Re:Enter the closed loop you cannot enter.
on
The Limits To Skepticism
·
· Score: 4, Informative
While labeled flamebait, this is something of a problem, even in less politicized fields of science. Most scientists are earnest truth-seekers, but a minority are not, and the peer-review system is not always robust to them. I work in an area of computer science that will never make Fox News, but even in this area things are sometimes suppressed for what's hard to describe as other than political reasons. At the very least, politically unpopular positions get all sorts of extra hoops to jump through that others don't--- e.g. if you're casting doubt on a position the journal editor or one of his friends staked his career on, better expect some random made-up requirements. If your paper scoops a large and well-funded group's work, there's a chance it'll be rejected by one of their friends, so they get to publication first--- and their publication might coincidentally borrow a few ideas or theorems from your rejected paper.
It's not all bad, and in fact most is probably good. But there are some very rotten parts of the scientific-publishing apparatus. It doesn't help that most journals are run by for-profit companies that are a bit shady themselves (Kluwer, Springer, etc.) who have no real interest in the quality of the science they publish or how to improve it. And it doubly doesn't help that the academic rat-race has gotten increasingly cut-throat, so people feel they need to resort to dirty tricks to get/keep a job, get tenure, get grants, etc.
Yeah, I agree it already happens, but I think there can be a qualitative difference if it gets hugely pervasive. In the case of free gadgets being sent out, you can at least plausibly still identify a "normal market", where people buy things based on price signals and preferences. It might be distorted somewhat by the reviews the free-gadget-recipients produce, but that's not really in principle different from any other disinformation that distorts markets; just another variety of misleading advertising.
But if the pricing variance is built into the market itself, it seems quite a bit worse to me. You no longer have a normal market with people trying to distort it, but no real market at all anymore.
Approaches like this are pretty direct attacks on why free markets work. Almost all classical and neoclassical economic theory assume things like the existence of a supply/demand price curve, availability of pricing information, etc. If you have some nutty system where price curves aren't really defined beyond an individual level, prices aren't widely available, etc., all the usual pricing signals, resource allocation by the "invisible hand", etc., get a lot more muddled, and probably begin to break down.
Of course, that's certainly a reason I can see Microsoft wanting it: finding ways to profit other than "make a good product and compete fairly on the open market" is their modus operandi.
I've gotten cash before, though usually small amounts. Back in 2004 I got a check for $9 out of some sort of music-CD price-fixing settlement.
This particular case appears not to be a class-action suit at all, though; it's a criminal investigation that imposed a fine. So there is no settlement to distribute, since it's not a civil suit with plaintiffs.
While true, does contemporary print media actually do that much? Flipping through my local paper, I don't see much of value in the illustrations. A few graphs, a few photos of politicians, a few photos of sports games. The graphs and infographics could be rendered fine in black-and-white, anyway, at least if the paper's got a competent graphic designer.
While ISPs do pay by bandwidth (though large ones just peer directly with no exchange of money involved), I'd suspect their bandwidth usage, which is the aggregate of their customers' bandwidth usage at any given time, is better predicted by customers' data usage in MB, than by the size of customers' local bandwidth channels. If a bunch of your customers on 512 KB/s links start using their phones twice as much, that's going to have a bigger impact on the ISP's bandwidth needs than if a bunch of your customers with 512 KB/s phones upgrade to 1024 KB/s phones.
(There are some cases where that wouldn't be true, like if usage is super-peaky with nearly everyone doing data transfer at the same time of day, but I suspect that's not the usual case.)
As a technical reason, I'm not sure why charging based on the bandwidth is superior, if you know that the vast majority of customers don't max out the connection most of the time. Charging by usage seems a little closer to capturing the proportion of resources a customer uses in that case.
There are other downsides to it, but they seem mostly like social ones, not technical ones. For example, people don't like feeling like they're being metered, and it has a chilling effect on a lot of online services if people have to worry about their bandwidth usage.
Clearly MySpace+Imeem doesn't form a monopoly in the area, so I can see why there'd be no reason to block the purchase. But don't companies that purchase their competitors have to at least pretend to be doing it for some reason other than simply to shut down and thereby get rid of a competitor? If the sole reason for buying a competitor is to get rid of them, isn't that roughly equivalent to paying them to leave a particular market, which would be illegal?
Under most U.S. states' rules for class actions, the Representative Plaintiffs (those who initiated the action and retained the attorneys) can be given something additional as a reward for initiating the case.
