Davis_Kingsley

• Davis_Kingsley 24 Sep 2019 9:38 UTC

  14 points

  on: Heads I Win, Tails?—Never Heard of Her; Or, Selec­tive Re­port­ing and the Tragedy of the Green Rationalists

  This post is ex­cel­lent. It is prob­a­bly the best post I have read on LessWrong in a long time. Thank you for writ­ing it!

• Davis_Kingsley 22 Sep 2019 14:27 UTC

  8 points

  on: Timer Toxicities

  In­ter­est­ingly, even some re­spected games do stuff that vi­o­lates this. For in­stance, Star­craft has queues but with sev­eral con­straints that make them far worse than they could be:

  • Some types of things you might want to build can­not be queued (you can­not, for in­stance, queue your Bar­racks to build an ad­don, only unit pro­duc­tion; if you’re re­search­ing up­grades at your Eng­ineer­ing Bay, you can­not queue level 2 weapons while already re­search­ing level 1 weapons).

  • Ad­ding some­thing to the queue costs the en­tire cost that it would nor­mally take to build that thing and does so up front rather than charg­ing you the price once the queued item ac­tu­ally starts build­ing. In other words, if I want to queue up three 50-min­eral Marines, that will cost me 150 min­er­als up front even though only one will ac­tu­ally be build­ing at first.

  Th­ese con­straints mean that us­ing the queue pe­nal­izes you, as queu­ing a unit means that you are es­sen­tially wast­ing re­sources; if you want to build three Marines, it’s much bet­ter to re­s­e­lect your Bar­racks and build a new Marine af­ter each com­pletes than it is to add three Marines to the queue, as queu­ing them “wastes” 100 min­er­als on Marines not ac­tu­ally be­ing built, which you could in­stead use on other things while wait­ing for the first Marine to build.

  I con­sider this ex­tremely an­noy­ing and es­sen­tially anti-com­pet­i­tive; Star­craft no­to­ri­ously has a high bar­rier to en­try in terms of pure APM (ac­tions per minute), since the user in­ter­face is­sues mean that you have to take very many ac­tions in or­der to play effi­ciently. I would be very in­ter­ested to see what the game would look like in a ver­sion where the in­ter­face was de­signed to be as effi­cient as pos­si­ble in­stead of adding ex­tra con­straints onto the player.

• Davis_Kingsley 17 Sep 2019 12:22 UTC

  12 points

  on: Who’s an un­usual thinker that you recom­mend fol­low­ing?

  Do­minic Cum­mings, a strate­gist who worked on Brexit. He is sur­pris­ingly ra­tio­nal­ist-al­igned and has very in­ter­est­ing thoughts at his blog.

• Davis_Kingsley 14 Sep 2019 10:05 UTC

  2 points

  in reply to: elityre's comment on: Rev­er­sal Tests in Ar­gu­ment and Debate

  I used this term be­cause I think the fun­da­men­tal move be­ing pointed to­wards is fairly similar (al­though ac­tu­ally I think the Bostrom/​Ord ap­pli­ca­tion of this method is in­cor­rect, which maybe means I should have come up with a differ­ent name!).

Rev­er­sal Tests in Ar­gu­ment and Debate

• Davis_Kingsley 25 Aug 2019 1:42 UTC

  5 points

  on: Read­ing Raph Koster (notes)

  Thanks for the link! I ended up read­ing a large num­ber of his ar­ti­cles. His thoughts on UO and Galax­ies were pre­dictably the most in­ter­est­ing to me—I definitely share his sense that the old “wild west” Ul­tima and the like was bet­ter and more al­ive than the more soul­less mod­ern games (though I didn’t ac­tu­ally play Ul­tima and maybe I’d change my tune af­ter be­ing ganked re­peat­edly by PKs… :P).

  I also find it in­ter­est­ing how suc­cess­ful Galax­ies was de­spite the fact that the com­bat sys­tem ap­par­ently never worked as in­tended and was ba­si­cally dys­func­tional! It kinda makes me won­der, what if Galax­ies had had the dev re­sources and bud­get of WoW? Would that be the new face of MMOs? (Some­times I’ve had similar thoughts re: Netrun­ner and MtG...)

  For me the most “wild west” ex­cit­ing al­ive game right now is EVE On­line, but the ac­tual game­play is some­thing I’m profoundly un­in­ter­ested in so I ba­si­cally live vi­car­i­ously through sto­ries of in­ter­est­ing hap­pen­ings.

• Davis_Kingsley 24 Aug 2019 8:52 UTC

  2 points

  in reply to: Benquo's comment on: [Er­ror com­mu­ni­cat­ing with LW2 server]

  Over­all, it seems to me that there are peo­ple try­ing to do the kind of trans­la­tional work Davis is ask­ing for, but the com­mu­nity is not, as a whole, ap­ply­ing the sort of dis­cern­ment that would de­mand such work.

