294 karma · joined April 8, 2013
It is not unusual for a $50 OS upgrade to necessitate hundreds of thousands of dollars of software and hardware upgrades. So we just don't upgrade the OS.
For all the rhetoric about the need for teacher accountability, school principals and district superintendents already have the power to fire teachers. Yes, it's true: Primary and secondary school-teachers do have a form a tenure, but it's not the same kind of tenure that university professors have. In primary and secondary education, tenure simply means that there is a more formalized process for firing a teacher. More hoops to jump through, if you will. But, principals can jump through them if they really want to.
The logical question to ask is, so why don't they? They don't fire the bad ones because, simply put, they need butts in seats. Or, rather, feet in shoes in front of the blackboard. If you have 4 teachers for 100 students in 3rd grade, firing one and increasing the average class size to 33 students isn't a tenable option. As a result, the principal must hire a new teacher to replace the one he just fired, and therein lies the rub.
There simply aren't good teachers on the market at the prices that school districts are willing to pay. That qualifier at the end of the sentence is important, so read it again. Many of my cohort from grad school were fantastic teaching assistants in the classroom (many were bad, too, but we'll save that discussion for another time). That's not to say that they would automatically make good primary school teachers, but if I had kids, I would wish that those excellent TAs were in the classroom. I'll bet that many people on HN know similarly talented people from their time in university, who were natural leaders and mentors.
Unfortunately, very few (read: none) of those people who are good at teaching are employed as teachers. Why not? It has nothing to do with teacher's unions or certification requirements or institutional bureaucracy or barriers to innovation, and everything to do with the labor market. Namely, a good engineer or scientist with a masters or doctorate can out-earn a teacher by a factor of at least two. Then there's the fact that an engineer employed as an engineer isn't required to restrict his use of paid vacation days to just a few weeks each year (a couple high-school teachers with whom I'm still in contact tell me that they've never been able to take their kids on vacation for spring break because their spring break didn't line up with their kid's spring break).
So, I think the solution is easy: Pay teachers a salary that is competitive with other high-skilled professions, and you'll get good teachers.
As for Trademark... what is the trademark here, and how would anyone be confused by an MLK-Hitler poster? I don't think anyone is going to buy an MLK-Hitler poster thinking that they are getting MLK-brand doughnuts (or even MLK-Hitler brand doughnuts).
1) Transformative factor: To say that an MLK-Hitler poster "adds new expression or meaning" to the original is to make the understatement of the year. I may not like the new aesthetics, but they are certainly there. Furthermore, the posters are powerful imagery (value is added) to those who hate MLK.
2) Amount: The posters only excerpt a single frame of a much longer video.
3) Nature of original: The video is not abstract art... it documents a historically- and culturally-significant event. This gives poster-makers broad latitude to do as they please.
4) Effect on market for original: Any argument that an MLK-Hitler poster is a substitute for video of the original speech is a non-starter and doesn't pass the giggle test. You would have to argue second-order effects, (that the posters ruin MLK's image, and therefore reduce demand for all things MLK) and courts haven't really gone for that (if they did, you couldn't publish an unfavorable movie review that showed a still from the movie). If anything, it probably inspires people to try to find the original and pay the (outrageous) licensing fees.
So an MLK-Hitler poster passes the fair-use test on every single count.
That kills off a quanta of energy from whatever phonon mode gave the kick (which cools the LED), and the electron emits a photon when it relaxes back into the valence band. The LED is no longer in thermodynamic equilibrium with the surrounding environment, so heat flows into the LED.
From Wikipedia[1]: 'Japan's ambassador to Switzerland, observed that "unconditional surrender" applied only to the military and not to the government or the people, and he pleaded that it should be understood that the careful language of Potsdam appeared "to have occasioned a great deal of thought" on the part of the signatory governments—"they seem to have taken pains to save face for us on various points."'
So it seems that both Japan and the Allies understood that unconditional surrender only applied to Japan's military.
Further: 'That day, Hirohito informed the imperial family of his decision to surrender. One of his uncles, Prince Asaka, then asked whether the war would be continued if the kokutai (national polity) could not be preserved. The emperor simply replied "of course."'
In other words, Japan was willing to continue fighting, even after the atomic bombings, if it didn't get to insist on a few conditions. That doesn't sound like unconditional surrender to me.
I should also point out that the parent of my original post asserted that any and all Japanese conditions were "showstoppers." I pointed out that, whether or not the emperor's seat was a condition explicitly writing into the surrender documents, at least one of Japan's surrender conditions, namely keeping the emperor in place, was clearly not a showstopper.
But that isn't true. The seat of the emperor was maintained, at least nominally, so it clearly wasn't a showstopper. The US may have demanded unconditional surrender before the atomic bombings, but it accepted a conditional surrender after.
Actually, I think owning an EV would make me less worried about multi-day blackouts. You can charge an EV from a gas or diesel generator or solar (or wind or...), but a gas car can only run on gas.
