That's right, and because they are already terrible people, these NSA and other
types need to be hunted down and imprisoned like we already do with
rapists and paedos.
Building automated systems that can hunt us down and survey us night and day is a kind of organized crime against humanity and should be recognized as such in the lawbooks. We can't stop everyone from turning to a life of crime, but we should make it clear that if you do, you can expect to live your life on the fringe of society with a large amount of imprisonment thrown in. Kids who are smart and good with technology don't get a free pass on evil deeds, even if they were picked on in school.
Depends, really. Right now the NSA has a lot of brand recognition, so if they market their tools with a recognisable logo they could dominate the global market in a month, tops. They should hire Bill Gates to head their malware product division, I'm sure they could make him an offer he can't refuse.
Porn works like that. Have you ever landed on a porn video site? Most of the videos have no story, and show even less acting ability or camera skills.
But you know what? Nobody cares! It's the same with 50 shades, people don't read it because it's art. Women read it to get ideas and phantasies. And to be honest most porn sites don't cater to women, so they have a limited choice in the matter.
http://bit.ly/1dgDo7d . Come on slashdot editors, do the legwork and link the article directly! Otherwise people will post a link in the comments, and who's to say it's not a goatse?
Anyway, I'm a little worried about the methodology. If you train on PG, and test on PG your generalization error will suffer. This is especially easy to get wrong when both the train and test set are constructed repeatedly with various thresholding rules, and the classifier features are (presumably) optimized during the research being conducted.
However many "problems" with performance today are I/O-based and not
calculation based. It's time for the storage systems to catch up [...]
Meh, that's a copout. IO has always been a bottleneck. Why do you think that Knuth spends so much time optimizing sorting algorithms for tapes? It's not a new issue, solve it by changing your algorithm (aka calculation).
The current generation of programmers are so used to doing cookie cutter work, gluing together lower level libraries that they do not understand in essentially trivial ways, that when they are faced with an actual mismatch between the problem and the assumptions of the lower level code, there is nothing they know how to do. Here's a hint: throw away the lower level crutches, and design a scalable solution from scratch. Most problems can be solved in many ways, and the solution that uses your favourite framework is probably never the most efficient.
I've sometimes seen the marble replaced with.... sticky tape. The beauty of sticky tape is that gravity in the physical model doesn't matter. In fact, the tape approximates locally a section of the tangent bundle, and since it is straight when stretched, it actually integrates the laws of motion for a geodesic on the surface, which is exactly what happens in relativity to an isolated particle that is not under the influence of any other force.
You can make any shape out of cardboard or paper, and put a long strip of sticky tape on it, to see what the trajectory will be, and you can twist and turn the model to better see what's happening. As an aside, if you make hundreds of little triangles and stick them edge to edge, you can approximate any embeddable two-surface to play with. You can also easily see how a "marble" would get trapped in orbit around a black hole, represented by a paper cylinder, if the tape ends up curving back on top of itself.
I've said similar before -- the same goes for their data mining techniques.
Sure it's being used inappropriately, but the fact they are able to collect,
store, and analyze such an humongous data set is really a marvel of computer
science.
No, it's unethical. That's like saying it was ok for Oppenheimer to develop the atomic bomb - what a marvel of physics. The result was that the world ended up facing nuclear annihilation.
Ethics matters. If you're a computer scientist developing data mining techniques to search all the world's data, you SHOULD think about the ethics of it, and either refuse or ALSO work on anti data mining protection systems. No government or company should have the power to search all the world's data. They WILL misuse it, and THEN we'll want those anti data mining systems to be ready.
However, in a statement the company said it listens to customers and announced that all the reported security bugs and suggestions would be fixed and implemented in the next revision of the software - using self modifying code that overwrites itself with random bits after 10 seconds.
The colony on the moon was actually a pretty good prediction. NASA was well on its way to achieve this, when funding was cut and priorities were changed for purely political reasons. And it's not even like NASA's mission was that expensive, compared to the trillions in defense spending that just end up being blown up (literally: a bomb is a way of blowing up the money that built it).
