Kenny

• Kenny 20 Sep 2019 16:55 UTC

  1 point

  in reply to: slikts's comment on: Is there an ex­ist­ing la­bel for the cat­e­gory of fal­la­cies ex­em­plified by “para­dox of tol­er­ance”?

  You seem to have a lot of as­sump­tions that prob­a­bly need to be ‘un­packed’ (made ex­plicit). For one, ‘ab­solutism fal­lacy’ isn’t ob­vi­ously a fal­lacy. Paci­fism can definitely be ‘ab­solute’ and, as I claimed in my an­swer to this ques­tion, I don’t even think that’s para­dox­i­cal.

  Are you try­ing to gather ‘rhetor­i­cal am­mu­ni­tion’ to defend ‘paci­fism’ and ‘tol­er­ance’ as prin­ci­ples, speci­fi­cally? I’m con­fused be­cause you seem to be deny­ing that ei­ther of those prin­ci­ples can even be in­ter­preted liter­ally or ‘ab­solutely’ and it seems ob­vi­ous to me that they can (and that peo­ple of­ten do so).

  I’m per­son­ally on-board with ‘game-the­o­ret­i­cal steel­man’ ver­sions of ‘paci­fism’ and ‘tol­er­ance’, but the ‘game-the­o­ret­i­cal steel­man­ning’, in my mind, nec­es­sar­ily in­volves all of my other val­ues, i.e. there aren’t ‘pure (but so­phis­ti­cated) non-ab­solute’ ver­sions of those prin­ci­ples to which ev­ery suffi­ciently ad­vanced thinker would read­ily agree. (For one, I sus­pect that the steel­manned ver­sion of those prin­ci­ples is in­evitably com­pli­cated and in­tri­cately de­tailed due to its in­ter­ac­tions with other val­ues, and to the vari­a­tion and gen­eral in­co­her­ence/​in­con­sis­tency of hu­man val­ues.)

• Kenny 20 Sep 2019 16:42 UTC

  1 point

  on: Is there an ex­ist­ing la­bel for the cat­e­gory of fal­la­cies ex­em­plified by “para­dox of tol­er­ance”?

  I don’t think they’re fal­la­cies.

  Some paci­fists re­ally do be­lieve that vi­o­lence should be avoided ab­solutely, even as a last re­sort. And that doesn’t even seem to be a para­dox, just a strat­egy with an ex­treme weak­ness.

  I think the ‘para­dox of tol­er­ance’ re­ally is a para­dox given that it’s not ob­vi­ous, for any­one abid­ing by the prin­ci­ple, how tol­er­ant they should be of in­tol­er­ance. Of course, any given non-ab­solute de­gree or ‘dis­tri­bu­tion’ of tol­er­ance could be ‘self-con­sis­tent’ so it’s not an un­avoid­able ‘gotcha’ by any means. But the sim­plest, most literal forms do definitely seem to be para­dox­i­cal, un­less it’s in­ter­preted en­tirely per­son­ally, e.g. ‘I should tol­er­ate ev­ery­thing and any­thing.’.

  My fa­vorite ex­am­ple – which I think is, in a sense, para­dox­i­cal – is the pre­cau­tion­ary prin­ci­ple. It’s definitely not ob­vi­ous that it shouldn’t ap­ply to peo­ple adopt­ing the prin­ci­ple it­self and, in fact, do­ing so is one rea­son why I re­ject it as a prin­ci­ple. It seems ob­vi­ous to me that the su­pe­rior prin­ci­ple is to ‘make the best de­ci­sions one can given the in­for­ma­tion, and at­ten­dant un­cer­tainty, available’.

  Gen­er­ally, I sus­pect that if the above prin­ci­ples, and similar ones, are ‘sharp­ened’ by mod­ify­ing them to “in­te­grate crit­i­cal re­jec­tion”, one would ar­rive at an en­tirely differ­ent (and more so­phis­ti­cated) prin­ci­ple like, e.g. ‘make the best de­ci­sions one can’.

• Kenny 19 Sep 2019 0:29 UTC

  1 point

  in reply to: Liron's comment on: The Power to Solve Cli­mate Change

  While the link “wash clothes > re­duce your per­sonal car­bon foot­print” is definite, my point is that the cat­e­gory of solu­tions that rely on the link “re­duce your per­sonal car­bon foot­print > solve car­bon change” are in­definite.

