Open Thread 98.5

This is the twice-weekly hidden open thread. As the off-weekend thread, this is culture-war-free, so please try try to avoid overly controversial topics. You can also talk at the SSC subreddit or the SSC Discord server.

This entry was posted in Uncategorized and tagged . Bookmark the permalink.

212 Responses to Open Thread 98.5

  1. A1987dM says:

    No April Fools’ post this year?

  2. Gareth Allen says:

    Anyone interested in a study group for “Valuation: Measuring and Managing the Value of Companies?” It’s a dense, technical textbook and comes with a workbook with practice problems. I’ve gone through the first four chapters and so far it’s an excellent resource for running a business or investing.

    Valuation amazon link

    I’ve been neglecting it and I could use some peer pressure to help me keep going through it, hence the study group. Doesn’t need to be too formal, just need to follow a schedule and email back and forth questions, etc.

    If there’s no interest in a study group then I could hire a tutor to help me with it, $20/hr to a charity of your choice.

  3. lemur says:

    Hi 🙂
    Not sure if it’s ok to ask here.
    I’m looking for a room from April 16 at most 50min to Embarcadero by public transportation. My budget is 2300.
    I just moved from Stockholm to work for Google, love maths and reading, and I think it would be great to live with people that read this (you’re all probably super smart).
    Please let me know!

  4. Andrew Hunter says:

    Here’s an interesting mathematical statement that I wish is true. I have no idea if it holds in general or not, but I’d like it to be a theorem.

    Let f : R^n -> R be strictly monotonic on each input and invariant to permutations on its inputs (that is, f(x_1, x_2, …, x_n) = f(x_s(1), x_s(2), …, x_s(n)) for all s in S_n.) Then there exist strictly monotonic g, h : R->R such that

    f(x) = h(\sum_i g(x_i))

    Note this is definitely not a theorem without strict monotonicity: f = max is non-strictly monotonic in each input and invariant to permutations, but no such factorization exists for that function. but f(0, 1) = 1 = f(1,1) makes the goal impossible (WLOG take g(0) = 0. For each x > 0 we have x = f(x, 0) = h(g(x) + g(0)) = h(g(x)), but also x = f(x,x) = h(2g(x)); monotonicity of h implies it is constant on [g(x), 2g(x)]. But then we can uniquely specify h on positive inputs (in the range of g) by specifying its value at integer powers of two (or points in the image of g contained in power of two ranges.) But this is a countable number of values; since f takes all real values, h must as well; contradiction.)

    Anyway, you can see why this is a plausible theorem but difficult to prove. I haven’t been able to disprove it either, though i wouldn’t be stunned if some simple but clever counterexample exists.

    Why do I care: because this statement is loosely equivalent to “if you believe in the veil of ignorance, you believe in utilitarianism.” I think there’s something elegant about this if true.

    • entobat says:

      Counterexample: f = “sum of the x_i plus their max”. This breaks even in the case n = 2. Proof is by contradiction.

      As in your proof, WLOG g(0) = 0. Now

      2y = f(0, y) = h(g(0) + g(y)) = h(g(y)).
      2y = f(2y / 3, 2y / 3) = h(2 g(2y / 3)).
      2y = f( 4y / 9, 7y / 9) = h(g(4y / 9) + g(7y / 9)).

      By strict monotonicity of h,

      [1] g(y) = 2 g(2y / 3)


      [2] 2 g(2y / 3) = g(4y / 9) + g(7y / 9)

      If y = 2w / 3, then [1] reads

      g(2w / 3) = 2 g(4w / 9)

      Substituting this into [2] with y in place of w gives

      3 g(4y / 9) = g(7y / 9)

      Counterexample: f = “sum of the x_i plus their max”. This breaks even in the case n = 2. Proof is by contradiction.

      As in your proof, WLOG g(0) = 0. Now

      2y = f(0, y) = h(g(0) + g(y)) = h(g(y)).
      2y = f(2y / 3, 2y / 3) = h(2 g(2y / 3)).
      2y = f( 4y / 9, 7y / 9) = h(g(4y / 9) + g(7y / 9)).

      By strict monotonicity of h,

      [1] g(y) = 2 g(2y / 3)


      [2] 2 g(2y / 3) = g(4y / 9) + g(7y / 9)

      If y = 2w / 3, then [1] reads

      g(2w / 3) = 2 g(4w / 9)

      Substituting this into [2] with y in place of w gives

      3 g(4y / 9) = g(7y / 9)

      Recall that [1] was

      2 g(2y / 3) = g(y)

      So we have both that multiplying the input by 7/4 gets you a factor of 3 increase, and that multiplying the input by 3/2 gets you a factor of 2 increase. But

      (3/2)^3 = 3.375 > 3.0625 = (7/4)^2


      8 g(y) = 2^3 g(y) = g( (3/2)^3 y) = g(3.375 y)
      > g(3.0625 y) = g( (7/4)^2 y) = 3^2 g(y) = 9 g(y)

      i.e. g(y) < 0.

      Since g(0) = 0, g(y) > 0 for y > 0, contradiction.

      • Andrew Hunter says:

        I suspect your proof is repairable, but isn’t there an error? [2] should be:

        [2] 2 g(2 y/ 3) = g(4y/9) + g(7y/9)

        and then your logic derives

        3g(4y/9) = g(7y/9)

        which isn’t an immediate contradiction.

        • entobat says:

          D’oh. This is what I get for doing math late at night.

          This is repairable; I will edit my original post to fix.

          Edit: original comment is repaired. Thanks for pointing out the error.

    • smocc says:

      My first thought is that it is likely because of representation theory. You have a group action of S_n on the vector space R^n, and the way to get the invariant submodule is to sum over permutations, so the projection of (x1, x2,…) into an invariant submodule is (x1+x2+…, x1+x2+…, …).

      Here’s a sketch of half a proof:
      We are looking for functions f such that p.f(v) = f (p^{-1}v) = f(v). If we can find v0 that is invariant under all group actions then and can consistently map all v to G(v)\propto v0 so that G(v) = g(v)v0, then f(v) = h(G(v)) “=” h(g(v)) is a valid function.

      span(v0) is the invariant submodule of R^n, and can be found with the projection operator P = \sum_p p that sums over all permutations. P sends any vector (x_1,x_2,x_3,…) to (x1+x2+x3+…)v0 where v0 = (1,1,1,…).

      The valid maps G are P preceded by any bijection that commutes with all group operations. A slight extension of Schur’s lemma should tell us that these bijections must be diagonal, ie that Y((x1,x2,x3,…)) = (y(x1),y(x2),y(x3),…).

      Putting this all together, we have mostly implied that valid functions are of the form h(\sum_i g(x_i)). There are definitely some steps missing, but I suspect this is roughly what a proof will look like.

      • smocc says:

        One issue with this is I don’t know where monotonicity comes in, but you have already shown that it is required.

    • littskad says:

      I don’t think it’s true even if you make smoothness assumptions. I’ll consider two dimensions here, and put the h on the other side, so that we have g(x)+g(y) = h(f(x,y)).

      Differentiating with respect to both x and y, the left-hand side becomes 0, so we get:
      0 = h”(f(x,y)) f_x f_y + h'(f(x,y)) f_xy,
      where I’ve used f_x for the derivative of f with respect to x, etc.

      Rearranging, we get:
      f_xy / (f_x f_y) = – h”(f(x,y)) / h'(f(x,y)) = H(f(x,y)),
      where H = – h” / h’

      If you then differentiate this first with respect to x, then with respect to y, you’ll get two equations which you can solve separately for H'(f(x,y)) in terms of f and its derivatives, which you can set equal to each other yielding a third-order PDE that f has to satisfy:
      f_xxy f_x (f_y)^2 – f_xy f_xx (f_y)^2 – f_xyy (f_x)^2 f_y + f_xy (f_x)^2 f_yy = 0.

      • smocc says:

        But if f is fully symmetric under all its arguments then f_xxy = f_xyy, f_x = f_y, and f_xx = f_yy, which trivializes your last constraint

  5. roastingcanopus says:

    Do people who’ve taken the Giving What We Can pledge plan on giving much more than 10%? I save a lot more than 10% every year and don’t see what I could possibly spend that much money on (other than a family, I guess) and I’ll be forced to give it away after I die.

    • keranih says:

      I know you asked specifically of GWWC (I am not one) but I tithe a bit more than 10%.

      Of the rest, you should most probably be saving about 15% for retirement and another 5-10% for non-retirement items (house or auto, or college for kids, or some other item or set of items.)

      If this, plus your daily living expenses, are not consuming all your income, good job! I counsel you to avoid tempting Murphy – resist the temptation of increasing your living standard to match your excess income.

    • says:

      For me it depends on how much money I eventually end up having and how stable my job is. I donate 10% now, while I’m in grad school, and save some more for a rainy day. If and when I get tenure, and start making more money, I intend to donate more.

  6. Andrew Hunter says:

    A fun but not particularly hard math puzzle (If you do a lot of set theory, this shouldn’t be hard; if you do more analysis I wonder if you’ll get trapped): construct an explicit bijection between the reals and the irrational numbers.

    I have an answer, but I wonder how close mine will be to others’: Cvpx fbzr pbhagnoyl vasvavgr fhofrg bs gur veengvbanyf, fnl gur fdhner ebbgf bs cevzrf, ynoryrq l_v. Cvpx fbzr rahzrengvba bs engvbanyf. Znc engvbanyf gb gur bqq vaqrkrq l; znc gur l gurzfryirf gb gur rira vaqrkrq l; rirelguvat ryfr vf gur vqragvgl.

    • entobat says:

      Mine is the same as yours, except I used angheny-ahzore zhygvcyrf bs cv vafgrnq bs fdhner ebbgf bs cevzrf.

      • RavenclawPrefect says:

        Similar here; V frag gur agu engvbany gb cv cyhf gur 2agu engvbany naq cv cyhf gur agu engvbany gb cv cyhf gur 2a+1gu engvbany.

    • shakeddown says:

      engvbanyf znc gb vagrtref znc gb cv^2a+1. Cbjref bs cv znc gb gurve fdhnerf, naq rirelguvat ryfr zncf gb vgfrys.
      Okay that’s basically an example of yours.

    • Universal Set says:

      I think it says something about me that my first thought was “Whfg hfr n Unzry onfvf naq fuvsg gur onfvf ryrzragf” before deciding that no, this didn’t count as explicit.

