The system was set up as a checks and balances system to keep the US policy from being dictated by people in large cities.
No, it wasn't. Well, preventing the people from large states (not cities) from dictating the choice was a small part of the rationale, but claiming it was the whole thing is like claiming that your house has plumbing so you can brush your teeth. Here's a decent (brief) overview of that rationale for and evolution of the electoral college: http://uselectionatlas.org/INF....
When we ascribe the power of offensiveness to a word, it decreases the likelihood it will wither away from the language.
Maybe, maybe not, but that has nothing to do with the case at hand, because no one is ascribing power of offensiveness to the word. The word absolutely has power of offensiveness, which arises from centuries of bloody horrific oppression. Nothing you can do will remove that history, so the best thing you can do is to make it clear that invoking the history is an anti-social action, not acceptable among decent people.
"Nature did it with meat" has no scientific basis. All Science has is interface observations.
What are you on about? Yes, science has only interface observations... we have only interface observations, and in fact we observe nothing whatsoever directly. Everything is perceived through layers of translation and understood through layers of theory. You're trying to argue that nothing is understandable, which is clearly false.
Getting back to the point, what we have observed is that humans are intelligent. We don't know in detail what that means but we can describe large categories of cognitive abilities that it includes, and so it's sensible to talk about non-human intelligences that can do the same thing.
The other crucial factor here is one that we can't observe but only assume... but it's an assumption that has been working pretty damned well in all of the rest of science, so there's no particular reason to expect it to fail here. That assumption is that every phenomenon we observe is the result of interactions of physical particles according to consistent and well-defined rules (even if some of those rules appear to be probabilistic in nature). That is, that there is nothing else other than matter and energy and the rules by which they interact.
In a nutshell, this second assumption is that the supernatural does not exist. That's not to say that we understand all of what is natural.
Your argument is that the supernatural does exist, and that something supernatural is essential to intelligence. That's a position you're perfectly welcome to take, just don't expect people to consider it scientific.
It doesn't actually say whether or not they have a dossier on me
Have you ever gotten a loan in the United States? And by "loan" I mean any sort of credit whatsoever... car loan, credit card, non-prepaid mobile phone, rented an apartment, had an electric bill, etc. If so, Equifax has a dossier on you. Just to clear up the first part of the question.
Should I rehash my fundamental principles of personal information protection? On Slashdot?
If you like. But they're irrelevant in this case, because you don't directly get to decide when your personal information is sent to Equifax (or Experian or Trans Union, the other two credit agencies). The people who do credit checks on you and to whom you pay bills for various services, some of which are loans, are the ones who send it. With your approval, mind you. It's buried in the fine print, and never says "Equifax", just mentions credit bureaus or credit reporting agencies or some similar phrasing.
Most likely the users. At least if the content is worth it.
Which means that 90% of the content on the web will disappear, and the remainder will all be paywalled and inaccessible to many people. There's good reason to dislike advertising, but there are also extremely good reasons that it has been the primary mechanism for funding broadly-distributed content for centuries. No other approach has proven to work remotely as well.
But, actually, none of that will happen. Instead, we'll just have an arms race between adblockers and adblocker-blockers, and the better-funded side will always be far enough ahead that most people will choose not to block, because it will prevent them from seeing much. Your proposal at the top of this thread, to randomize cookies to prevent effective tracking, would just spawn alternative tracking mechanisms. At the extreme, perhaps all sites will simply refuse visitors who aren't logged into one of their supported ad networks' tracking systems or who have an adblocker.
Honestly, that last option may well be the best. It would make tracking opt-in rather than opt-out, and would make the value proposition clear.
Next target hackers! We now know the former CSO wasn't the sharpest tool in the box. Rot is almost certainly there too.
Hackers don't need some additional notice or incentive to go after First Data. First Data is one of the biggest, tastiest and most potentially lucrative targets in the world. But you haven't heard that, because they do a very good job on security.
I worked several security projects at First Data when I was doing security consulting, and I was consistently impressed with quality of their people, systems and processes. I was also a little appalled at how many eggs are in the First Data basket. They issue and manage a large majority of the credit and debit cards in the United States. You almost certainly have a card they issued in your wallet, and they also generate your statements, process your payments and potentially even operate your bank's web site.
The largest project I worked for First Data was directly supervised by the NSA (in their role of protecting the nation's data infrastructure, not their role of spying on everyone -- two very different organizations within the NSA) because the security of First Data systems is essential to national security. They're that big and that important to the country's credit and banking infrastructure. More important than Equifax, I'd say.
