Kenny

• Kenny 26 Jan 2020 17:56 UTC

  1 point

  in reply to: Pattern's comment on: [Link] “Do­ing be­ing ra­tio­nal: polymerase chain re­ac­tion” by David Chapman

  Thanks!

• Kenny 24 Jan 2020 15:17 UTC

  4 points

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

  So, is your idea that, be­cause of the gen­eral prin­ci­ple (ha) of ‘in­tel­lec­tual char­ity’, we should – typ­i­cally at least, or maybe by de­fault – ha­bit­u­ally steel­man prin­ci­pled ar­gu­ments to au­to­mat­i­cally in­clude any jus­tified ex­cep­tions of which we’re aware?

  I think maybe it’d be bet­ter to sim­ply offer to, and query, our fel­low rea­son­ers about which jus­tified ex­cep­tions they ac­cept for any prin­ci­ples un­der dis­cus­sion.

  Mak­ing the prin­ci­ples pure or ab­solute is an at­tempt to make the re­quired judge­ment for­mu­laic in­stead, of­ten due to a cyn­i­cism about in­di­vi­d­ual judge­ment abil­ities of peo­ple, …

  It’s in gen­eral, in my ex­pe­rience any­ways, difficult to dis­t­in­guish be­tween cyn­i­cism and re­al­ism. Peo­ple re­ally are, or seem to be, pretty bad rea­son­ers in a lot of situ­a­tions. We re­ally do seem to be still, mostly, run­ning how-to-get-along-in-a-small-tribal-band soft­ware, par­tic­u­larly when do­ing moral rea­son­ing. Do you re­ally trust most peo­ple to make good moral judge­ments gen­er­ally? I’m on the fence, tho I do lean to a kind of Taoist ‘peo­ple are nat­u­rally good’ stance (‘at­ti­tude’). But I’m also reg­u­larly watch­ing for strong ev­i­dence of spe­cific peo­ple’s ac­tual moral de­ci­sions and rea­son­ing – and ‘cyn­i­cism’ isn’t always wrong.

• Kenny 24 Jan 2020 15:10 UTC

  6 points

  in reply to: Pattern's comment on: [Link] “Do­ing be­ing ra­tio­nal: polymerase chain re­ac­tion” by David Chapman

  I love this com­ment!

  In linkposts, of­ten the con­tent is quoted whole­sale (be­gin­ning to end) or par­tially (First X para­graphs or so). Although the post start­ing with a video might have made these op­tions more difficult than usual.

  I didn’t know that – thanks for the feed­back!

  Should I edit my post to quote the refer­enced con­tent?

  David Chap­man’s thoughts on AI might differ from yours.

  At suffi­cient de­tail, ev­ery­one’s thoughts about any­thing (suffi­ciently com­plex) differ from ev­ery­one else’s. But I don’t think David Chap­man and I have any fun­da­men­tal dis­agree­ments about AI.

  “AI have mem­ory prob­lems”

  Ooooh! That’s a perfectly con­cise form of a crit­i­cism I’ve thought about neu­ral net­work ar­chi­tec­tures for a long time. The net­works cer­tainly are a form of mem­ory them­selves, but not re­ally a his­tory, i.e. of dis­tinct and rel­a­tively dis­crete events or en­tities. Our own minds cer­tainly seem to have that kind of mem­ory and it seems very hard for an ar­bi­trary in­tel­li­gent rea­soner to NOT have some­thing similar (if not ex­actly some­thing like this).

  The quoted text you in­cluded is a perfect ex­am­ple of this kind of thing too; thanks for in­clud­ing it.

  Isn’t there ev­i­dence that hu­man brains/​minds have what is effec­tively a ded­i­cated ‘causal rea­son­ing’ unit/​mod­ule? It prob­a­bly also re­lies on the ‘thing mem­ory’ unit(s)/​mod­ule(s) too tho.

  • Per­haps the dis­tinc­tion is:

  • Ra­tion­al­ity is what you should do.

  • Meta-ra­tio­nal­ity is what you should do in or­der to “make ra­tio­nal­ity work”.

