The 80% figure, which is the AUC (area under the curve), refers to threshold tuning. In order to make that usable in the Real World, you'd have to crank it so that it has nearly zero false positives (and thus very few detected trolls) or else you'd have to make it flag posts non-fatally, perhaps with nearly impossible captchas, which immediately defeats its anti-troll utility (not to mention angering all of the falsely identified trolls!).
The article, like the paper itself, ends on this note:
Regarding the possibility of developing automated methods for identifying and even banning trolls, the researchers are circumspect, since 1 in 5 of users were misclassified by their analysis system, which otherwise claims to spot a persistent comment pest within as few as ten posts. “While we present effective mechanisms for identifying and potentially weeding antisocial users out of a community, taking extreme action against small infractions can exacerbate antisocial behavior (e.g., unfairness can cause users to write worse)“
In Time Warner's quarterly earnings call on 2013/08/07, Tuna Amobi of S&P U.S. Equity Research Services asked CEO Jeff Bewkes about piracy and Game of Thrones. The end of the call, as transcribed (and emphasized) by Business Insider:
Tuna Amobi:Game of Thrones has obviously had a phenomenal performance, but one other issue that has come up with regards to that title is the online piracy. I think by all accounts one of the highest pirated shows and I'm not aware what you guys have done to kind of address that. It seems that you have viewed it as kind of a compliment in terms of looking the other way so much. Is that the right way of thinking? Kind of a paradigm shift with the piracy and its impact on shows going forward that what you've done.
[...]
Time Warner CEO Jeff Bewkes: To end on Game of Thrones on HBO, I have to confess I think you're right. I have to admit, our first reaction to how much people want to watch it — now first of all it's got ratings of 14, 15 million — a lot of it is VOD [video on demand] on your TV system, an increasing amount of it is VOD on your [HBO]Go Service.
It's just really strengthening not just the image, but the engagement of our subs [subscribers] with HBO programming, it's also getting them familiar and more involved with using the video on demand capabilities of HBO and don't forget, the television part. The part where you go to your house and you turn on that big screen TV watching it over the video plan, also the HBO Go service where Game of Thrones is the leading introduction manual for how to use HBO Go which more and more people are doing.
Then go to people watching it who aren't subs, it's a tremendous word of mouth thing, the issue would be if they were doing it and because they could get it not subscribing, we don't see much of that.
Basically, we've been dealing with this issue for years with HBO, literally 20, 30 years, where people have always been running wires down on the back of apartment buildings and sharing with their neighbors.
Our experience is, it all leads to more penetration, more paying subs, more health for HBO, less reliance on having to do paid advertising — we don't do a whole lot of paid advertising on HBO, we let the programming and the views talk for us — it seems to be working.
If you go around the world, I think you're right, Game of Thrones is the most pirated show in the world. Well, you know, that's better than an Emmy. (laughter)
one of the best ways to route around a big firm's brand recognition is to buy special treatment in the form of promotions, product placement and the like (payola, after all, is how rock and roll circumvented major label contempt for the genre)
Promotions and product placement are not shoving unsolicited third-party ads into Google or throttling Netflix, they are buying ads from Google or getting characters in movies to use the promoted brand. If you happen to see a movie on Netflix in which the characters are talking about a show on Hulu, you'll know it works (and works well).
This has nothing to do with net neutrality, which is a far better tool at doing the opposite; a big player like Netflix can pay the ransom and get special treatment, a up-and-coming startup video streaming service can't pony up the resources to do that, but perhaps they can get a celebrity to name-drop their brand in an ad-libbed line of a hot movie.
This is the fallacy of small numbers, a.k.a. hasty generalization. There weren't many CS majors (of either gender) in the 80s, so the gender ratio will be less representative of a real trend (consider flipping ten coins. Your probability of getting 50% heads isn't as good as it would be if you flipped a thousand coins). Most of my software engineer peers who got degrees in that era actually studied other fields, such as math or electrical engineering.
That said, the drop from 10-15 years ago is completely valid and this is indeed a problem.