The fact that she was CSO for First Data changes my perception of the headline considerably. I can't see First Data hiring someone unqualified for a role like CSO. Security is way, way too important there, and they have a lot of people who know how to do security.
Alexa plays Audible books? In hindsight that's obvious. Of course it does, since Amazon owns Audible. Now I have to buy an Echo. I already have a Google Home (and like it a lot, even though they nerfed the shopping list, grumble), and I didn't think there was any reason for me to get an Echo instead. But clearly there is.
You just cost me $50. Though I expect I'll be happy I spent it.
Still takes a minute or so to get them clean enough that you'd want to touch something you don't want to get greasy. If you have to do this more than a couple of times per hour, being able to do it by voice is very nice.
You and the parent poster are mostly focused on what, IMO, is just a consequence of the problem: the non-specialised press, PR, public opinion, etc. You should ask yourselves another question: why are these people who aren't in a position to adequately understand partial outputs being given those at all? Because certain scientific subfields aren't precisely being managed in a too scientific way and this the real problem: objective correctness, critical attitudes, long-term expectations, growth of the scientific knowledge as a whole, etc. are being ignored on exchange of funding or some minutes of fame.
That does not follow. You're arguing that in order to be objective, have properly critical attitudes, etc. scientists need to keep their results secret, or at least avoid letting anyone not sufficiently versed in the relevant science to know about them. The one thing has nothing to do with the other.
Scientists should publish detailed papers with accurate abstracts. It's not their job to withhold information from others who they don't think are able to understand it. Indeed, trying to decide who should and shouldn't learn of their work would be a serious waste of time which could be better spent on research. And if scientists find it useful to promote their work in order to obtain funding, that demonstrates a flaw in the way the work is funded, but you're trying to call it a failing of the scientists.
In terms of achieving the security goals it sets out to achieve, it's exactly what you'd expect from the likes of David Chaum and Ron Rivest (the "R" in "RSA"), not to mention the other less famous but also eminent cryptographers involved. And it's also very practical, including having been tested in real-world elections, and refined based on lessons learned. The only real downside is that ballots are significantly more expensive than standard scan-tron sheets, because every one of them is unique, and because an additional ink "color" is needed for the invisible ink layer. But we're still talking fractions of a penny per ballot in volume. Well worth the cost, IMO. The invisible ink marker pens turn out not to cost any more than standard markers as long as you're buying at least a few thousand of them.
It's a good system. It should get much more attention. Which is why I post about it every time the question of elections show up on slashdot:-)
like any acceptable voting system, it's possible to fall back on manual counting
Absolutely, though it should be pointed out that the assurance provided by manual recounts is strictly weaker than the integrity guarantees provided by Scantegrity. Being able to do them is still good because everyone can understand a manual recount.
The problem isn't the message itself, it's the messengers.
I agree with that statement, but understood in its widest sense: not just the target audience, also a big proportion of people performing these studies. Finding reliable enough trends to get truly worthy insights into a wide variety of phenomena is certainly possible. The problem is that getting that ideal result in quite a few scenarios is really difficult; it requires lots of knowledge (including high quality information), objectivity and resources which are rarely available.
Science is an error correction process. To a first order approximation (and probably to a third-order approximation), every scientific paper ever published is wrong. The reason science works is that the iterated process of conjecture and criticism -- in the broadest meaning of "criticism", which includes experimental testing, and lots more -- gradually identifies and weeds out errors. So the science on any given topic asymptotically approaches correctness in the long run, but the initial efforts look more like a random walk as people begin to scope out the solution space.
The problem is that that vast majority of people fundamentally don't understand how science works, the press included. So the press packages early results and presents them as "science". And they are scientific results, and in many cases may represent the best current scientific knowledge on a topic... but "best current" isn't the same thing as "good". Even worse, because science works by proposing and testing answers to very precise questions, the press reports not what the scientists actually examined, but some generalization of what they examined.
So, the press reports generalizations of preliminary results as "science", and laypeople expect those generalizations to be authoritative Truth. But science basically never provides authoritative Truth, and certainly doesn't provide anything remotely close to it early on. This is clearly a recipe for regularly-repeated disaster, and is exactly why so many laypeople have simply given up on science, preferring pyramids and crystals and things that feel truthy.
The human brain sees pattern everywhere it looks too.
Yep. Pattern identification system identifies patterns, news at 11.
OTOH, the ones I find interesting are the cases where ML identifies patterns that humans might not be able to identify. Sometimes this is less interesting for the potential to use a machine to identify these patterns than for the indication that patterns exist where we might think they don't.
