It's probably time for Facebook to give up trying to be the morality police, because it isn't working. While nobody expects the social media giant to be perfect at policing its site for images and posts deemed "offensive", it's shown itself time and time again to be utterly incapable of getting this right at even the most basic level. After all, when the censors are removing iconic historical photos, tirades againstprejudice, forms of pure parody, and images of a nude bronze statue in the name of some kind of corporate puritanism, it should be clear that something is amiss.
Yet the armies of the absurd march on, it seems. Facebook managed to kick off the new year by demanding that an Italian art historian remove an image of a penis from her Facebook page. Not just any penis, mind you. It was a picture of a godly penis. Specifically, this godly penis.
That, should you not be an Italian art historian yourself, is a picture of a statue of the god Neptune. In the statue, which adorns the public streets of Bologna, Neptune is depicted with his heavenly member hanging out, because gods have no time for clothes, of course. Yet this carved piece of art somehow triggered a Facebook notice to the photographer, Elisa Barbari.
According to the Telegraph, Barbari got the following notification from Facebook. “The use of the image was not approved because it violates Facebook’s guide lines on advertising. It presents an image with content that is explicitly sexual and which shows to an excessive degree the body, concentrating unnecessarily on body parts. The use of images or video of nude bodies or plunging necklines is not allowed, even if the use is for artistic or educational reasons.”
Even were I to be on board with a Facebook policy banning nudity and, sigh, "plunging necklines" even in the interest of education or art -- which I most certainly am not on board with -- the claim that the image is explicitly sexual and focused on "body parts" is laughably insane. There's nothing sexual about the depiction of Neptune at all, unless we are to believe that all nudity is sexual, which simply isn't true. Also, the depiction focuses not on one body part, but on the entire statue. Nothing about this makes sense.
And that's likely because Facebook is relying on some kind of algorithm to automatically generate these notices. Confusingly, the site's own community standards page makes an exception for art, despite the notice Barbari received claiming otherwise.
Strangely, an exception is made for art. “We also allow photographs of paintings, sculptures, and other art that depicts nude figures.”
Except when it doesn't, that is. Look, again, nobody is expecting Facebook to be perfect at this. But the site has a responsibility, if it is going to play censor at all, to at least be good enough at it not to censor statues of art in the name of prohibiting too much skin.
I have to admit at the start of this post that I rather like Mark Cuban. Not his reality show shtick, but rather what I've seen from him in interviews and his positions. On matters of innovation and intellectual property, I've found him rather refreshing, even as we at this site have poked him on the topic of net neutrality. That admiration makes it all the more jarring when Cuban goes somewhat off the rails on unrelated matters.
For instance, Cuban, who owns the NBA's Mavericks organization, recently pulled the press credentials of two sports journalists. Due to the vacuum of explanation for the move, those that covered it were left to speculation to explain why Cuban would do something like this. That speculation amounted mostly to Cuban either punishing ESPN, the employer of both journalists, for moving Mavs beat journalists into a national coverage role and thereby decreasing the exposure of the team, or pulling an ego trip on the journalists over the type of coverage the team was receiving. The latter was a rather unfair and an all too easy potshot at Cuban, while the former didn't make a great deal of sense as one of the journalists hadn't been on the Mavs beat for at least a decade.
But when Cuban finally commented publicly on his motives, they were revealed to make even less sense, which I wouldn't have thought possible.
Mark Cuban and the Dallas Mavericks pulled the credentials of experienced ESPN reporters Marc Stein and Tim MacMahon this weekend under mysterious circumstances. There was no obvious motive for the Mavericks to do so, and Cuban hasn’t said much to clear it up—but he broke his silence this evening. Cuban told the Associated Press he banned MacMahon and Stein to stave off the advances of the encroaching robot sportswriter horde.
Lest you think this is an unfair representation of Cuban's position, here are his comments directly from an email he wrote to the Associated Press.
“Maybe I will be wrong but I see a direct path from the trends in coverage of games we are seeing over the last couple years to the automation of reporting on games and the curation of related content,” Cuban wrote in an email to the AP. “This isn’t a knock on wire services or their reporters. They are valued and valuable in sports coverage.”