(disclaimer: I did not listen to that story and I don't have stats at the ready to prove my observations)
They had one of these (not necessarily this vendor) on the floor of one of the wings of the Burlington Mall in Massachusetts 5+ years ago (it may still be there). It's a fun toy, but it has little practical applications beyond games and promotions. There's no reason this couldn't be on a wall or table though.
Restaurants: I see this technology as the future of table service at restaurants; consider your white tablecloth as your touchscreen, capable of breaking down into one screen per patron (the camera notes where people are seated) or one big screen for everybody to watch a video presentation. This becomes your menu. The camera can also note when you are running out of drink, when it's appropriate to bring out your next course, and when to clear your plates, which allows the wait staff to better optimize their time. Perhaps the bussers are even drones.
Gaming: A ceiling-mounted camera and projector are far cheaper than a coffee-table sized tablet, and you don't have to worry about spilling drinks on your tabletop destroying your system. This can replace board game equipment and other tabletop games and activities. Giant jigsaw puzzles and multi-day wargames can be saved and cleared to make room for something else, then resumed on demand.
It does not appear that this drone operator was making money himself. The FAA doesn't want a cut of the profit (even 100% of $0 is zero), so this is perhaps more complicated than it may seem.
That said, even if they were to demand a cut of Google's profits from the YouTube ads, the collection process would cost the FAA more than the take-home.
The FCC says this is "commercial" because the drone's videos were posted to YouTube and because YouTube has advertisements, even though the drone operator gets zero profit from those ads.
... what if the drone were flying a banner (and not recording a video)? Is that an advertisement? What if the banner said "Vote for Joe Candidate" and nothing else?
Yeh, seriously, Nevral's Lobster shows an exceptional lack of journalistic integrity by being an employee of dice.com and posting nothing but dice.com stories--WITHOUT A DISCLAIMER.
Hm, I hadn't realized that Nevral's Lobster was exclusively a Dice.com sock puppet. That's fine for the submissions, but not so fine for the accepted stories, which an editor (ideally) more affiliated with Slashdot than other Dice holdings should have curated and appended the standard disclaimer after Nevral's Lobster's quote.
Don't forget your journalistic rigor. I know it's so very often forgotten these days, but I've chosen Slashdot as one of my last "traditional" news outlets (in the sense that it the editors, including Nevral's Lobster, are paid to curate the content) because it used to be better about this. It is irresponsible of Slashdot to omit the fact that Dice owns Slashdot in the article summary.
I can't find references to the actual license text, but the expectation of paying royalties back to Epic certainly makes it non-free with respect to software freedom. This makes it incompatible in the same sense that the Creative Commons License's "noncommercial" clause is incompatible; most copyleft licenses insist on unrestricted redistribution (which would be broken by a requirement of paying royalties).
The video notes that this is "unprecedented," yet Epic's competitor Id Software used to release all of its engines as GPLv2 once they were ~two generations obsolete (e.g. Doom 3). No royalties expectations necessary.
All studies I've seen have suggested that more intervention, as early as possible, is ideal. The idea of play groups and other less formal types of socialization seems pretty good to me, perhaps it would serve as a better control for future studies (I'm not that well read, perhaps some research paper has already done this?).
The main point to all of this is that your son needs as much social opportunity as possible, and it needs to be NOW. That said, you can't really afford not to use as much of each option as you can. There is no opportunity to "fix" this later.
Opera has been using the same rendering (and JS?) engine (Blink) as Chrome for over 1.5 years (ever since Opera 15), so you may only have to run your preferred choice of Opera vs Chrome in addition to Firefox on Linux.
Firefox Hello bundles this kind of thing right into the web browser. I kind of like this idea for allowing basic functionality (think of the browser-based IM in Google and Facebook) and even extending that to voice and video (the way Google Hangouts does), but I'd ideally like to see a more powerful stand-alone client for people that want more than just a few casual conversations here and there. (This is an even better idea for Thunderbird, since your contact list lives there.)
Fortunately, we have pidgin, a stand-alone IM client with a great feature set and wonderful cross-platform support (Adium is merely an OS X implementation of Pidgin). Pidgin desperately needs help, as it hasn't successfully had an easy-to-use voice (let alone video) capability. I'm hoping that WebRTC (which powers Firefox Hello and, I think, Google Hangouts) can provide this, at least for using Firefox Hello and/or bridging between two Pidgin/Adium/Libpurple users.