  I’m not see­ing much differ­ence in ‘definit­ed­ness’ be­tween per­sonal be­hav­ior change and “world or­ga­ni­za­tion that sets per-coun­try tar­gets” un­less it’s just that solv­ing global warm­ing nec­es­sar­ily must in­volve global gov­ern­ment. I think there are a LOT more ex­am­ples of even ‘in­tra-coun­try tar­gets’ , for any­thing (not global warm­ing), be­ing effec­tively bul­lshit com­pared to ex­am­ples of the op­po­site. I’m think­ing of things like (the U.S.) The War on Drugs, but drug pro­hi­bi­tion gen­er­ally seems to fit pretty well.

  And more gen­er­ally, a lot of ‘gov­ern­ment’ solu­tions seem to be pretty in­definite. It’s a de­press­ingly com­mon fea­ture of gov­ern­ment. (And of course gov­ern­ments do im­ple­ment definite spe­cific poli­cies too, so it’s not an in­evitable failure.)

• Kenny 15 Sep 2019 5:32 UTC

  1 point

  in reply to: Zvi's comment on: Who To Root For: 2019 Col­lege Foot­ball Edition

  Foiled! I was go­ing to write a ‘scathing’ com­ment about this not be­ing ap­pro­pri­ate for cross-post­ing here.

  (I loved the post. I think it does a great job at ges­tur­ing at the kinds of things I’d ex­pect you to in­clude in an even­tual “fun­da­men­tals-level guide on sports”.)

• Kenny 11 Sep 2019 3:05 UTC

  1 point

  in reply to: ryan_b's comment on: [Link] Book Re­view: Refram­ing Su­per­in­tel­li­gence (SSC)

  A lot of the dis­tinc­tion be­tween a ser­vice and an agent seems to rest on the differ­ence be­tween think­ing and do­ing.

  That doesn’t seem right to me. There are sev­eral, po­ten­tially sub­tle differ­ences be­tween ser­vices and agents – the bound­ary (or maybe even ‘bound­aries’) are prob­a­bly neb­u­lous at high re­s­olu­tion.

  A good pro­to­typ­i­cal ser­vice is Google Trans­late. You sub­mit text to it to trans­late and it out­puts a trans­la­tion as text. It’s both think­ing and do­ing but the ‘do­ing’ is limited – it just out­puts trans­lated text.

  A good pro­to­typ­i­cal agent is AlphaGo. It pur­sues a goal, to win a game of Go, but does so in a (more) open-ended fash­ion than a ser­vice. It will con­tinue to play as long as it can.

  Down-thread, you wrote:

  I am aiming di­rectly at ques­tions of how an AI that starts with a only a robotic arm might get to con­trol­ling drones or trad­ing stocks, from the per­spec­tive of the AI.

  I think one thing to point out up-front is that a lot of cur­rent AI sys­tems are gen­er­ated or built in a stage dis­tinct from the stage in which they ‘op­er­ate’. A lot of ma­chine learn­ing al­gorithms in­volve a dis­tinct pe­riod of learn­ing, first, which pro­duces a model. That model can then be used – as a ser­vice. The model/​ser­vice would do some­thing like ‘tell me if an image is of a hot dog’. Or, in the case of AlphaGo, some­thing like ‘given a game state X, what next move or ac­tion should be taken?’.

  What makes AlphaGo an agent is that it’s model is op­er­ated in a mode whereby it’s con­tinu­ally fed a se­quence of game states, and, cru­cially, both its out­put con­trols the be­hav­ior of a player in the game, and the next game state its given de­pends on it’s pre­vi­ous out­put. It be­comes em­bed­ded or em­bod­ied via the feed­back be­tween its out­put, player be­hav­ior, and its sub­se­quence in­put, a game state that in­cludes the con­se­quences of its pre­vi­ous out­put.

  But, we’re still miss­ing yet an­other cru­cial in­gre­di­ent to make an agent truly (or at least more) dan­ger­ous – ‘on­line learn­ing’.

  In­stead of train­ing a model/​ser­vice all at once up-front, we could train it while it acts as an agent or ser­vice, i.e. ‘on­line’.