That's the entire theory of class-action lawsuits. Normal lawsuits are opt-in: the plaintiffs' attorneys can only claim to represent any plaintiffs that have explicitly retained them as representation in the case. Other plaintiffs therefore retain the right to sue separately, but unless they actually do so, they are not represented in the lawsuit. In class-action lawsuits, plaintiffs' attorneys can ask to have a class certified, and they will be taken to represent all members of the class, except insofar as members of the class opt out.
In theory at least, this has some pros for both sides (in addition to cons, of course). For consumers, the pro is that harm that would not be worth suing for individually will still be remedied, because it's worth suing for in aggregate: so a hypothetical evil corporation that did something illegal to cause $10 of damage to every person in the United States won't get away with it. For companies, the pro is that settling a class-action lawsuit settles it for all members of the class simultaneously, except those who explicitly opt out, so they can get the complaint over with once and for all.
Sure, if you were independently wealthy, you would probably not work for a company, but would find whatever it was you wanted to do, and do it on your own. But I don't see a reason it has to be a pure binary: independently wealthy people do whatever they want, everyone else works shit jobs just to get paid. If you have no options, sure; but many people have a range of choices of jobs, some of which they prefer more than others, and I don't see any reason to weight money above job enjoyment.
My dad managed to hang on at the top of the engineering ladder at a major oil/chemicals company for about 20 years after the first attempt to promote him, resisting an attempted promotion into the managerial ranks about every 2-4 years. A lot of companies, especially old-style companies, are set up with the assumption that everyone wants to climb out of the "working" ranks into the "management" ranks if they can, perhaps because that was more true when the working ranks involved more physical labor. It got a little easier to "stick" at his desired place when someone managed to dig up some sort of super-senior-engineer ranking that was rarely used, which let them give him a promotion without the usual promotion to management.
If the lower levels of management is okay with it, it can work, and they might even like it. Engineers who "should" be in management are essentially experienced enough to manage themselves, and maybe even de-facto manage a few of othe other team members, which can make the manager look good by making it easier for them to pretend they know what's going on--- at large companies, the lower level of management right above the engineers are often people who rotate in/out of jobs every 5 years or so, usually on a quest to move up the ranks to VP, so they honestly rarely have much idea what's going on or any historical perspective/experience.
There's a little bit of that already, in the form of rerouting traffic and instituting parking/stopping restrictions to try to mitigate the risk of a car bomb.
I might be missing something, but that doesn't sound particularly suspicious to me. Someone who owned a stake of Craigslist sold it; isn't that basically how stakes in companies work?
It does happen in some industries, and the law on it is pretty complex and murky (and varies by country). When it's below a certain threshold, so the minority stake doesn't exercise control over the company, and has representation basically only to ensure its rights as a minority shareholder are respected, it's considered a "passive investment" and not subject to the usual antitrust scrutiny that would ensue if, say, eBay actually tried to buy Craigslist (or buy a stake considered controlling). A lot of economists are a bit skeptical of just how passive such passive investments are, though. The keywords +"passive investment" +competitors bring up a whole pile of writing on the subject...
The biggest problem I've found with public speaking is that often the speaker really does not want to be giving a talk, and so it's kind of forced. Sure, they may be interesting people with a lot to say, but for this 30-minute, or 60-minute, or whatever it is slot, they've been assigned to do it, or are doing it for money, or are doing it for some other reason besides, "really have something I want to say to this audience".
I see this a lot as an academic--- in computer science, conferences have in a lot of areas displaced journals as the primary publication venues, so people mainly submit to a conference to get their paper published in the proceedings. Then the conference rolls around, and they have a 20- or 30-minute slot to fill. Some people really have something to say in that slot, but others, whether through inclination or the subject matter, really don't, and give essentially a very long and drawn-out version of, "just read the damn paper".
Of course, some technical improvements to talks---use of figures, non-monotone voice, etc.---can make them more engaging, but it doesn't really solve the underlying problem. In the academic-conference case, it basically transforms a boring 20-minute ad for the paper into a somewhat amusing 20-minute ad for the paper--- but still not a good talk.
A common view on Slashdot (though certainly not held by everyone) is less a position on copyrights than a position on DIY hackability: that it should be 1) possible; and 2) legal.
