However, social choice theory is a fairly precise discipline with a number of marvellous impossibility theorems, new voting procedures, etc., and there are also many socialogists who make fairly good statistical research. In a nutshell, your friend chose to focus on the crapiest areas of her discipline.
Really there is a simple reason for this. (Of course, it's not the only one but presumably the primary one.)
Tenure-track positions and funding are to a large extent determined on the basis of the number of publications weighted by the reputation of the journals, not by the quality of publications.
The idea is that good journals will reject bad papers, which doesn't work as well as is desirable due to the extreme amount of submissions the journals receive, which have to be reviewed by relatively small numbers of unpaid voluntary reviewers.
There are many ways this problem could be alleviated and I have no idea which would be the best one. For example, hiring comittees could be encouraged to only take a look at 10 papers chosen by the applicant and disregard all others including their total number. But it's doubtful they would follow this advice in practise. Or, "allowed" publications per average year could be limited to a minium of n and a maximum of m papers. So for example, to keep funding you need to publish (on average, over a larger period of time) at least 1 peer-reviewed article and no more than 3 per year in average. Sounds crazy and I don't know how to enforce this, but it would increase the quality of papers if m is chosen sufficiently low. Or, get more stringent peer reviewing, although it's a mistery how you'd obtain that in the current system. Perhaps open access journals with crowd reviewing/ranking and meta-moderation would work, as long as mechanisms are held in place to weed out sockpuppets and trolls - difficult, though.
Anyway, it's mostly the publication pressure, in terms of numbers, that causes bad publications.
Now back to work... I need to finish a hastily written paper.
What I don't understand about modern day US libertarians is that the classical economic theory they seem to mostly endorse contradicts their own central principles, and even in relatively obvious ways. For instance, the Principle of Diminishing Marginal Utility in combination with sum-utilitarianism implies that monetary transfers from the rich to the poor increase overall utility, game theory can be used to show that free markets lead to cartels, there are Pareto efficient states in which one person owns almost everything and the rest are slave workers, etc., none of which seems to be desirable to anyone. There are countless examples of this kind.
Notice that I'm not saying that classical economic theory is right, I'm just pointing out how baffling the main doctrines of libertarianism can appear to an outsider. But I'm sure I got it all wrong in the eyes of some libertarian economists...;-)
Really? Why is it their job to spy on Angela Merkel, if Obama could just read the newspaper or call her up to find out what she's thinking? Nobody says the NSA shouldn't spy on North Korea, but how about the right balance? Shouldn't they spy on the military infrastructure of enemy countries rather than close allies and trade partners?
Anyway, the bright side of this news is that the cooling down of relations between the US and EU countries might result in less violations of constitutional rights of citizens on both sides of the pond. Perhaps the illegal procedure of just asking a befriended agency for information about your own citizens will become less customary.
You've got an overly optimistic view of the German intelligence apparatus.
First of all, you forget to mention the real reason why the Germans likely do not spy much on the US: After WW2 US authorities recruited Nazis to build up the German intelligence agencies. Although these Nazis might have been willing to spy on Americans, they were under tight supervision by their new allied friends, and overall they were more eager to continue spying on the USSR. Their Nazi origins is also the reason why German intelligence agencies have always been spying on German leftists much more than on right-wing extremists inside the country. (The recent fairly direct involvement of the Verfassungssschutz in the protection of a German right-wing murder cell demonstrates this once more.) German intelligence agencies were practically built by the US, and are therefore still well-guarded against spying against the US, e.g. in the form of old ties, direct collaboration with the NSA, mauls, etc. In other words, they are historically 'entangled' with the US agencies.
