Ramiro

• Ramiro 29 Mar 2020 23:44 UTC

  3 points

  on: [Link] Ex­pert Com­mu­ni­ties and Public Revolt

  Robert analo­gized this illus­tra­tion to how pub­lic trust in ex­pert com­mu­ni­ties of economists plum­meted af­ter the Great Recession

  And yet, I guess cen­tral banks across the world are usu­ally shielded from poli­tics; and you can ob­serve more in­ter­na­tional co­op­er­a­tion and con­ver­gence in macroe­co­nomics (e.g., see how gov­ern­ments and cen­tral banks adopted similar mea­sures to provide stim­u­lus wor­ld­wide).

  So maybe we’ll have an analo­gous out­come for epi­demiol­o­gists—mis­trust from peo­ple, re­spect from poli­ti­ci­ans.

• Ramiro 23 Mar 2020 16:04 UTC

  1 point

  on: What promis­ing pro­jects aren’t be­ing done against the coro­n­avirus?

  Us­ing rapid di­ag­nos­tic tests for other res­pi­ra­tory in­fec­tions to ex­clude covid-19 di­ag­no­sis?

  Most Western coun­tries have reached the limits of their test­ing ca­pa­bil­ities; it’ll take a while to de­ploy the new serolog­i­cal an­ti­body rapid tests, and even so, we’ll likely have a huge de­mand for them. Is it fea­si­ble to use other point-of-care rapid tests for differ­ent res­pi­ra­tory in­fec­tions (par­tic­u­lalry In­fluenza, but I guess there might be rapid tests for Strep­to­coc­cus, too) to ex­clude Covid-19 di­ag­no­sis?

• Ramiro 23 Mar 2020 15:58 UTC

  1 point

  on: What promis­ing pro­jects aren’t be­ing done against the coro­n­avirus?

  Miti­ga­tion of nega­tive effects of so­cial dis­tanc­ing?

  The eco­nomic effects are already be­ing tar­geted (enough?). How­ever, could we im­prove mea­sures aiming at so­cial and psy­cholog­i­cal effects? For in­stance, could we have a sys­tem to use pub­lic spaces in shifts, in­stead of mak­ing ev­ery­one stay at home -so en­sur­ing ev­ery­one can en­joy some­time out­doors, with­out crowds?

• Ramiro 23 Mar 2020 11:54 UTC

  1 point

  in reply to: avturchin's comment on: Ubiquitous Far-Ul­travi­o­let Light Could Con­trol the Spread of Covid-19 and Other Pandemics

  I agree that pre­vent­ing ex­po­sure to viri­ons is a pri­or­ity, but I am con­cerned with in­door air qual­ity over­all, es­pe­cially if peo­ple are stay­ing in­doors for long pe­ri­ods: https://​​en.m.wikipe­dia.org/​​wiki/​​In­door_air_qual­ity?wprov=sfla1

• Ramiro 23 Mar 2020 11:50 UTC

  1 point

  in reply to: Davidmanheim's comment on: What promis­ing pro­jects aren’t be­ing done against the coro­n­avirus?

  Could you be more spe­cific or provide an ex­am­ple?

• Ramiro 20 Mar 2020 18:01 UTC

  1 point

  in reply to: avturchin's comment on: Ubiquitous Far-Ul­travi­o­let Light Could Con­trol the Spread of Covid-19 and Other Pandemics

  Epistemic sta­tus: not my ex­per­tise, I’m guess­ing.

  It’s hard for it to flow up­wards, and it’ll prob­a­bly dis­perse a lot (since it doesn’t re­pro­duce out­side a host, I guess this min­i­mizes the chance of be­ing in­fected)… but yeah, if your apart­ment is close to an in­fected per­son, there’s a chance that the wind will carry viri­ons to your apart­ment; that’s why hos­pi­tals are sup­posed to place in­fected peo­ple ac­cord­ing to the airflow.

  There’s prob­a­bly a trade-off be­tween prob­a­bil­ity of ex­ter­nal con­tam­i­na­tion vs. time viri­ons stay vi­able on sur­faces in an en­vi­ron­ment. It seems like, at least for other res­pi­ra­tory in­fec­tions, for most col­lec­tive en­vi­ron­ments, we should be more con­cerned about the lat­ter.

  What’s your opinion here? Of course, there’s a point where the ex­ter­nal en­vi­ron­ment be­comes so con­tam­i­nated (in a hos­pi­tal, or if ev­ery­one in your build­ing is in­fected) that you bet­ter in­su­late your per­sonal en­vi­ron­ment as best as you can.