The problem with your suggestion that money should be a gatekeeper system for scarce resources, is that undeserving people inherit money from their parents. So you always get an upper class of people with a guaranteed income (aka trust fund) who did nothing other than were born into the correct family. These people don't deserve to get first (and effectively guaranteed) access to scarce resources when other people can never do so, just because their parents happen to be poor.
*note: this is a guess, but it's based on a random supposition that - in the
last year - the governments we are discussing (US, UK, EU) have targeted less
than 700 completely innocent people in any given year using the NSAs (or UK
or EU equiv.) surveillance dragnets. If you have a list longer than that,
Are you out of your mind? In the last year these governments have targeted us all. Oh wait, you like to make a distinction between stalking everybody in general, and "targeting" specific individuals (who, by the way? If you can't name them, you shouldn't claim there's 700 or so).
That's a classic propaganda technique, using words in highly specific ways to obfuscate reality. We're not allowed to use the word "targeting" in its traditional meaning, akin to being in the cross hairs of a surveillance system. We now have to use it explicitly only in a very rare case, and if we no longer have the vocabulary to talk about the illegal surveillance that is happening, too bad!
The truth is that we have all been targeted. When you're facing a system that can target millions of people simultaneously, you're still a target.
The data is being collected right now, and is being stored for later, when the next Bush Jr wannabe gets to exploit it. He'll just sign an executive order to "target" anyone he doesn't like when the time comes, using all that existing data retroactively. And that's only possible because you're a target, *right now*.
I don't understand why that is such a big deal anyway. They are going to spam
me with ads one way or the other; at least if I find value in the product or
service being advertised, it's less of a waste of my time and perhaps it's
even a valuable proposition.
You're a slave. You're so used to being taken for a ride, you don't even know what the alternatives are like anymore.
Spamming is illegal. You can actually complain, and you can actually escalate to your provider's help desk, and you can actually write blogs naming and shaming spammers, and you can actually block ads in your browser, and you can actually do a whole lot of other specific things in specific cases.
But you'd have to get off your ass and do it, rather than promoting inaction and defeatism on slashdot.
The boundaries or human knowledge are easy to see if you're educated, and vague otherwise. The algorithm for identifying the boundary is therefore: 1) learn everything everyone knows about the subject. This takes about 7 - 14 years. 2) If after 1), you don't know the answer and you know or suspect that everyone else doesn't know the answer, then it's on the other side.
It's not rocket science, lots of people go through 1) and 2).
Wrong. Many CCTVs are related, and record data for a limited amount of time. There's a world of difference between a CCTV pointed at a carpark, and some glasshole slaking people around town. The most important difference is that an evil company named Google doesn't get to sift through the CCTV footage.
You're way off about the reason people object. Disliking being recorded is not the main problem, it's what can be done with the images, and where they might end up and who might see them. Nobody gives a fuck if a security guard sees you walking down a carpark, but if that picture is seen by people you know, or prospective employers, or any number of other uses that might harm you later, then tha's evil. And that is exactly what Google/Doubleclick (remember them?) wants to do.
When is using a graphical interface (especially X or the OS/X desktop) not using UNIX?
Quite often. The UNIX philosophy effectively requires programs to not make a distinction between human users, and scripts or robots. When a UNIX program produces output, it can be accepted directly as input to some other program with very little (awk)wardness (pun intended). When a UNIX program requires input, all of the requirements can be specified formally, using the command line or the standard input. Thus is makes no difference if a human operates the program, or a bash(1) script, or an expect(1) session.
By contrast, GUI interfaces are ONLY intended for humans using a particular combination of screen and pointer technology (think Desktop vs Tablet as the latest example of this), have no way to specify inputs in a script friendly way, and do not produce output that any other program can use, only a human being. Yes these limitations of GUI programs can be remedied in various ways by introducing command line paraphernalia, but this usually ends up in a haphazard implementation which is almost, but not quite, in sync with a limited subset of the GUI interface's capabilities, causing uncertainty as to which interface is the most capable.