  I would be very sur­prised if an AI in­stalled to con­trol a robotic arm would gain con­trol of drones or be able to trade stocks, but just be­cause I would ex­pect such an AI to not use on­line learn­ing and to be over­all very limited in terms of what in­puts with which it’s pro­vided (e.g. the po­si­tion of the arm and maybe a cam­era cov­er­ing its work area) and what out­puts to which it has di­rect ac­cess (e.g. a se­quence of arm mo­tions to be performed).

  Prob­a­bly the most dan­ger­ous kind of tool/​ser­vice AI imag­ined is an or­a­cle AI, i.e. an AI to which peo­ple would pose gen­eral open-ended ques­tions, e.g. ‘what should I do?’. For or­a­cle AIs, I think some other (pos­si­bly) key dan­ger­ous in­gre­di­ents might be pre­sent:

  • Knowl­edge of other or­a­cle AIs (as a plau­si­ble step­ping stone to the next in­gre­di­ent)

  • Knowl­edge of it­self as an or­a­cle AI (and thus an im­por­tant as­set)

  • Knowl­edge of its own effects on the world, thru those that con­sult it, or those that are oth­er­wise aware of its ex­is­tence or its output

• Kenny 8 Sep 2019 18:13 UTC

  1 point

  in reply to: Vanilla_cabs's comment on: Mis­take Ver­sus Con­flict The­ory of Against Billion­aire Philanthropy

  That’s a good ques­tion. Or, rather, of the sev­eral ways I can in­ter­pret it (ha), each seems in­ter­est­ing.

  I in­ter­pret your an­swer as be­ing hon­est and in good faith. I’d de­fault to the same were Re­ich to an­swer, if he were to an­swer like you did. I’d ex­pect most other promi­nent pub­lic crit­ics to deflect in some way.

  More gen­er­ally, I’d in­ter­pret similar an­swers from oth­ers writ­ing ‘against billion­aire philan­thropy’ as weak-mod­er­ate ev­i­dence of the same.

  As to how to more pre­cisely test that, I ad­mit that it’s prob­a­bly very tricky and thus I down­grade how “cru­cial” a test it re­ally is. Here’s one idea:

  Some billion­aire, one of those pre­vi­ously crit­i­cized in the man­ner un­der dis­cus­sion, an­nounces that, for ev­ery philan­thropic dona­tion they make, they’ll make ‘match­ing’ dona­tions to the rele­vant fed­eral, state, and mu­ni­ci­pal trea­suries to ‘offset’ the tax re­bate/​re­fund effect of the dona­tions.

  I’d ex­pect that, mostly, this would re­sult in heav­ier crit­i­cism and in­creas­ing sus­pi­cion. I’d ex­pect you, if asked, to mod­er­ate your own crit­i­cism or praise the offset­ting di­rectly.

  ‘Ideally’, we’d ask The Si­mu­la­tors of the Uni­verse, to re-run the uni­verse simu­la­tion and ‘mag­i­cally’ have some kind of tax law passed that re­moves the re­fund/​re­bate at some point be­fore some por­tion of billion­aire philan­thropic dona­tions and we could mea­sure the num­ber and ‘sen­ti­ment’ of crit­i­cisms.

  Real­is­ti­cally, we could prob­a­bly much much more crudely ap­prox­i­mate some­thing similar, but any com­par­i­sons would in­evitably be con­founded by all kinds of other things.

• Kenny 26 Aug 2019 15:09 UTC

  2 points

  in reply to: Spiracular's comment on: Paper Trauma

  GitLab’s Mark­down sup­ports charts and di­a­grams.

  I find ta­bles in Mark­down pretty easy to in­put, on a com­puter, be­cause I can for­mat them eas­ily in Vim, my fa­vorite and always-open text ed­i­tor.

• Kenny 22 Aug 2019 18:21 UTC

  4 points

  on: Do you do weekly or daily re­views? What are they like?

  I don’t do weekly, or longer-pe­riod, re­views any more – or at least I don’t com­mit to do­ing them on a sched­ule.

