Raemon

• Raemon 11 Dec 2019 17:07 UTC

  3 points

  in reply to: aaq's comment on: aaq’s Shortform

  Nod. I do agree with that.

• Raemon 10 Dec 2019 23:51 UTC

  5 points

  on: The Re­view Phase: Helping LessWrongers Eval­u­ate Old Posts

  The Re­view Phase is a bit of an evolv­ing pro­cess – I’m ex­pect­ing us to learn over the course of the month what sort of re­views are most helpful.

  One ex­plicit up­date I made since last week is shift­ing the Re­view Phase from “write up whether you think this post should be in­cluded in the book” to “fo­cus on pro­vid­ing in­for­ma­tion to other peo­ple who are eval­u­at­ing the post.”

  The “judge” mind­set seemed to be out­putting less use­ful con­tent than the “provide in­for­ma­tion to help eval­u­ate” mind­set.

  I do think in­clud­ing notes about what you think should be in­cluded in the book is still valuable, but is some­thing it makes more sense to do af­ter you’ve spent some time in “eval­u­ate and add in­for­ma­tion” mode.

• Raemon 10 Dec 2019 23:36 UTC

  5 points

  in reply to: Taryn East's comment on: The LessWrong 2018 Re­view [posts need at least 2 nom­i­na­tions]

  The pro­cess isn’t finished yet – it’ll hope­fully com­plete some­time in Jan­uary of next year.

• Raemon 10 Dec 2019 23:11 UTC

  3 points

  in reply to: Said Achmiz's comment on: Every­thing I ever needed to know, I learned from World of War­craft: Good­hart’s law

  The lat­est ver­sion of the “off­topic com­ment” fea­ture that the team had chat­ted about was a “col­lapse” fea­ture, where some com­ments are just forcibly col­lapsed with a flag, and this is just a generic tool that ad­mins and some au­thors have ac­cess to. Doesn’t re­ally re­quire any­thing au­to­matic, just, when you no­tice such a thread, you can close it. (It’s still ap­pear in the com­ment list, just col­lapsed as if it had low karma, pos­si­bly with a rea­son dis­played)

• Raemon 10 Dec 2019 22:54 UTC

  3 points

  in reply to: Raemon's comment on: Rae­mon’s Short­form Feed

  The LW team has been try­ing this out the “bolded un­read posts” a few days as an ad­min-only set­ting. I think pretty much ev­ery­one isn’t lik­ing it.

  But I per­son­ally am lik­ing the fact that most posts aren’t grey, and I’m find­ing my­self won­der­ing whether it’s even that im­por­tant to high­light un­read posts. Ob­vi­ously there’s some value to it, but:

  a) a post be­ing read isn’t ac­tu­ally that much ev­i­dence about whether I want to read it again – I find my­self click­ing on old posts about as of­ten as new posts. (This might be some­thing you could con­cretely look into with an­a­lyt­ics)

  b if I don’t want to read a post, mark­ing it as read is sort of annoying

  c) I still re­ally dis­like hav­ing most of my posts be grey

  d) it’s re­ally hard to make an “un­read” var­i­ant that doesn’t scream out for dis­pro­por­tionate at­ten­tion.

  (I sup­pose there’s also an op­tion for this to be a user-con­figurable set­ting, since most users don’t read so many posts that they all show up grey, and the few who do could maybe just man­u­ally turn it off)

• Raemon 10 Dec 2019 22:45 UTC

  3 points

  in reply to: Said Achmiz's comment on: Every­thing I ever needed to know, I learned from World of War­craft: Good­hart’s law

  Read­ing this thread in the fu­ture, I find my­self kinda wish­ing for ways com­ment threads like this could be auto-col­lapsed or re­solved or some­thing af­ter reach­ing their con­clu­sion.