The recent paper on ML "gaydar", where researchers trained a machine to identify sexual orientation from dating site photos is a potentially-fascinating example. In that one I suspect the algorithm was mostly keying on things like hairstyle and other indicators that the people in the photos might be deliberately using to advertise their orientation (this was a dating site), but there seems to be some evidence that facial structure also played a significant role. Of course, given that the most common genders and orientations (cisgendered heterosexuals) are clearly strongly correlated with certain characteristic facial structures, it's certainly reasonable to expect that the same genetic and development processes that determine gender and orientation and clearly affect face structure in the common cases should also affect face structure in the less common cases. But humans can't really see these patterns. Absent behavioral clues, people have very bad "gaydar". But that doesn't mean the patterns aren't there, just that we can't see them. If ML can see strong correlations between facial structure and orientation (and I don't think this one study proves that), that tells us something about the nature of sexual orientation and the degree to which it is expressed in the body.
OTOH, we're talking about such small amounts of data that I don't know why it's only kept for two months.
Keeping that data around if you have discontinued use of your phone is a security risk... WiFi passwords for instance. If you aren't using the device for that long a period of time there is a good chance you have gotten a new device. I'd rather the inconvenience of having to reconfigure some WiFi passwords and such rather than having someone pulling a dead phone out of the trash and potentially accessing my information.
Well, it's not like the data is stored unencrypted. But the Wifi passwords may well be the reason for the short lifetime.
Seems radically low. Some people go on foreign vacation for that long and don't use their phone.
One year would be reasonable.
If you don't use your phone for one year, you should have no expectation that the data is still there.
But two months = idiots that only looked at most common usage patterns.
Meh. Keep in mind that the data we're talking about is "device configuration, such as wallpaper, WiFi passwords, and default apps." We're not talking about your contacts, which are synced with Google Contacts, or your emails, which are synced with GMail, or your docs, spreadsheets, presentations or Drive files, which are synced with Drive, or your photos, which are synced with Google Photos, or... I could go on, but you get the idea.
Each of the Google apps backs its own data up independently, using your Drive storage, and all of that is separate from this "phone backup". The only thing the phone backup really does is streamline the process of setting up a new device, so all of your apps are auto-downloaded, WiFi passwords are auto-configured, etc.
OTOH, we're talking about such small amounts of data that I don't know why it's only kept for two months.
Bah. I'm responsible with my money, but if I were to take all of the credit offered me, heck, even if I were just to use all of the revolving credit I already have right now, there's no way I could make the payments.
You're right that banks are not stupid, though. The reason they're willing to offer me insane amounts of credit is because I have an 800+ credit score and they know that I will be responsible and won't overcommit myself. The credit-limiting algorithms are extremely generous for people with excellent credit histories, good jobs, etc. especially when the loan is well-secured. That is an excellent description of my co-workers who are buying $1.5M homes ($6K monthly payments) on net monthly incomes of $8K.
I'm actually an advocate of companies retaining most of their employees as FTEs. Current accounting rules and tax law doesn't make this as appealing as it used to be.
I think that's an understatement.
Let me note at the outset that what I'm going to say is thirdhand information which may not be correct. I'd appreciate correction from someone who actually does know. In fact, the potential to provoke such a correction is 80% of the reason I'm posting.
I've been told by people who I expect should know that federal labor law restricts corporations to having no more than two classifications of full-time employees with respect to benefits packages. The two levels allow companies to offer different packages to upper management and regular workers, with middle management usually being lumped in the second category. This structure makes a fair amount of sense for factory operations where you basically have the big bosses and the line workers. It doesn't work so well for companies that have a large professional workforce as well as a large pool of unskilled or semi-skilled labor.
Putting the professional workforce in the same category as upper management either limits the ability of the company to attract good executives or overcompensates the non-management professionals and middle managers. Putting it in the same category as the unskilled or semi-skilled labor either limits the ability of the company to attract good professionals and middle managers or overcompensates the unskilled and semi-skilled laborers.
So, the common solution is to have an executive class and a professional employee class and to contract out the unskilled and semi-skilled labor.
Liberal philosophy is largely around collecting more tax revenue because citizens are either too dumb or evil, lazy shits to know how to spend it on what's "best for the country".
This would be more convincing if there were any evidence that conservatives are less interested in spending taxpayer money. From my perspective (pragmatic libertarian) both the left and the right in the US seem interested in expanding spending programs without limit. The primary differences between them are (a) what they want to spend the money on and (b) fiscal responsibility. The liberals demonstrate greater fiscal responsibility, being willing to raise taxes to pay for their spendthrift ways. Liberals tax and spend, conservatives borrow and spend.