Secondly, it is highly unrealistic to claim that Germany does not do mass surveillance. Fact is that Siemens has been caught to build mass surveillance capabilities into telco equipment delivered to foreign states, and there are good reasons to believe that the BND delivers lots of data from some countries, e.g. in the Middle East. There are also good reasons to believe that German authorities helped, on the basis of an official surveillance act from the 70s (and inofficially from the start of the BRD), to install surveillance equipment in German landlines for the NSA. Roughly speaking, much of the BNDs SIGINT is probably delivered directly to the NSA - again, for historical reasons and because German agencies greatly benefit from the data they get from the US in return.
In summary, German intelligence agencies do mass surveilance and not just for themselves but also for the NSA, and this is the real reason why Chancellor Merkel is not saying anything. She's essentially a hypocrite liar about the whole issue, but since she's a person who dislikes lying openly she tried and still tries to remain silent about it as much as possible.
It's a subjective measure, based on long experience with someone and someone's writings. It's much easier to assess trust from personal contacts, but even then you can get thoroughly disappointed - just think of some types of failed marriages as an example.
The question is why would you personally have to trust Bruce Schneier? I don't have to, in order to enjoy his books and blog posts and make up my own mind. Has he recently asked you to hand over the masterpassword for you computer?
Regarding business with his company, overall reputation and a realistic assessment of threat scenarios is more important than personal trust. If you believe the NSA is your main adversary and you contemplate whether you should put all of your trust into Bruce Schneier as your sole savior, you might want to revise your "requirements".
According to my version of the live map, there is a mid-sized attack from the US to China and at the same time a gigantic attack on the US from outer space!!!
You've got your numbers all wrong. The largest known genocide in the history of mankind so far was the genocide on the native indian americans with an estimated death toll of 40 to 90 million people, and it was (mostly) commited by Christians. (That does not even include the parallel mass killings committed by slave traders, all of which were Christians, too.)
They are already 'outlawed' in pretty much the whole of Europe under existing laws, in the sense that you have the right to demand from a private person who films you to stop doing that, have a right to obtain the originals and a right to get all copies destroyed. These rights are rarely asserted in daily life nowadays, like when someone films people with his mobile phone. But regarding Google glasses, who are much more sneaky and evil in their typical use cases (such as filming women and biometrically identifying them from their Facebook pages), people will assert their rights much more often and it's not unreasonable to expect that European courts will outright ban these kinds of devices (i.e. all devices that do not clearly indicate when they are filming someone).
In other words, even in the unlikely event that they will catch on in the US, Google classes will likely never pose any serious problems in Europe.
So this amateur managed to "debunk" a paper based on his "intuitions". Quite frankly, I'm not impressed.
First of all, the field of psychology is vast and some parts of it shouldn't be called scientific. Given that, the approach criticized seems refreshingly precise.
Secondly, the amateur does not seem to be aware very much of mathematical psychology, because otherwise he wouldn't have had such clearcut "intuition" why in this particular case the model was wrong. As it has turned out often in the past, and much to the astonishment of mainstream psychologists, very simple ad hoc models sometimes predict psychological phenomena better than experts in the field. In fact, some simply analytic models have turned out to perform even better than experts who applied these formulas (meaning that the experts didn't trust them entirely and modified the results). A famous example is the comparison of predictions by experienced marriage counselors versus the formula number of sexual intercourses during last month-the number of fights/arguments between the partners - you don't want this number to be negative. Such simple models have also been successful in certain types of medical diagnosis, casting doubts on the expertise of practitioners and instigating a discussion about the ethics of using diagnostic formulas instead of assessment by a human expert. (For example, whom do you blame when the diagnosis by the formula is wrong, even though the formula statistically makes better predictions than humans?)
Overall, mathematical psychology has made tremendous advances during the past 30-40 years and so it's not as strange as the title story makes it appear when psychologists use modeling techniques, and it's actually a good sign if they take a look at other domains and attempt to find out whether modeling techniques from other domains can be used. It's also not uncommon for this to occur.
So I don't understand what is the alleged "bullshit" here. An approach turned out to be wrong, though probably not as wrong as, say, C.G. Jung's archetypes or Freud's notion of a "psychological force", but perhaps the author's were looking in the right direction and built relatively precise models. It's not as if all papers in theoretical physics are right and final for all eternity.
