17 February 2021

Declaration of allegiance

 All the cool kids are at substack these days.  So I've subscribed to Matt Yglesias and am following Noah Smith and (could you doubt it? would you believe me if I said I wasn't?) Scott Alexander *******.  I also still read Tyler C and Alex T's joint blog.  On recommendations from the above, I'm reading Zvi Mowshowitz, Jason Crawford, Susanna Viljanen, and Derek Lowe sporadically.  OK, nobody had to recommend Susanna to me.  So if I write something you haven't already heard eight million times, check to see which of the above can have inspired it.

I remain a proud and loyal subject of Her Majesty.   Not that anyone who knows me is wondering about that.

14 March 2020

No, It Is Not

Let's get this straight. Today is not π day. That day does not arrive for another 1021 years, two months, and -5 days.

Oh, and there was a rehearsal for it 1706 years, two months, and eight days ago. Sorry I missed it.

30 July 2017

Why We Can't Fix Rush Hour

I live in Mountain View, as I have done since before Google was founded. Even in the 1980s it was not much fun to drive on the Bay Area's freeways during rush hour; today, the delays are almost legendary.  A few naive youngsters still have the courage to ask why local (or state, or federal) governments don't Do Something About It. Well, they've opened a new stretch of freeway (I can remember when most of SR 85 wasn't there) and widened an old one, but commuting still sucks. All right, let's look at the causes of the trouble.

Rush hour occurs because many people start and end work at roughly the same time, but what makes it so horrible is that they live far away from where they work, and spend a long time getting there. Thus the people who leave home at 7:30 are still on the road when the people who leave at 8:00 start driving, and the road gets crowded, which makes it even slower. Come 8:15, the 7:30 people still haven't reached work, and more cars are pouring onto the roads. We have a positive feedback loop here. Now, a positive feedback loop can work both ways. Many years ago, I used to work within a mile of where I lived. I would walk (yes, walk) the whole way in ten minutes, and arrive in a pretty good mood. Imagine if most people could get to work in ten minutes! The people who used to leave home at 7:30 "to beat rush hour" could wait until 7:45, and would still be able to vacate the road in time for the 8:00 people to have full use of it. If commutes were short, we could get a lot more people to work, over the same amount of road.
Trouble is, most people don't want to live that close to work because they can get a bigger place five miles further away. Or a yet bigger place ten miles away. Or an even bigger place twenty miles away, with a real back yard. Mind you, this means that after eight hours at work, eight hours for sleep and personal hygiene, two hours commuting, and some yard work besides, they don't actually have much time to enjoy the bigger place (and are too tired and grumpy to do so after driving home), but hey, it's the American Dream, it's going to  make them rich, and they get a tax break on the mortgage. So when we build more roads and widen a few existing ones, people decide they can now live thirty miles away in a really spacious place, which they have even less time to enjoy (and more yard work to do).
What we have here is a negative feedback loop. We can build roads, or paint carpool lanes, or do anything that makes for shorter commutes, and after a year or two there will be an increase in land values in the places that used to be out of reach for commuting to the Googleplex. Some people will have decided that they can now afford that spacious house, and have not really reckoned with the cost of the extra commute in money, time, and temper.
What else is there to do? I don't imagine that more meetings will help much. Driverless cars may well increase the effective capacity of freeways, but such an increase will quickly fall victim to the negative feedback loop, pushing property values up until commutes are, once again, just barely bearable. The same applies to mass transit: a bus or a train is, from the commuter's point of view, a very large self-driving car. Ask people to take jobs closer to where they live? Good luck with that! Having the best possible job is important. (I must admit to commuting twenty miles. I do value my job.) Telepresence? It works for some jobs, some of the time; not enough, not yet.
One fix that is gaining some traction is denser housing, which I can see being built in many places. Simple geometry permits people to live closer to work (or to mass transit) if they live closer to each other, and perhaps they'll even get rich out of it. In fact, if rush hour gets worse, it would be hard not to get rich out of owning a place that allows of a short commute.
I still live in the house we bought twenty years ago, but I haven't traded up to a place with a spacious back yard. My kids don't need a back yard to play in: they have YouTube. Telepresence, you might call it.
[This is adapted from something I wrote on Quora a year or two ago.]