      Perhaps on a related note, another puzzle. Prove or disprove the existence of:
      (a) an uncountable family of sets of integers such that the intersection of any two members of the family is finite.
      (b) an uncountable family of sets of integers on which the subset relation is a total order.

      • zqed says:

        Let me try these too.

        (n) Nffbpvngr gb rnpu ahzore va gur havg vagreiny gur frg bs vavgvny frtzragf bs vgf qrpvzny rkcnafvba. Svavgr qrpvzny rkcnafvbaf pbeerfcbaq gb vagrtref. Vs gjb ahzoref ner qvssrerag, gura gurve qrpvzny rkcnafvbaf jvyy qvfnterr sebz fbzr cbvag ba, fb nal gjb bs gur erfhygvat frgf unir svavgr vagrefrpgvba.

        (o) Nffbpvngr gb rnpu erny ahzore k gur frg bs engvbanyf ovttre guna k. Fvapr gur engvbanyf ner pbhagnoyr, jr pna vqragvsl gurfr engvbanyf jvgu vagrtref. Fb jr unir frgf bs vagrtref, Jung’f zber, vs gur frg bs k pbagnvaf gur frg bs l gura k < l, fb guvf vf n gbgny beqre.

      • RavenclawPrefect says:

        Seen (b) before, it’s a fun one. As for (a);

        Tvira n pbhagnoyl vasvavgr frdhrapr bs mrebf naq barf (bs juvpu gurer ner pbagvahhz-znal), gnxr lbhe fhofrg bs A gb pbagnva mreb vs gur svefg ovg vf mreb naq bar vs gur svefg ovg vf bar, gjb/guerr/sbhe/svir vs gur svefg gjb ovgf ner mreb-mreb/mreb-bar/bar-mreb/bar-bar erfcrpgviryl, naq fb ba. Nal gjb qvfgvapg fgevatf qvssre nsgre svavgryl znal cynprf, nsgre juvpu gurve pbeerfcbaqvat fhofrgf jvyy arire vagrefrpg.

    • zqed says:

      My job description says legal analyst, let’s see how I deal with set theory.

      Va gur sbyybjvat, yrg a,c,d qrabgr angheny ahzoref,

      Fgrc 1: fraq rirel erny bs gur sbez (fdeg gjb)^a c/d gb (fdeg gjb)^(a+1) c/d naq yrnir rirel bgure erny svkrq. Fgrc 2: trg evq bs 0 ol fraqvat rirel ahzore bs gur sbez (fdeg gjb) a gb (fdeg gjb) (a+1) naq yrnir rirelguvat ryfr svkrq.

      • Brad says:

        It’d be funny if there was a class in law school:

        Legal Analysis (Seminar, 3L only), perquisites: Math 330 & 331 (Analysis I & II)

    • Sbe rirel engvbany ahzore n/o, jurer o > 0 naq tpq(n, o) = 1, yrg s(n/o) = n + 2o fdeg 2.

      Sbe rirel cnve bs vagrtref n naq o fhpu gung o > 0 naq tpq(n, o) = 1, yrg s(n + o fdeg 2) = n + (2o – 1) fdeg 2.

      Sbe rirel bgure erny ahzore k, yrg s(k) = k.

    • James says:

      V tbg fghpx orpnhfr V gubhtug lbh jnagrq n znccvat orgjrra nyy bs gur engvbanyf naq nyy bs gur ernyf, juvpu—fcbvyre nyreg!—V frrz gb erpnyy vf vzcbffvoyr.

    • Tatterdemalion says:

      Pick a sequence 0= x_0, x_1, x_2, … such that each of the x_i is in a different class of RR/QQ – for example, the square roots of the primes.

      If x – x_i is rational, take x to x – x_i + x_{i+1}. Otherwise, map x to x.

  7. Shion Arita says:

    There’s a pretty cool project going on right now to replicate the result of AlphaZero chess using distributed computing:

    LeelaZero is a neural network that’s learning to play chess from scratch, and there’s a similar project going on with go as well. Leela currently plays chess at the level of a strong amateur, and is very interesting to play against right now since it is unlike traditional chess engines it plays very ‘human-like’. When traditional engines are restricted to play a t a lower level that would be fair competition for a human, their moves are very strange and the mistakes they make are not things a human would. Playing against Leela feels a lot more natural and is very interesting.

    The project needs a lot of computing power to go, so if you’re interested, contributing your CPU or GPU time would be a great help.

  8. a reader says:

    I tried to make a beginnings quiz, although I couldn’t think of 10 texts known enough, only 7, even after including some well known texts that are not literature (at least, not intended as literature).

    What book/story begins with:

    1. “In the beginning was the Word”

    2. “A spectre is haunting Europe”

    3. “It is a truth universally acknowledged, that a single man in possession of a good fortune, must be in want of a wife.”

    4 . “It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness, it was the epoch of belief, it was the epoch of incredulity, it was the season of Light, it was the season of Darkness, it was the spring of hope, it was the winter of despair”

    5. “[girl name], light of my life, fire of my loins. My sin, my soul.”

    6. “During the whole of a dull, dark, and soundless day in the autumn of the year, when the clouds hung oppressively low in the heavens, I had been passing alone, on horseback, through a singularly dreary tract of country; and at length found myself, as the shades of the evening drew on, within view of the melancholy House of [name].”

    7. “All happy families are alike; each unhappy family is unhappy in its own way.”

    • johan_larson says:

      1. Gur Tbfcry bs Wbua
      2. ??
      3. V gubhtug vg jnf Frafr naq Frafvovyvgl, ohg vg gheaf bhg gb or Cevqr naq Cerwhqvpr.
      4. N Gnyr bs Gjb Pvgvrf
      5. Ybyvgn
      6. ??
      7. Naan Xneravan


    • Brad says:

      1) Tbfcry bs Wbua
      2) Veba Phegnva fcrrpu, Jvafgba Puhepuvyy.
      3) Wnar Nhfgva’f Cevqr naq Cerwhqvpr
      4) N Gnyr bs Gjb Pvgvrf, Qvpxraf
      5) No idea.
      6) Vasreab, Qnagr (ybj pbasvqrapr)
      7) N Ehffvna abiry, ohg V qba’g erzrzore juvpu bar. Naan xneravan znlor?


      • Orpheus says:

        Actually 2 is from gur pbzzhavfg znavsrfgb.

      • Deiseach says:

        Ah, I see why you think that’s Number Six, but the actual opening of that is:

        Ary zrmmb qry pnzzva qv abfgen ivgn
        zv evgebinv cre han fryin bfphen
        pué yn qvevggn ivn ren fzneevgn.

        Zvqjnl va gur wbhearl bs bhe yvsr
        V pnzr gb zlfrys va n qnex jbbq,
        sbe gur fgenvtug jnl jnf ybfg.

    • Deiseach says:

      I know these! 🙂

      1. Tbfcry bs Wbua

      2. Gur Pbzzhavfg Znavsrfgb

      3. Cevqr naq Cerwhqvpr

      4 . N Gnyr bs Gjb Pvgvrf

      5. Ybyvgn

      6. Gur Snyy bs gur Ubhfr bs Hfure

      7. Naan Xneravan

    • hls2003 says:

      These hit me in the right place.

      1. Tbfcry bs Wbua
      2. Gur Pbzzhavfg Znavsrfgb
      3. Cevqr naq Cerwhqvpr
      4. N Gnyr bs Gjb Pvgvrf
      5. Ybyvgn
      6. Snyy bs gur Ubhfr bs Hfure
      7. Naan Xneravan

    • Iain says:

      I recognized everything except 6, and could deduce the answer based on the clue. I was more confident in the author than the book for #7 (although my best guess was correct), and I brainfarted and came up with the other famous book by the same author (Qnf Xncvgny) for #2.

      6/7, but should have been 7/7.

    • The Nybbler says:

      The style gives 6 away. Number five was an easy guess. I made the same mistake as Brad with 2. Number 7 gets mentioned here a lot. I knew what number 1 wasn’t, but not what it was. I got 4,5,6, and 7.

    • j1000000 says:

      6/7. But not only did I not know #6, I’ve never even heard of it.

    • Nornagest says:

      7/7. And here’s three more famous opening lines to fill your list out:

      8. “It was a bright cold day in April, and the clocks were striking thirteen.”

      9. “We were somewhere around Barstow on the edge of the desert when the drugs began to take hold.”

      10. “‘There was a boy called Eustace Clarence Scrubb, and he almost deserved it.”

      An easy bonus: “The man in black fled across the desert, and the gunslinger followed.”

      • hls2003 says:

        8. Avargrra Rvtugl-Sbhe
        9. Srne naq Ybnguvat va Ynf Irtnf
        10. Iblntr bs gur Qnja Gernqre

        Gur Thafyvatre (or possibly Qenjvat bs gur Guerr? I’m pretty sure the former).

      • AG says:

        “It was a dark and stormy night.”

        • Nornagest says:

          Ohyjre-Ylggba originally, used many places since then. But it’s probably best known from N Jevaxyr va Gvzr these days.

    • 1. Gur Ovoyr
      2. Gur Pbzzhavfg Znavsrfgb
      3. Cevqr naq Cerwhqvpr
      4. N Gnyr bs Gjb Pvgvrf
      5. Ybyvgn
      6. ?
      7. Naan Xneravan

      (V tbg 1 jebat, vg’f npghnyyl gur Tbfcry bs Wbua engure guna Trarfvf)

      • Brad says:

        V unir gb vzntvar vg jnf n qryvorengr vzvgngvba bs trarfvf, ohg gurer vg’f “Va gur ortvaavat Tbq perngrq gur urniraf naq gur rnegu.”

    • fion says:

      I’m assuming 1. is gur ovoyr.

      2. is very easy. Probably the second-most quoted sentence of that book after the last one.
      “Jbexref bs gur jbeyq havgr! Lbh unir abguvat gb ybfr ohg lbhe punvaf!”

      4. sounds so familiar, but I can’t put my finger on it!

      7. also sounds familiar. I feel as though I’ve not read it, but I have heard it quoted.

