Archive for the 'glibness' Category

28
Sep
14

Scott Sumner makes three points

The first one is pretty subtle; the second is rather less so; and the third is about as subtle as global nuclear war.

17
Aug
14

Academic Bubbles and Those Damn Austrians

(That is, Austrian-school economists, not people from Austria, unless the latter are also the former.)

Here’s Scott Sumner on a natural experiment in hard-line grade deflation:

OK, so the Austrians are right in this case. But I still think they are wrong about 2006-09.

This in reaction to the unsurprising-at-first-glance result that students often choose their courses to get the best grades, and when a previously-profligate department tightens the GPA purse strings, those students will switch away from that department’s courses.  I say “at first glance” because grade scales are just arbitrarily-relabelled intervals; my undergraduate university awarded grades on a nine-point scale (but not a stanine), and my grad school’s 4-point scale actually went up to eleven 4.33.  Both were heavily weighted towards the top end of the interval in practice, although I suspect a handful of exceptionally deserving students scored 1s in both.

Anyway, do click through and read the whole thing.

03
Jul
14

The kids aren’t just all right, they’re out of your hair

Granted that this post is part of a lasing medium, but Elizabeth Nolan Brown gives us an article with a mildly optimistic title:

(Hat tip: Coyote Blog.)

For the better part of a decade now, folks have been fretting about “boomerang” kids, the 20- and sometimes 30-something children of boomers who’ve come flocking back to their parents’ nests under the duress of a poor economy.

[...]

The dire pronouncements tend to be based on U.S. Census Bureau data, which does show an increasing number of young adults—more than half of those under 25, according to the most recent data—to be living with their parents. But Derek Thompson at The Atlantictears through this gloomy prognosis with one simple fact: The Census counted students who live on college campuses as living in their parents’ homes.

That almost makes me want to start reading The Atlantic again.  Then again, I’ve been pretty happy since I stopped.  Let’s roll with happy.

Brown concludes:

Even this isn’t quite as scary—or at least not as singularly scary for young adults—when you put it into perspective. When (if) the job market improves, young adults will likely have an easier time slipping back into it than their older counterparts simply by virtue of being younger and cheaper, said Dan Schawbel, founder of Millennial Branding management and consulting firm.

Meanwhile millennials are only barely less employed than Gen X’ers, who make up 37 percent of unemployed Americans. The oldest Gen X’ers turn 50 next year, while the youngest hover around age 35. This is the generation in the prime of their “prime earning years.” Whither the concern for Gen X everybody?

(Emphasis added.)

I submit that “concern for Gen X”, much like “concern for Gen Y”, is being expressed in minimum wage increases, or advocacy therefor.  It ‘s after all pretty cheap for currently-employed Boomers to demand that new hires be paid more, even though this leads to fewer new hires under the drearily predictable logic of simple division.  Because I’m sure I’ll need to spell it out, here goes: If I’m an employer with a budget of $T to spend on new hires, and the minimum wage is $k, I will hire at most n = floor(T/k) people.  Increase k and n decreases, unless you somehow manage to increase T.  Any minimum-wage doves want to go publicly all-in on subsidizing big business?

(I’ll note in passing, because I haven’t been enough of a nerd yet today, that all three of those terms should be parameterized with respect to time.  Under what assumptions does it make sense to increase $T(t) to compensate for an increase in $k(t)?  What do those assumptions imply about current restrictive immigration policies?  Please show your work; you should be able to use LaTeX to mark up integrals.)

26
Jun
14

Misery lasers

So this is how outrage fatigue works, kind of.

Most lasers — actual devices that amplify light by stimulated emission of radiation — work by bouncing photons back and forth between mirrors through a lasing medium until — zot! — they slide through one partially-reflective mirror.  I’ll let you look this shit up on Wikipedia if you so desire, as I’m too drunk to muster the giveafuck myself.

Software lasers are a mildly fanciful generalization of the concept: They get packets bouncing back and forth between mutually-antagonistic routers (or mail servers, or whatever) until one or both of the “mirrors” melts into more or less euphemistic slag.  But the generalization is instructive.

