alextp

I think the argument is slightly more convoluted, so it survives your criticisms.

Let's imagine a society where there is one thing which everybody wants (say, food) which is controlled by one big corporation (say, monsanto) with few owners. There very well can be a parallel economy between people, but for the economy to function the few owners of this big corporation must want stuff from random people just as much as people want food. The moment this small number of people control something which far more people want than they want anything you will have an infinite sink; that is, each year a certain supply of everybody's wealth will be "eaten up" by the corporation without anything flowing back in return. With technology you can conceivably have a few of these (say, Larry Page or Mark Zuckerberg can't possibly want enough things to not be effectively huge money sinks in the short term), and if you have enough of these you will have a crippled economy.

About the McLovin ID thing, I don't think it's Google's responsability to police which ids are fake and which are not. As far as I know faking an ID is a misdemeanor or a felony in most US states [1] and some other countries (like Brazil), so if there is any complaint against a reporter and he is found to have submitted a fake ID he is in for a fine and maybe jail time. I'm personally ok with requiring a user to commit a crime to cause antisocial behavior, as this ensures that they can be prosecuted in some way (this is the same principle behind DMCA takedown notices, where the person sending the letter is liable for a number of crimes if they are not who they claim to be or if they do not in fact own the right to the media at hand).

[1] http://www.celticfringe.net/teenlife/fakeid.html ) and

The "high margin" part of the quoted sentence accounts for profit (after all margin is revenue - cost, roughly).

Except with antibiotics history did not quite play out like this, in that there were far more fatal diseases before the invention of antibiotics, and some superbugs can be controlled with different variations on the same basic technology.

I think google hires a lot of low-level systems programmers, as should anyone running thousands of datacenters. After all, the power and network savings you can get from bit-twiddling, when multiplied by the number of computers in these networks, is worth quite a lot in dollars.

It's really hard to understand all this crippling branding imposed by the manufacturers and the carriers in Android phones. Do you really want people to wince every time they see your brand?

Thanks :-)

Interesting. Do you have more information on both the texas/australia and the texas/germany comparisons?

Agreed, but only up to a point. To program it is fundamental to have a strong intuition as to what should happen, why, and how to get there. Doing that without some amount of manual computation is temerary.

In all honesty, reading this makes me want the writer to actually spend some time reading grocery store counter magazines or cosmopolitan, or even watching 5 straight hours of fox news or reality shows or whatever goes completely against his values.

Truth is, no one would come back to a website that showed them uninteresting things.

Microsoft research feels a lot like this, only with a more academic bent.

Personalization seems like an attractive solution, but it is not clear if personalization in this level is even achievable. Even assuming it is, there is an inherent bias problem: if you never get to see what the computer things you are not going to like, as soon as its model of you becomes slightly wrong, you will never have the chance of correcting it and getting to see the good stuff.

Regardless, I think it is undeniable that some clever filtering is necessary to keep like people together in an environment they can both agree to enjoy and respect, however that is. Personalization can do that without a priori fixating on things like highbrow-versus-lowbrow, etc, which can easily kill a community.

The method resolution order for instances of the same class is always the same for all methods. If the class doesn't implement that method, the interpreter just skips and goes to the next place in the call chain. All super does is call the next method in the chain. So if you have class A(object), class B(A), class C(A), and class D(B, C), the MRO of D's instances will be B C A object. Each time an instance of D calls super, it will call the next method in the chain, not necessarily B's or C's or A's, even if the call to super is in B's method. This is why super() receives the class as a parameter, so it can know from when in the object's mro to start looking up the next method.

I don't think it suffers much as the method resolution order (mro) is computed ahead-of-time when creating the class.

This is a really valuable list, specially because by reading journals and going to conferences one would not think these are the major problems (as most of these are too hard to be tackled directly and yet need simple solutions that would not translate to dozens of publications), but at the same time in talking to non-researchers there is too much clouding this clear picture.

I'm trying to focus more on these problems in my research, but it is no accident that they are still unsolved.

I think the main recent ideas to tackle these are: sampling instead of doing things exactly, random projections to compress information when needed, limiting the memory footpring of existing algorithms, and controlling shared state to increase distributedness. The hardest of these problems, how to act when what you do changes the world and the data you see, desperatly needs at least some guiding principles.

I think this is the most insightful reply to the original post. Indeed, thinking about "numbers of students teached" is a really crappy metric for the performance of a teacher; it's far better, for example, to use something like (average increase in standardized test scores per student) * (number of students) / (hours taught) or somethng similar (assuming we do believe in standardized tests), the only problem with this metric being how you evaluate which teacher is responsible for which proportion of the increase in test scores, but this can be solved with proper randomization.

The idea behind this is that one should consider that a bad teacher is far more likely to fail his students (as lon as he doesn't control the exams), while a good teacher will send his students to the best univerisities/high schools/jobs availiable.

> American idol - influence of music on society, what defines popular, changes in the music industry that make idol possible, etc.

In my experience, if you bring up american idol most people will more likely say something like "god, did you look at her hair??" than something like "I wonder how the perception of fame and importance will change in 21st century society after the music industry realizes the artist brand is meaningless". You must certainly have different social circles from mine, as while it is possible to talk highly about trash pop culture, most people are completely uninterested in doing so (i.e., they turn to trash to turn their minds off, not to reflect about society, life, the universe, and everything).