      • fion says:

        Uhu. 1. srryf yvxr n purng. :C Univat fnvq gung V frrz gb or va n zvabevgl sbe snyyvat sbe vg, fb…

        Bqqyl, nsgre purpxvat gurz, gur barf gung fbhaqrq snzvyvne ner npghnyyl abg snzvyvne gb zr ng nyy. Va snpg V guvax V bayl urneq 4. va n cerivbhf dhvm!

    • James says:

      7/7, I think. Thanks for finally providing one I could ace.

    • A1987dM says:

      1. is both what most people said and Gur Anzr bs gur Ebfr, if I remember correctly.

  9. Levantine says:

    I’m interested in your perception of trends in competencies in the US:
    for any time length you feel able to judge about,
    on any scope of activities – narrower or wider, professional or not – you feel like speaking about.

    I’m a European and I’ve heard some contradictory claims related to this.

  10. Universal Set says:

    Last fall, I posted some questions about getting into software from (math) academia, but saying that I didn’t think I was going to leave academia right away. A few months ago, I posted an update that I had in fact decided to leave academia for software after the spring semester, but that I didn’t want to move to NYC or the Bay.

    Continuing the pattern of my doing something I said I wasn’t planning to do, I’d like to report that I have in fact accepted a job offer to start as a software engineer this summer… at a medium-sized tech company in San Francisco. It’s a great opportunity that practically fell into my lap — or at least it feels that way to me. I’m excited about the job; the team I’ll be joining is working on some very interesting and difficult problems.

    If anyone has insights on moving to SF, I’m all ears. My wife would like to live in Sunset (it’s just the two of us right now, though we hope to start having kids in a year or so) within easy walking distance of the Judah St. Muni line (to make the commute to work not a hassle). I’m not sure how feasible this will be, though there did seem to be apartments available there when I checked last.

    • johan_larson says:

      I’ve been hearing complaints about the Bay area housing situation for a long time, and San Francisco seems to be the epicenter of the insanity. Think carefully before deciding to move there, particularly if you intend to raise a family. Commuting in from the east bay seems to be popular. People bitch about BART a lot, but I had no complaints about it when I lived there. I’ve heard bad things about Muni, though: asshole drivers, crammed buses, and real weirdos among the passengers.

    • skef says:

      A decade ago the N Judah was, by a significant margin, the least reliable of the Muni lines. It’s very possible that the longstanding kinks are worked out now, but you might look into that before committing to a location.

      And note that Muni buses are a garbage fire. Go to a stop on a line that runs “every fifteen minutes” and you’re likely to wait an hour before three buses pull up together.

  11. Egregious Philbin says:

    Hey all:

    Has anyone here taught English in a foreign country (commonly referred to as TEFL)? I’m curious about your experience.

    I know Scott himself has, given he recommended it in a 2013 post titled “Floor Employment“.

    About to graduate this semester with a BA – not in English – and I am, among other things, looking to travel, spice my CV before medical school, and meet foreign girls (had to). I’m unsure about the “you need a (however dubious) TEFL online certification which costs money and is totally complicit with credentialist norms” thing however, and would therefore prefer a way in straight after college. Any leads?

    • theredsheep says:

      It varies widely depending on country. I did ten months in Peru. There are varying levels of “English teaching” there, ranging from the really scuzzy up through the immensely prestigious. A lot of it gets done by rootless twentysomething backpackers just passing through and looking for a few bucks to pad their wallets before they move on. I did a monthlong course at Maximo Nivel in Cusco; I don’t know how necessary that really was, given the (to put it mildly) informality of the TEFL situation there.

      Unless you get in with the international schools there, or somewhere equally lah-di-dah, your employment will be technically illegal; foreign workers require a workers’ visa, which takes six months and sponsorship by a Peruvian company, etc. Not worth it for a fly-by-night mochilero stopgap. Go in on a tourist visa, spend your six months, go to the border and spend a day on the other side, come back in for another six months, repeat indefinitely. It’s illegal, and you have to commit a form of fraud to get paid, but nobody gives the smallest fraction of a damn. Peru’s two most recent presidents left office amidst allegations of receiving bribes from a Brazilian mining consortium. Same consortium for both. They’re fighting hard to end corruption, but it’s endemic and nobody begrudges an otherwise law-abiding gringo slipping in to do useful work.

      Most of my students were engineers and accountants at a mining company in Lima. I had a group of four or five. Lessons consisted of slavishly running them through a textbook, because I worked for one of the lower-tier English options. I doubt the lady who ran it cared about my Maximo Nivel certificate, except insofar as it reflected that I was serious about it and probably wouldn’t scarper without notice.

      Now, I eventually quit because I didn’t have the social skills to make it work; teaching contracts tend to evaporate on short notice, and to make real income you have to have several gigs going at once. I simply wasn’t up to the constant hustling for new contracts with new “institutes.” Plus I had a new baby, born in Lima, and my wife and I felt overwhelmed. However, if I could go back on a more regular financial basis I wouldn’t hesitate for a second. Lima was a wonderful place, for all its legal goofiness, with friendly people and great food. They have the best rotisserie chicken I’ve ever tasted.

      Again, the situation in other countries will be quite different. I did my ten months in 2011-2012, so the situation may have changed, but last I heard the biggest demand was in the Middle East, where they pay top dollar. Korea is also a popular option, where housing is commonly provided, but they’re more choosy than the laid-back Peruvian scene. There are online hangouts where you can talk these things over–I want to say Dave’s ESL Cafe, but I might be misremembering–and they’re usually pretty good for on-the-ground info.

      • Egregious Philbin says:

        Massively helpful. Thanks! Dave’s ESL cafe is my next stop.

        Oh man does the thought of being a fraudster in Peru make me nauseous. That gig of yours – teaching a group of miners (I’m looking for minors!) – seems so sketch; a good story in any case, I’m sure.

    • Aapje says:

      A teacher in Japan wrote a great blog that is still up. Although your experience may be a little different if you are not black.

      • Egregious Philbin says:


        Impressive you remember those posts from ’04.

        • quaelegit says:

          Oh if older stuff is okay, there’s a lot more information about Scott’s experiences on his old blog (my quick google search didn’t bring up any super relevant posts, but I know he had at least one that went into detail about requirements for different programs in several Asian countries. This will also be a decade old though.)

          And for pure entertainment similar to Aapje’s second link, look up the webcomic “Let’s Speak English” by Mary Cagle.

        • Aapje says:

          @Egregious Philbin

          I don’t expect I’ll ever forget reading about Japanese dick ninjas and butt assassins.


          She also draw a comic about kancho!

    • cassander says:

      I did something similar after graduating. I got my CELTA certification, largely because they said they helped place you with a job when you finished. They help they provided in that regard was pretty limited, but it was something. In terms of whether or not the education actually helped, eh, probably, especially if you’re nervous about being in front of the classroom, but I don’t think the certification was all that valuable if you’re a native english speaker. The demand for teachers there is the highest, especially (I’m told) if you’re white, so it’s the best place to go if you care about how much you make or don’t want to spend the time or money to get certified.

      I did my time in Latin America (Mexico, to be specific) and very much enjoyed it. Didn’t make very much money, but living costs were quite low, so I was quite comfortable as long as I remembered not to convert my paydays into dollars. A lot of people in my certification class wanted to go to europe, but this is apparently difficult. A lot of demand, but a lot of good non-native speakers to compete with and getting work visas can be tricky in some places.

      Do you have any specific questions?

    • b_jonas says:

      I thought had something about teaching English abroad, but even if you count the comment thread there are testimonies from two people only, so that’s a bust.

  12. A Definite Beta Guy says:

    A few actuaries lurk here, correct? How would you describe the job market at the moment? I’ve read it is quite saturated.

    I’m interested in making a mid-career jump, since my current career plans are….not progressing as well as I’d like. My math skills are rusty, but were reasonable when I last took anything academic 10+ years ago (basic linear regression in econometric, basic AP stats, your average college calc class and average college probability class….all A grades with all the minimal effort you’d expect of an 18 year old slacker). So it seems possible to study enough to pass the first 2 exams and then apply for positions, if that’s an available option.

    My background is accounting, around 7 years experience, and I’d come with recommendations from current managers, if that helps at all.

    Market is the Chicagoland area.

    Any actuaries here able to chip in with their thoughts?

    • SamChevre says:

      I’m an actuary.

      The job market is definitely tighter than it was a decade ago: that said, it’s still a good career, and if you are a fluent English speaker with two exams you have a very good chance of being hired. Seven years of accounting, though, might get you in with only one exam: we are generally short on accounting expertise.

      The one warning: the exams are brutal beyond sane imagining. I took my first exam in 2002, and got my FSA last month. (Yes, yes, I got distracted by work and everything else, but still: our oldest is 11, and I was already taking exams when I met my wife the first time.)

      • ShemTealeaf says:

        I’m also considering a career change, and I’m curious if you think the published salary estimates are accurate. The DW Simpson charts claim that a P&C actuary in my area (NY/NJ/CT) could expect to be making six figures within a couple years, and as much as $300k by the time they’re 2-3 decades in. For a job that is reputed to be fairly low stress and doesn’t require graduate school, that seems almost unrealistically high.

        • SamChevre says:

          I cannot say with good certainty, as I do not know what my colleagues make, and do not have a good sense for geographic differentials. That said, I think that the D W Simpson surveys are accurate, but optimistic, and a lot of people read them optimistically.

          Read them this way for maximum “in Sam’s experience” realism: figure you’ll pass one exam per year, and will make what Simpson shows as 25th percentile in the last year (so, for 3-5 years experience, you’ll get that in year 5), as a company actuarial student with typical responsibilities (you are getting part of your compensation in training). If you want more than that, you’ll be trading off for either a heavier workload (consulting)–which makes passing exams harder, a higher-stress workload, or once you have some experience, a management role. The higher salaries, in the 10 year and up categories, are for managers–and managing is a significantly different skill than the technical basics. Getting really good at either managing or actuarial work is a multi-year process.

          Put it this way: you can fairly reliably get 6 figures with either an FSA or a decade of experience. I would not bet on twice that with both.

          • ShemTealeaf says:

            That’s very interesting; thanks for the insight. If you don’t mind me asking, can you give me a sense of the geographic area you’re basing this off? I’m coming from a very high cost of living area (suburbs of NYC), so that seems lower than I’d expect.

          • SamChevre says:

            I started in Virginia market outside the DC suburbs, and I am in the Hartford market now.

            This username, at google’s mail, will reach me if you want to talk off-line.