  Agreed, yeah. This is maybe the main thing I’m get­ting at—I’m try­ing to shock peo­ple into re­al­iz­ing “hey, ev­ery­thing isn’t fine, things are go­ing wrong” and ap­ply­ing more dis­cern­ment to what’s go­ing on in the com­mu­nity.

• Davis_Kingsley 24 Aug 2019 8:50 UTC

  2 points

  in reply to: Benquo's comment on: [Er­ror com­mu­ni­cat­ing with LW2 server]

  There’s an im­por­tant part miss­ing in my cur­rent draft that has to do with the fact that much of the “es­o­teric” con­tent is in fact not be­ing openly pushed, but rather smug­gled into the com­mu­nity via other meth­ods; I think it’s very difficult to trans­late/​syn­the­size/​ground con­cepts if you aren’t be­ing told about them un­til they’ve already taken over rele­vant parts of the com­mu­nity.

  This is also why much of the post has a “sound the alarm” feel­ing. I think that if the com­mu­nity’s in­sti­tu­tions were op­er­at­ing more prop­erly, it would be much more re­silient to things be­ing “smug­gled in”, which in turn would mean that peo­ple try­ing to spread these ideas would have to make a stronger/​more rea­soned case for them in or­der to get trac­tion here.

  As for the “list” for­mat—this post is some­what based on a talk I gave at the CFAR alumni re­union last year which was much bet­ter-re­ceived than I’d an­ti­ci­pated. Sev­eral peo­ple told me they had similar con­cerns but wasn’t sure if it was just them or what, and if we’re try­ing to “get the shields back on­line” just warn­ing peo­ple that this is go­ing on may be suffi­cient to prompt at least some­what more care­ful think­ing.

• Davis_Kingsley 24 Aug 2019 6:18 UTC

  11 points

  in reply to: clone of saturn's comment on: Is LW mak­ing progress?

  Bayes oc­cu­pies a place of priv­ilege in the Se­quences and for some time had a place of priv­ilege in the CFAR cur­ricu­lum as well, but it didn’t much work and so we stopped. I’m not say­ing it’s bad in prin­ci­ple, I’m say­ing it em­piri­cally didn’t work as a class for in­tro work­shops de­spite our try­ing many differ­ent ap­proaches.

  You might dis­pute whether that’s a pos­i­tive change, but it’s cer­tainly a way in which I see “the state of the art” as hav­ing shifted over time.

• Davis_Kingsley 24 Aug 2019 5:00 UTC

  15 points

  on: Is LW mak­ing progress?

  As some­one who worked for CFAR full-time from early 2014 to mid-late 2016 and still teaches at CFAR work­shops fairly reg­u­larly as a con­trac­tor, I can tell you that there has definitely been progress from the “Se­quences era” style of ra­tio­nal­ity to what we are cur­rently teach­ing. Ear­lier ver­sions of the CFAR cur­ricu­lum were closer to the Se­quences and were also in my view worse (for in­stance, CFAR no longer teaches Bayes’s The­o­rem).

  Not all of this has been fully visi­ble to the pub­lic, though at least some of it is—for in­stance, here is Dun­can’s post ex­plain­ing Dou­ble Crux, one of CFAR’s “post-Se­quences” in­no­va­tions. I don’t think there are posts for ev­ery new tech­nique but I do quite think there’s progress be­ing made, some of which is re­flected on LW.

• Davis_Kingsley 23 Aug 2019 10:23 UTC

  2 points

  in reply to: Pattern's comment on: Davis_Kingsley’s Shortform

  In­ter­est­ingly that’s ac­tu­ally quite dis­puted—you linked to a reprint of the card that dis­plays as be­ing rated ⁵⁄₅, but the origi­nal print­ing of the card is ac­tu­ally rated 3.455 out of 5.

  (That said, it’s cer­tainly bet­ter than Book Burn­ing!)

• Davis_Kingsley 22 Aug 2019 22:05 UTC

  9 points

  in reply to: Timothy Underwood's comment on: Davis_Kingsley’s Shortform

  Go af­ter polyamory all you want—if there is data about prob­lems that are statis­ti­cally likely to show up, I’d like to know about that, so that I can try min­i­miz­ing the odds that we will ex­pe­rience them—and we do ex­pe­rience some of the stan­dard poly jeal­ousy prob­lems—but if you are go­ing to shoot a sa­cred cow, bring a high pow­ered rifle, not a squirt gun.

