Milan_Griffes

• Milan_Griffes 20 Jul 2019 2:18 UTC

  2 points

  in reply to: aarongertler's comment on: De­brief: “cash prizes for the best ar­gu­ments against psychedelics”

  (b), per­haps with a dash of (a) too

• Milan_Griffes 17 Jul 2019 14:52 UTC

  2 points

  in reply to: John_Maxwell_IV's comment on: De­brief: “cash prizes for the best ar­gu­ments against psychedelics”

  … I’m not sure this is such a bad setup over­all.

  
  

  Yeah it doesn’t seem ter­rible. It prob­a­bly misses a lot of up­side, though.

• Milan_Griffes 16 Jul 2019 14:22 UTC

  2 points

  in reply to: John_Maxwell_IV's comment on: De­brief: “cash prizes for the best ar­gu­ments against psychedelics”

  (This might be the clos­est thing I’ve seen to that so far.)

  
  

  Whoa, I didn’t know about this one. Thanks for the link!

• Milan_Griffes 16 Jul 2019 14:18 UTC

  6 points

  in reply to: Kit's comment on: De­brief: “cash prizes for the best ar­gu­ments against psychedelics”

  Thanks, I think I over­stated this in the OP (added a dis­claimer not­ing this). I still think there’s a thing here but prob­a­bly not to the de­gree I was hold­ing.

  In par­tic­u­lar it felt strange that there wasn’t much en­gage­ment with the trauma ar­gu­ment or the moral un­cer­tainty /​ moral hedg­ing ar­gu­ment (“psychedelics are plau­si­bly promis­ing un­der both longter­mist & short-ter­mist views, so the case for psychedelics is more ro­bust over­all.”)

  There was also ba­si­cally no en­gage­ment with the stud­ies I pointed to.

  All of this felt strange (and still feels strange), though I now think I was too strong in the OP.

• Milan_Griffes 16 Jul 2019 14:12 UTC

  6 points

  in reply to: Larks's comment on: De­brief: “cash prizes for the best ar­gu­ments against psychedelics”

  You asked for the best ar­gu­ments against psychedelics, not for counter-ar­gu­ments to your spe­cific ar­gu­ments in favour, so this doesn’t seem that sur­pris­ing.

  
  

  Fair enough. I think I felt sur­prised be­cause I’ve spent a long time think­ing about this & tried to give the best case I could in sup­port, and then sub­mis­sions for “best case against” didn’t seem to en­gage heav­ily with my “best case for.”

• Milan_Griffes 16 Jul 2019 14:10 UTC

  6 points

  in reply to: anonymous_ea's comment on: De­brief: “cash prizes for the best ar­gu­ments against psychedelics”

  1. I like that the origi­nal­ity of it. (It’s not just say­ing “the ev­i­dence base isn’t strong enough!”)

  2. The ob­jec­tion bet­ter ac­cords with my cur­rent wor­ld­view.

• Milan_Griffes 16 Jul 2019 14:08 UTC

  2 points

  in reply to: rohinmshah's comment on: De­brief: “cash prizes for the best ar­gu­ments against psychedelics”

  If you in­stead look at CFAR as a fun­nel for peo­ple work­ing on AI risk, the “ev­i­dence base” seems clearer.

  
  

  Do you know if there are stats on this, some­where?

  e.g. Out of X work­shop par­ti­ci­pants in 2016, Y are now work­ing on AI risk.

• Milan_Griffes 16 Jul 2019 14:05 UTC

  3 points

  in reply to: rohinmshah's comment on: De­brief: “cash prizes for the best ar­gu­ments against psychedelics”

  I agree if for CFAR you are look­ing at the met­ric of how ra­tio­nal their alumni are. If you in­stead look at CFAR as a fun­nel for peo­ple work­ing on AI risk, the “ev­i­dence base” seems clearer.

  
  

  Sure, I was point­ing to the ev­i­dence base for the tech­niques taught by CFAR & other ra­tio­nal­ity train­ing pro­grams.

  CFAR could be effec­tive at re­cruit­ing peo­ple into AI risk due to Schel­ling-point dy­nam­ics, with­out the par­tic­u­lar tech­niques it teaches be­ing effi­ca­cious. (I’m not sure that’s true, just point­ing out an or­thog­o­nal­ity here.)