          • ShemTealeaf says:

            Thank you; that definitely helps put the information in perspective.

  13. johan_larson says:

    Here is an excellent presentation about hominin fossils from the Rising Star cave in South Africa. The presentation is recent, from 2017, by an anthropology professor at with University of Wisconsin-Madison.

  14. Andrew Hunter says:

    Hey – any commenters here current or former employees of Jane Street Capital (one of our site advertisers)?

    I’d like to pick your brain(s) quickly; if we are in the same city I’ll buy it with coffee or cocktails, and if not with promises of same next time I’m where you are or donations to charity.

    Contact me:

  15. Well... says:

    Between dropping my kids off at their schools and driving to work I spend about an hour a day in my car, so to make the most of it I’m listening to the Great Courses lecture series on Islam by John Esposito. I’m 3 CDs in right now.

    If anyone here has listened to this series, would you like to discuss/share your thoughts on it?

    Some of the reviews of the course were helpful, in case nobody can answer but others are curious.

    • Well... says:

      I’ll kick it off and say it was interesting to hear Esposito describe Islam as being primarily concerned with social justice, and highlight the ways in which it was groundbreakingly progressive for its time in the ways in which it treated women. I’m agnostic about whether this is true though I err on the side of trusting Esposito the Islam expert; I was more interested in whether there are parallels between Islam and the SJ movement we have today, because a few started popping out in my mind.

      • John Schilling says:

        Progressive for its time, yes. I’ve often described Islam as a religion that has been stuck in the fourteenth century for the past fourteen centuries, and that’s not exactly accurate but it is fairly close. There was no sense of, e.g., women being equal to men, but a clear dissatisfaction with the kind and degree of inequality encoded in the laws and customs of 7th-century Arabia and so the new rules were a big step forward.

        The obvious parallel with contemporary Social Justice is the inflexibility with which the One True Path to reform was implemented. What else did you have in mind?

    • fraza077 says:

      I found it useful because it was very obviously pro-Islam, and have a pre-existing bias against it. I did feel at parts it was a little dishonest, glossing over the brutality and problematic scriptures, and rationalising in a way that echoed my frustrating experiences growing up with Christianity.

      I listened to it twice, because I found I retained less information than other things I listen to.

      Overall it’s a fascinating topic, and I would enjoy learning more about it, but as far as my attitudes towards Islam go, it’s pretty much the same “Some limited utility for society, but overall not factual and no longer suited to modern society, overall a social ill”.

  16. proyas says:

    I wonder if someone with knowledge of optics, telescopy and astronomy could answer some questions for me about how what the limits of telescope resolution are (theoretical and practical), and how those limits will affect our ability to see Earth-like planets in other star systems.

    For starters, are these estimates about gravitational microlensing accurate?

    ‘Imagine putting a telescope, radio or optical, onto a rocket and sending it to the Sun’s gravitational focus—roughly twenty times the distance of Pluto. When aimed back at the Sun, the telescope’s sensitivity will be increased by thousands or millions of times, depending on wavelength. Such an instrument would be capable of detecting even low-power signals (far weaker than your local top-forty FM station) from a thousand light-years. At the wavelengths of visible light, this setup would be able to find the street lighting of New York or Tokyo from a similar remove.

    Consequently, it’s indisputable that any extraterrestrials with the hardware necessary to engage in interstellar warfare will have the capability to heft telescopes to the comparatively piddling distance of their home star’s gravitational focus.

    The conclusion is simple: It’s too late to worry about alerting the aliens to our presence. That information is already en route at the speed of light, and alien societies only slightly more accomplished than our own will easily notice it. By the twenty-third century, these alerts to our existence will have washed across a million star systems.’

    • bean says:

      Even assuming gravitational lensing works as well as described, it’s incredibly poorly suited for volume search. You essentially only get what’s directly behind the sun, so to search, you have to steer the telescope through the entire sphere around the sun. Which is a huge task, and makes this a much better way of looking at things you already know are interesting than it is for looking for things in the first place.

      • Andrew Hunter says:

        wikipedia estimates the density of stars in our ‘hood as 0.004stars/ly^3. So in a 500 ly sphere around us there are ~200K stars; we can probably eliminate many of these as poor habitats for life-like-us, but let’s ignore that.

        If I have 500 research vessels with jump drive–quite a few but by no means unreasonable, and I spend a week imaging every star around me, I’m done in less than ten years.

        Now given that we’re talking about a civilization that’s presumably bordering on Kardashev I–certainly has the ability to move that far efficiently–that doesn’t strike me as unreasonable to have going as a regular thing, unless jump drive just doesn’t work on short ranges or is expensive **for a Type I civilization** (in which case, presumably they can’t make interstellar war cost-effectively against defenseless Earthlings with no useful resources.)

        I have no comments as to whether or not lensing works as well as described.

        • bean says:

          The OP says “rocket”, so I assumed rocketry was the means of moving the telescope around. Jump drive obviously changes the equation, and makes it something that isn’t obviously unreasonable.

          • Andrew Hunter says:

            I think the author is being imprecise; a rocket is how we would set up a lensed scope, but since we’re explicitly in the argument talking about a civilization with interstellar warfighting, we actually mean a small starship propelled however they do that (jump drive, Weber impellers, reactionless drives, whatever…) with the implicit claim being that if they can afford to send us their Great White Fleet, they can afford to routinely telescope us with this technique.

          • bean says:

            That’s not the impression I get from the OP. Even if we waive the specific mention of rockets, there’s also “That information is already en route at the speed of light, and alien societies only slightly more accomplished than our own will easily notice it.” Leaving aside a Road Not Taken-style breakthrough in physics, someone with a jump drive is not “only slightly more accomplished”. I agree that the bar for interstellar warfare is high, but I’m not sure that whoever wrote that realizes how high.

      • Nornagest says:

        You could set it in a polar orbit at the required distance and get decent coverage… if you’re willing to wait a few hundred thousand years.

    • John Schilling says:

      Shostak may be overestimating the resolution achievable by gravitational microlensing; Landis suggests that resolution will likely be limited to no more than ~10×10 pixels per planet due to focal blur. But the broader point that Earth’s technological civilization will be known to any advanced civilization within the light-cone of WWII at the latest is almost certainly correct.

      • Andrew Hunter says:

        I am eagerly waiting for the first time I find myself at a cocktail party with a radio astronomer (yes, really) because I have a question about sensitivity I’ve wondered about for years that I can’t find a straight answer for on the internet:

        Suppose there is an exact duplicate of Earth N light years away (and displaced N years backwards in time, so we’re just now receiving signals from their 2018 emissions.) For which values of N are you confident we’d know they exist (and could clearly recognize them as technologically produced?)

        It’s not clear to me how far out we go before the power of earth’s signals is indistinguishable from background noise. Popular press sometimes bandies the number 100ly around, but I have no evidence for or against that number or any clear source for the computation. (It’s also difficult to define “indistinguishable” and “background noise”–like, we can look at the power of terrestrial radio output, the power of the CMB, and give an inverse square, but I’m sure the spectrum are very different–can we notice that? Etc.)

        • John Schilling says:

          Do you want to know how far away contemporary Earth could be detected by a civilization with the same technology and the same (lack of) interest in SETI as contemporary Earth, or are you interested in the distance at which Earth could be detected by an advanced alien civilization? Generally, you’ll get answers to the first question by default, and often either blank stares or active who-is-this-UFO-nut disinterest if you insist on the second, but there are exceptions.

          This paper, from 1979, suggests that 1979-Earth could detect a copy of 1979-Earth (by having the Arecibo dish listen for signals from BMEWS radars) at a distance of 18 light-years, but that a SETI project proposed in 1971 (and cancelled by penny-pinching bureaucrats) could extend that to 250 light-years. More advanced technology could presumably increase the detection range, but we haven’t been sending out radio signals long enough for that to matter.

          • Andrew Hunter says:

            I was thinking the first, or at least the first + some interest in widefield searches affordable with Earth tech, not assuming any higher technology than what we actually have.

            A lot of discussion of this tends to be directional searches for beacons, which requires both ends to get lucky about who looks (and talks) when; what I was much more interested in was widefield detection of incidental widefield emission.

  17. proyas says:

    To what extent does America’s higher per capita electricity consumption relative to other rich countries owe to America using the 120V standard while almost all its peers use 220 or 230V as standard? More electricity is wasted during transmission at lower voltages.

    • CatCube says:

      I’d be surprised if this was a big effect. We don’t transmit at low voltages; the lines from the substation to your house are (IIRC) 13.5 kV, and it gets stepped down to 120/240 V on the transformer at the pole outside your house. So it’s really not going long distances at low voltages.

      Now, I don’t know what the distribution voltages in other countries are, and they could be using a higher distribution voltage to near your house. But that’s not really dependent upon the final consumer voltage, as a different transformer could make most any voltage pair work.

      Plus, something like half of all electricity consumption in the US is electric motors. Obviously, some of that is going to be refrigerators and the like, but I recall that it’s mostly commercial entities like factories, and commercial power distribution is usually something like 13 or 69 kV directly to the customer, and they’ll step it down to 120/240 in their own plant for offices. (Even lighting in factories is typically 240 or 480)

    • The Nybbler says:

      Not significant. Transmission and most distribution voltages are independent of the choice of end-user voltage. Only the lines downstream of the final service transformer are affected.

    • Eric Rall says:

      I suspect it’s mostly driven by electric heating. Comparing data for rich European countries (couldn’t find equivalent data for the US), the countries with higher-than-US per capita electricity usage are cold countries that get a substantial portion of their heat demand (space heating, central heating, and water heating) from electric heat. Finland looks like about 25%, Sweden about 30%, and Norway (the extreme outlier in per capital electricity usage) about 60%. By comparison, rich countries I checked with conspicuously lower than the US in per capital electricity usage use little or no electricity for heating: UK 8%, Germany looks like about 5%, and not enough to appear in the graph for France and the Netherlands.

      My experience is consistent with the US being in the Finland/Sweden range for electric heating: most single-family homes seem to use natural gas or fuel oil, but electric radiators, heat pumps, and electric water heaters are far from unheard of, especially in older houses and in multi-unit buildings.


      • The Nybbler says:

        Residential US electricity use is here. 16% of US electricity is used for heating (plus another 2% to run the blowers/pumps, which should probably count). But another 15% for cooling, which is probably a big difference; I believe residential cooling is much more common in the US than Europe.