Okay, this is bothering me for some time. A friend told me during the shutdown that the government could, as a last resort, create a trillion dollar coin (or some such amount) and start paying their debts without congressional oversight. He also claimed that this method would have had essentially no bad effects on the economy and was just a way to circumvent the debt ceiling. Is that true? Can someone explain to me in layman's terms how it would work?
Obviously, if that was so easy they'd have done it, so there is a catch, isn't there?
Does it run OS X 10.9? In the past it's always been easier to get images for Virtualmachine for development, but Virtualbox is easier to set up. Anyway, I'm not going to buy a Crapple(TM) just for testing my free cross-platform apps so I was wondering.
The problem is that sentences rarely occur in isolation, making encryption harder and harder if you want a bijective mapping to *texts* that make sense. In fact, at some time in a context there might not be enough texts that make sense to encrypt to unless you're willing to introduce a lot of redundancy.
The real reason why open source practically always beats closed source in security applications is that the authors have to presume that someone else will take a look at the code later and therefore want to avoid too messy and unclean coding. With closed source the temptation is simply too high to introduce dirty hacks and shortcuts, such as crappy PRNGs where cryptographically secure ones would be required, using no salt or using default initialization vectors - things that would be too embarrasing if anybody could discover them easily.
Closed source developers can avoid that by independent security auditing, frequent reviews and strict coding guidelines, but that costs a lot of money and is only done when there is an external incentive like having to fulfill some FIPS regulation. In many if not all cases you can and should give a shit about the claims of even the most reputable closed source vendors. They are very likely lying about one thing or another and their managers likely don't even know exactly what they are really selling and how it works (viz., doesn't work).
The SAS will be flown in by helicopter and they will massacre the terrorists until none of them is left. (Unles they forget their brew kit, of course, because they cannot fight without one.)
Seriously, you think it's okay if parents choose to let their kids become demented or even die because of an infection with measles that could have easily been prevented?
Smallpox and polio are practically extinct because of mandatory vaccination and tetanus is no longer a problem - do you have any idea what these diseases meant to people before vaccination was available? Please, do me a favor and look them up on Wikipedia to get an idea what you're talking about. You want to allow parents to "choose" that their children don't get a shot against lockjaw? So when they get a little scratch when playing in the garden they might die the most painful and horrible death imaginable?
The situation in publishing is very different than what you're imagining. Word is just how the text gets edited up to the point where page production starts. Then the whole thing is converted to InDesign or QuarkExpress.
Although you're generally right, I'd like to add that this is not always so. I've seen a number of smaller publishers and academic journals who use Word for the camera-ready result. It's easy to spot them...
TeX also has serious paragraph-(re-)flow limitations that make it inferior to some commercial typesetting systems. In a nutshell, if you need to fulfill criteria such as strict raggedbottom, which many publishers require, (La-)TeX typesetting can be really painful and/or occasionally yields bad results, because it cannot optimize text flow across many paragraphs and pages. At least that's my experience from typesetting books.
It would be nice to have a modern replacement, but lack of packages makes this unlikely. For example, lout should theoretically be better but last time I tried lacked essential functionality in practice. It's a chicken and egg problem.
That being said, even the worst output from TeX is still better than what text processors like Word produce. As incredible as it may sound, there are academic publishers and journals who use Word for producing camera ready manuscripts, which consequently look like shit. Word is really the cancer of publishing, it has caused me a lot of grief and stupid, unnecessary work, and I sincerely hope it will die sooner than later.
Slashdot Mirror
However, social choice theory is a fairly precise discipline with a number of marvellous impossibility theorems, new voting procedures, etc., and there are also many socialogists who make fairly good statistical research. In a nutshell, your friend chose to focus on the crapiest areas of her discipline.
Really there is a simple reason for this. (Of course, it's not the only one but presumably the primary one.)