“While it may seem counterintuitive to ban someone from covering us as a way of stopping automation, it really was my only option,” Cuban said. “As is evident by the AP partnership with Automated Insights, it’s not if but when.”
To be clear, the AP does have some automated coverage being done for some sporting events, chiefly minor league baseball games for which paid press attendance doesn't really make sense. But there is not automated recaps being done of NBA games at this time and it is difficult to imagine a rabid sports fan base being satisfied with auto-recaps of the professional teams in major leagues that they so love. Whatever Cuban is really worried about, I'm struggling to understand how it can really be about robotic coverage of Mavs games.
And even if his motivations really are that simple, in what world does it make sense to remove the credentials for two human reporters to stave off SkyNET's NBA coverage?
It’s not a legitimate concern right now, and even if Cuban has decided to take a stand against automated recappers, banning two human reporters is an obtuse solution. ESPN said that they will rely on wire services for recaps of games that they don’t send reporters to, so if anything, all Cuban is really doing is ensure that Mavericks game recaps are written by a wire service, such as the AP, and not ESPN. If his bone to pick is truly with robot sportswriters, ESPN is not the institution to take it out on.
It's a sort of off-brand version of the Streisand Effect, in which you take action to stop something and, by doing so, directly encourage that something to occur. For someone generally smart, this move doesn't resonate. Regardless, I for one welcome our new robotic sports journalist overlords.
I'm going to dispense with any introduction here, because the meat of this story is amazing and interesting in many different ways, so we'll jump right in. Blade Runner, the film based off of Philip K. Dick's classic novel, Do Androids Dream Of Electric Sheep, is a film classic in every last sense of the word. If you haven't seen it, you absolutely should. Also, if you indeed haven't seen the movie, you've watched at least one less film than an amazing artificial intelligence software developed by Terrance Broad, a London-based researcher working on his advanced degree in creative computing.
His dissertation, "Autoencoding Video Frames," sounds straightforwardly boring, until you realize that it's the key to the weird tangle of remix culture, internet copyright issues, and artificial intelligence that led Warner Bros. to file its takedown notice in the first place. Broad's goal was to apply "deep learning" — a fundamental piece of artificial intelligence that uses algorithmic machine learning — to video; he wanted to discover what kinds of creations a rudimentary form of AI might be able to generate when it was "taught" to understand real video data.
The practical application of Broad's research was to instruct the artificial neural network, an AI that is something of a simulacrum of the human brain or thought process, to watch Blade Runner several times and attempt to reconstruct its impression of what it had seen. In other words, the original film is the interpretation of the film through human eyes, while Broad's AI reconstructed what is essentially what the film looks like through the eyes of an artificial intelligence. And if that hasn't gotten your heartrate up a bit, then you and I live on entirely different planets.
The AI first had to learn to discern footage from Blade Runner from other footage. Once it had done that, Broad has the AI "watch" numerical representations of frames from the film and then attempt to reconstruct them into a visual medium.
Once it had taught itself to recognize the Blade Runner data, the encoder reduced each frame of the film to a 200-digit representation of itself and reconstructed those 200 digits into a new frame intended to match the original. (Broad chose a small file size, which contributes to the blurriness of the reconstruction in the images and videos I've included in this story.) Finally, Broad had the encoder resequence the reconstructed frames to match the order of the original film.
Broad repeated the "learning" process a total of six times for both films, each time tweaking the algorithm he used to help the machine get smarter about deciding how to read the assembled data. Here's what selected frames from Blade Runner looked like to the encoder after the sixth training. Below we see two columns of before/after shots. On the left is the original frame; on the right is the encoder's interpretation of the frame.
Below is video of the original film and the reconstruction side by side.
The blur and image issues are due in part to the compression of what the AI was asked to learn from and its response in reconstructing it. Regardless, the output product is amazingly accurate. The irony of having this AI learn to do this via Blade Runner specifically was intentional, of course. The irony of one unintended response to this project was not.