• Ramiro 19 Mar 2020 0:51 UTC

  2 points

  on: Ubiquitous Far-Ul­travi­o­let Light Could Con­trol the Spread of Covid-19 and Other Pandemics

  I like this idea. On the other hand, the other promis­ing en­vi­ron­men­tal mea­sure analysed by WHO, ven­tila­tion, seems very straight­for­ward and in­tu­itive—but it’s still ne­glected (I haven’t ob­served any em­pha­sis on that re­cently, at least). If peo­ple can’t open their win­dows and turn off the air con­di­tion­ing dur­ing epi­demics, I wouldn’t be very hope­ful con­cern­ing UV lamps.

[Question] Is rapid di­ag­nos­tic test­ing (RDT), such as for coro­n­avirus, a ne­glected area in Global Health?

• Ramiro 3 Mar 2020 12:36 UTC

  2 points

  on: Cortés, Pizarro, and Afonso as Prece­dents for Takeover

  First, thanks! I had no idea Afonso de Albu­querque’s con­quests had been so mar­velous. I mean, yeah, Camões and Pes­soa ded­i­cated him some verses, but it’s not very in­for­ma­tive.

  Se­cond,

  Why didn’t it hap­pen the other way around: some am­bi­tious lo­cal ruler talks to the con­quis­ta­dors, ex­ploits their in­ter­nal di­vi­sions, al­lies with some to defeat the oth­ers, and ends up on top

  Ac­tu­ally, it hap­pened some­times—na­tives played Euro­peans against each other in Africa and Brazil, where the ab­sence of cen­tral­ized gov­ern­ment (and bad ter­rain) made a quick takeover im­pos­si­ble.

• Ramiro 14 Feb 2020 21:42 UTC

  1 point

  in reply to: Cullen_OKeefe's comment on: FHI Re­port: The Wind­fall Clause: Distribut­ing the Benefits of AI for the Com­mon Good

  (Epistemic sta­tus: there must be some flaw, but I can’t find it.)

  Sure. But, let me be clearer: what drew my at­ten­tion is that, ap­par­ently, there seems to be no down-side for a com­pany to do this ASAP. My whole point:

  First, con­sider the “sim­ple” ex­am­ple where a sig­na­tory com­pany promises to donate 10% of its prof­its from a rev­olu­tion­ary AI sys­tem in 2060, a situ­a­tion with an es­ti­mated prob­a­bil­ity of about 1%; the pre­sent value of this obli­ga­tion would cur­rently amount to U$650 mil­lion (in 2010 dol­lars). This seems a lot; how­ever, I con­tend that, given in­vestors’ hy­per­bolic dis­count, they prob­a­bly wouldn’t be very con­cerned about it – it’s an un­likely event, to hap­pen in 40 years; more­over, I’ve checked with some ac­coun­tants, and this obli­ga­tion would (to­day) be prob­a­bly clas­sified as a con­tin­gent li­a­bil­ity of re­mote pos­si­bil­ity (which, un­der IAS 37, means it wouldn’t im­pact the com­pany’s bal­ance sheet – it doesn’t even have to be dis­closed in its an­nual re­port). So, I doubt such an obli­ga­tion would nega­tively im­pact a com­pany’s mar­ket value and prof­its (in the short-term); ac­tu­ally, as there’s no “bad mar­ket­ing”, it could very well in­crease them.

  Se­cond (all this pre­vi­ous ar­gu­ment was meant to get here), would it vi­o­late some sort of fi­du­ciary duty? Even if it doesn’t af­fect pre­sent in­vestors, it could af­fect fu­ture ones: i.e., sup­pos­ing the Clause is en­forced, can these in­vestors com­plain? That’s where things get messy to me. If the fi­du­ciary duty as­sumes a per­son-af­fect­ing con­cep­tion of du­ties (as law usu­ally does), I be­lieve it can’t. First, if the Clause were pub­lic, any in­vestor that bought com­pany shares af­ter the promise would have done it in full knowl­edge – and so wouldn’t be al­lowed to com­plain; and, if it didn’t af­fect its mar­ket value in 2019, even older in­vestors would have to face the ob­jec­tion “but you could have sold your shares with­out loss.” Also, given the pre­cise event “this com­pany made this dis­cov­ery in such-and-such way”, it’s quite likely that the event of the promise figures in the causal chain that made this pre­cise com­pany get this re­sult – it cer­tainly didn’t pre­vent it! Thus, even fu­ture in­vestors wouldn’t be al­lowed to com­plain.
  There must be some flaw in this rea­son­ing, but I can’t find it.

  (Could we con­vince start-ups to sign this un­til it be­comes trendy?)