        If you are looking for the reverse figure (percent of heating that’s electric), that I don’t have.

        • Eric Rall says:

          Yes, the latter is what I was looking for, since that’s the format I found for the other countries, but the data you found does seem to confirm that heating and cooling are a big part of the picture.

          Discounting the US per capita usage figure by 31% (not quite the right amount, since your figure is for residential usage and the wikipedia usage figures are overall) puts US usage a bit higher than France and Germany, but no longer the huge outlier we are among non-Scandinavian countries.

          Adding air conditioning to the picture explains the part I was wondering about: the Scandinavian countries are quite a bit colder than most of the US, so I wouldn’t expect us to use nearly the amount of electricity they do for heating even if electricity were our primary source of heat. But I’m given to understand that air conditioning is much more common in the US (especially the South) than in Europe.

          • Iain says:

            Indeed, Paul Krugman has pointed out that the growth in population in the South coincides neatly with the spread of air conditioning, and Lee Kuan Yew credited it as a key factor in Singapore’s success.

        • A1987dM says:

          Yeah, comparisons of energy consumption across countries with dissimilar climate are rather nontrivial to interpret.

  18. Nabil ad Dajjal says:

    Since the pros and cons of different voting systems come up fairly often, how do people feel about sortition as a means of selecting representatives?

    It seems like it should deal with a lot of the problems that a lot of people have about our representative democracy. You can’t have an entrenched political class without career politicians. Women and minorities are guaranteed proportional representation, as are the poor. Voter participation is irrelevant because you only need to poll a few hundred citizens at a time.

    Of course there are some obvious downsides too. If the pay and prestige are too low it will become a chore to be avoided, like jury duty is now. Randomly selected representatives won’t have any experience in government and will, all jokes aside, probably be much dumber on average than congresspeople.

    Maybe a good compromise position would be to go back to the old idea of a higher and lower house system. Pair a Senate consisting of one hundred telegenic, highly-vetted robots with a House of Representatives consisting of three hundred odd random people with valid mailing addresses.

    • Brad says:

      The problem, at least in the US and I think elsewhere, is that legislatures are too weak as compared to the executive and the bureaucracy. The founders got that one wrong. Something like this is bound to weaken it further as you get a incompetents regularly cycled through. I oppose term limits for similar reasons.

      • albatross11 says:

        This makes sense. It’s also an argument for why term limits for the executive branch make sense–the executive branch probably has too much power, and knowing that you and your cronies will have to leave office eventually probably decreases some of the temptation for abuse.

      • Garrett says:

        legislatures are too weak

        In the US, I have to disagree. By design they are the most powerful branch of the government. However, they have chosen to delegate much of their power via rule-making to the executive branch. Congress, with a single act (plus certain veto override), could dissolve virtually all of the Federal law enforcement apparatus, spending, and military. It’s not difficult, it’s merely unpopular, and would likely limit future career options.

        • DavidS says:

          In the same sense the UK Parliament is far stronger than the Government. In the UK the issue is the Government controls e.g. how Parliamentary time is used in a way that means Parliament can’t initiate much (short of bringing down a Government). But when people say Government is too strong and Parliament too weak, this usually depends on Parliamentarians choosing to support the Government. The greatest power of the British PM is his/her command of a Parliamentary majority.

        • beleester says:

          In a democracy, “difficult” and “unpopular” are one and the same. Getting 60 votes to override a filibuster isn’t trivial, let alone the 67 you need to override a veto.

          You’re right that, rules-as-written, Congress is more powerful than the President, but it’s also designed to act slowly, and to default to inaction when it’s uncertain.

          • Matt M says:

            With 2-year terms, shouldn’t the House be the one part of government most incentivized to move quickly and decisively?

          • Nornagest says:

            The opposite, if anything. 2-year terms mean your next election probably hinges on one big decision, if it’s contested at all. That incentivizes conservatism, not in the sense of being right-wing but in the sense of covering your ass.

          • BBA says:

            In a democracy, “difficult” and “unpopular” are one and the same.

            The most effective legislative leaders their respective parties have had in decades are Mitch McConnell and Nancy Pelosi.

            Not coincidentally, they’re also the political figures who inspire the most bipartisan hatred.

    • albatross11 says:

      Imagine adding a level of indirection here–we randomly select a set of people who will vote for people to send to the legislature. (If we assume a parliamentary system, this will give us two levels of indirection to select the executive.) I think this system would closely approximate what we get in normal elections where there is mandatory voting–it’s just a random poll. One important difference is that if there are only a few thousand voters who know their vote counts a lot, they may take the whole matter more seriously.

      My intuition is that just randomly selecting people for the legislature would get a lot of people who either weren’t interested enough in the mechanics of governing, or weren’t smart enough, or weren’t in a position in their life to do a good job (say, the mother of a couple toddlers or someone caring for a parent with Alzheimers). That would probably make most of them them dependent on their professional staffs, who would be making most of the actual decisions. (That’s probably largely true anyway.) But I don’t know for sure how we’d test that intuition.

      If my intuition is right, maybe we’d benefit from having an intermediate step–the candidates have to name themselves willing to take the job, and you maybe have some kind of minimal qualifications–something like 25-75, in reasonable health, no criminal record or involuntary commitments, etc.

      • DavidS says:

        Not sure it helps to replace voting with something which disenfranchises those with caring responsibilities or the ill: feels like it would mean the needs of those sort of people would be devalued.

        I can see the case for giving e.g. one in a thousand people the vote so those people think more. But I’m not convinced the main benefit of elections is that people thinking about who to vote for leads to good legislators which leads to good laws.

      • rlms says:

        The converse — randomly select a set of people who are eligible to run and allow the whole population to vote for those who choose to do so — would also be interesting.

    • Protagoras says:

      Classical Athens, of course, did this for a lot of positions. They did have, as albatross suggests, a vetting system to exclude some particularly undesirable candidates from the lottery; I don’t know exactly how that worked in Athens, but I would be concerned both that the system would produce too many incompetents without that and that in the modern world the process would be too likely to become corrupted. Still, I’d like to see some experiments along these lines to see if there is any way to make it work (maybe it would work better for some government positions than others).

      • DavidS says:

        Is your implication that it wasn’t corrupted in the ancient world?

        What sort of government positions was it used for (or do you think it could be)? Seems instinctively to me like having a lottery for ‘Minister in charge of X’ is fair crazier than for a legislature, as a Minister is one person so their own views wouldn’t be balanced out by others.

        • Protagoras says:

          In Athens, it was used for almost all government positions except the council of generals (who were elected and were effectively the supreme authority in Athens due to some combination of the legitimacy that derived from their popular support and the extraordinary powers they needed to cope with military emergencies). The Athenian system seems to have had many problems, and is also rather difficult to understand at this remove; I would be surprised if corruption in the vetting for eligibility for the lotteries was never a problem, but I don’t know of any specific ancient sources discussing it. As I understand it, the vetting was done by the leaders of the various demes (local government units), and procedures may very well have varied from deme to deme. Much that we would like to know has not been preserved over the intervening more than two thousand years, and the details of local government is not one of the areas people have been motivated to preserve records through the centuries.

          I guess I’d be most interested in seeing it used for lower level positions where the elections suffer from a lack of voter interest anyway (the kind where candidates often run unopposed), but I don’t have any detailed theory of exactly where it would work best, hence my interest in seeing experiments.

    • DavidS says:

      As with all discussions of voting systems I kind of want to know what’s trying to be achieved (e.g. is it about the legislature reflecting what the people would want if they could vote on every issue? If they could vote on every issue and were better informed? Or is it more about ensuring that political power isn’t concentrated in a single class, that we can have peaceful transfers of power etc. etc.

      If there’s a place for sortition I agree it has to be in a bicameral approach: e.g. with an elected house amending legislation and going into the detail and the sortition house tending to be more ‘yes or no’ (not sure if you’d prevent them amending at all to do this or would just follow from them having less time etc.)

      Another concern about it of course is it’s vulnerable to both the reality and the perception of bribery and corruption. It’s fairly obvious if elected reps vote directly against their platform, they have some accountability. Not so sortition.

      More generally I like political parties: they mean you can be held to account for what actually happened last time you were in power, rather than for the fineness of your rhetoric alone. Any system which removes them (or through coalitions blurs who is in charge) seems to me to lose that fairly major benefit. Also the fairly major psychological/social benefit of an election being a point where you can chuck out a government you don’t like or signal your support for one you do.

      • Nabil ad Dajjal says:

        Over the last few years I’ve come around to the idea that the incentives on a decision-maker matter more than whether or not they have expert credentials.

        An average person has more and more detailed information about their needs than any third party could possibly gather. That’s the core of the economic calculation problem as I understand it as well as James Scott’s concept of ‘metis’ or folk knowledge. And they have the most incentive to make sure that their needs are met, since they’re the ones who feel them the most acutely.

        Representative democracy fails in both respects. Representatives have at best second-hand information on the aggregate needs of their constituents and only a limited incentive to meet them. The voters themselves also have little incentive to inform themselves on what exactly their representatives are doing or to vote carefully as the marginal vote has essentially no chance of changing the outcome.

        Sortition is a way to pull things back to an individual level. Each of the selected representatives can vote in their own interests as individuals which they are in the best position to understand, and they can’t avoid being individually responsible for whether that choice helps or hurts them in the future. Because they’re statistically representative of the population, this should create overall better legislation.

        The big caveat is that this really only applies to domestic questions where people actually have hands-on experience. Foreign policy is something that affects ordinary people but usually in indirect ways that aren’t conducive to building up practical knowledge.

        • albatross11 says:

          Incentives are important, but they’re not everything.

          If you put me in the pilot’s seat of a helicopter in midair, or in the operating room where I’m supposed to perform an appendectomy on one of my kids, I have every incentive imaginable to do a good job, but I still won’t be able to, because I don’t know enough.

          We do indeed have a problem with decisionmakers whose incentives don’t align with ours, and we should fix that. But just aligning incentives doesn’t solve all our problems.

        • albatross11 says:

          Thinking more about this, it seems like randomly selecting legislators has the benefit that it makes it harder for a smallish ruling class to keep all power to themselves, but the cost that it makes it harder for anyone to *specialize* in being a legislator, at least in the randomly-selected house.