  Daily Review

  My daily re­view cen­ters around re­view­ing all of my ‘tasks’ in Habitica. I have a daily re­view task in Habitica that con­sists of the fol­low­ing steps (and is rep­re­sented as a check­list):

  • Re­view all tasks

  • Re­view cal­en­dar for to­day*

  • Pro­cess ev­ery un­read email

  • Re­view all of my ‘to be re­viewed’ projects

  * I’ve been ob­serv­ing some­thing like a ‘sec­u­lar Sab­bath’ (roughly) each Fri­day start­ing at 6p to Satur­day at the same time. On Fri­day, in the morn­ing or later dur­ing the day, I re­view any cal­en­dar items for the next day, Satur­day, too. I gen­er­ally avoid schedul­ing any­thing for Satur­day, but some­times I ei­ther have to do some­thing that day or want to any­ways.

  Re­view­ing tasks

  The goal is just to read (or skim) each task and think about them min­i­mally. If one is some­thing I can do in liter­ally 2-5 min­utes, some­times I’ll do it right then, but I don’t have to ac­tu­ally com­plete any­thing dur­ing the re­view.

  I’ll oc­ca­sion­ally delete old tasks that I’ve given up on ever do­ing. If I re­mem­ber that I’ve already com­pleted a task, I’ll mark it com­pleted and, if it’s part of a pro­ject, up­date the pro­ject info and, usu­ally, flag that pro­ject ‘to be re­viewed’.

  Tasks in Habitica can have check­lists but I’ve re­al­ized that I’m us­ing them too of­ten and that, in­stead of hav­ing one task with a check­list, I should more of­ten have sep­a­rate tasks. The key dis­tinc­tion to which is bet­ter is whether each step needs to be done to­gether, es­pe­cially in or­der, or whether each step can be done in­de­pen­dently. Laun­dry, for me, is a se­quence of steps that all need to be done, in or­der, to re­sult in me hav­ing clean clothes and other items. Dust­ing my house how­ever is some­thing that I re­ally can do room by room.

  One rea­son I re­ally like Habitica is that there are three differ­ent types of tasks: ‘to-dos’, ‘dailies’, and ‘habits’. To-dos are just like tasks in most any other task list sys­tem – some­thing that you can mark ‘com­pleted’ when its done. Dailies are tasks that re­peat – I mostly just use a ‘weekly’ sched­ule for spe­cific days, e.g. the daily task for my daily re­view re­peats ev­ery day ex­cept Satur­day. Habits are tasks that can be com­pleted, or ‘missed’, at any time. I’ve got one now to ‘re­mem­ber to ei­ther pump up the tires on your bike be­fore you ride it, or at least check that their pres­sure is fine’. I only have a hand­ful, or less, at any one time.

  Re­view­ing my calendar

  I use Google Cal­en­dar – mainly be­cause I can ac­cess it from my phone, so most any­where, and it’s free.

  I have sep­a­rate cal­en­dars and I use them to cat­e­go­rize items. My main cal­en­dar has my re­minders and events that I am ei­ther plan­ning on at­tend­ing (e.g. some­thing to which I need to ei­ther travel or com­mute) or in which I am plan­ning on par­ti­ci­pat­ing (e.g. a phone call). I’ve got a ‘fam­ily’ cal­en­dar for track­ing the sched­ules of fam­ily mem­bers or friends. I’ve got a ‘maybe’ cal­en­dar in which I put things like, e.g. fun events I might want to at­tend or the hours of my lo­cal rock climb­ing gym.

  There are two types of items: events and re­minders.

  For events on my main cal­en­dar, there are two broad types: all-day and with-times. All-day events are usu­ally just re­minders, e.g. I’m on va­ca­tion. For events with times, typ­i­cally I just need to de­cide whether I need to set an alarm on my phone, e.g. to get ready to leave to travel or com­mute to the event.

  For re­minders, I mostly just copy them to my Habitica to-do list; some­times I’ll just mark a few as com­pleted or delete them. I’m us­ing them much like what the Get­ting Things Done sys­tem terms a ‘tick­ler file’. I gen­er­ally add tasks that I need to do ‘later’ or on some kind of sched­ule as cal­en­dar re­minders; the idea be­ing that my task (‘to-do’) list in Habitica can be free of them un­til they’re due.

  Pro­cess­ing email

  My goal isn’t nec­es­sar­ily to read ev­ery email, com­pletely – just pro­cess each one (and then mark them as read, un­til I reach ‘in­box zero’). For long emails that I do want to read, I’ll ei­ther save a web ver­sion in a ‘read later’ app or add a task to Habitica to read or re­view the email. I’ll of­ten add a task in Habitica to re­spond to some­one if I can’t do so within a few min­utes right away.