• Raemon 10 Dec 2019 22:11 UTC

  5 points

  on: Karate Kid and Real­is­tic Ex­pec­ta­tions for Disagree­ment Resolution

  The origi­nal mo­ti­vat­ing ex­am­ple for this post doesn’t ac­tu­ally quite fit into the lens the post ori­ented around. The post fo­cuses on “dis­agree­ments be­tween par­tic­u­lar peo­ple”. But there’s a differ­ent set of is­sues sur­round­ing “dis­agree­ments in a zeit­geist.”

  Groups up­date slower than in­di­vi­d­u­als. So if you write a brilli­ant es­say say­ing “this core as­sump­tion or frame of your move­ment is wrong”, you may not only have to wait for in­di­vi­d­ual peo­ple to go “oh, I see, that makes sense”, but even longer for that to be­come com­mon knowl­edge, enough that so that you won’t re­li­ably see group mem­bers act­ing un­der the old as­sump­tions.

  (Again, this is an ob­ser­va­tion about the sta­tus quo, not about what is nec­es­sar­ily pos­si­ble if we grabbed all the low hang­ing fruit. But in the case of groups I’m more pes­simistic about things im­prov­ing as much, un­less the group has strong bar­ri­ers to en­try and strong re­quire­ments of “ac­tu­ally com­mit­ting se­ri­ous prac­tice to dis­agree­ment re­s­olu­tion.”)

• Raemon 10 Dec 2019 21:32 UTC

  5 points

  in reply to: aaq's comment on: aaq’s Shortform

  This cer­tainly seems im­por­tant (I do think this is a key value the com­mu­nity pro­vides). But it is im­por­tantly differ­ent from “the ra­tio­nal­ity con­tent of the com­mu­nity is di­rectly helpful for peo­ple-in-gen­eral.” If it were just “peo­ple who get you”, this wouldn’t ob­vi­ously be more or differ­ently im­por­tant than other ran­dom sub­cul­tures.

• Raemon 10 Dec 2019 1:23 UTC

  3 points

  on: K2 and D3 post

  Oh, it might have pat­tern-matched to spam ini­tially. But, I think it ex­ists now?

  https://​​www.less­wrong.com/​​posts/​​c5ay­cbSsSc38XWPEc/​​tak­ing-vi­tamin-d3-with-k2-in-the-morn­ing#bY6RdPQrYLakAzr3A

• Raemon 10 Dec 2019 1:22 UTC

  3 points

  in reply to: Chris_Leong's comment on: De­cou­pling vs Con­tex­tu­al­is­ing Norms

  The OP com­ment was op­ti­miz­ing for “im­prov­ing my un­der­stand­ing of the do­main” more than di­rect ad­vice of how to change the post.

  (I’m not nec­es­sar­ily ex­pect­ing the points and con­fu­sions there to re­solve within the next month – it’s pos­si­ble that you’ll re­flect on it a bit and then figure out a slightly differ­ent ori­en­ta­tion to the post, that dis­tills the var­i­ous con­cepts into a new form. Another pos­si­ble out­come is that you leave the post as-is for now, and then in an­other year or two af­ter mul­ling things over some­one writ­ers a new post do­ing a some­what differ­ent thing, that be­comes the new refer­ent. Or, it might just turn out that my cur­rent epistemic state wasn’t that use­ful. Or other things)

  Re: “Jum­bled”

  I think there’s sort of a two-step pro­cess that goes into nam­ing things (iron­i­cally, or ap­pro­pri­ate, which map di­rectly onto the post) – first figur­ing out “okay what ac­tu­ally is this phe­nomenon, and what name most ac­cu­rately de­scribes it?” and then, sep­a­rately, “okay, what sort of names are go­ing to re­li­ably go­ing to make peo­ple an­gry and dis­tract from the origi­nal topic if you ap­ply it to peo­ple, and are there al­ter­na­tive names that cleave closely to the truth?”