A misery laser comes to happen when two or more people of similar political persuasion come together to bitch and moan about Things These Days.  One person will bring up a topic that annoys their compatriots; the next will riff off of that topic to reinforce the notion that Things Are Going To Hell; the third will say “yes, and…” (and elaborate!), and then the malaise and frustration will reflect off of the side of the room and propagate back towards the first person.

Efforts to inject optimism into the system are, as you might expect, doomed to failure.  You might as well shine a flashlight crosswise through a lasing medium.

Eventually, an optical laser will shoot coherent light off in one or the other direction of its major axis.  A software laser will eventually fuck up one or the other mail server (or whatever the fuck services were having a spat).  And eventually a misery laser will lead to an emotional breakdown, or perhaps (mostly) minor acts of physical violence.

This is why I don’t talk politics any more.  And when people around me do, I fantasize about being on a motorcycle, on a twisty road, half a continent away.  Continents are good for that; there’s usually a twisty road somewhere way the fuck far away.

13
Nov
13

All linky, no thinky 2: The Linkening

So I’m joining the rising tide of anti-intellectualism that’s destroying Classical Liberal Arts Institutions, or whatever, and taking a course on reactive programming on Coursera (one of those MOOCs that’s destroying &c.).  Feels good to stretch my brain again; I’ve wanted an excuse properly to learn Scala for a while, and maybe this time around I’ll actually grok monads.  (If you’re wondering what “reactive programming” is, it’s writing Erlang in languages that aren’t Erlang.  So far as I can tell, at any rate.)

——

Is fairness a process thing or an outcome thing?  I suspect most of us’ll pick one until we come across an instance of the other we don’t like, at which point things go all Black Monolith and we club each other with femurs.

Money shot:

As I see it, many upper middle class parents desire their child to be slightly more successful than they are, and in related but not identical fields and ways.

Duh, you say, which tells me you haven’t read it.  “But why wouldn’t you prefer to hire a better worker?”  Why didn’t you buy a Bentley Mulsanne instead of a used Camry?  “So practical!”  Shut up, you’ve made my point.  Why hire a superstar developer for a gajillion dollars when all you need is someone to poke node.js with a stick?  “But assholes drive Bentleys!”  You think Mark Zuckerberg’s an asshole, don’t you?  “Huh?”  Just scroll down already.

The real insight here is into the minds of so-called “consumer advocates”.

Teetering dangerously close to reaggravating my outrage fatigue.

Oh look, a nice comforting hobby-horse.  Meta-analysis shows that “saturated fat is not the problem”.  No shit, buttercup.  Fat loss is widely correlated with improved cardiovascular health, and a fat loss diet is, de facto, high in saturated fat coming from your own god damn adipocytes.  Here’s the paper’s author giving me an enormous confirmation-bias boner:

Saturated fat has been demonised ever since Ancel Keys’s landmark “seven countries” study in 1970. This concluded that a correlation existed between the incidence of coronary heart disease and total cholesterol concentrations, which then correlated with the proportion of energy provided by saturated fat. But correlation is not causation. Nevertheless, we were advised to cut fat intake to 30% of total energy and saturated fat to 10%.” The aspect of dietary saturated fat that is believed to have the greatest influence on cardiovascular risk is elevated concentrations of low density lipoprotein (LDL) cholesterol. Yet the reduction in LDL cholesterol from reducing saturated fat intake seems to be specific to large, buoyant (type A) LDL particles, when in fact it is the small, dense (type B) particles (responsive to carbohydrate intake) that are implicated in cardiovascular disease.

We make kids go to school because it’s “good for them”, and everyone agrees that it’s “good for” kids to go to college.  So why not round them up at gunpoint, herd them into cattle cars, and send ‘em off to West Bumfuck State?

As odd as it may sound, the majority of time and resources of the FTC is not spent on punishing bad business practices as authorized in the FTC Act. The agency overwhelmingly concentrates on enforcing another act also passed in 1914, the Clayton Act, and specifically section 7, which prohibits mergers and acquisitions where the effect “may be substantially to lessen competition, or to tend to create a monopoly.”

This is why I don’t blog about politics any more:

Pierce, Rogers and Snyder find that political partisans are more upset about an election loss than a random sample of parents were upset by the Newtown shootings.