If approximately half the workforce is between 40 and 60 and half is between 20 and 40, then half your staff are managers. That would work if only two people report to each manager (branching factor of 2), but it seems very inefficient to have so many levels of management: log_2(#staff).
On the other hand, with a branching factor of 10 only 1/9th of your workforce are managers, so management positions only support those between 55 and 60.
The point of the original comment was that you can't have "move into management" as the default career track for people in their 40's. There just aren't enough management positions.
Yes, and this happens in the real world. You'll find that when an asset is deflationary (ie, its price in USD is increasing), buyers try to delay payment so they can hold onto their dollars for longer. The asset being deflationary is exactly the same thing as the currency being inflationary (in this simply two-component system): In the future the asset is worth more and dollars are worth less, so to get the best deal you want to pay with the least valuable dollars, which are future ones.
> Another argument...
Depends on the level of deflation and your expectations for the future. If Bitcoin replaced the world's currencies, it should be worth "a few" (handwavey order of magnitude) million USD per BTC[1]. Even if it just replaced the USD it would be worth hundreds of thousands of today's USD. And that assumes that all 21 million bitcoins are available for use in the economy. It is currently trading at about one hundred USD per BTC. Are you seriously telling me that you would trade a Bitcoin for a few cases of good beer right now and give up millions in future value? If Satoshi is holding on to orginal bitcoins right now he'd be stupid to spend them not because someone could blackmail him, but because he could be a multi-trillionare in a couple decades.
Let me put this another way: If bitcoin "only" deflated at 10% per year (the number you picked for beer), it would take around a century (100 years) to deflate enough to replace the world's currencies. You'd probably be dead before it's big enough to replace just the dollar. And 10% is huge. Imagine if we had 10%/yr inflation for a century. Would that be a good thing? Do you think 10% inflation could be maintained for that long without a vicious circle causing hyperinflation?
> Third argument: even if everyone is holding their money, what's the problem?
Because this amplifies the vicious cycle. Hoarding money pulls it out of the economy, which increases deflation. That makes people want to hoard their money. Which pulls it out of the economy, which...
> The minute someone needs something, they'd have to pay for it. They would prefer to consume less if cash appreciates and consume more if cash depreciates. But there is no magical threshold after which money becomes so expensive that everyone stupidly dies of starvation.
So this is interesting: It works fine if we pretend that our amplifier is relatively linear and has a healthy amount of negative feedback and plenty of phase margin. Unfortunately, things in the real world aren't quite so simple. What happens when all the farmers decide not to plant crops (or not beyond a sustenance level) one year because each one knows that he will be better off saving his money to buy seeds next year? Futures markets can help with this a little, but those are prone to bubbles, too. By the time the price signal trickles back through the feedback loop you already have people starving. Google for the MIT beer game. Oscillations "shouldn't" happen in that system, yet they do. The cause of the oscillations in the MIT beer game is delay in the feedback network. It's directly analogous to an op-amp circuit.
So, bitcoin might work in a soviet-style, centrally planned economy, where someone with a bird's-eye view can dictate the output of various industries. Maybe bitcoiners want central planning. I don't know. But, know that I would prefer not to have that.
[1] It was a while ago, but I calculated the approximate world money supply using numbers I found on wikipedia. It's something like the equivalent value of tens of trillions of USD (trillion=1e12), and there are at most tens of millions of bitcoins (million=1e6). 1e13/1e7 ~= 1e6. This is a very very rough estimate, but it shows that the current value of bitcoin isn't even close to being in the same county as the ballpark compared to what it needs to be to reach its apparent goals (world domination). If you really think that bitcoin is going to replace USD, buy as many as you can get your hands on for whatever price anyone asks because you're buying them for fractions of a penny on the dollar. (note: this is not actual investment advice and I am not an investment adviser)
Please don't waste your time typing the usual nonsense about how bitcoin can be divided into infinitesimally-small parts. Nobody is (or has ever been) seriously worried about that. It would be as if saying the primary problem with hyperinflation is that you need to put too many zeros on your bank notes. Imagine if someone asked about the feasibility of fiat currencies and the response was "No, we've planned for that: There will be extra space on the end of the bank notes to print more zeros." Absurd! Yet, that's the standard argument from bitcoin proponents when it comes to deflation (only in reverse).
The problem with bitcoin is that its designers believed that hyperinflation=bad necessarily implies that hyperdeflation=good. If that were the case, then bitcoin would be the perfect currency. Actually, both are fatal to an economy. Economies basically shut down when currencies start to deflate. It turns into a giant game of chicken: Nobody wants to spend his currency first because he knows it will be worth more later. First people delay luxury purchases, but before long you end up with a vicious circle and people wait until the very last second to purchase essentials like life-saving medicine. There is a way to stop this, of course: Print more money (Hey, that's kinda like how the solution to hyperinflation is to stop printing more money! Who'da thunk it?).
>Nation States don't want this, as it restricts their expansion.