          My impression is that Congress now doesn’t generally read or understand most of the bills they vote on, does a poor job of oversight of intelligence and other functions of government, doesn’t care to vote to declare any wars despite the plain language of the constitution on the matter, etc. I’m not 100% convinced that random selection of eligible voters would do all that much *worse*, but I’m not sure how we’d test that.

          • Matt M says:

            I think under the current system, getting “better” at “being a legislator” basically translates to getting better at winning elections, appeasing special interest groups, obfuscating what’s really going on in DC, etc.

            Random selection might be a win-win here.

          • John Schilling says:

            appeasing special interest groups

            Yes, of course. Everything is a special interest. Just about everything Congress can do, and just about everything Congress must do, is a thing that about 10% of the population cares about enough to actually want, 30% of the population doesn’t understand but, meh, the people who want it are part of my tribe but OK, and 60% regard as a “special interest” that is being appeased.

            About the only truly general interests are reducing violent crime, reducing taxes on the middle class, and waging quick, clean wars against Evil Dictators. Two of those things nobody knows how to do, and the third leads to disaster if that’s all you do.

            So building coalitions that get useful stuff done, in spite of the useful stuff being basically all “pandering to special interests”, takes real skill. As BBA notes, that means congressmen like Mitch McConnell and Nancy Pelosi, universally despised but consistently reelected because not having people like them is even worse.

            obfuscating what’s really going on in DC

            This, also, takes real skill. And is necessary so long as you’ve got an electorate daft enough to believe that Congress shouldn’t be “pandering to special interests”.

            You don’t specifically need legislators to have these skills. The permanent standing bureaucracy will certainly have them, and can use them to manage legislative as well as public expectations. See “Yes, Minister” for what this looks like in action, and expect it to be dialed up to eleven if you go even halfway to sortition in selecting your legislators.

        • A1987dM says:

          An average person has more and more detailed information about their needs than any third party could possibly gather.

          Given that possible third parties include Google and Facebook, I’m not sure that’s true anymore.

  19. Nancy Lebovitz says:

    Summary: Escolar (frequently sold as butterfish or white tuna) can make you pretty sick.

    I bought some smoked butterfish at the Reading Terminal Market– it was only $5 for about a pound, and I like smoked fish.

    I didn’t like it all that much– too salty and an odd flavor. I was trying to figure out whether rinsing it and cooking it with something was worth doing.

    Then I got some diarrhea which seemed vaguely different than usual– some of the details are TMI (Too Much Information) and I started thinking about what I’d been eating lately.

    I’d heard about white tuna (a sort of sushi) being hard on the digestive tract, so I was open to the possibility that fish might be a problem.

    Well! White tuna isn’t related to tuna, it’s butterfish. So is escolar. I will say a thing or two to the people at Reading Terminal Market– they’ve got a big fish shop with a neon sign over it that says something like EAT FISH BE HEALTHY.

    White tuna as sushi isn’t a hazard to me– the quantify in an assorted sushi plate isn’t enough to hit me, and it’s actually pretty tasty. It being labeled as white tuna is eroding my faith in humanity that little bit more, though.

    Substantial article. I got off easy, some people get a lot sicker. If you read the comments, you’ll find that people getting sick from escolar happens all over the world, except Italy and Japan where the fish is illegal. Pricey restaurants sometimes sell escolar (mislabeled, often enough) as a main dish.

    Teminology! There’s an English eel called butterfish.

    Mercifully, “black cod” is at least has black scales, but many species of sable aren’t black.

    I’m reminded of the bit in Stranger in a Strange Land which complains about English words having multiple meanings. The example was that red hair doesn’t resemble the color otherwise called red.

    Butterfish, the red-tailed hawk and turkey vulture of the sea.

    Which is not the same as butterfish.

    • Gossage Vardebedian says:

      I had a similar . . . episode . . . once. Sea bass at the local market for $10 a pound? What a deal! Yeah, no.

      I think it’s delicious, but evidently I am in the roughly 1/3 of people who can’t digest the particular fats the fish produces – which is something I read on the internet the day after I ate said fish and wondered what was going on. Don’t know if it’s true or not. I do know that if sea bass is being sold for around 10 bucks a pound, it isn’t sea bass. Which has led to a few interesting conversations at various fish counters over the years, though none in a while, as the recent stories about mislabeling fish have seemingly raised the profile of this issue for the better.

      • Nancy Lebovitz says:

        I’ve been buying bass (basso) for less than that at the Italian Market in Philadelphia and had no problems with it. Maybe it’s fresh water bass.

        It would be nice if there were a fast cheap DNA test which could identify fish.

        I don’t know if you read the comments on my first link, but there was one man who was so traumatized that he wouldn’t eat white fish unless he bought the whole fish so he could tell what he was getting.

        • Odovacer says:

          The way they currently genetically test fish (IIRC) is through DNA barcoding. They can identify species via differences in a short sequence in the Cox1 gene. Cox1 is a mitochondrial gene present in many eukaryotes. Basically, it is varied enough between species, but doesn’t have too much variance within a species, to be useful for genetically separating species.

          Plants and some other organisms are barcoded with different DNA sequences.

          As for fast and portable sequencing of DNA. Well, that’s coming sooner than you think.

    • Odovacer says:

      That’s not the only mislabeled fish out there! Oceana, an ocean conservation NGO, genetically tested seafood samples from restaurants and stores to see what was mislabeled.

      Key Results:

      • Mislabeling was found in 27 of the 46 fish types tested (59%).

      • Salmon, snapper, cod, tuna, sole, halibut and grouper were the top collected fish types.

      • Snapper (87%) and tuna (59%) were the most commonly mislabeled fish types.

      • Only seven of the 120 red snapper samples were honestly labeled.

      • Between one-fifth to more than one-third of the halibut, grouper, cod and Chilean seabass samples were mislabeled.

      • 44% of all the grocery stores, restaurants and sushi venues visited sold mislabeled seafood.

      • 84% of the white tuna samples were actually escolar, a species that can cause serious digestive issues for some individuals who eat more than a few ounces

      They have the following recommendations when buying seafood:

      • Ask questions. Consumers should ask more questions, including what kind of fish it is, if it is wild or farm raised, and where, when and how it was caught.

      • Check the price. If the price is too good to be true, it probably is, and you are likely purchasing a completely different species than what is on the label.

      • Purchase the whole fish. When possible, consumers can purchase the whole fish, which makes it more difficult to swap one species for another.

      • Nancy Lebovitz says:

        I don’t care nearly as much about getting something cheap cut into cylinders and paying scallop prices as I do about getting sick, and escolar is the only health hazard on your list.

        This being said, it seems to be implied that there is such a thing as real white tuna sushi. I wonder what that might be.

    • Nancy Lebovitz says:

      Hey! We’re supposed to be techno-optimists.

      Find a supplement which enables people to digest escolar comfortably.

      Genetically engineer safe-for-humans escolar.

      Make an AI which is directed to preserve humans and escolar and human pleasure in safe escolar and let it fight it out with a paper-clipper.

  20. Nancy Lebovitz says:

    The Death of Stalin— the whole movie. The site didn’t require signing in. I have no idea why this is available for a movie that’s in the theaters.

    • Montfort says:

      I admit I’m not certain, but I strongly suspect that whoever owns the film’s rights didn’t authorize this.

    • Andrew Hunter says:

      I saw it in theater this weekend and quite liked it. Not as good as Iannuci’s earlier movie In the Loop (two hours of british pols screaming obscenities at each other in a wonderful farce) but very beautifully shot and well acted.

      I wasn’t sure what the point of the movie was, honestly. A bunch of things happened, and I can see the motivations of various characters, but the closest thing to the plot’s central throughline (Orevn Zhfg Qvr) takes a long time to get established, and then just…happens.

      Still recommended.

      • James says:

        I feel like the stakes being higher (gulags and assassinations) than in In The Loop makes me enjoy the petty feuding and pecking-order politicking much more.

      • Well... says:

        I’m about 2/3s through it. I’ve started mentally substituting out Stalin/themes of Stalinism for objects of political correctness. I don’t think it’s controversial (and therefore not too CWy) to point out the movie works perfectly after this substitution, and actually becomes cogent and relevant.

        • Well... says:

          Having now finished the movie and thinking more about it, I think it was actually cogent and relevant for other reasons — ones superordinate to political correctness per se. Namely, the way mystical religious-style behavior combines with group psychology. The whole movie is a sort of Abilene Paradox/car crash in slow motion. Given the de jure status of religion and hierarchy in Soviet Russia, Death of Stalin chooses an extreme (but hilarious)** backdrop against which to illustrate this mixture of social phenomena.

          **Should that be “hilarious (but extreme)”?

    • Yakimi says:

      The film was first released in the United Kingdom in October of 2017.

    • James says:

      Great movie.

  21. johan_larson says:

    Quiz time.

    Each prompt is a quote from a major film. Name the film.

    1. Uh, we had a slight weapons malfunction, but uh… everything’s perfectly all right now. We’re fine. We’re all fine here now, thank you. How are you?

    2. Zed’s dead, baby. Zed’s dead.

    3. Teldar Paper, Mr. Cromwell, Teldar Paper has 33 different vice presidents each earning over 200 thousand dollars a year. Now, I have spent the last two months analyzing what all these guys do, and I still can’t figure it out.

    4. There was a nuclear war. A few years from now, all this, this whole place, everything, it’s gone. Just gone. There were survivors. Here, there. Nobody even knew who started it.

    5. “The Statue of Liberty is kaput” – that’s disconcerting.

    6. Not in the mood? Mood’s a thing for cattle and loveplay, not fighting!

    7. Quite frankly, I didn’t even want to use you guys, with your dip and velcro and all your gear bullshit. I wanted to drop a bomb.

    8. Do you know what will happen to me if the bridge is not built on time?

    9. This isn’t flying, this is falling with style!

    10. In teenage emoji, eggplants are dicks.

    • fion says:

      You sucked me in with the first one, but most of them are beyond me.

      1. n arj ubcr

      4. sounds a bit like gur grezvangbe, but I can’t be sure.

      6. sounds familiar. Taking a stab in the dark: qhar

      9. gbl fgbel

    • James says:

      I only got the really easy ones, 3 and 9.