  I track all of my fi­nan­cial ac­tivity in YNAB, a nice bud­get­ing app so any email re­ceipts get en­tered there im­me­di­ately.

  For some emails, I’ll up­date the info for any re­lated pro­jects, add tasks in Habitica, or add some­thing to my cal­en­dar.

  Re­view­ing ‘to be re­viewed’ projects

  I’m us­ing GitLab – a free ac­count on the offi­cial ‘hosted’ in­stance. I’ve got a lot of ‘pro­jects’ (GitLab’s term), most of them per­tain­ing to code, but sev­eral just for main­tain­ing info about var­i­ous pro­jects. I mostly use a sin­gle ‘pro­ject’ named “@misc”.

  I’ve been us­ing (soft­ware-de­vel­op­ment-fo­cused) is­sue track­ers for at least a decade now and GitLab’s my fa­vorite so far. The main rea­sons why I like it more than any oth­ers I’ve tried is that it uses Mark­down, and its Mark­down di­alect is fan­tas­tic, and that its got a sep­a­rate de­scrip­tion for each is­sue (whereas some track­ers only have com­ments). Mark­down, es­pe­cially GitLab’s di­alect, al­lows me to eas­ily quote emails, link to web pages (and en­tire sets of them from open tabs with a nice Chrome ex­ten­sion), and main­tain check lists of tasks. Hav­ing an is­sue de­scrip­tion sep­a­rate from com­ments let’s me main­tain a nice overview of a pro­ject and a sin­gle list (or tree) of tasks (or, more of­ten, a board out­line thereof).

  Each (‘real world’) pro­ject gets an ‘is­sue’ in the GitLab ‘pro­ject’. I reg­u­larly edit the is­sue de­scrip­tion so that it con­tains an up-to-date overview and out­line of tasks. I add com­ments with info, quotes, links, and mini sub-pro­jects and their tasks.

  I as­sign an is­sue to my­self to mark it as ‘to be re­viewed’. Dur­ing my daily re­view, my goal for each as­signed is­sue is mainly to re­view the pro­ject for that is­sue and de­ter­mine what the next task is to be done. Once I’ve de­ter­mined the next task, I make sure I add it to Habitica (and I link the task in Habitica to the is­sue in GitLab). If I ex­pect to work fairly in­ten­sively on a pro­ject short-term, I’ll leave the is­sue as­signed to me; oth­er­wise, I un-as­sign it to my­self. I also use GitLab, and the same ac­count too, for work, so I’ll usu­ally have one or two work is­sues as­signed to me as well and, be­cause I usu­ally fo­cus on a sin­gle work pro­ject at a time, I’ll leave the cur­rently ac­tive is­sue or is­sues as­signed to my­self un­til I’m ei­ther finished or stuck wait­ing for some kind of out­side in­put.

  Other components

  Alarms

  I use the stan­dard Alarm app on my phone (an iPhone) a lot. I’ve got a few stan­dard, re­peat­ing alarms – ‘wakeup’, re­view my ‘roughly sched­uled’ tasks – but I also use it liber­ally for any­thing I want to re­mem­ber to do. I’ll use the timer fea­ture if I’m do­ing some­thing like cook­ing but, be­cause (in the stan­dard app any­ways) there’s only one timer, I mostly de­fault to us­ing alarms be­cause I can la­bel them, e.g. ‘Check the dryer’, ‘Leave to go _’, or ‘Get ready for phone call with X in Y min­utes’.

  Capturing

  I of­ten use email – i.e. I email my­self – about new tasks, pro­jects, or ‘refer­ence ma­te­rial’ I want to be able to quickly find later. (I use Gmail mainly be­cause its search is fan­tas­tic.) Some­times I’ll add tasks di­rectly to Habitica or pro­jects di­rectly as an is­sue in GitLab, but email is much more fric­tion­less and, be­cause I ha­bit­u­ally pro­cess my un­read email ev­ery day, I’m con­fi­dent I’ll cre­ate tasks or GitLab is­sues later if I send my­self an email.

  I’ve got a cou­ple of ‘logs’ in sep­a­rate notes in my phone’s stan­dard Notes app. I some­times think about writ­ing my own lit­tle (web) apps but that would be a lot of work and reg­u­lar text, tho struc­tured fairly reg­u­larly, is prob­a­bly not much worse, and (of course) already pos­si­ble (and eas­ily too).