  (my pro­cess for gen­er­at­ing names in that risk offend­ing is some­thing like a multi-step Bab­ble and Prune, where I gen­er­ate names aiming to satis­fice on “a good ex­pla­na­tion of the true phe­nomenon” and “not likely to be un­nec­es­sar­ily dis­tract­ing”, un­til I have a name that satis­fies both crite­ria)

  I haven’t tried gen­er­at­ing a max­i­mally good name for Jum­bled yet since I wasn’t sure this was even carv­ing re­al­ity the right way.

  But, like, it’s not an ac­ci­dent that ‘jum­bled’ is more likely to offend peo­ple than ‘con­tex­tu­al­ized’. I do, in fact, think worse of peo­ple who have jum­bled com­mu­ni­ca­tion than de­liber­ately con­tex­tu­al­ized com­mu­ni­ca­tion. (com­pare “Virtue Sig­nal­ling”, which is an im­por­tant term but is ba­si­cally an in­sult ex­cept among peo­ple who have some kind of prin­ci­pled un­der­stand­ing that “Yup, it turns out some of the things I do had un­flat­ter­ing mo­tives and I’ve come to en­dorse that, or en­dorse my cur­rent [low] de­gree of pri­ori­tiz­ing chang­ing it.”)

  I am a con­ver­sa­tion con­se­quen­tial­ist and think it’s best to find ways of po­litely point­ing out un­flat­ter­ing things about peo­ple in ways that don’t make them defen­sive. But, it might be that the cor­rect carv­ing of re­al­ity in­cludes some un­flat­ter­ing de­scrip­tions of peo­ple and maybe the best you can do is min­i­mize dis­trac­tion-dam­age.

• Raemon 9 Dec 2019 23:58 UTC

  3 points

  in reply to: Kaj_Sotala's comment on: The LessWrong 2018 Re­view [posts need at least 2 nom­i­na­tions]

  Here’s a re­view of mine that I think is pretty rep­re­sen­ta­tive of the sort of re­view that I, per­son­ally, am most ex­cited about.

• Raemon 9 Dec 2019 23:15 UTC

  8 points

  on: De­cou­pling vs Con­tex­tu­al­is­ing Norms

  This post seems to be mak­ing a few claims, which I think can be eval­u­ated sep­a­rately:
  
  1) De­cou­pling norms ex­ist
  2) Con­tex­tu­al­iz­ing norms ex­ist
  3) De­cou­pling and con­tex­tu­al­iza­tion norms are use­ful to think as op­po­sites (ei­ther as a di­chotomy or spec­trum)

  (i.e. there are enough peo­ple us­ing those norms that it’s a use­ful way to carve up the dis­cus­sion-land­scape)

  There’s a range of “strong” /​ “weak” ver­sions of these claims – de­cou­pling and/​or con­tex­tu­al­iza­tion might be prin­ci­pled norms that some peo­ple ex­plic­itly en­dorse, or they might just be clusters of ten­den­cies peo­ple have some­times.

  In the com­ments of his re­sponse post, Zack Davis noted:

  It’s cer­tainly pos­si­ble that there’s a “gen­eral fac­tor” of con­tex­tu­al­iz­ing—that peo­ple sys­tem­at­i­cally and non-op­por­tunis­ti­cally vary in how in­fer­en­tially dis­tant a re­lated claim has to be in or­der to not cre­ate an im­pli­ca­ture that needs to be ex­plic­itly can­celed if false. But I don’t think it’s ob­vi­ous, and even if it’s true, I don’t think it’s ped­a­gog­i­cally wise to use a poli­ti­cally-mo­ti­vated ap­peal-to-con­se­quences as the cen­tral case of con­tex­tu­al­iz­ing.

  And, read­ing that, I think it may ac­tu­ally the op­po­site – there is gen­eral fac­tor of “de­cou­pling”, not con­tex­tu­al­iz­ing. By de­fault peo­ple are us­ing lan­guage for a bunch of rea­sons all jum­bled to­gether, and it’s a rel­a­tively small set of peo­ple who have the de­liber­ately-de­cou­ple ten­dency, skill and/​or norm, of “check­ing in­di­vi­d­ual state­ments to see if they make sense.”