  I’ll fully ad­mit that I don’t have for­mal statis­ti­cal data, but I think the point is worth mak­ing any­way as a po­ten­tial warn­ing. My in­tent is mostly to warn new­com­ers about pat­terns I’ve seen rather than to shame peo­ple already in the com­mu­nity; I cer­tainly do not claim that ev­ery poly re­la­tion­ship is abu­sive or the like.

  (Part of the rea­son I posted this in short­form rather than as a top-level post is be­cause I was hop­ing to just send a quick warn­ing rather than post­ing some­thing more for­mal and de­tailed!)

• Davis_Kingsley 22 Aug 2019 22:01 UTC

  10 points

  in reply to: G Gordon Worley III's comment on: Davis_Kingsley’s Shortform

  Link?

  The origi­nal site that the post was on has been taken down, but here’s a paste­bin of the rele­vant text (posted with the con­sent of the origi­nal au­thor, one para­graph re­moved to pre­serve the au­thor’s pri­vacy). I should add that my own view of poly is con­sid­er­ably more nega­tive than that of this au­thor, but even the linked post is sig­nifi­cantly more nega­tive than the Bay Area ra­tio­nal­ist com­mu­nity norms tend to be.

  Also, I slightly worry that what you’re see­ing is skewed be­cause the kind of peo­ple who are will­ing to try polyamory are un­usu­ally open about their per­sonal lives. That is, I think there’s a lot of drama go­ing on with monogamy, too, and it just hap­pens that peo­ple who choose to be monog­a­mous also have a cul­ture that bet­ter keeps drama se­cret un­til it is too big to keep se­cret any­more, so it sim­ply looks more com­mon in polyamory be­cause it is less hid­den.

  I think that there is cer­tainly a lot of drama with monogamy as well, and I agree that some as­pects of this can be un­der the sur­face. That be­ing said, I think there are some as­pects of poly that tend to ex­ac­er­bate/​lead to drama while there are some as­pects of monogamy that tend to miti­gate/​avoid it.

  I’ll give a ba­sic ex­am­ple. Let’s say that there is a cou­ple in a com­mit­ted re­la­tion­ship, and one mem­ber of the cou­ple starts get­ting closer to a third party. They be­come more and more emo­tion­ally close un­til even­tu­ally this bond seems stronger than the origi­nal re­la­tion­ship and the origi­nal cou­ple splits up.

  Now, this could eas­ily hap­pen ei­ther in monogamy or in polyamory—you could say that it’s the story of an “emo­tional af­fair” that turns into a real af­fair and splits up a monog­a­mous cou­ple; you could also say that it’s the story of a sec­ondary re­la­tion­ship in a polyamorous situ­a­tion that turns into a pri­mary re­la­tion­ship and splits up an old pri­mary re­la­tion­ship. In point of fact I have seen cases that seem to fit this de­scrip­tion in both monog­a­mous and polyamorous situ­a­tions.

  The key differ­ence, though, is that monog­a­mous norms tend to work against such things, while polyamorous norms tend to en­courage and even di­rectly sup­port such things (un­til per­haps things go too far and it may be too late). The per­son who starts to form a close bond with some­one already in an es­tab­lished monog­a­mous re­la­tion­ship may be chided or dis­cour­aged by the com­mu­nity; the per­son who starts to form a close bond with some­one already in an es­tab­lished polyamorous pri­mary re­la­tion­ship is less likely to be chided and may even be sup­ported.

  In some cases I be­lieve I have seen this lead to ma­jor strife or even di­vorce in a way that I sus­pect peo­ple would have dis­endorsed and tried to avoid if they were bet­ter equipped to pre­dict; un­for­tu­nately, they were deal­ing with polyamory norms that made them ex­tra vuln­er­a­ble to this sort of thing.

• Davis_Kingsley 21 Aug 2019 0:08 UTC

  4 points

  in reply to: Dagon's comment on: Davis_Kingsley’s Shortform

  An in­ter­est­ing ex­am­ple of a re­lated thing is “charm” effects, where you pick be­tween mul­ti­ple things you can do with a card, all of them are sort of bad for the price, but the flex­i­bil­ity makes it worth it over­all. Some­times peo­ple fo­cus a lot on “raw effi­ciency”, but when it’s be­tween you pick­ing the best of sev­eral in­effi­cient op­tions or your op­po­nent pick­ing the worst of sev­eral highly effi­cient op­tions, the former tends to be much bet­ter.

• Davis_Kingsley 21 Aug 2019 0:04 UTC

  9 points

  in reply to: G Gordon Worley III's comment on: Davis_Kingsley’s Shortform

  My ex­pe­rience has been some­what differ­ent. I think that if you look at the ac­tual re­sults of what’s go­ing on re: poly in the ra­tio­nal­ist com­mu­nity it’s fairly ev­i­dent that things are go­ing wrong—there are large amounts of drama and prob­lems, well be­yond that in other com­mu­ni­ties that I cur­rently par­ti­ci­pate in (even other com­mu­ni­ties in the Bay Area etc.).