I understand that some conservatives are fiscally irresponsible intentionally, on the "starve the beast" theory that eventually it will become impossible to borrow more and that at that point we'll be forced to cut spending. They offer no explanation as to why, when that notional day arrives, it will be the liberals' programs that get cut rather than their own. I speculate that conservatives simply believe that their programs are so essential that they will not get cut, but I'm sure that liberals would -- and will -- say that the liberal programs are the ones that are essential and must not be cut.
Now that would be a plugin! One where people can choose the (ad) cookies to share with the other users of the plugin, basically rendering any and all data collected absolutely worthless because nobody can ever know anymore who used what ad cookie to visit a page.
Interesting idea. It's not so certain that it would be a clear win for users, though. Inability to target ads means that ads are worth less, all else being equal. This leads advertisers to try to make their ads worth more by making them attention grabbing (bigger, brighter, blinkier, self-playing video, etc.), or to simply pay less for ad space. If they pay less for ad space, then site owners are incented to increase the amount of ad space on their sites, or else to stop depending on advertising revenue and just paywall.
I understand the concern about online ad networks tracking users. There is a benefit to users as well as a cost, though, and you should consider how to replicate that benefit when thinking about how to eliminate the cost.
There will always be a need for 'mom and pop' convenience stores, and 7-11, and what-have-you.
I agree with this. What's not so clear is that there will always be a need for people to staff such stores. Their idea of using a small box on the street may be foolish (or may not... it may just augment the stores, so you don't have to go all the way to the corner for some stuff), but automated checkout from unattended stores seems quite likely to become the norm at some point.
I'm not sure what "other schemes" you're referring to.
The ones listed as anonymizing at the end of the Wikipedia article you referenced. At this point, I don't know quite what you're talking about.
Until/unless you read the paper, we can't really communicate about this effectively. It's linked from the Wikipedia article, but I'll provide it here as well: http://www.usenix.org/event/ev...
If it's the idea about having lists of candidates by letter, and varied orders on the ballot
It's not. You can either vary orders on the ballots or not. Makes no difference to this system, and voters mark their ballots by filling in the bubble next to their selection, so there's no unusual opportunity for confusion.
The scheme to fool the bribers and/or threateners appears to be to fill out several ballots, vote with one, and keep all the receipts. Again, that complicates the voting process.
That would not provide any way to either inform or fool bribers and/or threateners. All the ballots are different and showing multiples would change nothing.
Hand-counting the vote becomes immensely more difficult. Rather than just going through ballots and seeing which candidate was chosen for each, each ballot has to be looked up, and the lookup process has to be secure.
No, it doesn't. Ballots can be hand-counted just by looking at bubbles next to names.
It appeared to hand that off to other schemes, which don't look to me like they'll work.
I'm not sure what "other schemes" you're referring to.
If I can verify how my vote was counted, that gentleman sitting over there with the large wrench, or that other gentleman with the roll of $20 bills, can ask me to verify it.
Yes you can, and no you can't, respectively. What you can verify is that your receipt (containing a random code that was associated with your ballot selection) was included in the public receipts, and you can verify that the public receipts are tallied correctly, but you can't directly verify that your receipt indicates a vote for Trump rather than Clinton.
The relationships between ballots, candidates and codes is pre-committed in some published tables. When you vote, you can take some extra ballots, reveal all of the codes and validated them against the published tables. Since these "test ballots" are taken from the same pool as the actual ballot (you should take several and pick one at random for your real ballot), any discrepancies in the bindings between codes and candidates would be probabilistically revealed. There are some other validation steps that are taken by election officials and representatives of the candidates; read the paper.
The bottom line is that all you can validate is that your ballot receipt is found in the list of published receipts. The other elements of the process confirm that the presence of your receipt means that your vote was counted, and counted correctly.
This is all laid out fairly clearly in the paper. Note also that the paper was published nearly a decade ago and was authored by some of the biggest names in cryptographic security, which means there has been plenty of time and plenty of motivation for people to find weaknesses in the design. None have been published.
Slashdot Mirror
The system was set up as a checks and balances system to keep the US policy from being dictated by people in large cities.
No, it wasn't. Well, preventing the people from large states (not cities) from dictating the choice was a small part of the rationale, but claiming it was the whole thing is like claiming that your house has plumbing so you can brush your teeth. Here's a decent (brief) overview of that rationale for and evolution of the electoral college: http://uselectionatlas.org/INF....