  Upon re­flec­tion, this is ac­tu­ally more in line with the origi­nal Nerst ar­ti­cle, which used the terms “Low De­cou­pling” and “High De­cou­pling”, which less strongly con­veys the idea of “con­tex­tu­al­izer” be­ing a co­her­ent thing.

  On the other hand, Nerst’s origi­nal post does make some claims about Klein be­ing the sort of per­son (a jour­nal­ist) who is “defini­tively a con­tex­tu­al­izer, as op­posed to just ‘not a de­cou­pler’”, here:

  While sci­ence and en­g­ineer­ing dis­ci­plines (and an­a­lytic philos­o­phy) are pop­u­lated by peo­ple with a knack for de­cou­pling who learn to take this norm for granted, other in­tel­lec­tual dis­ci­plines are not. In­stead they’re largely com­posed of what’s op­po­site the sci­en­tist in the gallery of brainy archetypes: the liter­ary or artis­tic in­tel­lec­tual.

  This crowd doesn’t live in a world where de­cou­pling is stan­dard prac­tice. On the con­trary, cou­pling is what makes what they do work. Novelists, po­ets, artists and other sto­ry­tel­lers like jour­nal­ists, poli­ti­ci­ans and PR peo­ple rely on thick, rich and am­bigu­ous mean­ings, as­so­ci­a­tions, im­pli­ca­tions and al­lu­sions to evoke feel­ings, im­pres­sions and ideas in their au­di­ence. The words “artis­tic” and “liter­ary” refers to us­ing idea cou­plings well to sub­tly and in­di­rectly push the au­di­ence’s mean­ing-but­tons.

  To a low-de­cou­pler, high-de­cou­plers’ abil­ity to fence off any threat­en­ing im­pli­ca­tions looks like a lack of em­pa­thy for those threat­ened, while to a high-de­cou­pler the low-de­cou­plers in­sis­tence that this isn’t pos­si­ble looks like naked bias and an in­abil­ity to think straight. This is what Har­ris means when he says Klein is bi­ased.

  Although they’re in­ter­wo­ven, I think it might be worth dis­t­in­guish­ing some sub­claims here (not nec­es­sar­ily made by Nerst or Leong, but I think im­plied and worth think­ing about)

  • There ex­ist a class of gen­eral sto­ry­tel­ling contextualists

  • There ex­ist PR-peo­ple/​poli­ti­ci­ans/​ac­tivists who wield con­tex­tual prac­tice as a tool or weapon.

  • There ex­ist “prin­ci­pled con­tex­tu­al­iz­ers” who try to evenly come to good judg­ments that de­pends on con­text.

  My Epistemic State

  Em­piri­cal Questions

  There’s a set of fairly con­crete “em­piri­cal” ques­tions here, which are ba­si­cally “if you do a bunch of fac­tor anal­y­sis of dis­cus­sions, would de­cou­pling and/​or con­tex­tu­al­iza­tion and/​or any of the spe­cific con­tex­tual-sub­cat­e­gories listed above be ma­jor pre­dic­tive power?”

  The ex­per­i­ments you’d run for this might be ex­pen­sive but not very con­fus­ing.

  I would cur­rently guess:

  • “De­cou­pling fac­tor” definitely ex­ists and is meaningful

  • Sto­ry­tel­ling con­tex­tu­al­ists ex­ist and are mean­ingful (though not nec­es­sar­ily es­pe­cially use­ful to con­trast with de­cou­plers)

  • PR-ists who wield con­text as tool/​weapon definitely ex­ist (and de­cou­pling is of­ten rele­vant to their plans, so they have de­vel­oped tools that al­low them to mod­u­late the de­gree to which de­cou­pling fits into the con­ver­sa­tional frame)

  • I think I could name a few peo­ple at least at­tempt­ing to be “fair, prin­ci­pled con­tex­tu­al­ists”, at least in some cir­cum­stances. I am less con­fi­dent that this is a real thing be­cause the al­ter­na­tive “se­cretly they’re just re­ally effec­tive or sub­tle PR-ists, ei­ther in­ten­tion­ally or not” is a pretty vi­able alternative