From that perspective, the GPL's use of copyright law is good, because it forces companies to both release their code and permit derivative works of it. Other uses of copyright law might be bad, because they attempt to make DIY modifications illegal: either prohibiting sharing your modifications (conventional copyright law) or prohibiting you from even making them at all (DMCA reverse-engineering restrictions). And of course there are non-copyright impediments to DIY modifications as well, like DRM and simple obfuscation, all of which are also sort of unpopular on Slashdot.
Looking at a historical example, lets take a war. Men will want to know HOW an invasion was performed, what sort of strategy was used, the units involved and the outcomes. Women would be much more interested in what drove the country to invade, how it impacted the country being invaded.
It's a plausible hypothesis, but what strikes me as odd is that the male-dominated fields that study wars in the modern era mainly focus on the 2nd, which you consider a female interest. I see it as more of an era thing: 19th-century historians of war mostly focused on strategy, generals, treaties, etc., while modern historians are much more likely to write books on motivations, cultural factors, impacts on civilian experience, etc. Taking that approach even further, folks like Bourdieu and Foucault were male as well..,
While I agree Ada Lovelace and Grace Hopper are important pioneers, I'm not sure it's a particularly large number--- you've right there, with your 2-entry list, basically exhausted the list of prominent female pioneers of geekdom, while the equivalent male list (Alan Turing, Steve Wozniak, Donald Knuth, Edsger Dijkstra,...) goes on for a while longer. Sure, there a few others once you move down the list into "famous among specialists"--- Radia Perlman (inventor of the spanning-tree protocol), say. But you can't even put together a list of 20 without digging pretty deep.
More than that--- Perl 6 was announced 9 1/2 years ago! Even O'Reilly's Perl 6 Essentials is now 6 1/2 years old, and some Perl 6 books are into 2nd editions.
Yahoo does have some built-in market share that'll let them struggle on for a while as long as their search isn't terrible, which is what they might be hoping for from the Bing backend. They have huge installed bases of users of lots of their other services (email, mailing lists, now Flickr), and the Yahoo search bar shows up at the top of the page for some of those services (though not, at least for now, on Flickr)--- check out the top of the Yahoo Groups page, for example. If they can keep those services all steady with reasonably large user bases, they'll have a decent stream of search users even if nobody purposely goes to search.yahoo.com.
Slashdot Mirror
While labeled flamebait, this is something of a problem, even in less politicized fields of science. Most scientists are earnest truth-seekers, but a minority are not, and the peer-review system is not always robust to them. I work in an area of computer science that will never make Fox News, but even in this area things are sometimes suppressed for what's hard to describe as other than political reasons. At the very least, politically unpopular positions get all sorts of extra hoops to jump through that others don't--- e.g. if you're casting doubt on a position the journal editor or one of his friends staked his career on, better expect some random made-up requirements. If your paper scoops a large and well-funded group's work, there's a chance it'll be rejected by one of their friends, so they get to publication first--- and their publication might coincidentally borrow a few ideas or theorems from your rejected paper.
It's not all bad, and in fact most is probably good. But there are some very rotten parts of the scientific-publishing apparatus. It doesn't help that most journals are run by for-profit companies that are a bit shady themselves (Kluwer, Springer, etc.) who have no real interest in the quality of the science they publish or how to improve it. And it doubly doesn't help that the academic rat-race has gotten increasingly cut-throat, so people feel they need to resort to dirty tricks to get/keep a job, get tenure, get grants, etc.
Yeah, I agree it already happens, but I think there can be a qualitative difference if it gets hugely pervasive. In the case of free gadgets being sent out, you can at least plausibly still identify a "normal market", where people buy things based on price signals and preferences. It might be distorted somewhat by the reviews the free-gadget-recipients produce, but that's not really in principle different from any other disinformation that distorts markets; just another variety of misleading advertising.
But if the pricing variance is built into the market itself, it seems quite a bit worse to me. You no longer have a normal market with people trying to distort it, but no real market at all anymore.
Approaches like this are pretty direct attacks on why free markets work. Almost all classical and neoclassical economic theory assume things like the existence of a supply/demand price curve, availability of pricing information, etc. If you have some nutty system where price curves aren't really defined beyond an individual level, prices aren't widely available, etc., all the usual pricing signals, resource allocation by the "invisible hand", etc., get a lot more muddled, and probably begin to break down.
Of course, that's certainly a reason I can see Microsoft wanting it: finding ways to profit other than "make a good product and compete fairly on the open market" is their modus operandi.