07 December 2016

Is the USA short of SWEs?

Reposted from Quora, to be sure.  Lightly edited.


Is there actually a STEM shortage that requires foreign workers? If so why are Americans lacking in these fields?

TL;DR: the industry that employs these damn’ foreigners also provides some of the fastest progress and best value in the USA.

Many excellent Americans, including a girlfriend of mine, a crush of mine, and others I’ve known, have gone into the legal and medical professions even though they mostly had the brains to go into STEM. Why? Those two professions have high barriers to entry, reinforced in the USA by government-mandated licencing that gives foreign credentials much less weight than domestic. This boosts the incomes of Americans in those professions artificially, for which they express their gratitude to their lawmakers every election season in the customary manner $$$.

The only STEM field I know well, software, has no such barriers or licences that matter, and will accept anyone who can demonstrate the needed ability. I’ve worked alongside engineers from Australia, Brazil, Canada, Denmark, Egypt, Estonia, France, Germany, Greece, Hungary, Iceland, India, Indonesia, Ireland, Israel, Japan, Korea, Latvia, Lebanon, Mexico, … Turkey, Ukraine, Vietnam, but nobody from Western Sahara, Yemen, nor Zimbabwe. (I myself am from England.) Do I need to spell it out? Oh: I just did.

I expect even the Canadian would have found it tough to qualify as an attorney in most states of the US. Almost for sure, the others would have found it very tough. So my ex-girlfriend and others command famine prices for their talent, and the famine is caused, just like real famines where people starve to death, by government policy. Does anyone care to suggest that the USA spends too little on medical care and lawsuits?

Now, if you think Google or Amazon or Apple or Facebook are charging famine prices for the software they provide, or that American software is less good than what other countries build, please shout it from the rooftops. The only criticism I hear (living in Silicon Valley) is that these giants pay their engineers too much and thereby drive house prices up, causing gentrification etc. Certainly house prices have climbed. Don’t tell me that people who buy a three bedroom house for a million bucks are underpaid.  But these foreigners are paying such prices, out of their income from work.


Footnote: if you want to shout from the rooftops about privacy on the Web, you have every right to do so, but the aforementioned companies would not offer you any better privacy if they were forced to hire fewer engineers at higher pay.

08 November 2016

27 August 2016

battleships: a widespread misconception

Does this look as if I'm reposting another answer that I wrote for quora.com? It should. Enjoy.

Why has the concept of battleships/pocket battleships been wiped out of modern military warfare?
The vulnerability of unescorted battleships to air attack is not the reason. A battleship could be part of a carrier group, protected by the carrier's air umbrella, and be at no more risk than the carrier itself, besides being a smaller target. (Test run of this argument: the advent of torpedoes did not make battleships obsolete, but did mean that they needed a screen of destroyers.)
The true reason is that big guns are no longer the longest-range weapons afloat, nor do they deliver the heaviest warheads. Since the 1940s, naval battles are fought at ranges of hundreds of km using aircraft and guided missiles, or if at close range, stealthily using torpedoes and depth charges. I hope I don't need to elaborate on why it's hard to fire a 16-inch shell stealthily from underwater. More importantly, the long ranges of aircraft and missiles have made 16-inch shells as irrelevant as pikes and cutlasses became during the 19th century. If an opposing ship is close enough to you that you can use these weapons, the fight is already over. As for thick armour, conventional bombs can be built far bigger and deadlier than a 16-inch shell, so a ship that could "just sit there and take it" would have to be about as big as a shopping mall or football stadium, and would not be much more manoeuvrable. The best defence (for a warship) really is attack, or sometimes escape.

