gwillen

• gwillen 30 Jul 2019 21:06 UTC

  5 points

  in reply to: Raemon's comment on: Short­form Beta Launch

  I do see the “show par­ent” icon, now that you have pointed it out, and I have viewed a short-form post with miss­ing com­ments, ex­panded the com­ments, then ex­plic­itly gone back and looked for the com­ments I knew were miss­ing. I would say that it is ex­tremely sub­tle. I couldn’t find it, even when I was look­ing for it, un­til I knew ex­actly which com­ment to look for it on.

  I think (ab­sent a more-code solu­tion) a rea­son­able thing would be to re­place the tiny sub­tle icon with text like “(… par­ent com­ment omit­ted, click to show …)” (pre­sum­ably on a line by it­self above the cur­rent top line.)

  Right now I claim there is re­ally no in­di­ca­tion that a com­ment is miss­ing un­less the reader is ex­tremely fa­mil­iar with the in­ter­face, and even then it’s tiny and would be easy to miss even if you knew where to look.

• gwillen 30 Jul 2019 21:04 UTC

  2 points

  in reply to: Richard_Kennaway's comment on: Open Thread July 2019

  I agree that it only makes toy de­mos, but it definitely goes be­yond WIMP. It’s a simu­la­tion of the sort of in­ter­face one might ex­pect in a fu­ture where ev­ery sur­face is a screen—it’s a janky, ex­tremely-low-fidelity simu­la­tion, which holds to­gether barely well enough to serve the pur­pose, but it does serve, and it’s an in­ter­est­ing way to try out this in­ter­ac­tion style.

  A sur­pris­ingly large part of Dy­namic Land is ac­tu­ally (in some sense) self-host­ing. There is a small core/​ker­nel that is ex­trin­sic, in­clud­ing e.g. the parser for the lan­guage. But a lot of low-level func­tion­al­ity is in­deed im­ple­mented in­side the in­ter­preter (as I re­call, they use a sort of pseudo-Lua, crossed with Smal­ltalk-like agent stuff—my vague rec­ol­lec­tion is that there’s some­thing like a Lua in­ter­preter with a cus­tom pre­pro­ces­sor un­der­ly­ing it.)

• gwillen 30 Jul 2019 1:09 UTC

  2 points

  in reply to: Raemon's comment on: Short­form Beta Launch

  I think dis­play­ing this in a non-mis­lead­ing fash­ion would be pretty easy—just dis­play the divs for any in­ter­me­di­ate elided com­ments, but with no con­tent in them. Then the nest­ing will be visi­ble. You could dis­play their con­tent as ‘...’ or ‘show com­ment’ or what­ever if you wanted, but I think the min­i­mal solu­tion with­out that would to­tally solve the prob­lem where I claim the cur­rent dis­play is su­per mis­lead­ing about the con­text of com­ments.

  (The most se­vere ex­am­ple of the thing I’m point­ing to would be some­thing like:

  >  A: I hate pup­pies, kick them all!
  >      B: You are a ter­rible per­son!
  >          C: I agree com­pletely.
  

  be­ing ren­dered as:

  >  A: I hate pup­pies, kick them all!
  >      C: I agree com­pletely.
  

  with no di­rect in­di­ca­tion that any­thing is miss­ing. The ex­am­ple that prompted me to com­plain was not at this level of sever­ity but the out-of-con­text re­ply was jar­ring and con­fus­ing.)

• gwillen 30 Jul 2019 1:02 UTC

  7 points

  in reply to: Raemon's comment on: Short­form Beta Launch

  Face­book does not in any way have the spe­cific prob­lem I’m point­ing to here, which is that the nest­ing struc­ture of the com­ments is mis­rep­re­sented. Face­book never elides a com­ment while show­ing a di­rect child of that com­ment. Of course, face­book only sup­ports a sin­gle level of com­ment nest­ing, so it does some­times elide a com­ment while show­ing a same-level re­ply. But I think that does not ap­pear mis­lead­ing, both be­cause (1) fb users are thus ac­cus­tomed to a re­ply-look­ing com­ment NOT be­ing a re­ply to the thing it’s dis­played un­der, and (2) the ‘show more’ link is DIRECTLY in the place where the eye goes search­ing for the miss­ing com­ment, not in some un­re­lated lo­ca­tion down be­low.