Tenure-track positions and funding are to a large extent determined on the basis of the number of publications weighted by the reputation of the journals, not by the quality of publications.
The idea is that good journals will reject bad papers, which doesn't work as well as is desirable due to the extreme amount of submissions the journals receive, which have to be reviewed by relatively small numbers of unpaid voluntary reviewers.
There are many ways this problem could be alleviated and I have no idea which would be the best one. For example, hiring comittees could be encouraged to only take a look at 10 papers chosen by the applicant and disregard all others including their total number. But it's doubtful they would follow this advice in practise. Or, "allowed" publications per average year could be limited to a minium of n and a maximum of m papers. So for example, to keep funding you need to publish (on average, over a larger period of time) at least 1 peer-reviewed article and no more than 3 per year in average. Sounds crazy and I don't know how to enforce this, but it would increase the quality of papers if m is chosen sufficiently low. Or, get more stringent peer reviewing, although it's a mistery how you'd obtain that in the current system. Perhaps open access journals with crowd reviewing/ranking and meta-moderation would work, as long as mechanisms are held in place to weed out sockpuppets and trolls - difficult, though.
Anyway, it's mostly the publication pressure, in terms of numbers, that causes bad publications.
Now back to work... I need to finish a hastily written paper.
What I don't understand about modern day US libertarians is that the classical economic theory they seem to mostly endorse contradicts their own central principles, and even in relatively obvious ways. For instance, the Principle of Diminishing Marginal Utility in combination with sum-utilitarianism implies that monetary transfers from the rich to the poor increase overall utility, game theory can be used to show that free markets lead to cartels, there are Pareto efficient states in which one person owns almost everything and the rest are slave workers, etc., none of which seems to be desirable to anyone. There are countless examples of this kind.
Notice that I'm not saying that classical economic theory is right, I'm just pointing out how baffling the main doctrines of libertarianism can appear to an outsider. But I'm sure I got it all wrong in the eyes of some libertarian economists... ;-)
It appears that my aggressive spam filter is now a "dark corner of the Internet."
Really? Why is it their job to spy on Angela Merkel, if Obama could just read the newspaper or call her up to find out what she's thinking? Nobody says the NSA shouldn't spy on North Korea, but how about the right balance ? Shouldn't they spy on the military infrastructure of enemy countries rather than close allies and trade partners?
Anyway, the bright side of this news is that the cooling down of relations between the US and EU countries might result in less violations of constitutional rights of citizens on both sides of the pond. Perhaps the illegal procedure of just asking a befriended agency for information about your own citizens will become less customary.
another entry into the rapidly expanding and increasingly competitive wearables marketplace that no customer is interested in
TFTFY
You've got an overly optimistic view of the German intelligence apparatus.
First of all, you forget to mention the real reason why the Germans likely do not spy much on the US: After WW2 US authorities recruited Nazis to build up the German intelligence agencies. Although these Nazis might have been willing to spy on Americans, they were under tight supervision by their new allied friends, and overall they were more eager to continue spying on the USSR. Their Nazi origins is also the reason why German intelligence agencies have always been spying on German leftists much more than on right-wing extremists inside the country. (The recent fairly direct involvement of the Verfassungssschutz in the protection of a German right-wing murder cell demonstrates this once more.) German intelligence agencies were practically built by the US, and are therefore still well-guarded against spying against the US, e.g. in the form of old ties, direct collaboration with the NSA, mauls, etc. In other words, they are historically 'entangled' with the US agencies.
Secondly, it is highly unrealistic to claim that Germany does not do mass surveillance. Fact is that Siemens has been caught to build mass surveillance capabilities into telco equipment delivered to foreign states, and there are good reasons to believe that the BND delivers lots of data from some countries, e.g. in the Middle East. There are also good reasons to believe that German authorities helped, on the basis of an official surveillance act from the 70s (and inofficially from the start of the BRD), to install surveillance equipment in German landlines for the NSA. Roughly speaking, much of the BNDs SIGINT is probably delivered directly to the NSA - again, for historical reasons and because German agencies greatly benefit from the data they get from the US in return.