  While these two things can be com­bined un­der one um­brella, mak­ing defi­ni­tions smaller:

  • In­creases clar­ity (of dis­cus­sion)

  • Makes it eas­ier to talk about components

  • Makes it clear when all mem­ber­ship crite­rion for a cat­e­gory have been met.

  • Might help with teach­ing/​retention

  As I men­tioned, or im­plied, in this post, I’m in­differ­ent about the ter­minol­ogy. But I like all of your points and think they’re good rea­sons to make the dis­tinc­tion that Chap­man does. I’m go­ing to con­sider do­ing the same!

[Question] How much do we know about how brains learn?

• Kenny 14 Dec 2019 0:23 UTC

  3 points

  in reply to: Douglas_Knight's comment on: Matthew Walker’s “Why We Sleep” Is Rid­dled with Scien­tific and Fac­tual Errors

  I ap­pre­ci­ate any de­bunk­ing, at least a lit­tle.

  I found this of in­ter­est, if for no other rea­son than to trust claims made by the au­thor of the ‘de­bunked’ book less. The de­tails about the spe­cific claims de­bunked were also (mildly) in­for­ma­tive.

[Link] “Do­ing be­ing ra­tio­nal: polymerase chain re­ac­tion” by David Chapman

• Kenny 13 Dec 2019 22:05 UTC

  1 point

  in reply to: chaosmage's comment on: Elon Musk is wrong: Rob­o­taxis are stupid. We need stan­dard­ized rented au­tonomous tugs to move cus­tomized owned un­pow­ered wag­ons.

  Good an­swer!

  I was think­ing about peo­ple liv­ing in de­tached homes in res­i­den­tial neigh­bor­hoods, i.e. places where I would ex­pect lo­cal poli­tics to pre­vent car parks (‘park­ing lots’ in my col­lo­quial­isms) from be­ing built at all.

• Kenny 13 Dec 2019 21:57 UTC

  1 point

  in reply to: jkaufman's comment on: Ar­gu­ing about housing

  There’s prob­a­bly (at least) some­thing to that idea. I imag­ine com­mer­cial con­struc­tion is similarly con­strained as res­i­den­tial. It’s pretty com­mon to hear that com­mer­cial rents are high in the places where res­i­den­tial rents are too.

• Kenny 2 Dec 2019 23:34 UTC

  5 points

  on: Book Re­view: De­sign Prin­ci­ples of Biolog­i­cal Circuits

  Thanks for this post!

  I was ex­cited to read the book re­viewed just based on the first few sen­tences!

• Kenny 2 Dec 2019 20:39 UTC

  1 point

  in reply to: chaosmage's comment on: Elon Musk is wrong: Rob­o­taxis are stupid. We need stan­dard­ized rented au­tonomous tugs to move cus­tomized owned un­pow­ered wag­ons.

  Spe­cial­ized cab­ins seem like they would hurt this idea – where would peo­ple store all of their cab­ins?

• Kenny 28 Nov 2019 21:57 UTC

  1 point

  on: When do you start look­ing for a Bos­ton apart­ment?

  I’m re­ally con­fused. I’m used to the NYC rental mar­ket, par­tic­u­larly Brook­lyn, and, aside from lin­ing up apart­ment-mates, the rule is that you look for a new apart­ment right be­fore you’re read­ing to move. I can’t even re­mem­ber see­ing apart­ments listed for rent months in ad­vance, tho I wouldn’t be en­tirely sur­prised that it hap­pens, e.g. for stu­dents.

  Where are you get­ting your list­ings and how can you tell when the lease is in­tended or ex­pected to start from the list­ing?

• Kenny 8 Nov 2019 3:32 UTC

  3 points

  in reply to: gjm's comment on: ML is an in­effi­cient market

  Ep­silon value maybe?

  In a cer­tain sense, this is a triv­ial claim. Ob­vi­ously some peo­ple are go­ing to have ‘lo­cal’ com­pet­i­tive ad­van­tages but noth­ing speaks louder than suc­cess so word gets out even­tu­ally.