An interesting discussion on how humans can add value to computer programs when those programs are really, really good.  The context there is chess, which is a pretty well-understood game of finite complexity.  I claim that humans have been doing this for decades in software development, whose practical complexity is limited only by what you can convince your publisher is actually possible.  Worried about computers taking over your job?  Computers have taken over mine on the regular over the past two decades, and as a result I keep getting better and more interesting jobs.

“Creative destruction” is something that most people who aren’t raging anarchocapitalists like to write off as abstract, idealistic propaganda.  Fortunately, Bryan Caplan is a raging an-cap, and he’s set it all out in time-series graphs so you can actually see it.

I have to admit, I threw this in just for the shock value.  But see previous no-think-link about college being good for kids.

Why do altruists help people?  Because they want to be seen helping people.  This should surprise precisely no-one.

Rob Ford lol.

——

Tune in next time for part 3, when we’ll discover whether this series is better-on-evens (Star Trek) or better-on-odds (Back to the Future)… or just shit (The Fast and the Furious).

06
Nov
13

All linky, no thinky

Old News edition.

(Take “capitalism” to mean “competition in relatively unfettered markets”.)

I’m sick of people shitting on Twitter, especially now that it’s established its utility as a low transaction cost (that part’s important, reread it until you understand why) ad-hoc communications medium.  For example, this would have been mired in delusional architecture and Kafkaesque specifications-by-committee if Twitter hadn’t made it possible for Translink’s customer-service folks to simply tell everyone who was interested why the bus is late today.  And if you don’t think web development is all that hard, I invite you to try to buy American health insurance.  Software will fuck you up.

Turns out the market’s response to the horrible, eschatological, cataclysmic government shutdown was… “Huh, did you say something?  Government what now?  Sorry, I was focusing on things that might affect my bottom line.”  You’d think a Mad Max scenario would qualify, which should make you wonder whether it was likely to happen in the first place.

“Treasury bonds default” has gone from “don’t be absurd” to “well, of course it’s an absurd idea, but…“.  This does not make me any happier.  There’s also this:

Addendum: By the way, we used to read that an attack of the bond market vigilantes would be good for the economy, but it seems this is no longer the case when the vigilantes are led by Republicans.  Hint: an attack of the bond market vigilantes is not good for the economy.

Curiously, it turns out that most of the things teachers’ unions champion are of great benefit to teachers but not so much to students.  Don’t read the comments, they’ll give you cancer.

An immediate consequence is that developing countries are turning into service economies at substantially lower levels of income.

“Deindustrialization”, if you haven’t bothered to click through, is the act of turning manufacturing jobs that make anti-sweatshop activists righteously indignant into call-centre jobs that make anti-sweatshop activists with bricked iPhones righteously indignant.  Which is progress, I suppose, because if their iPhones are bricked they can’t tweet you handwringing nonsense when you’re busy trying to set your fantasy football lineup.

Quick: Who among you would be so generous as to take in a family of complete strangers with a sick child, give them a room for the night, and serve them breakfast?  Oh, uh, spoiler warning.

“And then Bryan Caplan stood up and just started Bryan Caplanning at everyone.”

More later.

15
Jul
13

Alpha-testing a massively incomplete autoregulatory bastardization of 5/3/1

…that’s what I’m doing in the gym these days.

So as I’ve mentioned before, I really like the idea of “cybernetic periodization”, which means “lifting as much as you can on any given day, but no more” rather than “replacing your body with robot parts” as you might expect.  I’ve started lifting five days a week, which is great because it means (wait for it…) I get to lift five days a week, but it’s not so great in that some of those days I’m varying degrees of beat up and can’t exactly push for PRs.  (Fun fact: This happens independently of how often I lift, because my job’s kind of engrossing and every once in a while it rises to dominate my life.)  So, what to do?

Well, I’ve had my best gains on Jim Wendler’s 5/3/1.  It’s based on the following principles:

  1. Emphasize compound barbell lifts
  2. Start too light
  3. Progress slowly
  4. Set PRs

I’m on board with 1-3, but 4 tends to give me fits.  I take it too literally, and my ego gets in the way.  So, wat do?