I've gotten cash before, though usually small amounts. Back in 2004 I got a check for $9 out of some sort of music-CD price-fixing settlement.
This particular case appears not to be a class-action suit at all, though; it's a criminal investigation that imposed a fine. So there is no settlement to distribute, since it's not a civil suit with plaintiffs.
While true, does contemporary print media actually do that much? Flipping through my local paper, I don't see much of value in the illustrations. A few graphs, a few photos of politicians, a few photos of sports games. The graphs and infographics could be rendered fine in black-and-white, anyway, at least if the paper's got a competent graphic designer.
While ISPs do pay by bandwidth (though large ones just peer directly with no exchange of money involved), I'd suspect their bandwidth usage, which is the aggregate of their customers' bandwidth usage at any given time, is better predicted by customers' data usage in MB, than by the size of customers' local bandwidth channels. If a bunch of your customers on 512 KB/s links start using their phones twice as much, that's going to have a bigger impact on the ISP's bandwidth needs than if a bunch of your customers with 512 KB/s phones upgrade to 1024 KB/s phones.
(There are some cases where that wouldn't be true, like if usage is super-peaky with nearly everyone doing data transfer at the same time of day, but I suspect that's not the usual case.)
As a technical reason, I'm not sure why charging based on the bandwidth is superior, if you know that the vast majority of customers don't max out the connection most of the time. Charging by usage seems a little closer to capturing the proportion of resources a customer uses in that case.
There are other downsides to it, but they seem mostly like social ones, not technical ones. For example, people don't like feeling like they're being metered, and it has a chilling effect on a lot of online services if people have to worry about their bandwidth usage.
in a daze 'cause i found juice
Clearly MySpace+Imeem doesn't form a monopoly in the area, so I can see why there'd be no reason to block the purchase. But don't companies that purchase their competitors have to at least pretend to be doing it for some reason other than simply to shut down and thereby get rid of a competitor? If the sole reason for buying a competitor is to get rid of them, isn't that roughly equivalent to paying them to leave a particular market, which would be illegal?
Under most U.S. states' rules for class actions, the Representative Plaintiffs (those who initiated the action and retained the attorneys) can be given something additional as a reward for initiating the case.
It doesn't look like there's any settlement money for normal Facebook users.
The settlement provides that the $9.5 million will be spent on:
That's the entire theory of class-action lawsuits. Normal lawsuits are opt-in: the plaintiffs' attorneys can only claim to represent any plaintiffs that have explicitly retained them as representation in the case. Other plaintiffs therefore retain the right to sue separately, but unless they actually do so, they are not represented in the lawsuit. In class-action lawsuits, plaintiffs' attorneys can ask to have a class certified, and they will be taken to represent all members of the class, except insofar as members of the class opt out.
In theory at least, this has some pros for both sides (in addition to cons, of course). For consumers, the pro is that harm that would not be worth suing for individually will still be remedied, because it's worth suing for in aggregate: so a hypothetical evil corporation that did something illegal to cause $10 of damage to every person in the United States won't get away with it. For companies, the pro is that settling a class-action lawsuit settles it for all members of the class simultaneously, except those who explicitly opt out, so they can get the complaint over with once and for all.
Sure, if you were independently wealthy, you would probably not work for a company, but would find whatever it was you wanted to do, and do it on your own. But I don't see a reason it has to be a pure binary: independently wealthy people do whatever they want, everyone else works shit jobs just to get paid. If you have no options, sure; but many people have a range of choices of jobs, some of which they prefer more than others, and I don't see any reason to weight money above job enjoyment.
My dad managed to hang on at the top of the engineering ladder at a major oil/chemicals company for about 20 years after the first attempt to promote him, resisting an attempted promotion into the managerial ranks about every 2-4 years. A lot of companies, especially old-style companies, are set up with the assumption that everyone wants to climb out of the "working" ranks into the "management" ranks if they can, perhaps because that was more true when the working ranks involved more physical labor. It got a little easier to "stick" at his desired place when someone managed to dig up some sort of super-senior-engineer ranking that was rarely used, which let them give him a promotion without the usual promotion to management.
If the lower levels of management is okay with it, it can work, and they might even like it. Engineers who "should" be in management are essentially experienced enough to manage themselves, and maybe even de-facto manage a few of othe other team members, which can make the manager look good by making it easier for them to pretend they know what's going on--- at large companies, the lower level of management right above the engineers are often people who rotate in/out of jobs every 5 years or so, usually on a quest to move up the ranks to VP, so they honestly rarely have much idea what's going on or any historical perspective/experience.