Banks are secure because they lock your account when you fail to log in ~three consecutive times. Doesn't matter over what time period or what IP address you are using.
This is rather aggressive; somebody can lock your account with knowledge of your username, but it makes sense. One trick I use: my financial usernames are rather passwordlike (in that you're not going to guess them easily).
Who is to say that we'd even be able to conceive these "robots" as anything but another form of life? They'd have to tell us that they were manufactured, and given the required self-sufficiency of space travel, said "manufacture" would probably be rather akin to what we call "reproduction." All of these lines are blurred when talking about sufficiently advanced technology and science.
Neil deGrasse Tyson's video pleas We Stopped Dreaming and its follow-up A New Perspective proposed we increase NASA spending to 1% of the US Federal Budget (current spending: 0.5%)
suggests we could go to Mars and innovate the way we did in the 70s, so there's a long way to go (a 2% boost leaves us 98% shy of Tyson's goal).
NASA is already trying to plan a manned mission to Mars or an asteroid in the future. It would be nice if they were funded for it.
Almost one-fourth of video ads and 11 percent of display ads are viewed by fake consumers created by cyber crime networks seeking to take a chunk of the billions of dollars spent on digital advertising
I think getting "clicks" from actual targeted customers is a non-problem in the face of all this other fraud. When it comes to security research (my field), more information pretty much always leads to better verdicts. It's therefore quite reasonable that you want to crawl an extra step deep in order to vet a page you're on. This isn't even unprecedented; think of the browser link prefetching, which anticipates where you'll click and downloads content ahead of time.
our educational system is still probably the best at producing software engineers
First, the tech industry is not just software engineers.
Second, while the current US educational system is very good at producing people who can drive good design, it's not so great at producing people who can implement it. The raw technical chops, especially with respect to understanding of advanced mathematics, is a rarity here in the US compared to (e.g.) much of the EU.
This has been proposed before, but perhaps not strongly enough or from an important enough source, which is too bad because it solves practically all of our worries.
The premise is simple: the tech industry doesn't have enough good workers because our education system is not well suited to producing the necessary skill sets. Therefore, allow qualified talent to come in and fill that gap. Tax employers based on their salaries (for this to work, salaries must be lower rather than having the same salaries with extra deductions). The collected taxes would be directed to improve our educational systems (K-12 as well as public universities) so that this problem goes away. In time, it won't be worthwhile for an employer to consider this type of talent acquisition because qualified US citizens would be more readily available and would cost less (due to not requiring this proposed tax).
(Sorry if I got some of that terminology wrong; I'm not in HR nor do I deal with immigration paperwork.)
Nissan sells 5,000,000 cars per year and made US$3 billion in profit last FY. Nissan makes good cars that sell well, pretty much the antithesis of American car corporations, so they're quite safe.
Yes, you are currently correct, but I'm talking about before the Leaf was released.
The story was that they had invested all of their research into batteries and then made a major play to be the first to market for plug-ins (be they electric vehicles or hybrids). The Leaf uses their advanced batteries and serves as a demonstration of a very basic electric car (the Leaf started as merely a Versa converted to be an EV). With the Leaf's success, Nissan is on its way to having the same kind of dominance in hybrid/electric car batteries that Toyota has in regenerative braking (which is leased by many competitors).
Warren, in the event you're reading the comments (and at a threshold low enough to see this...),
Can you speak more to why you listed so many cookbooks? I can understand René Redzepi's Work In Progress because, as Amazon notes, "it includes a personal journal written by René himself over a full year in which he explores creativity, innovation, and the meaning and challenges of success," but are all of the books on cuisine in the same light? Or are you an avid amateur chef? You've definitely given me something to seek out the next time I go to Iceland.
Contrast this with Nissan, another Japanese automobile manufacturer, which has invested so deeply into battery technology that if the Leaf were to fail, it's quite likely that they'd become a battery company. (A while back, I read (or watched?) a really compelling article/documentary on Nissan's battery research. It concluded that Nissan was gambling so heavily on both its own future with the Leaf and the future of automobiles as being electric that the company would likely stop making cars if the Leaf were to fail. Sorry I can't find a good citation to that.)