      I work with someone called Ed whose last name begins with a Z and everyone calls him ‘Ed Z, baby. Ed Z.’ in tribute to 3. Only works with British English pronunciation, obviously.

      I misread the prompt as being about ‘opening lines’ and thought that #10 was a fantastic opening line to a film.

    • S_J says:

      I recognize number 1 and number 9 right away.

      1. Fgne Jnef, gur svyz yngre fhogvgyrq nf “Rcvfbqr VI: N Arj Ubcr”
      9. Gbl Fgbel

      Looking up the rest, I’ve seen the film quoted in number 6, but I didn’t remember the quote.
      I ought to have guessed numbers 3, 4, and 8.
      If I’d seen them, I would have likely remembered the films which provided numbers 2 and 5.

      • fion says:

        Aaah! That’s so weird!

        I naq V ner guvegrra yrggref ncneg, fb jura lbh ebg13 gurz lbh punatr fbzr ebzna ahzoref vagb bgure ebzna ahzoref! Funzr vg qbrfa’g jbex sbe K, Y, Q, P be Z. :/

    • Iain says:

      I got 1, 4, 8, and 9 correct.

    • sandoratthezoo says:

      How the hell has nobody answered #2 yet? Guys, it’s Chyc Shpgvba.

    • quaelegit says:

      I got 1 and 4. (I’m proud of 4 because I haven’t seen it but guessed from context.)

      I’ve heard 9 a lot but didn’t know it came from this movie!

      My (very tentative guesses) for 5: Cynarg bs gur Ncrf? Vaqrcraqrapr Qnl?

      • johan_larson says:

        Neither of those guesses is right. Let me give you a hint: it’s a war movie.

        • quaelegit says:

          Then I definitely don’t know it! 😛

          (I looked it up, and I’m actually familiar with the movie and have seen most the famous scene, but I don’t remember any lines from the movie.)

    • Nornagest says:

      I can only get 1 and 2.

    • johan_larson says:

      The answers are:

      1. Fgne Jnef: N Arj Ubcr
      2. Chyc Svpgvba
      3. Jnyy Fgerrg
      4. Gur Grezvangbe
      5. Fnivat Cevingr Elna
      6. Qhar
      7. Mreb Qnex Guvegl
      8. Gur Oevqtr ba gur Evire Xjnv
      9. Gbl Fgbel
      10. Oybpxref

  22. a reader says:

    Anybody wants to share with me Jordan Peterson’s Self Authoring Suite Special! (2 for 1) – for $15?

    The suite is complete, contains Past, Present (Virtues & Faults) and Future and allegedly can help someone improve his life and fight akrasia (we’ll see if it’s true):

    The Past Authoring Program helps you remember, articulate and analyze key positive and negative life experiences.

    The Present Authoring Program has two modules. The first helps you understand and rectify your personality faults. The second helps you understand and develop your personality virtues.

    The Future Authoring Program helps you envision a meaningful, healthy and productive future, three to five years down the road, and to develop a detailed, implementable plan to make that future a reality.

    There is a study (for the Future part), having J. Peterson as co-author, that seemed to prove its results – but there is also a replication where it didn’t work.

    The promotional 2 persons suite costs 30 USD (like the normal 1 person suite) so it will be 15 USD/person. If you are interested, please post an email (rot13 if you want to protect it from spambots) and I will send you my Paypal to send me 15 USD and after, I will send you the second username and password of the suite (the password can be changed).

    (Please, don’t deviate the thread towards JP’s involvement in culture wars.)

  23. James says:

    Jordan Peterson (him again?) says somewhere that artists are great at creating value, but terrible at capturing it. In other words, the fact that artists are not, in general, well-remunerated does not reflect the actual value of their works. Do we tend to agree? What is the marginal value of one more artist? (I’m treating ‘artist’ as a monolithic category, but of course it could include all sorts of professions: musicians, writers, actors…)

    This is a question of some consequence to me as someone whose ambitions almost entirely lie not with his day job but with his musical work outside of it.

    • Nabil ad Dajjal says:

      Art as a whole is very valuable and the best artists probably don’t capture even a fraction of the value they create. But I suspect that the marginal artist creates very little value or may even destroy value.

      For example, it seems that the popularity of musicians follows a power law distribution. The greatest musicians of the modern era have brought joy to hundreds of millions of people; an amount of value which is difficult to imagine and of which they likely received only a small portion. But the median garage band or guy banging on drums in the subway is a public nuisance; the value they create is worth a noise citation.

      Music is a great hobby but too many people are lured into a career in music by the long tail of prestige. My little brother almost made that mistake himself, until an older musician took him aside and explained that a merely above average musician like him had no hope of success in the music industry. The only people who should be encouraged to pursue music professionally are the top fraction of a percent.

      • DavidS says:

        On the second para: do you mean classical? Because while I agree on the power law in terms of success etc. I’m not sure that the most successful musicians in many areas are actually the most talented in any straightforward sense (unless you define it in a circular way)

      • Chalid says:

        The greatest musicians of the modern era have brought joy to hundreds of millions of people; an amount of value which is difficult to imagine and of which they likely received only a small portion

        The question is, if great musician A had never existed, to what extent would some other great musician B have just filled their niche and brought everyone equivalent amounts of joy?

        • I think of this in terms of writing, that being my art. What excuse do I have for writing Brothers when there are books out there by Heinlein and Bujold and Cherryh that are better than anything I am going to write?

          The answer in that case, I think, is that none of them has written a book with the characteristics of the books I write. Nor have the three of them together written more books than a reader who likes them is going to read in a lifetime. My writing isn’t better—in some reasonable sense it is worse—than theirs, but it is enough different to provide those readers who have the relevant tastes a different and potentially valuable experience.

          There’s a quote somewhere to the effect that an author is someone who believes there is not enough of some particular sort of literature and aims to make more of it.

          • theredsheep says:

            I’d add that authors aren’t always good in the same way. Heinlein was great at a lot of things; I’ve never heard anyone suggest that deep, believable characters was one of them. Bujold, on the other hand, is all about character, but not as good as Heinlein at really digging into scientific or social hypotheticals. The sci-fi or fantasy aspects of her stories are, at times, very nearly background or props.

      • Matt M says:

        But the median garage band or guy banging on drums in the subway is a public nuisance; the value they create is worth a noise citation.

        I’m not sure about this one. Maybe I’m weird but I like almost all music, or at least, I like it better than a lack of music entirely. Even the crappy street performers, the vast majority of them I’d rather have around than not have around.

        • quaelegit says:

          I agree with you on street performers, but actually I don’t mind silence and in fact tend to dislike piped-in music at malls and such. I think I was trained to appreciate live performances by all the concerts my dad took me to. Also I’m very volume sensitive, and live street music at least doesn’t tend to be uncomfortably loud.

      • James says:


        Well, for my part, I hope to at least stay on the right side of creating vs. destroying value 🙂

        You’re right about the power law distribution, especially so in the age of mass reproduction. I’m also not especially talented in the narrow musical sense. I came to it fairly late and am by no means a prodigy or anything. But I guess I aspire to do something that appeals by its (mild) originality and innovation, rather than by particularly talented ploughing of a well-known, well-understood furrow. But of course it remains to be seen whether it will really work out this way.

        In any case, I’m not doing it to make lots of cash. Like I said, I have a day job and I expect I probably always will (though perhaps not always full-time). Really I’m just in it for the admiration of women.

        And besides, in many ways I enjoy the work; there’s a gentle-but-deep pleasure in slowly bringing something beautiful into being. I spend a lot of time on it, but it’s just time that I’d be spending anyway.

    • smocc says:

      An example I thought of the other day — The guy who invented the “I♥NY” logo was certainly not compensated in proportion to the share the concept has in people’s minds. (Just looked him up, his name is Milton Glaser and he did the design pro bono, so definitely not). Even though the logo itself is copyrighted, and New York state tries to stop copycats, people use the basic idea everywhere.

      On the other hand, I couldn’t think of how he could have profited from it. It seems like the sort of idea that people love to use, but also that very few people would pay for the right to use it if it cost money. (Or they would do it illegally, as they do now)

      • rlms says:

        It might have a lot of mindshare, but I don’t think that necessarily equates to a lot of value. I don’t think many people would pay a significant premium for an I♥NY shirt over whatever the alternative was before that logo was invented.

    • ohwhatisthis? says:

      Idk. I only listen to mega pop stars. They tend to have millions of millions.

  24. johan_larson says:

    Over the long weekend I had a chance to rewatch a couple of favorite films about the financial industry: The Big Short and Margin Call.

    Any others worth a look?

  25. Nornagest says:

    I watched a few episodes of Netflix’s Altered Carbon series. tl;dr: it’s not very good.

    Cyberpunk’s a pretty dated genre at this point: it’s a cliche to say that we can read about state-sponsored hacking, bioterror threats, and military robots in the New York Times, and that’s true as far as it goes, but more importantly cyberpunk never grew past a fairly limited set of themes (simulation, artificiality, economic inequality, “what does it mean to be human?”). Those were pretty well mined out as early as Snow Crash (1992), and subsequent authors dealt with that by taking them as given. They became part of the cultural vocabulary: an establishing shot of a noodle shop in a rainy, neon-lit cityscape is all the background we need, allowing authors to foreground their newer, more interesting ideas. This is the basic assumption of post-cyberpunk, of which the Takeshi Kovacs books (Altered Carbon the book is the first) are a typical example.

    Altered Carbon the series is straight-up old-school cyberpunk, though, and that’s the problem. The production values are high, the sets are gorgeous, the acting is tolerable to good, and there’s nothing terribly wrong with most of the casting choices. And it’s not an unusually unfaithful adaptation, really. But all the changes it makes — mostly to character roles and backstories — work to retool it from a fairly nuanced Philip Marlowe whodunnit that just happens to take place in a world with body-swapping technology, to a cheap morality play stretched over a neo-noir skeleton. Flawed human beings turn into Captain Planet villains. Kovacs’ skillset and attitude no longer lines up well with his backstory. Many characters’ motivations flat-out don’t make sense. The timeline’s extended by a few hundred years for no good reason, which renders the legislative and Neo-Catholic subplots much less reasonable than they were in the original.

    None of this would necessarily have been a deal-breaker — in 1985. But now we’ve seen it all before. And there’s really no excuse in a 10-episode series, where there’s plenty of time for nuance. I’d skip it, unless you really need your fix of gritty spacefuture for some reason.