  But I’m in­trigued, so maybe this is not liter­ally val­ue­less.

• Kenny 8 Nov 2019 1:36 UTC

  8 points

  on: Maybe Ly­ing Doesn’t Exist

  This is a great post! A lot of these points have been ad­dressed, but this is what I wrote while read­ing this post:

  It’s not im­me­di­ately clear that an ‘ap­peal to con­se­quences’ is wrong or in­ap­pro­pri­ate in this case. Scott was ex­plic­itly con­sid­er­ing the policy of ex­pand­ing the defi­ni­tion of a word, not just which defi­ni­tion is bet­ter.

  If the (chief) pur­pose of ‘cat­e­gories’ (i.e. words) is to de­scribe re­al­ity, then we should only ever in­vent new words, not mod­ify ex­ist­ing ones. Chang­ing words seems like a strict loss of in­for­ma­tion.

  It also seems pretty ev­i­dent that there are ul­te­rior mo­tives (e.g. poli­ti­cal ones) be­hind overt or covert at­tempts to change the com­mon shared mean­ing of a word. It’s cer­tainly ap­pro­pri­ate to ob­ject to those mo­tives, and to ob­ject to the con­se­quences of the de­sired changes with re­spect to those mo­tives. One com­mon rea­son to make similar changes seems to be to ex­ploit the cur­rent valence or ‘mood’ of that word and use it against peo­ple that would be oth­er­wise im­mune based on the cur­rent mean­ing.

  Some cat­e­gory bound­aries should re­flect our psy­chol­ogy and the his­tory of our ideas in the lo­cal ‘cat­e­gory space’, and not be con­stantly re­vised to be bet­ter Bayesian cat­e­gories. For one, it doesn’t seem likely that Bayesian ra­tio­nal­ists will be de­cid­ing the op­ti­mal cat­e­gory bound­aries of words any­time soon.

  But if the word “ly­ing” is to ac­tu­ally mean some­thing rather than just be­ing a weapon, then the in­group and the out­group can’t both be right.

  This is con­fus­ing in the sense that it’s ob­vi­ously wrong but I sus­pect in­tended in a much more nar­row sense. It’s a demon­strated facts that peo­ple as­sign differ­ent mean­ings to the ‘same words’. Be­sides oth­er­wise un­re­lated homonyms, there’s no sin­gle unique global com­mu­nity of lan­guage users where ev­ery word means the same thing for all users. That doesn’t im­ply that words with mul­ti­ple mean­ings don’t “mean some­thing”.

  Given my cur­rent be­liefs about the psy­chol­ogy of de­cep­tion, I find my­self in­clined to reach for words like “mo­ti­vated”, “mis­lead­ing”, “dis­torted”, &c., and am more likely to frown at uses of “lie”, “fraud”, “scam”, &c. where in­tent is hard to es­tab­lish. But even while frown­ing in­ter­nally, I want to avoid tone-polic­ing peo­ple whose word-choice pro­ce­dures are cal­ibrated differ­ently from mine when I think I un­der­stand the struc­ture-in-the-world they’re try­ing to point to.

  You’re a filthy fuck­ing liar and you’ve twisted Scott Alexan­der’s words while know­ingly ig­nor­ing his larger point; and un­der cover of valu­ing ‘epistemic ra­tio­nal­ity’ while lev­er­ag­ing your priv­ileged com­mand of your cult’s cant.

  [The above is my satire against-against tone polic­ing. It’s not pos­si­ble to main­tain valuable com­mu­ni­ca­tion among a group of peo­ple with­out polic­ing tone. In par­tic­u­lar, LessWrong is great in part be­cause of it’s tone.]

• Kenny 8 Nov 2019 1:23 UTC

  2 points

  in reply to: Said Achmiz's comment on: Maybe Ly­ing Doesn’t Exist

  This is a bad ex­am­ple, be­cause whether some­thing is a crime is, in fact, fully de­ter­mined by whether “we” (in the sense of “we, as a so­ciety, ex­press­ing our will through leg­is­la­tion, etc.”) de­cide to la­bel it a ‘crime’.