• Ramiro 14 Feb 2020 5:18 UTC

  1 point

  on: FHI Re­port: The Wind­fall Clause: Distribut­ing the Benefits of AI for the Com­mon Good

  I won­der how such a com­mit­ment would ac­tu­ally im­pact a com­pany’s bal­ance sheet.

  In the ex­am­ple of Wind­fall Clause worth $649.34 mil­lion (in 2010 dol­lars), I guess that, ac­cord­ing to IAS 37, it would be con­sid­ered a con­tin­gent li­a­bil­ity of re­mote pos­si­bil­ity—and so wouldn’t even need to be dis­closed by the com­pany.

  More­over, due to hy­per­bolic dis­count, it would prob­a­bly be per­ceived as much less costly than $650m. (and I thought time prefer­ences were evil...)

• Ramiro 11 Feb 2020 19:44 UTC

  2 points

  in reply to: brunoparga's comment on: Growth and the case against ran­domista development

  Poli­cies that claim for “more ed­u­ca­tion”, in Brazil at least, tend to em­pha­size a com­pletely differ­ent skil­lset: far leftist-bi­ased his­tory, far leftist-bi­ased ge­og­ra­phy, far leftist-bi­ased so­ciol­ogy, far leftist-bi­ased philos­o­phy, arts and cul­ture (there’s this per­cep­tion that “more cul­ture” is some sort of panacea), and “crit­i­cal think­ing”, which is usu­ally code for “op­pos­ing pro-growth poli­cies”.

  I do agree math & sci­ence are re­ally want­ing in the 3rd world, that they’re more fun­da­men­tal for growth, and that we should fo­cus on them. How­ever, I dis­agree with the di­ag­no­sis; I be­lieve the rea­son stu­dents are com­par­a­tively worse in hard sci­ences is, well, that they’re rel­a­tively harder—they re­quire train­ing and com­pe­tence, from stu­dents AND teach­ers. If the prob­lem were that we im­ple­mented leftist pro-cul­ture poli­cies, in­stead of im­prove hard sci­ences learn­ing, we should at least ob­serve im­prove­ments in some other ca­pa­bil­ities—e.g., they should be able to read, in­ter­pret, and ex­pose ar­gu­ments on why, e.g., ev­ery­thing bad was caused by colo­nial­ism, pa­tri­archy, etc.

  I think we have a more com­plex in­ad­e­quate equil­ibria: bad teach­ers in unions defend­ing their in­ter­ests, stu­dents from ter­rible back­grounds, tal­ented peo­ple avoid­ing teach­ing (if you know calcu­lus, why would you want to try to teach poor kids for a low salary?), and, of course, gov­ern­ments fo­cused on what­ever will win votes in the next elec­tion.

  I do agree that any pro­posal on chang­ing ed­u­ca­tional poli­cies will meet a back­lash, es­pa­cially from hu­man­i­ties, and that it will of­ten carry a leftist taste—but we shouldn’t fo­cus on this back­lash, that’s not the cause of illiter­acy, nor in­nu­mer­acy. When we frame the is­sue as “the prob­lem is that ed­u­ca­tion is dom­i­nated by marx­ist think­ing”, we’re just un­nec­es­sar­ily poli­ti­cis­ing it.

• Ramiro 11 Feb 2020 13:35 UTC

  10 points

  in reply to: brunoparga's comment on: Growth and the case against ran­domista development

  You should first find out how to make peo­ple (jus­tifi­ably) trust those poli­cies.

  Some­times I won­der if we’re in some sort of stale­mate here. A can say: “eco­nomics show that, un­less you adopt pro-busi­ness poli­cies – e.g., lower your taxes, slash la­bor and con­sumer reg­u­la­tions – in­vestors will avoid this coun­try.” And B replies: “so­cial sci­ence shows that, un­less you adopt re­dis­tribu­tive poli­cies – e.g., tax the rich, pro­tect work­ers and con­sumers – peo­ple won’t sup­port the gov­ern­ment.” Of course, that’s even worse when A and B iden­tify them­selves as be­long­ing to spe­cific classes—then it’s more a poli­ti­cal bar­gain than a de­bate on eco­nomics. I’d like to know more about how de­vel­oped coun­tries ac­tu­ally faced this co­nun­drum – as far as I know, very badly: 30 years later, the 80’s ne­oliberal poli­cies are still the core of de­bates. But the differ­ence be­tween de­vel­oped and de­vel­op­ing coun­tries re­gard­ing so­cial trust (and trust in the gov­ern­ment) is truly re­mark­able; I won­der what’s the di­rec­tion of causal­ity here.