I started off by identifying seven lifts I care about.  You can guess what these are: squat, front squat, press, bench, snatch, clean, and deadlift.  I took recent maxes on each of these from my training logs — not special “test day” maxes, not “six months ago I could lift this” maxes, not “on a good day I ought to be able to hit” maxes, but stuff I’ve done recently for clean singles.  Then I took 85% of those.

That’s my target lift for the day.  It’s a lift I ought to be able to hit for a single no matter what life throws at me, provided that I’m actually healthy enough to get to the gym and not spread cholera.

So on squat day, for example, I’ll work up to a single at my target weight.  This is not taxing; in fact, it’s basically a warmup.  From there, I proceed by feel.  Did that single feel snappy and smooth?  Add five or ten pounds and do another single, then reassess.  Did it feel slow, grindy, or awkward?  Maybe some technique issues need addressing?  Hit another single, and reassess.  They’re just singles, so they go by quickly — lots of lifts in a short time.  Changing the weights is usually enough rest.  Keep going as long as it feels good, adding weight when reasonable.

Did that single say “fuck this, I’m done”?  Move on to backoff sets.  Backoff sets are free-form, just get them done.  Some days it’s five sets of three at the target weight or above; some days it’s one set of five at 50% and get the fuck out.  The only rule here is “do some backoffs”.

If the day’s big lift is a squat, I just squat.  If it’s a press of some sort, I try to superset in chins or DB rows.  If it’s a snatch or a clean, I throw in a Klokov press or a jerk after each rep until I can’t any more, then I don’t worry about it.  Supersetting is, again, something I don’t think too hard about.

Deadlifts are a special case in that I don’t do ‘em as a major lift, I work them in after presses.  And I don’t do backoffs.  My guess is that I’m demanding enough of my recovery capacity without lots of deadlift volume on the regular, and putting them after the main lift limits the load I can use.

Once I’ve finished the main lift (or superset), I work in whatever else I need to complete the “push, pull, squat” trifecta.  If I squatted, I’ll do a push/pull superset like dips and chins.  If I pressed or pulled, I’ll do something squat-like, counting deadlifts (as above) and their variations (I particularly like snatch-grip deadlifts in here).  Again, I don’t think too hard about it, I just get it done.  Eight sets of five feels about right?  Okay.  Five sets of three?  Okay.

After that, I do some accessory work, conditioning, grip, whatever.  I try to make sure I get out of the gym in less than an hour, which usually leaves time for a few sets of curls or rotator-cuff work and a bunch of wrist curls (believe it or not, fellow nerds, they’ll make typing all day a much less painful experience).  If I’m having a bad day, my rest periods are long, and I barely get into accessory work?  No big deal, I hit the important things.  If I’m having a good day and squatting heavy singles forever, again, no big deal, I hit the important thing.  As Matt Perryman says: the more you lift, the less bad days matter.  You can always come back the next day and hit it again.

Progression is simple.  The target for any given lift is supposed to be a weight I can hit no matter how bad I’m feeling, as long as I’m feeling good enough to go lift.  If I hit (or exceed?) that weight for four workouts in a row, I’ll add five or ten pounds.  After four workouts I should have a decent idea of how much I can increase the target, if at all.

Does it work?  I don’t know yet.  I’m not competing in anything, so I’m not paying attention to (actual or calculated) maxes on any of my lifts.  The idea behind this is to maximize “total training effect” — that if I get into the gym and lift often enough, I’m going to get stronger.  So far my bench (of all things) has been feeling strong, and I’ve set a massive PR on reps on overhead squats — which, unsurprisingly, left me floored for the next couple workouts.  Mostly, lifting this way just feels fun and comfortable, and I’m finding ways to improve my lifts incrementally week by week.

I’m not suggesting any of you go out and do the same, or that I’ve found the next big thing, or even that I wouldn’t be better off just lifting on 5/3/1.  But so far I’m pretty pleased with this, and I thought I’d give it some air time.




anarchocapitalist agitprop

Be advised

I say fuck a lot

Categories

Archives

Statistics FTW


Follow

Get every new post delivered to your Inbox.