Why? This is nonsense. People who are unmotivated to begin with won't be good programmers, because programming involves lots of reading of boring reference manuals, lots of attention to detail, and mind-numbing repeats of the same running program, over and over, to drill down exactly where it's not doing exactly what's required.
Being a mental masochist is not sufficient for being a good programmer, but it is certainly necessary. Anyone who isn't even willing to master a semi-boring CS class doesn't pass muster.
You'd rather have the wild west than the tamed east coast? Fancy yourself a gunslinger rather than a little old lady? Meh, too many analogies.
It all depends where you are in the economic food chain. Regulation makes sense for people who don't have the individual power, or the time, or the expertise, to defend themselves. That's the majority of people, including me, and probably including yourself. There's always a handful of people who have the money and the connections to control or intimidate everyone who tries to fuck with them. Those people thrive in an anarchic free for all.
Beat me to it. The internet will "remember" you so long as the information about you is perceived to have value.
Which basically means as long as you live, and probably half way through your childrens' lives. And that is exactly the problem in the first place. Other people and companies having hoards of information about you is bad, as it given them power over you. For as long as you live.
Meh, 1984 came out 65 years ago, did you skip it? In case you haven't read it, they had two way TVs on practically every wall to both spy on people, and send propaganda messages at random times. Orwell didn't live in and Americanized world full of commercial ads, but in fact ads are a form of propaganda, just not political.
Meh, there are plenty of countermeasures we can develop.
On the "nice" side, all legally sold cameras should have a certified DRM system coupled with a well engineered peer to peer wifi protocol that will ask all devices within a radius of 50m if any pictures are allowed to be taken. Then everybody can wear a tiny ring on their finger which broadcasts a yes/no response. If you want to get fancy, if the ring is GPS enabled and records the time and place of any mandatory query, and if "legal" photographs must have a time and date embedded, it becomes easy to check after the fact if a particular picture was "legally" taken: do a search on the records for anyone who was within a 50m radius of the photograph's time and place, and verify if all the recorded responses were affirmative.
On the "dark" side, all such cameras need some way to connect to the internet. That means they are vulnerable to viruses. We simply develop sophisticated viruses that attack cameras and wipe them, or just futz with the settings so they are randomly out of focus etc. The nice thing about viruses is that everyone can carry them on their smart phones, and there's plausible deniability. What? Your camera got fucked with when you came close to me? I'm sorry, I didn't know about them virus thingamajigs. I don't do tech. It will be a never ending measure/countermeasure race, but I'm sure we can find dedicated hackers who are willing to take on this burden.
It all comes down to the following question: are the spyware industry giants like Google willing to self limit their technology in a foolproof way? (I don't know if that's possible, even for them) If not, the tech community can sabotage it for them, free of charge. It'll be fun, like open source.
Anybody who wears nearly impossible to detect spyglasses and gets found out has their nose punched in. That is just the price to be paid. Same with glass-holes, don't play the game if you can't take the blame.
Slashdot Mirror
Building automated systems that can hunt us down and survey us night and day is a kind of organized crime against humanity and should be recognized as such in the lawbooks. We can't stop everyone from turning to a life of crime, but we should make it clear that if you do, you can expect to live your life on the fringe of society with a large amount of imprisonment thrown in. Kids who are smart and good with technology don't get a free pass on evil deeds, even if they were picked on in school.
Depends, really. Right now the NSA has a lot of brand recognition, so if they market their tools with a recognisable logo they could dominate the global market in a month, tops. They should hire Bill Gates to head their malware product division, I'm sure they could make him an offer he can't refuse.
But you know what? Nobody cares! It's the same with 50 shades, people don't read it because it's art. Women read it to get ideas and phantasies. And to be honest most porn sites don't cater to women, so they have a limited choice in the matter.