The presumption that Hydrogen Fuel Cells will be the "next" car fuel (after either gas or after electric) is still quite a strong one. I've seen it painted (iirc, by the documentary Who Killed the Electric Car?) as something the oil companies latched onto because it competed with electric cars (which are ready now) and because hydrogen fuel cell cars are still quite a distant future prospect.
This paper has not been peer reviewed. Read with that in mind (peer review is academic currency).
(That said, it doesn't get much more prestigious than Science. It's merely too early to bring this to bear. Perhaps it will eventually get accepted, reviewed, and then published. Only at that point can it be considered good research.)
When writing within a popular series (e.g. X-Men or Hellblazer), there are certain hard limits in what liberties you can take. As a mundane example, you can't kill characters without planning out a large arc that builds up to it and/or quickly bringing them back, all with editorial approval from up on high.
What would you write within a popular series if only you could get permission to do it?
Slashdot Mirror
The 80% figure, which is the AUC (area under the curve), refers to threshold tuning. In order to make that usable in the Real World, you'd have to crank it so that it has nearly zero false positives (and thus very few detected trolls) or else you'd have to make it flag posts non-fatally, perhaps with nearly impossible captchas, which immediately defeats its anti-troll utility (not to mention angering all of the falsely identified trolls!).
The article, like the paper itself, ends on this note:
one of the best ways to route around a big firm's brand recognition is to buy special treatment in the form of promotions, product placement and the like (payola, after all, is how rock and roll circumvented major label contempt for the genre)
Promotions and product placement are not shoving unsolicited third-party ads into Google or throttling Netflix, they are buying ads from Google or getting characters in movies to use the promoted brand. If you happen to see a movie on Netflix in which the characters are talking about a show on Hulu, you'll know it works (and works well).
This has nothing to do with net neutrality, which is a far better tool at doing the opposite; a big player like Netflix can pay the ransom and get special treatment, a up-and-coming startup video streaming service can't pony up the resources to do that, but perhaps they can get a celebrity to name-drop their brand in an ad-libbed line of a hot movie.
http://www.npr.org/blogs/money/2014/10/21/357629765/when-women-stopped-coding
I'd encourage you to listen to the story as well.
This is the fallacy of small numbers, a.k.a. hasty generalization. There weren't many CS majors (of either gender) in the 80s, so the gender ratio will be less representative of a real trend (consider flipping ten coins. Your probability of getting 50% heads isn't as good as it would be if you flipped a thousand coins). Most of my software engineer peers who got degrees in that era actually studied other fields, such as math or electrical engineering.
That said, the drop from 10-15 years ago is completely valid and this is indeed a problem.
(disclaimer: I did not listen to that story and I don't have stats at the ready to prove my observations)
They had one of these (not necessarily this vendor) on the floor of one of the wings of the Burlington Mall in Massachusetts 5+ years ago (it may still be there). It's a fun toy, but it has little practical applications beyond games and promotions. There's no reason this couldn't be on a wall or table though.
Restaurants: I see this technology as the future of table service at restaurants; consider your white tablecloth as your touchscreen, capable of breaking down into one screen per patron (the camera notes where people are seated) or one big screen for everybody to watch a video presentation. This becomes your menu. The camera can also note when you are running out of drink, when it's appropriate to bring out your next course, and when to clear your plates, which allows the wait staff to better optimize their time. Perhaps the bussers are even drones.
Gaming: A ceiling-mounted camera and projector are far cheaper than a coffee-table sized tablet, and you don't have to worry about spilling drinks on your tabletop destroying your system. This can replace board game equipment and other tabletop games and activities. Giant jigsaw puzzles and multi-day wargames can be saved and cleared to make room for something else, then resumed on demand.
It does not appear that this drone operator was making money himself. The FAA doesn't want a cut of the profit (even 100% of $0 is zero), so this is perhaps more complicated than it may seem.
