UTC

  1 point

  on: The Global Pri­ori­ties of the Copen­hagen Consensus

  I’d be in­ter­ested in see­ing why they rate malaria so much lower, at least in rel­a­tive terms, than most of the EA com­mu­nity does. That’s prob­a­bly a good clue to the differ­ences in method­ol­ogy, and a short­cut to figur­ing out whose meth­ods yield more ac­cu­rate pri­ori­ties.

  P.S. I’m not sur­prised that mea­sur­ing the UN de­vel­op­ment goals is un­pro­duc­tive; a lot of them are ob­vi­ously dis­trac­tions. Pri­ori­ties re­search only adds value when it’s non-ob­vi­ous whether some­thing should be a pri­or­ity. Once it’s clear that the goal is garbage, move on.

• Bluefalcon 21 Dec 2018 2:09 UTC

  17 points

  on: Re­sponse to a Dy­lan Matthews ar­ti­cle on Vox about bipartisanship

  Even if your policy views are cor­rect, hav­ing friends on the other side of the aisle will do won­ders for your pre­dic­tive abil­ities, which should in­fluence how you vote in party pri­maries where electabil­ity is at is­sue. I’m a staunch Demo­crat cur­rently liv­ing in a similarly liberal area but born and raised in a much more con­ser­va­tive area. I was always more bullish on Trump’s odds than my friends here, and ev­ery time I hear them say they can’t un­der­stand how he got elected or how he’s still pop­u­lar with the base, I won­der what other eas­ily avoid­able mis­takes they’re cur­rently mak­ing. And there’s no spe­cial magic in my im­proved pre­dic­tive abil­ity; I just talk to my grandma reg­u­larly.

• Bluefalcon 18 Dec 2018 19:45 UTC

  2 points

  on: Col­lege and Earn­ing to Give

  Or let them pay for their own col­lege. The rates they’ll get charged will take your in­come into ac­count, so they’ll have to get a lot of loans. But the loans are pretty man­age­able and es­sen­tially risk-free as long as in­come-based re­pay­ment re­mains an op­tion. Now, if they are com­mit­ted EAs and will be earn­ing to give, this doesn’t ac­com­plish as much since this comes out of their fu­ture earn­ings. But even if there’s a 95% prob­a­bil­ity that they turn out to be earn­ing to give as EAs, that’s still bet­ter than the ~100% odds with re­spect to you. And the in­ter­est rates on those loans are lower than av­er­age stock mar­ket re­turns, so the smart fi­nan­cial move is always to get the loans and pay them off as slowly as pos­si­ble rather than pay higher up front costs (as­sum­ing they’re rea­son­ably risk tol­er­ant, as young peo­ple should be but es­pe­cially when the goal is to max­i­mize good done rather than max­i­mize their own com­fort) (this is a lit­tle more com­pli­cated to see when we’re talk­ing about earn­ing to give. but you must be as­sum­ing a larger dis­count rate for the value of dona­tions than what you could earn by in­vest­ing and giv­ing later; oth­er­wise you’d be do­ing that in­stead of giv­ing now. So if giv­ing the money away now is worth sac­ri­fic­ing the ~10% an­nual gains you could make on it, then it’s even more worth sac­ri­fic­ing ~7%/​year in in­ter­est pay­ments).

• Bluefalcon 11 Dec 2018 18:44 UTC

  2 points

  on: Giv­ing more won’t make you hap­pier

  I don’t think life eval­u­a­tion is the right mea­sure­ment of hap­piness for this though. I’m pretty ego­tis­ti­cal and more in­come would definitely make me hap­pier, well be­yond the thresh­old here. I make about 100k/​year and would definitely be more satis­fied with my life if I made 200k. But, that’s purely from a “feel­ing like a suc­cess”/​so­cial sta­tus stand­point. I have no re­al­is­tic lifestyle use for 200k. I can’t even figure out how to use 100k on lifestyle be­cause the things peo­ple typ­i­cally buy are re­ally bor­ing. So even though mak­ing more money would make me hap­pier, spend­ing more of it on my­self wouldn’t. Pos­i­tive af­fect is bet­ter but may still be af­fected too much by life eval­u­a­tion. So nega­tive af­fect is prob­a­bly the best mea­sure, since it’s more re­lated to the pos­si­bil­ity of suffer­ing due to in­suffi­cient money.

• Bluefalcon 9 Dec 2018 10:23 UTC

  2 points

  in reply to: Michael_S's comment on: [Link] Vox Ar­ti­cle on Eng­ineered Pathogens/​ Global Catas­trophic Biorisks

  Maybe I should’ve asked you the ques­tion I just asked on an­other post in­stead: as some­one in­ter­ested in min­i­miz­ing x-risk, who should I sup­port for Pres­i­dent? Or bet­ter yet, who has a good com­pila­tion of can­di­dates’ records on x-risk-re­lated is­sues, so I can make my own de­ci­sion?

• Bluefalcon 9 Dec 2018 10:00 UTC

  4 points

  on: Ex­is­ten­tial risk as com­mon cause

  As some­one with an in­ter­est in gov­ern­ment and rel­a­tively new to the con­cept of x-risk, I have a semi-ur­gent ques­tion: who should I sup­port for Pres­i­dent? I will prob­a­bly have to get in­volved with a cam­paign in some way or an­other in the next few months to max­i­mize my odds of get­ting a de­cent ap­point­ment af­ter the elec­tion. There’s plenty of in­ter­est group rat­ings, po­si­tion state­ments etc. out there on en­vi­ron­men­tal is­sues but I can’t find much that would be of prac­ti­cal use on the other types, which seem to be more se­ri­ous at least in ag­gre­gate and per­haps in­di­vi­d­u­ally too. I could try com­piling my own rat­ings but I know far less than a lot of the peo­ple in this com­mu­nity, so if some­one has already figured out or is in the pro­cess of figur­ing out where the can­di­dates stand on the risks they have ex­per­tise in, I would greatly ap­pre­ci­ate it. Doesn’t have to be like stan­dard in­ter­est group rat­ings and maybe shouldn’t be. E.g. the fact that some­one has a hawk­ish tem­per­a­ment to­ward China and that would make them more prone to start­ing an arms race is prob­a­bly more im­por­tant to AI safety than the speci­fics of any tech­nol­ogy-re­lated votes they’ve taken.