Or we can take off our tin-foil hats and admit that Bitcoin sucks and was designed by someone with little understanding of economics or the history of money. In order to be a useful currency, the total number of bitcoins needs to track the size of the portion of the economy that uses bitcoins. Bring me a cryptocurrency that was designed by someone who is more interested in making a solid, practical currency than rigidly following his ideology, then we can talk. Nobody in power is in any way threatened by bitcoin. It will implode by itself without any help from the outside (though I'm sure it will get some helpful nudges in that direction).
I'll close with this: Part of engineering a good system is making prototypes and seeing what doesn't work. Another part is looking at competing products and seeing where other engineers went wrong. Blind, religious adherence to the principle of deflation is only going to cripple the adoption of alternate currencies.
First, human trafficking is illegal, but (apparently) very profitable. It was an unstated assumption that those people who engage in these sorts of activities do so primarily out of a motive for profit. In other words, that human trafficking could be largely eliminated by making it unprofitable.
In order for traffickers to enjoy the considerable fruits of their (slaves') labors, they need to invent a plausible-seeming and, more importantly, legal source for their wealth. This is called money-laundering. To do otherwise would invite the attention of the local constabulary.
What the GP is suggesting is that a viable means to curtailing human trafficking is to make money laundering difficult. This makes it difficult to use the profits of the crime, which reduces the incentive to commit the crime in the first place.
The basic idea is a form of synchronous detection or lock-in amplification: The shutters are arranged so that a point source is occluded at a fixed frequency, while an area source will be only partially occluded at any given time. Due to the way it is designed, area sources should add little to the time-varying component of the photodetector signal. The lock-in amplifier is essentially a very narrow notch filter (centered at the "star occlusion frequency"), so it completely rejects the DC component from the sky.
There is a little more to it than what I'm writing, mainly because you need to not just detect the point source, but also estimate the star's position within the field of view (which IIRC is done by comparing the phase of the time-varying star signal to a reference point on the shutters).
STEM graduates (both undergrad and grad) wake up the day after graduation and start looking for a job. What do they find? Well, if they are smart (and, they probably are, because they just graduated with a STEM degree) and have one or two other skills outside the STEM degree, they find that there are people who are willing to pay them an incredible amount of money to do "not-science" and a middling amount of money to do "is-science." There are actually plenty of US-born STEM graduates created each year who leave their field to do something more lucrative.
Take a look as these statistics from the American Physical Society: http://www.aps.org/careers/statistics/index.cfm
Almost a third of people graduating with a Bachelor's in Physics work in some field that is unrelated to science, technology, math or engineering! In other words, by just paying market-clearing salaries, we could increase the number of physics graduates entering the workforce each year by a whopping 50%! That could happen tomorrow, not twenty years from now when today's kindergarteners finally get through the pipeline.
Pushing more students into a STEM track won't reduce the number of unfilled STEM positions: It will reduce the average salary of the STEM graduates who leave for other fields. With enough pushing, you might, maybe, one day push the salary low enough that graduates remain in STEM. Or, you could just increase the salary of those unfilled STEM positions.
So then what? Japanese-Americans give up all their possessions and live the hills? Where would they get food, guns and ammunition? After the war, there would be no way for them to reintegrate into society.
As for the Black Panthers, I would argue that images of them carrying weapons in the media did more harm than good to their cause. Today few people remember them as a community service organization and most remember them as a sort of quasi-terrorist death squad. A recent presidential candidate was even intimated to have had some sort of nebulous association with them as a way of tarnishing his reputation (though I hear he won anyway).
So I'm underwhelmed by arguments that guns are going to stop a tyrannical government. They may stop one that is completely inept, but a government with even modest propaganda resources will have no trouble playing one group of citizens against the other. Firearms enthusiasts aren't going to be portrayed as 'freedom fighters' by the media: They will be called 'terrorists' or 'cop killers.'
Not that I'm against gun ownership (though I don't personally own one). I just don't live under the misapprehension that it's going to magically 'solve government.'
No, but it's not because they aren't good enough. In grad school, I TA'd a physics class for pre-med students, which were widely considered to be one of the most miserable and obnoxious groups to teach. Each student felt entitled to an A, even though the department mandated that only 25% of students could receive an A. In reality, probably more than 80% of the class earned an A, and I would have felt comfortable that most of the rest were smart enough to one day give me medical advice.
Yet, less than 45% of people who applied to one or more medical schools [1] are admitted (and this is a self-selected group: those without 4.0 GPAs, three extra-curriculars, and a side-business of helping old ladies across the street generally don't apply). I would estimate that only about the top 15% to 25% of the class actually applied. So, from undergrad pre-med to MD, there is about a (hand-wavy) 90% attrition rate of people who are probably smart enough and probably willing to work hard enough, and who have the desire to become doctors.
> Increasing the number of doctors produced locally would seem to me to be a much easier way of increasing supply.
I agree, but the medical profession is not willing to allow this to happen. Each medical school has a quota, and nobody wants to rock the boat. The trouble is, there isn't really any way to force medical schools to admit more students.
[1]http://en.wikipedia.org/wiki/Medical_school_in_the_United_St...
Edit: steel works -> steelworkers