  I think it’s still a good ex­am­ple, per­haps be­cause of what you pointed out. It seems pretty clear to me that there’s a some­times sig­nifi­cant differ­ence be­tween the le­gal and col­lo­quial mean­ings of ‘crime’ and even big­ger differ­ences for ‘crim­i­nal’.

  There are many le­gal ‘crimes’ that most peo­ple would not de­scribe as such and vice versa. “It’s a crime!” is in­evitably am­bigu­ous.

• Kenny 8 Nov 2019 1:18 UTC

  1 point

  in reply to: jessicata's comment on: Maybe Ly­ing Doesn’t Exist

  It’s im­por­tant to be very clear on what ac­tu­ally hap­pened (incl. about vi­o­la­tions), AND to avoid pun­ish­ing peo­ple. Truth and rec­on­cili­a­tion.

  I think this a very much un­der­rated av­enue to im­prove lots of things. I’m a lit­tle sad at the thought that nei­ther are likely with­out the loom­ing threat of pos­si­ble pun­ish­ment.

• Kenny 8 Nov 2019 1:15 UTC

  1 point

  in reply to: Raemon's comment on: Maybe Ly­ing Doesn’t Exist

  I think we, and oth­ers too, are already con­struct­ing rules, tho not at as a sin­gle grand tax­on­omy, com­pleted as a sin­gle grand pro­ject, but piece­meal, e.g. like com­mon law.

  There have been re­cent shifts in ideas about what counts as ‘epistem­i­cally neg­li­gent’ [and that’s a great phrase by the way!], at least among some groups of peo­ple with which I’m fa­mil­iar. I think the peo­ple of this site, and the greater di­as­pora, have much more stringent stan­dards to­day in this area.

• Kenny 4 Nov 2019 19:21 UTC

  1 point

  in reply to: Viliam's comment on: Link: An ex­er­cise: meta-ra­tio­nal phe­nom­ena | Meaningness

  I think that links to Chap­man’s texts should con­tain some dis­claimer that “ra­tio­nal­ity” as defined by Chap­man is some­thing com­pletely differ­ent from “ra­tio­nal­ity” as defined by Less Wrong.

  I am of many minds about this. Some­times I feel as you’ve ex­pressed; that Chap­man un­der­sells ‘ra­tio­nal­ity’ and mis­rep­re­sents its pos­si­bil­ities. Cer­tainly LW!ra­tio­nal­ity is (mostly) aware of his spe­cific crit­i­cisms. But I still find his writ­ing im­mensely in­sight­ful as-is. And given that his au­di­ence is very differ­ent than LW, I’m in­clined to ac­cept his writ­ing as-is too.

  As for him us­ing ‘ra­tio­nal­ity’ differ­ently – that gen­eral phe­nom­ena (of words be­ing used differ­ently by differ­ent peo­ple) is some­thing that I’m all too aware of, among all the things I read and all the con­ver­sa­tions I have. I cer­tainly don’t find his writ­ing as painful to read as oth­ers.

  And maybe we should add dis­claimers to all of our pages point­ing out that our use of ‘ra­tio­nal­ity’ is idiosyn­cratic (with re­spect to ev­ery­one else in the world). I don’t think there’s a good solu­tion to this.

  I agree that “The LW!ra­tio­nal­ity already con­tains its own meta.” but I think Chap­man has a point that meta-ra­tio­nal­ity is some­thing dis­tinct from (‘reg­u­lar’) ra­tio­nal­ity. Hence the util­ity of a lot of the ad­vice that both Chap­man and the LW se­quence writ­ers provide.

  Chap­man warns peo­ple against go­ing from straw ra­tio­nal­ism to nihilism (un­less they ac­cept the Bud­dhism-in­spired wis­dom). But I don’t see nihilism pro­moted on Less Wrong. We have “some­thing to pro­tect”. And the sto­ries of “beisut­sukai” are ob­vi­ously writ­ten to in­spire.

  Maybe that’s miss­ing from LW? I agree that LW doesn’t pro­mote nihilism, but maybe it should do more to help oth­er­wise-in­tel­li­gent peo­ple avoid it.