> One way in which you might participate is to explore what the other person finds so fascinating about the topic. You may think its idiotic but they clearly don't, so rather than try to convince them of your point of view, find out theirs.

In my experience most people who follow the big fad of the day (royal wedding, big brother, american idol, etc) are never really fascinated by it. If you ask them "what fascinates you about american idol?", the answer you're most likely to hear is "wtf, what do you mean?". Many of these things are designed to be attention-grabbing and to encourage chattering about their consequences, even if there is absolutely nothing interesting or actually at stake in that.

In this sort of situation I either (a) go talk to actually interesting people whenever someone starts blabbing about fad-of-the-day or (b) pretend to find it interesting and listen to what the person is saying only to be made bored very soon.

The free version is important as a trial: I would never have bought instapaper if it were not for it. What I think, though, is that perhaps what most paid apps need is a way to do "shareware": allow the user to try for a day or a week, and then let apple block the app from working until the user buys it. There are a lot of apps that I would buy under this model that I don't due to fear/uncertainty/doubt (and in-app purchases for features don't really work for the same reason that I never know if those features are actually worth the cash, unless it's "pay me to stop bugging you with ads", which is really annoying).

This is a great strategy. It also trivially extends to show that there is no such thing as "too many customers, need to hire out", as you can always increase the price.

The one problem is that this doesn't transfer easily to things that are likely to be undoable no matter the resources (i.e., "I want 100% uptime on my blog", or "the project needs to get done in a month, and a saw a guy on the internet claiming he could do it in a weekend", etc).

Microsoft Research seem very evaluatable (they routinely sell their work to product tems, and things like the algorithms driving the ad network from bing came out of MSR) in terms of the bottom line, and Google's research is so integrated with the usual engineering it's hard to know which is which (and hence I doubt it fares any worse than the rest of google in terms of contributing directly to the bottom line).

> We already know the equations that we'd need to use do general intelligence, it's just that they're not computable with finite computer power, so we'd have to do some approximations, and at present it's not realistic because the approximation schemes we know of would work too slowly.

If you do know the equations, please tell me, as all models that I know of have fundamental failures that prevent them from generalizing to arbitrary problems even without considering computation.

This feels slightly untrustworthy. What if there are still external references to subfields of a struct? I think it's probably safer to just write explicit constructors (that allocate and initialize) and destructors (to free properly) for the things you use.

A possible counterpoint to this is that in some big cities you do have forms of anonymous camaraderie. I live in the largest city in south america (são paulo), and once I moved here my biggest surprise was how intimate this place could be in comparison with the much smaller places I had lived in before. While in the subway other people are complete strangers and won't receive a cursory greeting, if you do ask someone for help or information it's highly likely that you will get a very polite answer in return, no matter how weird, agressive, poor, or shabby looking you are (I've seen from bums to well-dressed business-people being treated the same way when asking for information in the subway). Also any reasonably self-contained place in this city creates familiarity--I know my butcher by name, the people from the fruit and vegetable market (there are a few in every neighborhood) know me and my wife, I talk to the owners of the nightclubs and restaurants I go to, it is perfectly acceptable to talk to strangers when lining up for the movies (but not necessarily inside the movies, although I've seen people making friendships in this way), etc. The secret seems to be that to create intimacy and trust you need to build some form of a smaller spaces where people can be identified and recognized inside the "big city". Continuing with the city analogy, as long as there are spaces where you will face consequences for misbehaving, social trust is built.

This suggests that a way to improve HN is to add more explicit punishments for bad behavior, preferably moderated.

How is your experience in the US? Is it mostly NYC, New England, and California (or big cities in other places)? From what I hear rural america (which controls a disproportionate share of the voting power) tends to be more xenophobic, specially around the border with mexico.

I agree.

I think that github is such a good tool to interact with git repos that if they made a version that can work locally (the main difference is explicitly dealing with the index and the current tree) I'd use it to manage git projects in a heartbeat.

Actually, google translate is the best academic system there is. They never lost a single nist evaluation (for an example see http://www.itl.nist.gov/iad/mig//tests/mt/2005/doc/mt05eval_... ), as far as I know, and they consistently publish intermediate results and details in their models.

If you go by your logic nobody ever writes software, as most of the building blocks were really invented by other people or adapted from other people's inventions.

And I say this as an NLP researcher.

> If we force everyone with wealth and power to live like the masses, then the only way to improve their lot would be to improve all of society or emigrate.

I think this is called comunism (as Marx intended, not as was executed in the USSR, Cuba, and any other place in practice).

The problem with this is that if you don't allow for the existence of official channels the rich can use to improve their lot then there is no point in being rich, so most people will either not bother and work as little as they can to live or find unofficial channels where they can have power to improve their lives (see the communist parties in the former USSR and China).

Of course, communism can arguably work, but how to do so is still an open problem.

Indeed. Friends of mine report downloading and watching dvdrips of movies that still have to arrive to the cinemas here in brazil. Sometimes the distribution channels go crazy for some odd reason, and you find things like sunshine cleaning (from 2008) premiering in brazil in 2011.

It is surprisingly easy to forget about poverty if you live shielded from it. You see a lot of this in brazillian urban upper-middle class, specially in the south.