• Milan_Griffes 16 Jul 2019 14:00 UTC

  2 points

  in reply to: John_Maxwell_IV's comment on: De­brief: “cash prizes for the best ar­gu­ments against psychedelics”

  Why are pop­u­lar­ity-con­test dy­nam­ics harm­ful, pre­cisely?

  
  

  A similar sort of thing is a big part of the rea­son why Eliezer had difficulty ad­vo­cat­ing for AI safety, back in the 2000s.

De­brief: “cash prizes for the best ar­gu­ments against psychedelics”

• Milan_Griffes 14 Jul 2019 15:48 UTC

  3 points

  in reply to: Elizabeth's comment on: Cash prizes for the best ar­gu­ments against psychedelics be­ing an EA cause area

  Easy money :-)

[Link] “Re­vis­it­ing the In­sights model” (Me­dian Group)

• Milan_Griffes 12 Jul 2019 22:42 UTC

  7 points

  on: EA Fo­rum Prize: Win­ners for May 2019

  New this month: Two users who have a re­cent his­tory of strong posts and com­ments (Larks and Khor­ton)

  
  

  Could you say more about the pro­cess by which Larks & Khor­ton were added to the ros­ter of peo­ple who have a vote?

  (I’m pretty sure I’ve been com­ment­ing & post­ing at the roughly same ca­dence as them. No one ap­proached me about this, so I’m cu­ri­ous about the pro­cess here.)

• Milan_Griffes 12 Jul 2019 20:55 UTC

  6 points

  on: In what ways and in what ar­eas might it make sense for EA to adopt more a more bot­toms-up ap­proach?

  
  

  My sense is that there’s a lot of causal/​top down plan­ning in EA.

  
  

  My quick thought here is that EA cur­rently has a very strong “eval­u­a­tive” func­tion (i.e. strong at as­sess­ing the pro /​ con of ex­ist­ing ideas), and a weak “gen­er­a­tive” func­tion (i.e. weak at com­ing up with new ideas).

  I’m bullish on in­creas­ing EA gen­er­a­tivity from the pre­sent mar­gin.

• Milan_Griffes 12 Jul 2019 16:55 UTC

  4 points

  in reply to: William_MacAskill's comment on: A philo­soph­i­cal in­tro­duc­tion to effec­tive altruism

  Thanks, this is helpful.

  
  

  I am talk­ing about obli­ga­tions in this In­tro­duc­tion (rather than ‘op­por­tu­ni­ties’)

  Could you say a bit more about why you chose to go with the ‘obli­ga­tions’ fram­ing?

[Link] “Why Re­spon­si­ble AI Devel­op­ment Needs Co­op­er­a­tion on Safety” (OpenAI)

• Milan_Griffes 11 Jul 2019 20:43 UTC

  6 points

  on: A philo­soph­i­cal in­tro­duc­tion to effec­tive altruism

  From my quick read of your Nor­ton In­tro­duc­tion, it seems like you’re ar­gu­ing for moral re­al­ism be­ing a pre­req­ui­site to EA. (Words like “duty” and “com­mand” make me think this.)

  Is that right?

[Link] “The AI Timelines Scam”

• Milan_Griffes 10 Jul 2019 22:53 UTC

  2 points

  in reply to: cole_haus's comment on: Do we know how many big as­ter­oids could im­pact Earth?

  Thanks, su­per helpful!

  Do you hap­pen to know how promis­ing it could be to work on in­no­vat­ing on new meth­ods of dis­cov­ery and track­ing things like Damo­cloids?

  An un­known num­ber of those guys be­ing out there is scary :-/​

• Milan_Griffes 10 Jul 2019 22:06 UTC

  6 points

  in reply to: Tetraspace Grouping's comment on: If physics is many-wor­lds, does ethics mat­ter?

  Thanks, I haven’t thought about this enough to say with con­fi­dence, but it seems plau­si­ble that many-wor­lds im­plies de­ter­minism such that this is re­ally a ques­tion about de­ter­minism /​ liv­ing in a de­ter­minis­tic sys­tem.