  (EDIT: Another thing I re­al­ized while com­pos­ing my other com­ment: If you use the ‘re­ply’ but­ton on face­book, and you’re re­ply­ing to a re­ply to a com­ment, so that you gen­er­ate a same-level re­ply, it will be pre­fixed with the name of the per­son you’re re­ply­ing to. So at worst I will see a com­ment start­ing with “B: (re­ply to b)” nested un­der a com­ment by per­son A, which is an­other cue that the in­ter­me­di­ate com­ment has been hid­den.)

• gwillen 30 Jul 2019 0:54 UTC

  8 points

  in reply to: ryan_b's comment on: Open Thread July 2019

  I have been! I thought it was an in­ter­est­ing ex­per­i­ment, and I re­ally hope they have an­other com­mu­nity day so I can visit again. I think there’s prob­a­bly a lot to learn from it, and I think there are things that you can only learn effec­tively by try­ing out weird ex­per­i­ments in real life to see “how this feels”. But I don’t re­ally ex­pect any­thing to di­rectly come of it—the pro­ject is pretty janky, and while it’s a fan­tas­tic plat­form for tiny cute de­mos, I don’t think any con­crete part of it (other than the gen­eral sense of “this is an in­spira­tion to try to go recre­ate this neat type of in­ter­ac­tion in a more ro­bust way”) is re­ally use­ful.

• gwillen 30 Jul 2019 0:22 UTC

  2 points

  on: Short­form Beta Launch

  Bug re­port copied from a FB com­ment I made:

  I was brows­ing short­form and no­ticed that a com­ment was elided (pre­sum­ably for low score) in a way that made its chil­dren ap­pear to be chil­dren of its par­ents. This is re­ally su­per duper mis­lead­ing. LW should find some way to in­di­cate that this is not the case, other than the ex­is­tence of a tiny “show more com­ments” link that isn’t nec­es­sar­ily even very nearby.

  [I have no rea­son to be­lieve that this was speci­fi­cally a short­form thing—I as­sume it will do this el­se­where too. This is just where I dis­cov­ered it.]

• gwillen 19 May 2019 0:47 UTC

  4 points

  in reply to: Raemon's comment on: The Village and The Mission

  I like your defi­ni­tion of the mis­sion—I haven’t heard it de­scribed in that way /​ that de­gree of de­tail be­fore, and I tend to agree with it. I’m not sure how uni­ver­sally agreed it is, but I would cer­tainly ad­vo­cate for your vi­sion of it.

• gwillen 13 Oct 2018 2:34 UTC

  23 points

  in reply to: Alicorn's comment on: Nyoom

  My ex­pe­rience with phys­i­cal ther­a­pists has been that they are good at find­ing soft tis­sue (mus­cle, ten­don, lig­a­ment) is­sues; and they gen­er­ally provide ex­er­cises for you to do at home which are sup­posed to help your spe­cific is­sues. (They do not ex­pect you to pay close at­ten­tion to your move­ment at all times; nor­mal peo­ple can’t do that and they’re at least that re­al­is­tic.) I sup­pose they might ask you to do some­thing like it for a short time to help with di­ag­no­sis, but not as a form of long-term treat­ment. (Although as an­other caveat, if there’s a spe­cific thing you are do­ing that they think is caus­ing ma­jor prob­lems, they may sug­gest you stop do­ing it. For ex­am­ple, I sig­nifi­cantly changed my sleep­ing po­si­tion to fix some back prob­lems I was hav­ing.)

  Pay­ment-wise, I be­lieve you may be able to get this cov­ered by in­surance if you get a refer­ral to a PT through your pri­mary care physi­cian. And no­tably, in the state of Cal­ifor­nia I be­lieve a PT can’t legally see you with­out such a refer­ral any­way, from my own re­search when I was go­ing to see one.

  Your physi­cian may want to test you for other things that might cause your is­sues, if they can think of any com­mon ones, e.g. vi­tamin defi­cien­cies. They may want you to get a gen­eral-pur­pose blood test for Ba­sic Defi­cien­cies in Im­por­tant Stuff, just to check. Your in­surance ought to cover all of this, sub­ject to any co­pays/​de­ductibles etc., which I un­der­stand could still be ex­pen­sive de­pend­ing on your in­surance.