18 March 2016

The theism question

I posted this some weeks ago on Quora, but thought a wider audience might care to see it. I've edited some material that isn't very relevant, and fooled with the punctuation.


I did not and do not think that theism is necessarily irrational. It follows rationally enough, from premises that I regard as made-up.
<digression>
Please bear with me. I owe this observation to a perceptive former student of my father's.
The private schools in England, where I was educated, had spent the 19th century, and probably the 18th, churning out empire-builders: teaching boys to be the military officers and civilian administrators who would turn "Wider still and wider // Shall thy bounds be set" from poem into reality. After 1947, such people were no longer required, but high technology certainly was, as was culture, so the schools set themselves a new task: to churn out creative and innovative individuals. Eccentricity had long been tolerated, but now it was to be encouraged. Traditions and ancient assumptions were still important, but they could be questioned. And the questions deserved thoughtful answers.
</digression>
As a young boy, I knew I had to look both ways before crossing the street, and I knew why. I knew that it was customary, when eating, to hold the knife in the right hand, because that was the dominant hand, but since I was left-handed, I often held it in my left hand. I knew that the school began each day with a hymn and a prayer, then a few announcements by the head teacher. I didn't need to ask why the announcements were made. And I never got a serious answer about why the hymn was sung and the prayer whispered. Tradition, to be sure, and pretty-sounding tunes, but that was all. I tried to sing in tune.
When I got older and could ask better questions, there was the same shortage of answers ... of thoughtful answers. God the creator? Sorry, we more or less understand how the Earth formed from a disc of gas and dust. God the lawgiver? The school makes the rules for us boys, and Parliament makes them for grown-ups. Redeemer? My parents love me; I'm not a Sinner with a capital S, just occasionally naughty. Celestial cop? not when the other boys are getting away with X, Y, and Z every day. And so on.
I don't ask anyone to believe that my thinking on this subject was watertight, or that it is so today. But a massive weight of evidence pointed, in my mind, to theism being an excuse to sing pretty songs and occasionally threaten people with divine vengeance for doing things that, for the most part, weren't especially evil. Oh, and give comfort to the insecure by telling them that, yes, in the end the good guys will win. But my limited knowledge of history already had shown that the only way to make the good guys always win was to use a flexible definition of "good". When I was a boy, we were deep in the Cold War, whose least unlikely outcome seemed to be that a modus vivendi would solidify over generations whereby the USSR didn't actually threaten us, nor we them. There did not seem to be much that gods could do to help. Certainly they had not yet done so. (I hadn't read or heard of Judges 1:19.)
Summary: theism as a hypothesis could account for only a few things, which already had more detailed secular explanations.
There also seem[ed] to be real advantages accruing to people who "had no need of that hypothesis". If you used physics instead of Genesis 1, you not only knew why the Earth and Sun were here, but you could build aircraft. If Parliament made the laws, it could change them in response to new developments (when I was born, England had only the feeblest of laws against drunken driving, and none against advertising tobacco products. Just to be clear, on principle I believe that if a product can legally be sold, it should also legally be advertised, but that's a separate argument.) As for the good guys winning, we beat Hitler with a little help from the USA, we beat Napoleon with a little help from Russia, and the USSR could be contained if not beaten. Except in Vietnam. Yes, I supported that war, like the parents in The Free Electric Band Lyrics .  Being an atheist did not make me all-wise, nor does it now.
One more thing. My schools emphasised Latin and Greek, and those subjects brought classical mythology in their wake, which was fun. I was perhaps better acquainted with the Twelve Olympians than with the Three in One And One In Three. It did not escape my attention that the gods of Homer, with their all-too-human frailties, were now deemed worthy of a place on the fiction shelves. Clearly it had been possible for intelligent, cultured, civilised, technological people to sincerely worship deities that existed only in their imaginations. Equally clearly, people were much the same two or three millennia later as they had been in those times. QED.