In summary, German intelligence agencies do mass surveilance and not just for themselves but also for the NSA, and this is the real reason why Chancellor Merkel is not saying anything. She's essentially a hypocrite liar about the whole issue, but since she's a person who dislikes lying openly she tried and still tries to remain silent about it as much as possible.
Once they have received their payment from Apple, which is about NOW.
Brushed metal look was the beginning of Apple's long, painful end.
It's a subjective measure, based on long experience with someone and someone's writings. It's much easier to assess trust from personal contacts, but even then you can get thoroughly disappointed - just think of some types of failed marriages as an example.
The question is why would you personally have to trust Bruce Schneier? I don't have to, in order to enjoy his books and blog posts and make up my own mind. Has he recently asked you to hand over the masterpassword for you computer?
Regarding business with his company, overall reputation and a realistic assessment of threat scenarios is more important than personal trust. If you believe the NSA is your main adversary and you contemplate whether you should put all of your trust into Bruce Schneier as your sole savior, you might want to revise your "requirements".
According to my version of the live map, there is a mid-sized attack from the US to China and at the same time a gigantic attack on the US from outer space!!!
I like it.
You've got your numbers all wrong. The largest known genocide in the history of mankind so far was the genocide on the native indian americans with an estimated death toll of 40 to 90 million people, and it was (mostly) commited by Christians. (That does not even include the parallel mass killings committed by slave traders, all of which were Christians, too.)
They are already 'outlawed' in pretty much the whole of Europe under existing laws, in the sense that you have the right to demand from a private person who films you to stop doing that, have a right to obtain the originals and a right to get all copies destroyed. These rights are rarely asserted in daily life nowadays, like when someone films people with his mobile phone. But regarding Google glasses, who are much more sneaky and evil in their typical use cases (such as filming women and biometrically identifying them from their Facebook pages), people will assert their rights much more often and it's not unreasonable to expect that European courts will outright ban these kinds of devices (i.e. all devices that do not clearly indicate when they are filming someone).
In other words, even in the unlikely event that they will catch on in the US, Google classes will likely never pose any serious problems in Europe.
So this amateur managed to "debunk" a paper based on his "intuitions". Quite frankly, I'm not impressed.
First of all, the field of psychology is vast and some parts of it shouldn't be called scientific. Given that, the approach criticized seems refreshingly precise.
Secondly, the amateur does not seem to be aware very much of mathematical psychology, because otherwise he wouldn't have had such clearcut "intuition" why in this particular case the model was wrong. As it has turned out often in the past, and much to the astonishment of mainstream psychologists, very simple ad hoc models sometimes predict psychological phenomena better than experts in the field. In fact, some simply analytic models have turned out to perform even better than experts who applied these formulas (meaning that the experts didn't trust them entirely and modified the results). A famous example is the comparison of predictions by experienced marriage counselors versus the formula number of sexual intercourses during last month-the number of fights/arguments between the partners - you don't want this number to be negative. Such simple models have also been successful in certain types of medical diagnosis, casting doubts on the expertise of practitioners and instigating a discussion about the ethics of using diagnostic formulas instead of assessment by a human expert. (For example, whom do you blame when the diagnosis by the formula is wrong, even though the formula statistically makes better predictions than humans?)
Overall, mathematical psychology has made tremendous advances during the past 30-40 years and so it's not as strange as the title story makes it appear when psychologists use modeling techniques, and it's actually a good sign if they take a look at other domains and attempt to find out whether modeling techniques from other domains can be used. It's also not uncommon for this to occur.
So I don't understand what is the alleged "bullshit" here. An approach turned out to be wrong, though probably not as wrong as, say, C.G. Jung's archetypes or Freud's notion of a "psychological force", but perhaps the author's were looking in the right direction and built relatively precise models. It's not as if all papers in theoretical physics are right and final for all eternity.