That said, even if they were to demand a cut of Google's profits from the YouTube ads, the collection process would cost the FAA more than the take-home.
The FCC says this is "commercial" because the drone's videos were posted to YouTube and because YouTube has advertisements, even though the drone operator gets zero profit from those ads.
... what if the drone were flying a banner (and not recording a video)? Is that an advertisement? What if the banner said "Vote for Joe Candidate" and nothing else?
Yeh, seriously, Nevral's Lobster shows an exceptional lack of journalistic integrity by being an employee of dice.com and posting nothing but dice.com stories--WITHOUT A DISCLAIMER.
Hm, I hadn't realized that Nevral's Lobster was exclusively a Dice.com sock puppet. That's fine for the submissions, but not so fine for the accepted stories, which an editor (ideally) more affiliated with Slashdot than other Dice holdings should have curated and appended the standard disclaimer after Nevral's Lobster's quote.
Dear Slashdot editors,
Don't forget your journalistic rigor. I know it's so very often forgotten these days, but I've chosen Slashdot as one of my last "traditional" news outlets (in the sense that it the editors, including Nevral's Lobster, are paid to curate the content) because it used to be better about this. It is irresponsible of Slashdot to omit the fact that Dice owns Slashdot in the article summary.
I can't find references to the actual license text, but the expectation of paying royalties back to Epic certainly makes it non-free with respect to software freedom. This makes it incompatible in the same sense that the Creative Commons License's "noncommercial" clause is incompatible; most copyleft licenses insist on unrestricted redistribution (which would be broken by a requirement of paying royalties).
The video notes that this is "unprecedented," yet Epic's competitor Id Software used to release all of its engines as GPLv2 once they were ~two generations obsolete (e.g. Doom 3). No royalties expectations necessary.
All studies I've seen have suggested that more intervention, as early as possible, is ideal. The idea of play groups and other less formal types of socialization seems pretty good to me, perhaps it would serve as a better control for future studies (I'm not that well read, perhaps some research paper has already done this?).
The main point to all of this is that your son needs as much social opportunity as possible, and it needs to be NOW. That said, you can't really afford not to use as much of each option as you can. There is no opportunity to "fix" this later.
Opera has been using the same rendering (and JS?) engine (Blink) as Chrome for over 1.5 years (ever since Opera 15), so you may only have to run your preferred choice of Opera vs Chrome in addition to Firefox on Linux.
Firefox Hello bundles this kind of thing right into the web browser. I kind of like this idea for allowing basic functionality (think of the browser-based IM in Google and Facebook) and even extending that to voice and video (the way Google Hangouts does), but I'd ideally like to see a more powerful stand-alone client for people that want more than just a few casual conversations here and there. (This is an even better idea for Thunderbird, since your contact list lives there.)
Fortunately, we have pidgin, a stand-alone IM client with a great feature set and wonderful cross-platform support (Adium is merely an OS X implementation of Pidgin). Pidgin desperately needs help, as it hasn't successfully had an easy-to-use voice (let alone video) capability. I'm hoping that WebRTC (which powers Firefox Hello and, I think, Google Hangouts) can provide this, at least for using Firefox Hello and/or bridging between two Pidgin/Adium/Libpurple users.
Banks are secure because they lock your account when you fail to log in ~three consecutive times. Doesn't matter over what time period or what IP address you are using.
This is rather aggressive; somebody can lock your account with knowledge of your username, but it makes sense. One trick I use: my financial usernames are rather passwordlike (in that you're not going to guess them easily).
Who is to say that we'd even be able to conceive these "robots" as anything but another form of life? They'd have to tell us that they were manufactured, and given the required self-sufficiency of space travel, said "manufacture" would probably be rather akin to what we call "reproduction." All of these lines are blurred when talking about sufficiently advanced technology and science.
Neil deGrasse Tyson's video pleas We Stopped Dreaming and its follow-up A New Perspective proposed we increase NASA spending to 1% of the US Federal Budget (current spending: 0.5%) suggests we could go to Mars and innovate the way we did in the 70s, so there's a long way to go (a 2% boost leaves us 98% shy of Tyson's goal).