When we ascribe the power of offensiveness to a word, it decreases the likelihood it will wither away from the language.
Maybe, maybe not, but that has nothing to do with the case at hand, because no one is ascribing power of offensiveness to the word. The word absolutely has power of offensiveness, which arises from centuries of bloody horrific oppression. Nothing you can do will remove that history, so the best thing you can do is to make it clear that invoking the history is an anti-social action, not acceptable among decent people.
"Nature did it with meat" has no scientific basis. All Science has is interface observations.
What are you on about? Yes, science has only interface observations... we have only interface observations, and in fact we observe nothing whatsoever directly. Everything is perceived through layers of translation and understood through layers of theory. You're trying to argue that nothing is understandable, which is clearly false.
Getting back to the point, what we have observed is that humans are intelligent. We don't know in detail what that means but we can describe large categories of cognitive abilities that it includes, and so it's sensible to talk about non-human intelligences that can do the same thing.
The other crucial factor here is one that we can't observe but only assume... but it's an assumption that has been working pretty damned well in all of the rest of science, so there's no particular reason to expect it to fail here. That assumption is that every phenomenon we observe is the result of interactions of physical particles according to consistent and well-defined rules (even if some of those rules appear to be probabilistic in nature). That is, that there is nothing else other than matter and energy and the rules by which they interact.
In a nutshell, this second assumption is that the supernatural does not exist. That's not to say that we understand all of what is natural.
Your argument is that the supernatural does exist, and that something supernatural is essential to intelligence. That's a position you're perfectly welcome to take, just don't expect people to consider it scientific.
It doesn't actually say whether or not they have a dossier on me
Have you ever gotten a loan in the United States? And by "loan" I mean any sort of credit whatsoever... car loan, credit card, non-prepaid mobile phone, rented an apartment, had an electric bill, etc. If so, Equifax has a dossier on you. Just to clear up the first part of the question.
Should I rehash my fundamental principles of personal information protection? On Slashdot?
If you like. But they're irrelevant in this case, because you don't directly get to decide when your personal information is sent to Equifax (or Experian or Trans Union, the other two credit agencies). The people who do credit checks on you and to whom you pay bills for various services, some of which are loans, are the ones who send it. With your approval, mind you. It's buried in the fine print, and never says "Equifax", just mentions credit bureaus or credit reporting agencies or some similar phrasing.
You're actually telling me that when you're having sex or a wank in bed, you're fine with Amazon listening to that
Amazon doesn't listen to that, a fact which is easily verified by watching network traffic.
You watch network traffic while wanking?
Far be it from me to kink shame, but I don't even recall ever seeing rule 34 about that one. Job well done!
Logs, man, logs.
Most likely the users. At least if the content is worth it.
Which means that 90% of the content on the web will disappear, and the remainder will all be paywalled and inaccessible to many people. There's good reason to dislike advertising, but there are also extremely good reasons that it has been the primary mechanism for funding broadly-distributed content for centuries. No other approach has proven to work remotely as well.
But, actually, none of that will happen. Instead, we'll just have an arms race between adblockers and adblocker-blockers, and the better-funded side will always be far enough ahead that most people will choose not to block, because it will prevent them from seeing much. Your proposal at the top of this thread, to randomize cookies to prevent effective tracking, would just spawn alternative tracking mechanisms. At the extreme, perhaps all sites will simply refuse visitors who aren't logged into one of their supported ad networks' tracking systems or who have an adblocker.
Honestly, that last option may well be the best. It would make tracking opt-in rather than opt-out, and would make the value proposition clear.
Next target hackers! We now know the former CSO wasn't the sharpest tool in the box. Rot is almost certainly there too.
Hackers don't need some additional notice or incentive to go after First Data. First Data is one of the biggest, tastiest and most potentially lucrative targets in the world. But you haven't heard that, because they do a very good job on security.
I worked several security projects at First Data when I was doing security consulting, and I was consistently impressed with quality of their people, systems and processes. I was also a little appalled at how many eggs are in the First Data basket. They issue and manage a large majority of the credit and debit cards in the United States. You almost certainly have a card they issued in your wallet, and they also generate your statements, process your payments and potentially even operate your bank's web site.
The largest project I worked for First Data was directly supervised by the NSA (in their role of protecting the nation's data infrastructure, not their role of spying on everyone -- two very different organizations within the NSA) because the security of First Data systems is essential to national security. They're that big and that important to the country's credit and banking infrastructure. More important than Equifax, I'd say.
The fact that she was CSO for First Data changes my perception of the headline considerably. I can't see First Data hiring someone unqualified for a role like CSO. Security is way, way too important there, and they have a lot of people who know how to do security.