  Roughly sched­uled tasks

  In Habitica, I’ve got three tags for tasks that are ‘roughly sched­uled’: ‘morn­ing’, ‘to­day’, and ‘tonight’. I’ve got alarms on my phone for each tag. I’ve com­mit­ted to re­view­ing any tasks with the rele­vant tag some­time around when the alarm is sched­uled. I don’t have to com­plete all, or even any, of those tasks; just re­view them. I, of course, try to do any that need to be done.

• Kenny 22 Aug 2019 4:00 UTC

  1 point

  in reply to: jessicata's comment on: Power Buys You Dis­tance From The Crime

  By-the-way, this is a fan­tas­tic com­ment and would make a great post pretty much by it­self (with maybe a lit­tle con­text about that to which it’s re­ply­ing).

• Kenny 22 Aug 2019 1:06 UTC

  1 point

  in reply to: jessicata's comment on: Power Buys You Dis­tance From The Crime

  en­act­ing con­flict in the course of dis­cussing conflict

  … seems to be ex­actly why it’s so difficult to dis­cuss a con­flict the­ory with some­one already con­vinced that it’s true – any dis­cus­sion is nec­es­sar­ily an at­tack in that con­flict as it in effect pre­sup­poses that it might be false.

  But that also makes me think that maybe the best rhetor­i­cal counter to some­one en­act­ing a con­flict is to ex­plic­itly claim that one’s un­con­vinced of the truth of the cor­re­spond­ing con­flict the­ory or to ex­plic­itly claim that one’s de­cou­pling the cur­rent dis­cus­sion from a (or any) con­flict the­ory.

• Kenny 21 Aug 2019 21:23 UTC

  1 point

  in reply to: johnswentworth's comment on: Power Buys You Dis­tance From The Crime

  I don’t think it’s use­ful to talk about ‘con­flict the­ory’, i.e. as a gen­eral the­ory of dis­agree­ment. It’s more use­ful in a form like ‘Marx­ism is a con­flict the­ory’.

  And then a ‘con­flict the­o­rist’ is some­one who, in some con­text, be­lieves a con­flict the­ory, but not that dis­agree­ments gen­er­ally are due to con­flict (let alone in all con­texts).

  So, from the per­spec­tive of a ‘work­ing class ver­sus cap­i­tal class’ con­flict the­ory, pub­lic choice the­ory is ob­vi­ously a weapon used by the cap­i­tal class against the work­ing class. But other pos­si­ble con­flict the­o­ries might be neu­tral about pub­lic choice the­ory.

  Maybe what makes ‘con­flict the­ory’ seem like a sin­gle thing is the prevalence of Marx­ism-like poli­ti­cal philoso­phies.

• Kenny 21 Aug 2019 19:21 UTC

  1 point

  in reply to: limerott's comment on: Per­mis­sions in Governance

  In do-oc­ra­cies, gen­er­ally the ‘re­vealed prefer­ences’ of the group mem­bers is pretty ob­vi­ous. The things the ‘group wants’ are read­ily re­vealed to be those things that the group mem­bers ac­tu­ally act to achieve or ac­quire.

  And, as a mat­ter of how do-oc­ra­cies form ini­tially, they typ­i­cally ‘ac­crete’ around a sin­gle per­son or a small group of peo­ple that are already ac­tively work­ing on some­thing. Think of a small open source pro­gram­ming pro­ject. Usu­ally the pro­ject is started by a sin­gle per­son and what­ever they ac­tu­ally work on is what they ‘want’ to work on. Often, when other peo­ple sug­gest changes, the ini­tial per­son (who is likely still the ‘pro­ject leader’) will re­spond along the lines of “Pull re­quests wel­come!”, which is ba­si­cally equiv­a­lent to “Feel free to work on the changes your­self and send them to me to re­view.”. And, some­times, a new con­trib­u­tor will work on the changes first, be­fore even dis­cussing the pos­si­bil­ity. And then, af­ter sub­mit­ting the changes to re­view, the pro­ject leader or other par­ti­ci­pants might ob­ject to the changes, but, by de­fault, any­one is free to make changes them­selves (tho typ­i­cally not any­one can ac­tu­ally make changes di­rectly to the ‘au­thor­i­ta­tive ver­sion’).