  The most ob­vi­ous ex­am­ple is that one of the peo­ple who was most in­volved in bring­ing polyamory to the early ra­tio­nal­ist com­mu­nity ended up get­ting di­vorced af­ter a lot of poly drama, left the com­mu­nity, and wrote a post about how poly is ac­tu­ally con­trary to long-term ro­man­tic goals. Un­for­tu­nately, since the au­thor had left the com­mu­nity, most peo­ple didn’t read the post.

  How­ever, most peo­ple are not ac­tu­ally look­ing back at the ev­i­dence IMO—it’s just be­come an in­stalled and un­ex­am­ined norm.

• Davis_Kingsley 20 Aug 2019 23:24 UTC

  8 points

  on: Davis_Kingsley’s Shortform

  There are a pair of things in the ra­tio­nal­ist com­mu­nity which I like to call “The Two Bad Polys”—polypha­sic sleep and polyamory. Both seem ap­peal­ing to many peo­ple and have been ex­per­i­mented with pretty widely in the com­mu­nity de­spite be­ing quite harm­ful; I strongly ad­vise against try­ing ei­ther. In prac­tice they seem to lead to lots of prob­lems for most peo­ple who try them.

  (At­tri­bu­tion note: I’m not sure whether I was the first to come up with this term to de­scribe the pair—I think the two were first referred to as a dan­ger­ous pair by some­one else but I might have come up with this par­tic­u­lar name for them.)

• Davis_Kingsley 20 Aug 2019 1:18 UTC

  10 points

  on: Davis_Kingsley’s Shortform

  Strat­egy mini-post:

  One thing that tends to be weak in strat­egy games is “op­po­nent’s choice” effects, where an abil­ity has mul­ti­ple pos­si­ble effects and an op­po­nent chooses which is re­solved. Usu­ally, each effect is stronger than what you would nor­mally get for a card with that price, but in prac­tice these cards are of­ten quite weak.

  For in­stance, the Magic: the Gather­ing card “Book Burn­ing” looks quite strong in the­ory, as it ei­ther does 6 dam­age or mills 6 cards (both strong effects that might well be worth more than the card’s cost, since this was a set where hav­ing cards in your grave­yard was quite rele­vant). How­ever, in fact it is quite weak, be­cause in prac­tice you will always get the effect that is less rele­vant; if the op­po­nent has life to spare they’ll take dam­age, and if the mill is no longer rele­vant they’ll let you mill in­stead.

  This pat­tern holds true across mul­ti­ple games. In Le­gend of the Five Rings, Levy is similarly weak de­spite the fact that a card that did only one of its effects would likely be over­pow­ered, as one effect or the other is likely to be much less rele­vant at any point in the game and the op­po­nent can always choose the less rele­vant effect.

• Davis_Kingsley 17 Aug 2019 0:26 UTC

  14 points

  on: Davis_Kingsley’s Shortform

  One con­cept peo­ple talk about in game de­sign is “pen­du­lum swing”, where some­thing that is too pow­er­ful or too weak is over­cor­rected in bal­ance patches and be­comes the op­po­site—some­thing too pow­er­ful be­comes too weak, while some­thing too weak be­comes too pow­er­ful.

  A similar con­cept can be pre­sent in other sec­tors as well—of­ten, notic­ing one prob­lem can lead to an over­cor­rec­tion that brings you the op­po­site prob­lem. For in­stance, an early stage or­ga­ni­za­tion might no­tice that they aren’t sys­tem­atic enough in their pro­cesses, over­cor­rect, and be­come too rigid and doc­tri­naire.

  
  

  (Dun­can Sa­bien uses this con­cept of pen­du­lum swing a lot, and while I was aware of it prior to his use he’s done a lot to bring it to at­ten­tion as a rele­vant ra­tio­nal­ity con­cept.)

Davis_Kingsley’s Shortform

• Davis_Kingsley 16 Aug 2019 0:23 UTC

  7 points

  in reply to: habryka's comment on: Beliefs Are For True Things

  Yeah, that’s why I said it ad­dressed this point “among oth­ers”—my sum­mary of the poem’s mes­sage would be some­thing like “There are time­less prin­ci­ples of moral­ity and com­mon sense that are fun­da­men­tally true; when what’s fun­da­men­tally true be­comes un­fash­ion­able and peo­ple be­lieve what’s pop­u­lar or sounds good in­stead, dis­aster ul­ti­mately en­sues.”

  My post refers pri­mar­ily to the sec­ond part of that mes­sage (be­liefs are for true things, re­ject this at your peril) rather than the first part.