• gwillen 9 Oct 2018 11:56 UTC

  2 points

  in reply to: ChristianKl's comment on: Things I Learned From Work­ing With A Mar­ket­ing Advisor

  I like the word ‘syn­ergy’ for this, be­cause peo­ple liter­ally use it as a generic ex­am­ple of a mean­ingless busi­ness word, but—when not be­ing abused—it ac­tu­ally has a very pre­cise mean­ing. (Value cre­ated by com­bi­na­tion; the part of the whole which ex­ceeds the sum of the parts. You could al­most trans­late it as “bind­ing en­ergy.” Or per­haps by anal­ogy “bind­ing value.”)

  I didn’t re­ally fully grasp the im­por­tance of the con­cept un­til I was spend­ing some time think­ing about the ques­tion: Why merge or spin off com­pa­nies? Doesn’t a free-mar­ket merger or spinoff trans­ac­tion re­sult in trad­ing a busi­ness for an amount of cash worth the same amount as the busi­ness? Syn­ergy (or, I sup­pose, anti-syn­ergy, though I’ve never heard the word used) -- bind­ing en­ergy, pos­i­tive or nega­tive—is the an­swer.

• gwillen 23 Jun 2018 1:14 UTC

  10 points

  on: Does any­one use ad­vanced me­dia pro­jects?

  I am ex­tremely in­ter­ested in this area and would love to work with other in­ter­ested in­di­vi­d­u­als in this area, or be di­rected to any rele­vant re­sources.

  (Bret Vic­tor and Alan Kay are both, as far as I know, re­spon­si­ble in part for this pro­ject, which I’ve vis­ited and seems pretty in­ter­est­ing: https://​​dy­nam­i­cland.org/​​ . I am not re­ally ob­ject-level con­vinced that what they are cur­rently build­ing is a fruit­ful di­rec­tion, but it’s at least quite cool, and it’s the right kind of new di­rec­tion/​idea.)

• gwillen 23 Jun 2018 1:10 UTC

  12 points

  on: Ben Hoff­man’s donor recommendations

  I am cu­ri­ous about the rea­son­ing be­hind “giv­ing un­so­lic­ited mon­e­tary gifts to friends or ac­quain­tances (peo­ple known to you per­son­ally, not via mass me­dia) who seem like they’re ba­si­cally benev­olent and com­pe­tent, have high marginal value for money, and aren’t ask­ing for any.” This doesn’t seem like a very EA recom­men­da­tion—any­body I know per­son­ally is al­most cer­tainly among the rich­est in­di­vi­d­u­als in the world, and even if they have the high­est marginal value for money of any­body I know, it’s prob­a­bly still quite low rel­a­tive to e.g. re­cip­i­ents of GiveDirectly grants. Right?

  (I’m ask­ing more to un­der­stand the rea­son­ing be­hind the sug­ges­tion, and whether Ben feels this is in­deed an “EA” recom­men­da­tion—not to challenge that this may well be a great thing to do, ei­ther way.)

• gwillen 5 Jun 2018 6:36 UTC

  1 point

  in reply to: gwillen-test's comment on: Open Thread June 2018

  Test re­ply-re­ply.

• gwillen 5 Jun 2018 6:35 UTC

  −4 points

  on: Open Thread June 2018

  Test com­ment. Please don’t up­vote etc. etc.

  (Sorry mod­er­a­tors, I as­sume you deleted my other ones, but I can’t re­ally try to de­bug no­tifi­ca­tion break­age with­out cre­at­ing more.)

• gwillen 1 Jun 2018 18:48 UTC

  30 points

  in reply to: Vaniver's comment on: Meta-Hon­esty: Firm­ing Up Hon­esty Around Its Edge-Cases

  I made a pro­posal for a mod­er­a­tor tool that seems like it might have been helpful to this thread, partly in re­sponse to your brack­eted text, and I’d be cu­ri­ous to hear your thoughts. https://​​github.com/​​LessWrong2/​​Less­wrong2/​​is­sues/​​610

• gwillen 22 May 2018 0:39 UTC

  9 points

  in reply to: stardust's comment on: The Berkeley Com­mu­nity & The Rest Of Us: A Re­sponse to Zvi & Benquo

  The win­ter sols­tice last year used the same venue it had used the pre­vi­ous year, but the venue im­posed a new, lower re­stric­tion on the max­i­mum num­ber of at­ten­dees, due to some new in­ter­pre­ta­tion of the fire code or some­thing. As a re­sult, tick­ets did sell out. (I wasn’t close enough to or­ga­ni­za­tion last year to know how last-minute the change was, but my im­pres­sion was that there was some scram­bling in re­sponse.)