• Kenny 21 Aug 2019 18:57 UTC

  1 point

  in reply to: kithpendragon's comment on: Per­mis­sions in Governance

  It’s usu­ally not so much that pay­roll or taxes are “legally tuned” to the benefit of larger cor­po­ra­tions but that com­ply­ing with all of the rele­vant laws and reg­u­la­tions is a rel­a­tively large ‘fixed cost’ that can be more eas­ily born by a larger or­ga­ni­za­tion. Even some­thing like ini­tially se­lect­ing a pay­roll com­pany, or mon­i­tor­ing (and po­ten­tially switch­ing to an­other) pay­roll com­pany is some­thing more eas­ily, and less costly, performed by a ded­i­cated HR pro­fes­sional, let alone a group of pro­fes­sion­als in an HR de­part­ment, whereas lots of small busi­nesses don’t even have a full-time, ded­i­cated HR per­son.

• Kenny 21 Aug 2019 18:51 UTC

  1 point

  in reply to: limerott's comment on: Per­mis­sions in Governance

  An ex­am­ple of a (rel­a­tively) high cost of com­pli­ance is the re­cen­tish EU GDPR. Large com­pa­nies will be able to com­ply (rel­a­tively) more eas­ily than small com­pa­nies so the effect of the reg­u­la­tion is to priv­ilege large com­pa­nies over small (or smaller) ones , i.e. “keep the fixed pool of re­sources from be­ing di­vided be­tween too many peo­ple”, where the pool of re­sources in this case are po­ten­tial cus­tomers of on­line busi­nesses (or even just users of on­line sites or ser­vices).

  And more gen­er­ally, for al­most ev­ery law and reg­u­la­tion, it’s eas­ier for larger com­pa­nies or or­ga­ni­za­tions to ‘pay com­pli­ance costs’ so ev­ery new law or reg­u­la­tion effec­tively pe­nal­izes smaller com­pa­nies or or­ga­ni­za­tions.

  Note that this is ba­si­cally never con­sid­ered, let alone ad­ver­tised, as a de­liber­ate effect of any law or reg­u­la­tion.

  You’re right that so­cial af­fili­a­tion is of­ten used, in effect any­ways, to me­di­ate ac­cess to re­sources, but I’ve never en­coun­tered any­one de­scribing the ini­ti­a­tion or main­te­nance of af­fili­a­tion as be­ing a ‘com­pli­ance cost’, tho it’s not an in­apt anal­ogy and might op­er­ate pretty similarly. I think it’s rel­a­tively un­com­mon for so­cial af­fili­a­tion to in­volve ex­plicit rules tho, which dis­t­in­guishes it from what is typ­i­cally de­scribed as ‘com­pli­ance’.

• Kenny 21 Aug 2019 18:37 UTC

  1 point

  in reply to: Raemon's comment on: Per­mis­sions in Governance

  I’ve used cal­en­dar re­minders for ex­actly this.

  (And ‘ping’ by-the-way.)

• Kenny 20 Aug 2019 18:04 UTC

  1 point

  in reply to: Dagon's comment on: Mis­take Ver­sus Con­flict The­ory of Against Billion­aire Philanthropy

  I’ve come around to the “con­flict-vs-mis­take fram­ing” in par­tic­u­lar be­cause “ev­ery sig­nifi­cant dis­agree­ment has el­e­ments of both”.

  It must be the case, in some sense any­ways, that ev­ery ‘con­flict the­ory’ be­gins its (epistemic) ex­is­tence as a ‘mis­take the­ory’ and is thus, hope­fully, at least some­what amenable to be­ing con­sid­ered ‘mis­taken’ later given suffi­cient con­trary ev­i­dence.

  In gen­eral too, con­flict the­o­ries seem to have a ‘memetic’ ad­van­tage in be­ing ‘epistem­i­cally to­tal­i­tar­ian’, i.e. sub­sum­ing all sub­se­quent ev­i­dence (un­til the ex­is­tence of the con­flict is it­self later con­sid­ered mis­taken).

  It’s also true that some­thing like philan­thropy could be both net-pos­i­tive for ev­ery­one and net-nega­tive for a par­tic­u­lar poli­ti­cal coal­i­tion.