  Con­cep­tual Question

  I have a re­main­ing con­fu­sion, which is some­thing like “what ex­actly is a con­tex­tu­al­izer?”. I feel like I have a crisp defi­ni­tion of “de­cou­pling”. I don’t have that for con­tex­tu­al­iz­ers. Are the three sub­cat­e­gories listed above re­ally ‘rel­a­tives’ or are they just three differ­ent groups do­ing differ­ent things? Is it mean­ingful to put these on a spec­trum with de­cou­plers on the other side?

  mr-hire sug­gests:

  “How much you and oth­ers are will­ing to think about the con­se­quences of what is said sep­a­rate from its’ truth value.”

  Which sounds like a plau­si­bly good defi­ni­tion, that maybe ap­plies to all three of the sub­cat­e­gories. But I feel like it’s not quite the nat­u­ral defi­ni­tion for each in­di­vi­d­ual sub­cat­e­gory. (Rather, it’s some­thing a bit down­stream of each cat­e­gory defi­ni­tion)

  “Jum­bled” vs “Con­tex­tual”

  “High de­cou­pling” and “low de­cou­pling” are still pretty con­fus­ing terms, even if you get rid of any no­tion of “low de­cou­pling” be­ing a co­gent thing. It oc­cured to me, writ­ing this re­view, that you might re­place the word “con­tex­tual” with “jum­bled”.

  Con­tex­tual im­plies some de­gree of prin­ci­pled norms. Jum­bled points more to­wards “the per­son is us­ing lan­guage for a ran­dom mish­mash of strate­gies all thrown to­gether.” (Poli­ti­ci­ans might some­times be best de­scribed as “jum­bled”, and some­times as “prin­ci­pled” [but, not nec­es­sar­ily good prin­ci­ples, i.e. ‘I will de­liber­ately say what­ever causes my party to win’]).

  ...

  That’s what I got for now.

  What links here?

  • Raemon's comment on The LessWrong 2018 Review by Raemon (9 Dec 2019 23:58 UTC; 3 points)

• Raemon 8 Dec 2019 23:08 UTC

  5 points

  in reply to: Raemon's comment on: The LessWrong 2018 Re­view [posts need at least 2 nom­i­na­tions]

  One of the key ideas here is that I’d like posts to have got­ten some­one to “look into the dark”. If the post wasn’t as use­ful as it seemed, how would we know? If 10 years from now you no longer en­dorsed the post, why might that be?

• Raemon 8 Dec 2019 23:05 UTC

  5 points

  in reply to: Kaj_Sotala's comment on: The LessWrong 2018 Re­view [posts need at least 2 nom­i­na­tions]

  There should be a post com­ing up soon that goes into more ex­am­ples of how to do Re­views. It’s a bit tough ques­tion be­cause differ­ent posts benefit from differ­ent types of re­views.

  A thing that I think is com­monly use­ful is ask­ing “what are the ac­tual claims this post is mak­ing”, and list­ing them suc­cinctly, and writ­ing up some thoughts about how we could ac­tu­ally em­piri­cally check if those claims are true. (Even if we don’t ac­tu­ally run the ex­per­i­ment, I think op­er­a­tional­iz­ing what ob­ser­va­tions we’d ex­pect in the world is helpful for eval­u­at­ing when/​why/​whether the post is valid)

• Raemon 8 Dec 2019 22:53 UTC

  3 points

  in reply to: Raemon's comment on: Karate Kid and Real­is­tic Ex­pec­ta­tions for Disagree­ment Resolution