Also plays audio books from Audible
Alexa plays Audible books? In hindsight that's obvious. Of course it does, since Amazon owns Audible. Now I have to buy an Echo. I already have a Google Home (and like it a lot, even though they nerfed the shopping list, grumble), and I didn't think there was any reason for me to get an Echo instead. But clearly there is.
You just cost me $50. Though I expect I'll be happy I spent it.
You're actually telling me that when you're having sex or a wank in bed, you're fine with Amazon listening to that
Amazon doesn't listen to that, a fact which is easily verified by watching network traffic.
It's called GoJo, dumbass.
Or Fast Orange, if you prefer.
I like Fast Orange, the gritty kind.
Still takes a minute or so to get them clean enough that you'd want to touch something you don't want to get greasy. If you have to do this more than a couple of times per hour, being able to do it by voice is very nice.
You and the parent poster are mostly focused on what, IMO, is just a consequence of the problem: the non-specialised press, PR, public opinion, etc. You should ask yourselves another question: why are these people who aren't in a position to adequately understand partial outputs being given those at all? Because certain scientific subfields aren't precisely being managed in a too scientific way and this the real problem: objective correctness, critical attitudes, long-term expectations, growth of the scientific knowledge as a whole, etc. are being ignored on exchange of funding or some minutes of fame.
That does not follow. You're arguing that in order to be objective, have properly critical attitudes, etc. scientists need to keep their results secret, or at least avoid letting anyone not sufficiently versed in the relevant science to know about them. The one thing has nothing to do with the other.
Scientists should publish detailed papers with accurate abstracts. It's not their job to withhold information from others who they don't think are able to understand it. Indeed, trying to decide who should and shouldn't learn of their work would be a serious waste of time which could be better spent on research. And if scientists find it useful to promote their work in order to obtain funding, that demonstrates a flaw in the way the work is funded, but you're trying to call it a failing of the scientists.
In terms of achieving the security goals it sets out to achieve, it's exactly what you'd expect from the likes of David Chaum and Ron Rivest (the "R" in "RSA"), not to mention the other less famous but also eminent cryptographers involved. And it's also very practical, including having been tested in real-world elections, and refined based on lessons learned. The only real downside is that ballots are significantly more expensive than standard scan-tron sheets, because every one of them is unique, and because an additional ink "color" is needed for the invisible ink layer. But we're still talking fractions of a penny per ballot in volume. Well worth the cost, IMO. The invisible ink marker pens turn out not to cost any more than standard markers as long as you're buying at least a few thousand of them.
It's a good system. It should get much more attention. Which is why I post about it every time the question of elections show up on slashdot :-)
like any acceptable voting system, it's possible to fall back on manual counting
Absolutely, though it should be pointed out that the assurance provided by manual recounts is strictly weaker than the integrity guarantees provided by Scantegrity. Being able to do them is still good because everyone can understand a manual recount.
The problem isn't the message itself, it's the messengers.
I agree with that statement, but understood in its widest sense: not just the target audience, also a big proportion of people performing these studies. Finding reliable enough trends to get truly worthy insights into a wide variety of phenomena is certainly possible. The problem is that getting that ideal result in quite a few scenarios is really difficult; it requires lots of knowledge (including high quality information), objectivity and resources which are rarely available.
Science is an error correction process. To a first order approximation (and probably to a third-order approximation), every scientific paper ever published is wrong. The reason science works is that the iterated process of conjecture and criticism -- in the broadest meaning of "criticism", which includes experimental testing, and lots more -- gradually identifies and weeds out errors. So the science on any given topic asymptotically approaches correctness in the long run, but the initial efforts look more like a random walk as people begin to scope out the solution space.
The problem is that that vast majority of people fundamentally don't understand how science works, the press included. So the press packages early results and presents them as "science". And they are scientific results, and in many cases may represent the best current scientific knowledge on a topic... but "best current" isn't the same thing as "good". Even worse, because science works by proposing and testing answers to very precise questions, the press reports not what the scientists actually examined, but some generalization of what they examined.
So, the press reports generalizations of preliminary results as "science", and laypeople expect those generalizations to be authoritative Truth. But science basically never provides authoritative Truth, and certainly doesn't provide anything remotely close to it early on. This is clearly a recipe for regularly-repeated disaster, and is exactly why so many laypeople have simply given up on science, preferring pyramids and crystals and things that feel truthy.
The human brain sees pattern everywhere it looks too.
Yep. Pattern identification system identifies patterns, news at 11.