There's a little bit of that already, in the form of rerouting traffic and instituting parking/stopping restrictions to try to mitigate the risk of a car bomb.
No doubt so the Elders of Zion (who live in the clouds) can read your data. ;-)
I might be missing something, but that doesn't sound particularly suspicious to me. Someone who owned a stake of Craigslist sold it; isn't that basically how stakes in companies work?
I think Kijiji is more popular than Craigslist in Canada in general, while Craigslist tends to dominate U.S. markets. Not sure why.
It does happen in some industries, and the law on it is pretty complex and murky (and varies by country). When it's below a certain threshold, so the minority stake doesn't exercise control over the company, and has representation basically only to ensure its rights as a minority shareholder are respected, it's considered a "passive investment" and not subject to the usual antitrust scrutiny that would ensue if, say, eBay actually tried to buy Craigslist (or buy a stake considered controlling). A lot of economists are a bit skeptical of just how passive such passive investments are, though. The keywords +"passive investment" +competitors bring up a whole pile of writing on the subject...
The biggest problem I've found with public speaking is that often the speaker really does not want to be giving a talk, and so it's kind of forced. Sure, they may be interesting people with a lot to say, but for this 30-minute, or 60-minute, or whatever it is slot, they've been assigned to do it, or are doing it for money, or are doing it for some other reason besides, "really have something I want to say to this audience".
I see this a lot as an academic--- in computer science, conferences have in a lot of areas displaced journals as the primary publication venues, so people mainly submit to a conference to get their paper published in the proceedings. Then the conference rolls around, and they have a 20- or 30-minute slot to fill. Some people really have something to say in that slot, but others, whether through inclination or the subject matter, really don't, and give essentially a very long and drawn-out version of, "just read the damn paper".
Of course, some technical improvements to talks---use of figures, non-monotone voice, etc.---can make them more engaging, but it doesn't really solve the underlying problem. In the academic-conference case, it basically transforms a boring 20-minute ad for the paper into a somewhat amusing 20-minute ad for the paper--- but still not a good talk.
A common view on Slashdot (though certainly not held by everyone) is less a position on copyrights than a position on DIY hackability: that it should be 1) possible; and 2) legal.
From that perspective, the GPL's use of copyright law is good, because it forces companies to both release their code and permit derivative works of it. Other uses of copyright law might be bad, because they attempt to make DIY modifications illegal: either prohibiting sharing your modifications (conventional copyright law) or prohibiting you from even making them at all (DMCA reverse-engineering restrictions). And of course there are non-copyright impediments to DIY modifications as well, like DRM and simple obfuscation, all of which are also sort of unpopular on Slashdot.
It's a plausible hypothesis, but what strikes me as odd is that the male-dominated fields that study wars in the modern era mainly focus on the 2nd, which you consider a female interest. I see it as more of an era thing: 19th-century historians of war mostly focused on strategy, generals, treaties, etc., while modern historians are much more likely to write books on motivations, cultural factors, impacts on civilian experience, etc. Taking that approach even further, folks like Bourdieu and Foucault were male as well..,
While I agree Ada Lovelace and Grace Hopper are important pioneers, I'm not sure it's a particularly large number--- you've right there, with your 2-entry list, basically exhausted the list of prominent female pioneers of geekdom, while the equivalent male list (Alan Turing, Steve Wozniak, Donald Knuth, Edsger Dijkstra, ...) goes on for a while longer. Sure, there a few others once you move down the list into "famous among specialists"--- Radia Perlman (inventor of the spanning-tree protocol), say. But you can't even put together a list of 20 without digging pretty deep.
More than that--- Perl 6 was announced 9 1/2 years ago! Even O'Reilly's Perl 6 Essentials is now 6 1/2 years old, and some Perl 6 books are into 2nd editions.
Yahoo does have some built-in market share that'll let them struggle on for a while as long as their search isn't terrible, which is what they might be hoping for from the Bing backend. They have huge installed bases of users of lots of their other services (email, mailing lists, now Flickr), and the Yahoo search bar shows up at the top of the page for some of those services (though not, at least for now, on Flickr)--- check out the top of the Yahoo Groups page, for example. If they can keep those services all steady with reasonably large user bases, they'll have a decent stream of search users even if nobody purposely goes to search.yahoo.com.