• Ramiro 11 Feb 2020 12:33 UTC

  1 point

  on: Ramiro’s Shortform

  Does any­one know or have a se­ri­ous opinion /​ anal­y­sis on the Euro­pean cam­paign to tax meat? I read some news at Le Monde, but noth­ing EA-level se­ri­ous­ness. I mean, it seems a pretty good idea, but I saw no data on pos­si­ble im­pact, prob­a­bil­ity of adop­tion, pos­si­ble ways to con­tribute, or even pos­si­ble side-effects?

  (not the best com­par­i­son, but worth not­ing: in Brazil a surge in meat prices caused an in­fla­tion peak in de­cem­ber and cor­roded the gov­erne­ment’s sup­port—yeah, peo­ple can tol­er­ate poli­ti­ci­ans med­dling with crim­i­nals and fas­cism, as long as they can have bar­be­cue)

• Ramiro 31 Jan 2020 19:36 UTC

  1 point

  in reply to: Aaron Gertler's comment on: Ramiro’s Shortform

  Too bad I don’t have a Face­book ac­count any­more… I’d ap­pre­ci­ate if some­one else (whou found it use­ful, of course) could raise this sub­ject in those groups.

  (man, do I miss the memes!)

  Or I could just post it as a Ques­tion in this fo­rum, to get more visi­bil­ity.

  Thanks.

• Ramiro 29 Jan 2020 16:58 UTC

  1 point

  in reply to: Aaron Gertler's comment on: Ramiro’s Shortform

  I’m think­ing on both: ad­e­quately cat­e­go­riz­ing pa­pers may have an in­di­rect im­pact on how other schol­ars se­lect their biblio­graph­i­cal refer­ences; and the vol­un­teer ed­i­tors them­selves may ac­quire (or an­ti­ci­pate its ac­qui­si­tion—I sup­pose that, if a pa­per is re­ally good, you’ll likely end up find­ing it any­way) knowl­edge of their cor­re­spond­ing do­mains.

  Of course, per­haps the an­swer is “it’s already hard enough to catch up with the posts on such-and-such sub­jects in the EA and ra­tio­nal­ist com­mu­nity, and read the stan­dard liter­a­ture, and do origi­nal work, etc. - and you still want me to work as a quasi-librar­ian for free?”

• Ramiro 26 Jan 2020 16:37 UTC

  1 point

  on: [Link] “Mo­ral un­der­stand­ing and moral illu­sions”

  Ku­dos, spe­cially for your con­cern about copyrights

• Ramiro 25 Jan 2020 14:03 UTC

  1 point

  in reply to: MichaelA's comment on: Po­ten­tial down­sides of us­ing ex­plicit probabilities

  Oh, I only apol­o­gised be­cause, well, if we start dis­cussing about catchy para­doxes, we’ll soon lose the track of our origi­nal point.

  But if you en­joy it, and since it is a rele­vant sub­ject, I think peo­ple use 3 broad “strate­gies” to tackle St. Peters­burg para­doxes and the like:

  [epistemic sta­tus: low, but it kind makes sense]

  a) “economist”: “if you use a bounded ver­sion, or takes time into ac­count, the para­dox dis­ap­pears: just ap­ply a log­a­r­ith­mic func­tion for diminish­ing re­turns...”

  b) “philoso­pher”: “un­bounded util­ity is weird” or “be­ware, it’s Pas­cal’s Wager with ob­jec­tive prob­a­bil­ities!”

  c) “statis­ti­cian”: “the prob­lem is this prob­a­bil­ity dis­tri­bu­tion, you can’t ap­ply cen­tral limit /​ other the­o­rem, or the in­differ­ence prin­ci­ple, or etc., and calcu­late its ex­pec­ta­tion”

• Ramiro 24 Jan 2020 17:26 UTC

  2 points

  on: Com­par­ing Four Cause Areas for Found­ing New Charities

  Has ac­tive agents, namely to­bacco com­pa­nies, that push against in­ter­ven­tions in this space

  I won­der if that’s so bad: con­sid­er­ing we are play­ing a zero-sum game against this com­pa­nies, each $ we make them spend to defend them­selves against pub­lic poli­cies will im­pact the price of their product—and, given price-elas­tic­ity, will de­ter con­sump­tion.

• Ramiro 24 Jan 2020 17:20 UTC

  3 points

  in reply to: John_Maxwell's comment on: Cor­po­rate cam­paigns af­fect 9 to 120 years of chicken life per dol­lar spent

  I won­der if com­pa­nies could be held li­able for some­thing like mis­lead­ing ad­ver­tise­ment in case of vol­un­tar­ily failing to fulfill their com­mit­ments. I googled it a lit­tle bit; there’s a grow­ing con­cern about legally en­forc­ing CSR com­mit­ments, but I still can’t make my mind on this pos­si­bil­ity (and its de­sir­a­bil­ity: could it back­fire?)