  I do still think there’s a lot of le­gi­t­i­mately hard stuff here. In the past year, in some de­bates with Habryka and with Ben­quo, I found a ma­jor com­po­nent (of my own up­dat­ing) had to do with giv­ing their per­spec­tives time to mull around in my brain, as well as some kind of aes­thetic com­po­nent. (i.e. if one per­son says “this UI looks good” and an­other per­son says “this UI looks bad”, there’s an as­pect of that that doesn’t lend it­self well to “de­bate”. I’ve spent the past 1.5 years think­ing a lot about Aes­thetic Dou­ble­crux, which much of this se­quence was lay­ing the ground­work for)

• Raemon 8 Dec 2019 22:51 UTC

  4 points

  on: Karate Kid and Real­is­tic Ex­pec­ta­tions for Disagree­ment Resolution

  Some­what re­ply­ing to both romeo and ben­dini el­sethread:

  Ben­dini:

  Disagree­ments aren’t always triv­ial to re­solve, but you’ve been ac­tively de­bat­ing an is­sue for a month and zero progress has been made, ei­ther the re­s­olu­tion pro­cess is bro­ken or some­one is do­ing some­thing be­sides putting max­i­mum effort into re­solv­ing the dis­agree­ment.

  Romeo:

  I pro­pose an al­ter­na­tive model. Peo­ple don’t re­solve dis­agree­ments be­cause there are no in­cen­tives to re­solve them. In fact the in­cen­tives of­ten cut the other way.

  I definitely have a sense that ra­tio­nal­ists by de­fault aren’t that great at dis­agree­ing for all the usual rea­sons (not in­cen­tivized to, don’t ac­tu­ally prac­tice the men­tal moves nec­es­sary to do so pro­duc­tively), and kinda the whole point of this se­quence is to go “Yo, guys, it seems like we should ac­tu­ally be able to be good at this?”

  And the prob­lem in part is that that this re­quires mul­ti­ple ac­tors – my sense is that a sin­gle per­son try­ing their best to listen/​learn/​up­date can only get halfway there, or less.[1]

  [1] The ex­act na­ture of “what you can ac­com­plish with only one per­son try­ing to pro­duc­tively-dis­agree de­pends on the situ­a­tion.” It may be that that par­tic­u­lar per­son can come to what­ever the truest-nearby-be­liefs rea­son­ably well, but if you need agree­ment, or if the other per­son is the fi­nal de­ci­sion maker, “one-per­son-com­ing-to-cor­rect-be­liefs” may not solve the prob­lem.

  Com­ing to Cor­rect Beliefs vs Poli­ti­cal Debate

  I think one of the things go­ing on is that it takes a bit of vuln­er­a­bil­ity to switch from “ad­ver­sar­ial poli­ti­cal mode” (a com­mon de­fault) to “ac­tu­ally be open to chang­ing your mind.” There is a risk that if you try earnestly to look at the ev­i­dence and change your mind, but your part­ner is just push­ing their agenda, and you don’t have some skills re: “re­silience to so­cial pres­sure”, then you may be sort of just ced­ing ground in a poli­ti­cal fight with­out even suc­cess­fully im­prov­ing truth­seek­ing.

  (some­times, this is a le­gi­t­i­mate fear, and some­times it’s not but it feels like it is, and notic­ing that in the mo­ment is an im­por­tant skill)

  I’ve been on both sides of this, I think. Some­times I’ve found my­self feel­ing re­ally frus­trated that it feels like my dis­cus­sion part­ner isn’t listen­ing or will­ing to up­date, and I find my­self sort of lean­ing into an ag­gres­sive voice to try and force them to listen to me. And then they’re like ’Dude, you don’t sound like you’re ac­tu­ally will­ing to listen to me or up­date” and then I was sheep­ishly like… “oh, yeah you’re right.”

  It seems like hav­ing some kind of mu­tally-trustable-pro­ce­dure for mu­tual “disar­ma­ment” would be helpful.