  This year a new venue is be­ing sought that can bet­ter ac­com­mo­date the num­ber of peo­ple who want to at­tend.

• gwillen 22 May 2018 0:15 UTC

  15 points

  in reply to: Qiaochu_Yuan's comment on: Vi­sions of Sum­mer Solstice

  Yikes, ok. “Party-hop­ping” makes it sound like they didn’t ac­tu­ally buy tick­ets or know what the event was, but just came in off the street. Per­haps this is a case for ticket en­force­ment /​ bounc­ers? I’m re­luc­tant to sug­gest a rem­edy based on a sin­gle ex­am­ple, since it sounds like the peo­ple you’re talk­ing about ar­rived to­gether, and so ar­guably con­sti­tute a sin­gle dat­a­point among them. But hav­ing some­one des­ig­nated to deal with “prob­lem peo­ple” is ar­guably a good idea even from zero dat­a­points—usu­ally this is the sort of thing a code of con­duct might spell out, but I think the most ba­sic step is hav­ing a per­son with ex­tra cy­cles who knows it’s their re­spon­si­bil­ity to deal with this kind of thing (and who knows they have the full au­thor­ity to eject peo­ple, with right of ap­peal if the or­ga­niz­ers want to take those, but with­out hav­ing to in­volve a com­mit­tee.)

• gwillen 21 May 2018 19:08 UTC

  14 points

  in reply to: Qiaochu_Yuan's comment on: Vi­sions of Sum­mer Solstice

  I kind of wish it were sim­ple to just go “hm, can you iden­tify these peo­ple to me so that I can make my own judge­ment?” But there are prob­a­bly sev­eral rea­sons that’s im­prac­ti­cal in­clud­ing (I ex­pect) that you don’t know them by name.

  It seems sur­pris­ing to me that, in such a large crowd, there could be mul­ti­ple peo­ple man­ag­ing to make you un­com­fortable (as op­posed to, by con­trast, man­ag­ing to make some­one some­where un­com­fortable, but not all the same some­one, if that makes sense.) I am definitely won­der­ing whether I didn’t en­counter them, or en­coun­tered them and they didn’t have the same im­pact on me. I’m cu­ri­ous if you are able to de­scribe the be­hav­iors you saw, and also if you have a sense of broader con­text of like, do you think they were hav­ing this effect on lots of peo­ple? Do you feel like other ra­tio­nal­ist events have similar is­sues or is this some­thing unique to this one (maybe be­cause it’s ad­ver­tised more widely? Did you have a sense of where they came from?)

• gwillen 19 Apr 2018 4:30 UTC

  6 points

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

  (A pos­si­ble down­side I see is that it might some­how do the op­po­site—that vot­ing will feel like some­thing that is re­in­forced in a con­di­tion­ing sense, so that users with more vot­ing power will get more re­in­forcers since they do click->re­ward more times, and that this will ac­tu­ally give them a habit of want­ing to ap­ply the max­i­mum vote more than they oth­er­wise would be­cause it feels satis­fy­ing to vote re­peat­edly. This isn’t clearly a lot worse than the situ­a­tion we have now, where you always vote max­i­mum with no op­tion.)

• gwillen 19 Apr 2018 4:28 UTC

  6 points

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

  I liked the idea I think you men­tioned in an ear­lier thread about this, where each click in­creases vote weight by one. It’s con­cep­tu­ally very sim­ple, which I think is a good prop­erty for a UI. It does in­volve more clicks to ap­ply more vot­ing power, but that doesn’t seem bad to me. How of­ten does one need to give some­thing the max­i­mum amount of votes, such that ex­tra clicks are a prob­lem? It seems to me this would tend to de­fault to giv­ing ev­ery­one the same vot­ing power, but al­low users with more karma to sum­mon more vot­ing power with very slightly more effort if they think it’s war­ranted. That feels right to me.

• gwillen 1 Apr 2018 20:16 UTC

  11 points

  on: An­nounc­ing: Karma 2.0

  Post­ing be­cause my ego feels com­pel­led to see how big my text is.