NASA is already trying to plan a manned mission to Mars or an asteroid in the future. It would be nice if they were funded for it.
For example, Fraud from bots represents a loss of $6 billion in digital advertising @Reuters says
I think getting "clicks" from actual targeted customers is a non-problem in the face of all this other fraud. When it comes to security research (my field), more information pretty much always leads to better verdicts. It's therefore quite reasonable that you want to crawl an extra step deep in order to vet a page you're on. This isn't even unprecedented; think of the browser link prefetching, which anticipates where you'll click and downloads content ahead of time.
First, the tech industry is not just software engineers.
Second, while the current US educational system is very good at producing people who can drive good design, it's not so great at producing people who can implement it. The raw technical chops, especially with respect to understanding of advanced mathematics, is a rarity here in the US compared to (e.g.) much of the EU.
This has been proposed before, but perhaps not strongly enough or from an important enough source, which is too bad because it solves practically all of our worries.
The premise is simple: the tech industry doesn't have enough good workers because our education system is not well suited to producing the necessary skill sets. Therefore, allow qualified talent to come in and fill that gap. Tax employers based on their salaries (for this to work, salaries must be lower rather than having the same salaries with extra deductions). The collected taxes would be directed to improve our educational systems (K-12 as well as public universities) so that this problem goes away. In time, it won't be worthwhile for an employer to consider this type of talent acquisition because qualified US citizens would be more readily available and would cost less (due to not requiring this proposed tax).
(Sorry if I got some of that terminology wrong; I'm not in HR nor do I deal with immigration paperwork.)
Yes, you are currently correct, but I'm talking about before the Leaf was released.
The story was that they had invested all of their research into batteries and then made a major play to be the first to market for plug-ins (be they electric vehicles or hybrids). The Leaf uses their advanced batteries and serves as a demonstration of a very basic electric car (the Leaf started as merely a Versa converted to be an EV). With the Leaf's success, Nissan is on its way to having the same kind of dominance in hybrid/electric car batteries that Toyota has in regenerative braking (which is leased by many competitors).
Warren, in the event you're reading the comments (and at a threshold low enough to see this...),
Can you speak more to why you listed so many cookbooks? I can understand René Redzepi's Work In Progress because, as Amazon notes, "it includes a personal journal written by René himself over a full year in which he explores creativity, innovation, and the meaning and challenges of success," but are all of the books on cuisine in the same light? Or are you an avid amateur chef? You've definitely given me something to seek out the next time I go to Iceland.
"Prius" is Latin for "before" while "Mirai" is Japanese for "Future." Kind of sets a bold statement; an old language for hybrids and a new language (Japanese roughly dates back to the 8th century) for the purported future of cars ... which still has yet to be determined.
Contrast this with Nissan, another Japanese automobile manufacturer, which has invested so deeply into battery technology that if the Leaf were to fail, it's quite likely that they'd become a battery company. (A while back, I read (or watched?) a really compelling article/documentary on Nissan's battery research. It concluded that Nissan was gambling so heavily on both its own future with the Leaf and the future of automobiles as being electric that the company would likely stop making cars if the Leaf were to fail. Sorry I can't find a good citation to that.)
The presumption that Hydrogen Fuel Cells will be the "next" car fuel (after either gas or after electric) is still quite a strong one. I've seen it painted (iirc, by the documentary Who Killed the Electric Car?) as something the oil companies latched onto because it competed with electric cars (which are ready now) and because hydrogen fuel cell cars are still quite a distant future prospect.
This paper has not been peer reviewed. Read with that in mind (peer review is academic currency).
(That said, it doesn't get much more prestigious than Science. It's merely too early to bring this to bear. Perhaps it will eventually get accepted, reviewed, and then published. Only at that point can it be considered good research.)
When writing within a popular series (e.g. X-Men or Hellblazer), there are certain hard limits in what liberties you can take. As a mundane example, you can't kill characters without planning out a large arc that builds up to it and/or quickly bringing them back, all with editorial approval from up on high.
What would you write within a popular series if only you could get permission to do it?
Let's not confuse extreme metallicity (the rare star containing nearly zero hydrogen or helium) with an all-metal body.