1. He has done the US a great service in the long run.
2. He is personally more trustworthy than the people in the US government and the intelligence community who have been caught lying already.
My 2 cents.
Okay, this is bothering me for some time. A friend told me during the shutdown that the government could, as a last resort, create a trillion dollar coin (or some such amount) and start paying their debts without congressional oversight. He also claimed that this method would have had essentially no bad effects on the economy and was just a way to circumvent the debt ceiling. Is that true? Can someone explain to me in layman's terms how it would work?
Obviously, if that was so easy they'd have done it, so there is a catch, isn't there?
Does it run OS X 10.9? In the past it's always been easier to get images for Virtualmachine for development, but Virtualbox is easier to set up. Anyway, I'm not going to buy a Crapple(TM) just for testing my free cross-platform apps so I was wondering.
First, it is opposed by more people than favor it (there's a large enough group of people who are undecided to make a huge difference, though).
Are you sure these people are not just against Obamacare?
The problem is that sentences rarely occur in isolation, making encryption harder and harder if you want a bijective mapping to *texts* that make sense. In fact, at some time in a context there might not be enough texts that make sense to encrypt to unless you're willing to introduce a lot of redundancy.
The real reason why open source practically always beats closed source in security applications is that the authors have to presume that someone else will take a look at the code later and therefore want to avoid too messy and unclean coding. With closed source the temptation is simply too high to introduce dirty hacks and shortcuts, such as crappy PRNGs where cryptographically secure ones would be required, using no salt or using default initialization vectors - things that would be too embarrasing if anybody could discover them easily.
Closed source developers can avoid that by independent security auditing, frequent reviews and strict coding guidelines, but that costs a lot of money and is only done when there is an external incentive like having to fulfill some FIPS regulation. In many if not all cases you can and should give a shit about the claims of even the most reputable closed source vendors. They are very likely lying about one thing or another and their managers likely don't even know exactly what they are really selling and how it works (viz., doesn't work).
The SAS will be flown in by helicopter and they will massacre the terrorists until none of them is left. (Unles they forget their brew kit, of course, because they cannot fight without one.)
Seriously, you think it's okay if parents choose to let their kids become demented or even die because of an infection with measles that could have easily been prevented?
Smallpox and polio are practically extinct because of mandatory vaccination and tetanus is no longer a problem - do you have any idea what these diseases meant to people before vaccination was available? Please, do me a favor and look them up on Wikipedia to get an idea what you're talking about. You want to allow parents to "choose" that their children don't get a shot against lockjaw? So when they get a little scratch when playing in the garden they might die the most painful and horrible death imaginable?
What the fuck is wrong with you people?
The situation in publishing is very different than what you're imagining. Word is just how the text gets edited up to the point where page production starts. Then the whole thing is converted to InDesign or QuarkExpress.
Although you're generally right, I'd like to add that this is not always so. I've seen a number of smaller publishers and academic journals who use Word for the camera-ready result. It's easy to spot them ...
TeX also has serious paragraph-(re-)flow limitations that make it inferior to some commercial typesetting systems. In a nutshell, if you need to fulfill criteria such as strict raggedbottom, which many publishers require, (La-)TeX typesetting can be really painful and/or occasionally yields bad results, because it cannot optimize text flow across many paragraphs and pages. At least that's my experience from typesetting books.
It would be nice to have a modern replacement, but lack of packages makes this unlikely. For example, lout should theoretically be better but last time I tried lacked essential functionality in practice. It's a chicken and egg problem.
That being said, even the worst output from TeX is still better than what text processors like Word produce. As incredible as it may sound, there are academic publishers and journals who use Word for producing camera ready manuscripts, which consequently look like shit. Word is really the cancer of publishing, it has caused me a lot of grief and stupid, unnecessary work, and I sincerely hope it will die sooner than later.