Anyway, I'm a little worried about the methodology. If you train on PG, and test on PG your generalization error will suffer. This is especially easy to get wrong when both the train and test set are constructed repeatedly with various thresholding rules, and the classifier features are (presumably) optimized during the research being conducted.
Meh, that's a copout. IO has always been a bottleneck. Why do you think that Knuth spends so much time optimizing sorting algorithms for tapes? It's not a new issue, solve it by changing your algorithm (aka calculation).
The current generation of programmers are so used to doing cookie cutter work, gluing together lower level libraries that they do not understand in essentially trivial ways, that when they are faced with an actual mismatch between the problem and the assumptions of the lower level code, there is nothing they know how to do. Here's a hint: throw away the lower level crutches, and design a scalable solution from scratch. Most problems can be solved in many ways, and the solution that uses your favourite framework is probably never the most efficient.
You can make any shape out of cardboard or paper, and put a long strip of sticky tape on it, to see what the trajectory will be, and you can twist and turn the model to better see what's happening. As an aside, if you make hundreds of little triangles and stick them edge to edge, you can approximate any embeddable two-surface to play with. You can also easily see how a "marble" would get trapped in orbit around a black hole, represented by a paper cylinder, if the tape ends up curving back on top of itself.
No, it's unethical. That's like saying it was ok for Oppenheimer to develop the atomic bomb - what a marvel of physics. The result was that the world ended up facing nuclear annihilation.
Ethics matters. If you're a computer scientist developing data mining techniques to search all the world's data, you SHOULD think about the ethics of it, and either refuse or ALSO work on anti data mining protection systems. No government or company should have the power to search all the world's data. They WILL misuse it, and THEN we'll want those anti data mining systems to be ready.
However, in a statement the company said it listens to customers and announced that all the reported security bugs and suggestions would be fixed and implemented in the next revision of the software - using self modifying code that overwrites itself with random bits after 10 seconds.
The colony on the moon was actually a pretty good prediction. NASA was well on its way to achieve this, when funding was cut and priorities were changed for purely political reasons. And it's not even like NASA's mission was that expensive, compared to the trillions in defense spending that just end up being blown up (literally: a bomb is a way of blowing up the money that built it).
The problem with your suggestion that money should be a gatekeeper system for scarce resources, is that undeserving people inherit money from their parents. So you always get an upper class of people with a guaranteed income (aka trust fund) who did nothing other than were born into the correct family. These people don't deserve to get first (and effectively guaranteed) access to scarce resources when other people can never do so, just because their parents happen to be poor.
Are you out of your mind? In the last year these governments have targeted us all. Oh wait, you like to make a distinction between stalking everybody in general, and "targeting" specific individuals (who, by the way? If you can't name them, you shouldn't claim there's 700 or so).
That's a classic propaganda technique, using words in highly specific ways to obfuscate reality. We're not allowed to use the word "targeting" in its traditional meaning, akin to being in the cross hairs of a surveillance system. We now have to use it explicitly only in a very rare case, and if we no longer have the vocabulary to talk about the illegal surveillance that is happening, too bad!
The truth is that we have all been targeted. When you're facing a system that can target millions of people simultaneously, you're still a target. The data is being collected right now, and is being stored for later, when the next Bush Jr wannabe gets to exploit it. He'll just sign an executive order to "target" anyone he doesn't like when the time comes, using all that existing data retroactively. And that's only possible because you're a target, *right now*.
You're a slave. You're so used to being taken for a ride, you don't even know what the alternatives are like anymore.
Spamming is illegal. You can actually complain, and you can actually escalate to your provider's help desk, and you can actually write blogs naming and shaming spammers, and you can actually block ads in your browser, and you can actually do a whole lot of other specific things in specific cases.
But you'd have to get off your ass and do it, rather than promoting inaction and defeatism on slashdot.
It's not rocket science, lots of people go through 1) and 2).