• Kenny 20 Aug 2019 17:50 UTC

  1 point

  in reply to: sludgepuddle's comment on: Mis­take Ver­sus Con­flict The­ory of Against Billion­aire Philanthropy

  I’m not sure it’s re­ally pos­si­ble to reach any con­flict the­o­rists if you think their the­o­rized con­flict is a mis­take.

  It seems like part of the prob­lem in do­ing so is that the the­o­rized con­flicts are (at least) im­plic­itly zero-sum. I’d think it’s pretty ob­vi­ous, that at least ‘in the­ory’, billion­aire philan­thropy could be net-pos­i­tive for ‘The Peo­ple’, but it’s hard to even imag­ine how one would go about con­vinc­ing some­one of that if they’re already con­vinced that (al­most) ev­ery­one’s ac­tions are at­tacks against the op­pos­ing side(s), e.g. philan­thropy is ‘re­ally just’ a way for billion­aires to se­cure some other kind of (in­di­rect) benefit to them­selves and their class.

• Kenny 20 Aug 2019 17:43 UTC

  2 points

  in reply to: Benquo's comment on: Mis­take Ver­sus Con­flict The­ory of Against Billion­aire Philanthropy

  It’s pretty hard to tell what you find hard to rec­on­cile in the two quotes.

  ‘Poli­tics as war’ is the same as ‘differ­ent sides fight for their own self-in­ter­est’, e.g. “whether the State ex­ists to en­rich the Elites or to help the Peo­ple”.

  The ‘hon­est mis­takes’ per­spec­tive would be that any par­tic­u­lar policy might be good or bad, or what­ever mix thereof, and dis­agree­ments about that would be due to differ­ent be­liefs and NOT due to sim­ply sup­port­ing one’s side.

• Kenny 20 Aug 2019 17:36 UTC

  1 point

  in reply to: Vanilla_cabs's comment on: Mis­take Ver­sus Con­flict The­ory of Against Billion­aire Philanthropy

  I think there’s a cru­cial test that could be performed, rel­a­tive to your ideas (in this thread) any­ways – how much of the ‘against billion­aire philan­thropy’ do you think is due to the tax re­bate/​re­fund? I think it’s close to zero.

  (And I don’t have a prob­lem with crit­i­ciz­ing any philan­thropy but I don’t have a prob­lem with billion­aires giv­ing large amounts gen­er­ally.)

• Kenny 17 Jul 2019 0:36 UTC

  1 point

  on: What are good re­sources for learn­ing func­tional pro­gram­ming?

  I’m a ‘lowly prac­ti­tioner’ and I’ve only used func­tional pro­gram­ming lan­guages a mod­est amount in my pro­fes­sional ex­pe­rience. Most of the other an­swers seem to be, mostly, fo­cus­ing on newer func­tional pro­gram­ming lan­guages, of which Haskell seems to be the ‘coolest’ one cur­rently, and that’s not one I’ve learned be­yond very cur­sory ‘skim­ming’.

  One met­ric by which I’m judg­ing this ques­tion as par­tic­u­larly great is that it bugs me – I’m not sure how to an­swer it, even ‘in prin­ci­ple’, for at least the ‘what’ and ‘how’ books. (For the ‘why’ book, I think SICP should be a great re­source even tho it’s not strictly func­tional pro­gram­ming.)

  What would a good ‘what’ book or re­source for func­tional pro­gram­ming look like? One rea­son I’m con­fused about this is that I’d ex­pect a good ‘what’ re­source to be spe­cific to an in­di­vi­d­ual pro­gram­ming lan­guage, but then would it still be a good ‘what’ re­source for func­tional pro­gram­ming in gen­eral?

  Similarly for any ‘how’ re­source – what would one look like that isn’t tied to a spe­cific lan­guage? Or not tied to any lan­guage at all? I’m prob­a­bly ‘typ­i­cally-mind­ing’ this, even from the per­spec­tive of an ex­pe­rienced pro­gram­mer, but I’m strug­gling to think of enough ‘how’ ma­te­rial spe­cific to func­tional pro­gram­ming to fill a book-sized re­source. Work­ing Effec­tively with Le­gacy Code is a great ‘how’ book (cov­er­ing ex­actly what the ti­tle im­plies), but I can’t think of, off the top of my head, how many ‘how’ ques­tions there could be for func­tional pro­gram­ming speci­fi­cally.