  And more gen­er­ally, (in­tel­li­gent) peo­ple re­ally do get stuck at “straw ra­tio­nal­ity” (“level 4″), i.e. ‘trapped’ in the spe­cific for­mal­isms of which they’re aware and in which they can ‘op­er­ate’. We don’t wor­ship sci­ence, but lots of other peo­ple sure seem to do so.

  I think the best ‘trick’ LW!ra­tio­nal­ity in­cor­po­rated into its ‘canon’ is the idea of in­stru­men­tal ra­tio­nal­ity. Cou­pled with a con­se­quen­tial­ism scoped to our ‘en­tire fu­ture light cone’, that idea alone acts like a source of in­tel­lec­tual free en­ergy ca­pa­ble of push­ing us out of any par­tic­u­lar for­mal­ism when (we sus­pect) it’s not good enough for our pur­poses. But it’s not clear, to me any­ways, that that it­self is ‘ra­tio­nal’. (It is LW!ra­tio­nal, ob­vi­ously.)

  Also, I’m not sure if “the Bud­dhism-in­spired wis­dom” was dis­mis­sive, but I re­ally en­joy his writ­ing about Bud­dhism (and it’s mostly pub­lished on other sites of his). From what I’ve read of that, he’s not a Bud­dhist – cer­tainly not a ‘tra­di­tional’ (or folk) Bud­dhist. He seems mostly in­ter­ested in very spe­cific schools, has his own idiosyn­cratic in­ter­pre­ta­tions, wants a bet­ter ‘mod­ern syn­the­sis’ draw­ing on his fa­vored in­sights, and is ac­tively ex­per­i­ment­ing with var­i­ous prac­tices for his own pur­poses. He definitely re­jects ‘woo’ (and his fa­vorite schools seem to be rel­a­tively light on that any­ways). But there’s a lot of in­sight available too. Just off the top of my head – the tantric Bud­dhist “prac­tice of views” char­nel ground and pure land. Tra­di­tional ra­tio­nal­ity, i.e. straw ra­tio­nal­ity, is pretty dis­mis­sive of emo­tions. LW!ra­tio­nal­ity is much bet­ter. Chap­man is min­ing pop­u­lar re­li­gion and philos­o­phy, in par­tic­u­lar the branches of Bud­dhism he likes, for in­ter­est­ing and some­times-use­ful info, of­ten per­tain­ing to emo­tions and what to do about them.

  So, iron­i­cally, from my per­spec­tive, it is like if straw ra­tio­nal­ity is level 4, and Chap­man’s “mean­ing­ness” is level 5, then Less Wrong would be level 6. (Yeah, I can play this game, too.)

  How se­ri­ously are you play­ing this game (ha)? Some­what se­ri­ously, we’re definitely around (or aiming for) his level 5. You’ve pointed out a lot of ‘meta-ra­tio­nal’ ad­vice from this site (and most that’s sev­eral years old now too). What would level 6 be, to you (be­sides 5 + 1)?

• Kenny 21 Oct 2019 17:01 UTC

  1 point

  in reply to: jkaufman's comment on: Ther­mal Mass Thermos

  My bad – I read that fol­low-up and was dis­ap­pointed in the last sen­tence:

  To be safe, though, I’m go­ing to keep us­ing the ther­mal mass ther­mos ap­proach.

  It doesn’t seem like the ther­mal mass ther­mos is strictly nec­es­sary “to be safe”, but I un­der­stand your ex­tra abun­dance of cau­tion.

Link: An ex­er­cise: meta-ra­tio­nal phe­nom­ena | Meaningness

• Kenny 18 Oct 2019 17:19 UTC

  3 points

  on: Ther­mal Mass Thermos

  I was dis­ap­pointed that Jeff didn’t con­clude, based on the de­tailed ev­i­dence he found, that it’s prob­a­bly fine to just pack the rice in a reg­u­lar un-mod­ified ther­mos.

  Or am I way off in sum­ma­riz­ing his ev­i­dence this way?