    • cassander says:

      I’ll grant you that there’s absolutely nothing new in the show, but I still found it quite enjoyable, and with the exception of xbinpf fvfgre I think your characterization of the antagonists a bit harsh.

      To be fair, though, there is a dire lack of tolerably gritty space future on TV right now, particularly anything that even pretends to try to take transhumanism seriously. What else do we have? That idiotic Titan movie that just released on netflix?

      Besides, who can resist watching James Purefoy chew the scenery?

      • Nornagest says:

        I may as well mention something I did like, which is the Raven. That eerily chipper Poe avatar is probably my favorite character in the show, its motivations do make sense, and it fills a real gap in the cast — the Hendrix in the book didn’t have anywhere near the personality or the screentime. If only all the Netflix additions went so well.

        • Wrong Species says:

          I liked his character too but I find it annoying when a “good AI” exists solely to be helpful to humans. It’s understandable in older works but it comes across as hokey now. Even Star Trek moved pass that kind of thing in Voyager. Why wouldn’t an AI have its own motivation? Why do they care so much about pleasing humans?

        • cassander says:

          he was great. and I just love the idea that the these incredible technological constructs were scattered around the world abandoned because they were a fad that people had gotten tired of.

    • Andrew Hunter says:

      I haven’t read the book (I am queued for Seattle Public Library’s copies), but I disagree pretty strongly. It isn’t an all time great, but I’m thoroughly enjoying the show. As you say, the production values are insane, and that is really something valuable: it’s a thoroughly beautiful exploration of an interesting world.

      I’d say the plot is just an excuse to show us what a world with these constraints might be like. (I don’t know if I agree with their answers, but I like to see it.)

      Parts are definitely crass as hell–tits and ass galore do sell–but I’ve found it generally worth watching.

    • The Nybbler says:

      IMO, it’s a little over half of a decent story. Goes to hell when they bring in the Big Bad, cool as the entrance itself is.

    • James says:

      Nice little write-up; thanks.

      I did think while watching the new Blade Runner that I probably should have seen (the remade) Ghost in the Shell because I like cyberpunk so much that I think I could enjoy even mediocre works in the genre. So maybe this goes on that pile for me.

      You’re right about it being dated, of course. The Neuromancer trilogy is now old enough to feel enjoyably retro.

    • Orpheus says:

      I really wanted (and still want!) to like cyberpunk, but I almost never seem to find anything that is good. The genre as a whole seems to be over fond of some very, very silly tropes, e.g. syfying shit up to make it look cool without regard for functionality (the biggest victim of this by far is computer screens/TVs. Why would I ever want a transparent monitor??).
      The only medium I know that gets cyberpunk mostly right is anime (e.g. Serial Experiments Lain, Psycho Pass etc.). What are some good cyberpunk works?

      • beleester says:

        Snow Crash is pretty good. It’s got technology that balances “useful” and “cool” (a skateboard that has smart wheels that allow it to roll over potholes, for instance), and it’s got a society that balances “corporate dystopia” and “actually functions surprisingly well.” – the Mafia might run your town’s protection racket, but they also run the local pizza company, and they make sure they deliver on time.

        And I’ll give it huge props for avoiding the “jack your brain into a computer” trope, which always struck me as one of cyberpunk’s dumbest tropes, since it always ends with someone getting their brain fried over the internet. The Metaverse is basically a VR version of Second Life, and it’s clearly a frontend for a computer system rather than some mystical other world. And you can access it with VR goggles, like a normal person, instead of cables going into the skull.

        • quaelegit says:

          Although the eponymous “Snow Crash” is (partly) for frying people’s brains over the internet…

          Depending on which silly tropes annoy Orpheus, Snow Crash might be exactly the wrong book because it takes some of those silly tropes, amps them up to 11 and runs with them. The “runs with them part” is really fun, but not if you’re already thinking “this whole thing is really stupid”.

          I do think Stephenson does a better job trying to justify/make sense of/provide context for a lot of the craziest parts of the world than most SciFi books (I haven’t actually read much cyberpunk so I don’t want to say compared to that genre specifically).

          • beleester says:

            Okay, yeah, it does go a little off the rails when we get to the ancient language that speaks directly to the brainstem. But that’s such a weird twist on the idea it almost feels like a different genre.

        • Nornagest says:

          One thing I liked about Snow Crash is that it doesn’t fall back on the cyberpunk trope of setting everything in an anonymous megalopolitan city somewhere between Hong Kong and New York. It’s very oriented to place: it takes place mostly in the LA suburbs, and it looks and feels like the LA suburbs, just moreso and with sporadic exchanges of sniper fire between rival fast-food franchises.

  26. Nancy Lebovitz says:

    This essay argues that rather than the human race having a history of increasing authoritarism (until recently) and technological capability, there’s actually been quite a lot of back and forth so far as hierarchy is concerned, and it’s possible to have early cities with little hierarchy.

    It’s pretty long and leisurely, but I recommend section 4 for the descriptions of human cultures which are very different from the usual narrative of humanity going from little egalitarian tribes to agriculture and empires in a straight line. Recommended for world-builders as well as those who just want the truth about the world.

    • Nornagest says:

      This brings up good evidence that the evolution of hierarchy doesn’t correspond closely to the agricultural transition, but I don’t see any good evidence in it that stable non-hierarchical structures appeared with any regularity in large-scale settled societies after agriculture had become well established. Despite spending most of parts 1 and 2 stacking the deck. The only concrete examples it gives are either very early and correspondingly small-scale (e.g. Çatalhöyük), or extremely speculative (e.g. Teotihuacan). And I think the picture of Tlaxcala as a republican polity has been brought up here before, with the consensus being that its governmental customs were probably not much different from the other Nahua city-states.

  27. fion says:

    When Americans say room-mate, do they literally mean living in the same bedroom as somebody or do they just mean sharing a house? I would tend to say house-mate or flat-mate for the latter so I always assumed room-mate meant the former. But I’ve never lived in the same bedroom as somebody who hasn’t either been a sibling or a romantic partner and to do so seems a bit weird. Is this a cultural difference or am I just misunderstanding how you use the word?

    • James says:

      Yeah, I’ve been thrown by this as well. They just say ‘roommate’ where we would say ‘housemate’.

    • Nabil ad Dajjal says:

      Generally roommate means flatmate or housemate. The exception to that is in college dorms, where you might actually share a bedroom with your roommate.

    • smocc says:

      In general it means someone who shares the same apartment as you, but people definitely do share bedrooms, at least in college. When you want to specify that the person shares not just your apartment but your bedroom you awkwardly try to emphasize roommate or say something unwieldy like “my roommate who I actually share a room with”

    • John Schilling says:

      Traditionally, Americans have been expected to move into private quarters upon completing their education and/or military service, and not share quarters with anyone else until they get married and start a family of their own. More recently, “get married” has been relaxed to allow for shacking up with a lover, possibly in anticipation of marriage. But that still only leaves two scenarios where an adult is going to be sharing a home with someone who isn’t family, and both of those (college dorm mates and cohabitating lovers) usually involve sharing a bedroom as well. So “roommate” was a perfectly good term, if flawed by the fact that some college dorms have private bedrooms.

      Prior to WWII, the standard lodging for people who couldn’t afford private apartments but weren’t outright homeless was a boarding house or SRO, where each person gets their own bedroom and the other people renting rooms in the same building aren’t mates of any kind. There was an acute housing shortage immediately following WWII, but American economic prosperity in that era resulted in good jobs and cheap housing for (almost) everyone before we had time to invent a language for unrelated adults forced by unromantic circumstance to cohabit.

      There has always been a freakish minority living in things like rationalist group houses by choice, but never one big enough to get the language rewritten to accommodate them. Only since the 2008 recession, I think, has it been common for unrelated American adults who aren’t sexual partners to share a house or apartment much past college graduation. Earlier than that in places like Manhattan and San Francisco, due to high housing costs, but again not enough to change the language.

      • James says:


        I live in a medium-sized city in England. I live in a shared house and I feel like most of my (bohemian) friends do, too, though I think I’m the only one of my (yuppy) colleagues who does (and I might even looked on slightly askance by my colleagues for it).

        At university, only a very small number of halls of residence—mostly very old and expensive ones—featured actual shared rooms (as opposed to shared flats). (I saw the inside of one of them, in which a lovely posh girl lived, once and—thanks to a nasty little faux pas—only once.)

        • John Schilling says:

          The bohemian parenthetical may be relevant; living in shared apartments is I think a long bohemian tradition on account of being the only way for Starving Artists ™ to afford living in a Cool City ™. Americans tend to be less bohemian than Europeans for reasons including but not limited to not being on the same continent as Actual Bohemia.

          • quaelegit says:

            >Americans tend to be less bohemian than Europeans for reasons including but not limited to not being on the same continent as Actual Bohemia.

            Haha. Although the term “Bohemian” meaning “weird artsy types” has been separate from the geographic demonym since at least the 1840s

        • quaelegit says:

          >At university, only a very small number of halls of residence—mostly very old and expensive ones—featured actual shared rooms

          Heh, it’s the opposite in the U.S. Most first year students share a bedroom with 1 to 3 other students (if they are living “on campus” rather than renting their own apartment or staying with parents). Depending on the college, sophomores and upperclassmen may move off campus (mostly renting apartments nearby), or staying in the dorms (and possibly getting single bedrooms).

          I think my use of “apartment” above is equivalent to the British use of “flat” but I might be missing nuances.

      • fion says:

        Thanks for your input. It hadn’t really occurred to me that sharing houses was weird. I guess the reason is a combination of me being a student and so being used to it and me being young such that post-2008 is really a very long time for me.

        I still think it’s weird that American students share bedrooms, though. 😛

      • Only since the 2008 recession, I think, has it been common for unrelated American adults who aren’t sexual partners to share a house or apartment much past college graduation

        You earlier said “completing their education,” which isn’t quite the same thing. When I was a grad student I normally shared an apartment with another grad student, and I think that was common.

        • John Schilling says:

          Yes, “completing their education” was me being careful about what I meant to say, “college graduation” was dialing back the precision and pedantry because I’d already made that point. You’re right that people pursuing post-baccalaureate degrees often share living quarters, but I think until fairly recently that has been a sufficiently tiny fraction of the population as to not affect the language we use to describe people’s living arrangements.