• Raemon 8 Dec 2019 21:32 UTC

  3 points

  in reply to: mr-hire's comment on: Affor­dance Widths

  For in­stance, the con­cept of “so­ciety isn’t made nice for hu­mans” is not new, but hav­ing moloch and in­ad­e­quate equil­ibria as con­cepts still pushed for­ward the discourse

  Nod. And in par­tic­u­lar, I saw this post as some­thing like “tak­ing the con­cept of ‘priv­ilege’, and flesh­ing it the gears of one par­tic­u­lar facet of it.” (Priv­ilege also be­ing a con­cept that’s in­ter­wo­ven with some broader nar­ra­tives or poli­ti­cal ma­neu­ver­ing that I don’t fully en­dorse, but is nonethe­less have found quite use­ful)

• Raemon 8 Dec 2019 4:17 UTC

  5 points

  in reply to: romeostevensit's comment on: Karate Kid and Real­is­tic Ex­pec­ta­tions for Disagree­ment Resolution

  This is an in­ter­est­ing com­ment. Some thoughts af­ter re­flect­ing a bit:
  
  Awhile ago you wrote a com­ment say­ing some­thing like “de­liber­ate prac­tice de­liber­ate prac­tice un­til you get re­ally good iden­ti­fy­ing good feed­back loops, and work­ing with them.” I found that fairly in­spiring at the time.

  I didn’t ever re­ally ded­i­cate my­self to do­ing that thor­oughly enough to have a clear opinion on whether it’s The Thing. I think I put maybe… 6 hours into de­liber­ate prac­tice, with in­tent to pay at­ten­tion to how de­liber­ate prac­tice gen­er­al­ized. I got value out of that that was, like, com­men­su­rate with the “se­ri­ous prac­tice” noted here (i.e. if I kept that up con­sis­tently I’d prob­a­bly skill up at a rate of 5-10% per year, and at the time that I did it, my out­put in the do­mains-in-ques­tion was maybe 10-20% higher, but more costly), but it re­quired set­ting aside time for hard cog­ni­tive la­bor that feels in short sup­ply.

  There were at least some do­mains (a par­tic­u­lar videogame I tried to de­liber­ate prac­tice), that seemed very sur­pris­ingly hard to im­prove at.

  I do have a gen­eral sense (from this past year as well as pre­vi­ous ex­pe­rience) that in many do­mains, there are some rapid ini­tial gains to be had for the first 20 hours or so.

  None of this feels like “things are ac­tu­ally eas­ier than de­scribed in the karate kid es­say.” I would agree with the claim “the karate kid es­say sort of im­plies you just have to try hard for a long time, and ac­tu­ally many of the gains come from ac­tu­ally hav­ing mod­els of how things work and you should be able to tell if you’re im­prov­ing.” But that doesn’t make things not hard.

  It seems plau­si­ble that if you gain the gen­er­al­ized De­liber­ate Prac­tice skill a lot of things be­come much eas­ier, and that it’s the cor­rect skill to gain early in the skill-tree. But, like, it’s still pretty hard yo.

  I also agree that most peo­ple aren’t ac­tu­ally even try­ing to get bet­ter at dis­agree­ment, and if they were do­ing that much at all that’d make a pretty big differ­ence. (“years” is what I think the de­fault ex­pec­ta­tion should be among peo­ple that aren’t re­ally try­ing)

• Raemon 7 Dec 2019 17:40 UTC

  3 points

  in reply to: BrienneYudkowsky's comment on: Brien­neYud­kowsky’s Shortform

  The Less­wrong com­ment guidelines say, “Aim to ex­plain, not per­suade.” Is this a method by which we cut out our own chests?

  I‘m cu­ri­ous how this ques­tion parses for Vaniver

• Raemon 7 Dec 2019 2:35 UTC

  5 points

  in reply to: Rob Bensinger's comment on: Rae­mon’s Short­form Feed

  I ini­tially wanted “bold ev­ery­where” be­cause it helped my brain re­li­ably parse things as “this is a bold line” in­stead of “this is a line with some bold parts but you have to hunt for them”. But, af­ter ex­per­i­ment­ing a bit I started to feel­ing hav­ing bold el­e­ments semi-ran­domly dis­tributed across the lines made it a lot bus­ier.