You're way off about the reason people object. Disliking being recorded is not the main problem, it's what can be done with the images, and where they might end up and who might see them. Nobody gives a fuck if a security guard sees you walking down a carpark, but if that picture is seen by people you know, or prospective employers, or any number of other uses that might harm you later, then tha's evil. And that is exactly what Google/Doubleclick (remember them?) wants to do.
Quite often. The UNIX philosophy effectively requires programs to not make a distinction between human users, and scripts or robots. When a UNIX program produces output, it can be accepted directly as input to some other program with very little (awk)wardness (pun intended). When a UNIX program requires input, all of the requirements can be specified formally, using the command line or the standard input. Thus is makes no difference if a human operates the program, or a bash(1) script, or an expect(1) session.
By contrast, GUI interfaces are ONLY intended for humans using a particular combination of screen and pointer technology (think Desktop vs Tablet as the latest example of this), have no way to specify inputs in a script friendly way, and do not produce output that any other program can use, only a human being. Yes these limitations of GUI programs can be remedied in various ways by introducing command line paraphernalia, but this usually ends up in a haphazard implementation which is almost, but not quite, in sync with a limited subset of the GUI interface's capabilities, causing uncertainty as to which interface is the most capable.
Being a mental masochist is not sufficient for being a good programmer, but it is certainly necessary. Anyone who isn't even willing to master a semi-boring CS class doesn't pass muster.
Whoa! They could call it Just In Time(TM) Spying.
That's not research, that's development. Research is extending the boundaries of human knowledge.
Duh. Of course you first take their glasses off from behind, and smash them on the ground. Give people a modicum of credit, not everybody is a moron.
It all depends where you are in the economic food chain. Regulation makes sense for people who don't have the individual power, or the time, or the expertise, to defend themselves. That's the majority of people, including me, and probably including yourself. There's always a handful of people who have the money and the connections to control or intimidate everyone who tries to fuck with them. Those people thrive in an anarchic free for all.
It's not your call if "the internet" remembers you. And that's the problem.
Which basically means as long as you live, and probably half way through your childrens' lives. And that is exactly the problem in the first place. Other people and companies having hoards of information about you is bad, as it given them power over you. For as long as you live.
Meh, 1984 came out 65 years ago, did you skip it? In case you haven't read it, they had two way TVs on practically every wall to both spy on people, and send propaganda messages at random times. Orwell didn't live in and Americanized world full of commercial ads, but in fact ads are a form of propaganda, just not political.
On the "nice" side, all legally sold cameras should have a certified DRM system coupled with a well engineered peer to peer wifi protocol that will ask all devices within a radius of 50m if any pictures are allowed to be taken. Then everybody can wear a tiny ring on their finger which broadcasts a yes/no response. If you want to get fancy, if the ring is GPS enabled and records the time and place of any mandatory query, and if "legal" photographs must have a time and date embedded, it becomes easy to check after the fact if a particular picture was "legally" taken: do a search on the records for anyone who was within a 50m radius of the photograph's time and place, and verify if all the recorded responses were affirmative.
On the "dark" side, all such cameras need some way to connect to the internet. That means they are vulnerable to viruses. We simply develop sophisticated viruses that attack cameras and wipe them, or just futz with the settings so they are randomly out of focus etc. The nice thing about viruses is that everyone can carry them on their smart phones, and there's plausible deniability. What? Your camera got fucked with when you came close to me? I'm sorry, I didn't know about them virus thingamajigs. I don't do tech. It will be a never ending measure/countermeasure race, but I'm sure we can find dedicated hackers who are willing to take on this burden.
It all comes down to the following question: are the spyware industry giants like Google willing to self limit their technology in a foolproof way? (I don't know if that's possible, even for them) If not, the tech community can sabotage it for them, free of charge. It'll be fun, like open source.
Anybody who wears nearly impossible to detect spyglasses and gets found out has their nose punched in. That is just the price to be paid. Same with glass-holes, don't play the game if you can't take the blame.