OTOH, the ones I find interesting are the cases where ML identifies patterns that humans might not be able to identify. Sometimes this is less interesting for the potential to use a machine to identify these patterns than for the indication that patterns exist where we might think they don't.
The recent paper on ML "gaydar", where researchers trained a machine to identify sexual orientation from dating site photos is a potentially-fascinating example. In that one I suspect the algorithm was mostly keying on things like hairstyle and other indicators that the people in the photos might be deliberately using to advertise their orientation (this was a dating site), but there seems to be some evidence that facial structure also played a significant role. Of course, given that the most common genders and orientations (cisgendered heterosexuals) are clearly strongly correlated with certain characteristic facial structures, it's certainly reasonable to expect that the same genetic and development processes that determine gender and orientation and clearly affect face structure in the common cases should also affect face structure in the less common cases. But humans can't really see these patterns. Absent behavioral clues, people have very bad "gaydar". But that doesn't mean the patterns aren't there, just that we can't see them. If ML can see strong correlations between facial structure and orientation (and I don't think this one study proves that), that tells us something about the nature of sexual orientation and the degree to which it is expressed in the body.
OTOH, we're talking about such small amounts of data that I don't know why it's only kept for two months.
Keeping that data around if you have discontinued use of your phone is a security risk... WiFi passwords for instance. If you aren't using the device for that long a period of time there is a good chance you have gotten a new device. I'd rather the inconvenience of having to reconfigure some WiFi passwords and such rather than having someone pulling a dead phone out of the trash and potentially accessing my information.
Well, it's not like the data is stored unencrypted. But the Wifi passwords may well be the reason for the short lifetime.
ISPs get money from their customers to connect them. Hosters get money from their customers to host content. What else do you need?
How do content producers get money to produce content? To take one specific example, who would pay for slashdot's hosting?
Seems radically low. Some people go on foreign vacation for that long and don't use their phone.
One year would be reasonable.
If you don't use your phone for one year, you should have no expectation that the data is still there.
But two months = idiots that only looked at most common usage patterns.
Meh. Keep in mind that the data we're talking about is "device configuration, such as wallpaper, WiFi passwords, and default apps." We're not talking about your contacts, which are synced with Google Contacts, or your emails, which are synced with GMail, or your docs, spreadsheets, presentations or Drive files, which are synced with Drive, or your photos, which are synced with Google Photos, or... I could go on, but you get the idea.
Each of the Google apps backs its own data up independently, using your Drive storage, and all of that is separate from this "phone backup". The only thing the phone backup really does is streamline the process of setting up a new device, so all of your apps are auto-downloaded, WiFi passwords are auto-configured, etc.
OTOH, we're talking about such small amounts of data that I don't know why it's only kept for two months.
Bah. I'm responsible with my money, but if I were to take all of the credit offered me, heck, even if I were just to use all of the revolving credit I already have right now, there's no way I could make the payments.
You're right that banks are not stupid, though. The reason they're willing to offer me insane amounts of credit is because I have an 800+ credit score and they know that I will be responsible and won't overcommit myself. The credit-limiting algorithms are extremely generous for people with excellent credit histories, good jobs, etc. especially when the loan is well-secured. That is an excellent description of my co-workers who are buying $1.5M homes ($6K monthly payments) on net monthly incomes of $8K.
So... what's your strategy for funding the web, since you clearly want to remove all advertising from it?
I'm actually an advocate of companies retaining most of their employees as FTEs. Current accounting rules and tax law doesn't make this as appealing as it used to be.
I think that's an understatement.
Let me note at the outset that what I'm going to say is thirdhand information which may not be correct. I'd appreciate correction from someone who actually does know. In fact, the potential to provoke such a correction is 80% of the reason I'm posting.
I've been told by people who I expect should know that federal labor law restricts corporations to having no more than two classifications of full-time employees with respect to benefits packages. The two levels allow companies to offer different packages to upper management and regular workers, with middle management usually being lumped in the second category. This structure makes a fair amount of sense for factory operations where you basically have the big bosses and the line workers. It doesn't work so well for companies that have a large professional workforce as well as a large pool of unskilled or semi-skilled labor.
Putting the professional workforce in the same category as upper management either limits the ability of the company to attract good executives or overcompensates the non-management professionals and middle managers. Putting it in the same category as the unskilled or semi-skilled labor either limits the ability of the company to attract good professionals and middle managers or overcompensates the unskilled and semi-skilled laborers.
So, the common solution is to have an executive class and a professional employee class and to contract out the unskilled and semi-skilled labor.
Liberal philosophy is largely around collecting more tax revenue because citizens are either too dumb or evil, lazy shits to know how to spend it on what's "best for the country".
This would be more convincing if there were any evidence that conservatives are less interested in spending taxpayer money. From my perspective (pragmatic libertarian) both the left and the right in the US seem interested in expanding spending programs without limit. The primary differences between them are (a) what they want to spend the money on and (b) fiscal responsibility. The liberals demonstrate greater fiscal responsibility, being willing to raise taxes to pay for their spendthrift ways. Liberals tax and spend, conservatives borrow and spend.
I understand that some conservatives are fiscally irresponsible intentionally, on the "starve the beast" theory that eventually it will become impossible to borrow more and that at that point we'll be forced to cut spending. They offer no explanation as to why, when that notional day arrives, it will be the liberals' programs that get cut rather than their own. I speculate that conservatives simply believe that their programs are so essential that they will not get cut, but I'm sure that liberals would -- and will -- say that the liberal programs are the ones that are essential and must not be cut.
Now that would be a plugin! One where people can choose the (ad) cookies to share with the other users of the plugin, basically rendering any and all data collected absolutely worthless because nobody can ever know anymore who used what ad cookie to visit a page.
Interesting idea. It's not so certain that it would be a clear win for users, though. Inability to target ads means that ads are worth less, all else being equal. This leads advertisers to try to make their ads worth more by making them attention grabbing (bigger, brighter, blinkier, self-playing video, etc.), or to simply pay less for ad space. If they pay less for ad space, then site owners are incented to increase the amount of ad space on their sites, or else to stop depending on advertising revenue and just paywall.
I understand the concern about online ad networks tracking users. There is a benefit to users as well as a cost, though, and you should consider how to replicate that benefit when thinking about how to eliminate the cost.
There will always be a need for 'mom and pop' convenience stores, and 7-11, and what-have-you.
I agree with this. What's not so clear is that there will always be a need for people to staff such stores. Their idea of using a small box on the street may be foolish (or may not... it may just augment the stores, so you don't have to go all the way to the corner for some stuff), but automated checkout from unattended stores seems quite likely to become the norm at some point.
The ones listed as anonymizing at the end of the Wikipedia article you referenced. At this point, I don't know quite what you're talking about.
Until/unless you read the paper, we can't really communicate about this effectively. It's linked from the Wikipedia article, but I'll provide it here as well: http://www.usenix.org/event/ev...
If it's the idea about having lists of candidates by letter, and varied orders on the ballot
It's not. You can either vary orders on the ballots or not. Makes no difference to this system, and voters mark their ballots by filling in the bubble next to their selection, so there's no unusual opportunity for confusion.
The scheme to fool the bribers and/or threateners appears to be to fill out several ballots, vote with one, and keep all the receipts. Again, that complicates the voting process.
That would not provide any way to either inform or fool bribers and/or threateners. All the ballots are different and showing multiples would change nothing.
Hand-counting the vote becomes immensely more difficult. Rather than just going through ballots and seeing which candidate was chosen for each, each ballot has to be looked up, and the lookup process has to be secure.
No, it doesn't. Ballots can be hand-counted just by looking at bubbles next to names.
It appeared to hand that off to other schemes, which don't look to me like they'll work.
I'm not sure what "other schemes" you're referring to.
If I can verify how my vote was counted, that gentleman sitting over there with the large wrench, or that other gentleman with the roll of $20 bills, can ask me to verify it.
Yes you can, and no you can't, respectively. What you can verify is that your receipt (containing a random code that was associated with your ballot selection) was included in the public receipts, and you can verify that the public receipts are tallied correctly, but you can't directly verify that your receipt indicates a vote for Trump rather than Clinton.
The relationships between ballots, candidates and codes is pre-committed in some published tables. When you vote, you can take some extra ballots, reveal all of the codes and validated them against the published tables. Since these "test ballots" are taken from the same pool as the actual ballot (you should take several and pick one at random for your real ballot), any discrepancies in the bindings between codes and candidates would be probabilistically revealed. There are some other validation steps that are taken by election officials and representatives of the candidates; read the paper.
The bottom line is that all you can validate is that your ballot receipt is found in the list of published receipts. The other elements of the process confirm that the presence of your receipt means that your vote was counted, and counted correctly.
This is all laid out fairly clearly in the paper. Note also that the paper was published nearly a decade ago and was authored by some of the biggest names in cryptographic security, which means there has been plenty of time and plenty of motivation for people to find weaknesses in the design. None have been published.