Last week, Warner Bros. issued a DMCA takedown notice to the video streaming website Vimeo. The notice concerned a pretty standard list of illegally uploaded files from media properties Warner owns the copyright to — including episodes of Friends and Pretty Little Liars, as well as two uploads featuring footage from the Ridley Scott movie Blade Runner.
Just a routine example of copyright infringement, right? Not exactly. Warner Bros. had just made a fascinating mistake. Some of the Blade Runner footage — which Warner has since reinstated — wasn't actually Blade Runner footage. Or, rather, it was, but not in any form the world had ever seen.
Yes, Warner Bros. DMCA'd the video of this project. To its credit, it later rescinded the DMCA request, but this project has fascinating implications for the copyright process and its collision with this kind of work. For instance, if the automatic crawlers looking for film footage snagged this automatically, is that essentially punishing Broad's AI for doing its task so accurately that its interpretation of the film so closely matched the original? And, at a more basic level, is the output of the AI even a reproduction copy of the original film, subjecting it to the DMCA process, or is it some kind of new "work" entirely? As the Vox post notes:
In other words: Warner had just DMCA'd an artificial reconstruction of a film about artificial intelligence being indistinguishable from humans, because it couldn't distinguish between the simulation and the real thing.
Other comments have made the point that if the video is simply the visual interpretation of the "thoughts" of an artificial intelligence, then how is that copyrightable? One can't copyright thoughts, after all, only the expression of those thoughts. If these are the thoughts of an AI, are they subject to copyright by virtue of the AI not being "human?" And I'm just going to totally leave alone the obvious subsequent question as to how we're going to define human, because, hell, that's the entire point of Dick's original work.
Broad noted to Vox that the way he used Blade Runner in his AI research doesn't exactly constitute a cut-and-dried legal case: "No one has ever made a video like this before, so I guess there is no precedent for this and no legal definition of whether these reconstructed videos are an infringement of copyright."
It's an as yet unanswered question, but one which will need to be tackled. Video encoding and delivery, like many other currently human tasks, is ripe for the employment of AI of the kind that Broad is trying to develop. The closer that software gets to becoming wetware, questions of copyright will have to be answered, lest they get in the way of progress.
It wasn't so long ago that we were discussing the problems with the United States Treasury Department's list of scary names and how it was being used to prevent completely innocent folks from using online services. The ultimate point of that post was that casting broad nets in which to turn suspicious eyes without applying any kind of checks or common sense was a recipe for calling a whole lot of people terrorists that aren't actually terrorists.
Bruce Francis transferred some money to his dog walker to pay for services to his pit bull, and wrote the dog's name, "Dash," in the notes field. The processors at Chase Bank thought that Dash might be a sneaky way of spelling Daesh (which is the mocking, insulting nickname used by critics to refer to "ISIS"), decided that this was possible terrorist money-laundering, and stopped the payment, froze his account, and notified the Treasury Department that he was a suspected terrorist.
It's hard to know exactly where to begin, but let's dig in, shall we? We can start with the fact that Daesh is a name for ISIS, chiefly used by those opposed to it. It's an acronym of sorts, but it came into being because it sounds like a couple of similar Arabic words that carry heavy negative connotations. In fact, I've heard researchers and opponents of ISIS insist that we should all call them Daesh specifically because it pisses them off so much. In other words, it's fairly unlikely that an ISIS member or supporter is going to call the group Daesh.
And it's less likely that they're going to misspell their terrorist benefactors' group name. It would be like a neo-Nazi funding a group and writing "Not-zees" on the memo line on the check. Sure neo-Nazis are dumb, but if you care enough about the group to fund it, you probably know how to spell the group's name, right? Dash being a fairly common word only pounds home the question of why didn't someone just pick up a phone and call Mr. Francis for an explanation.
Because, really, why would someone funding a terrorist group note that on a paper check? That's the stupidest part of all this. The folks over at Chase apparently thought a guy named Bruce Francis was funding ISIS, calling it Daesh but misspelling the word, and was noting that in the memo on the check. That's epic levels of dumb. Which is why I had assumed Francis would be royally pissed off about this. But, nope, we Americans have been conditioned at this point, which is how you get this:
Despite the inconvenient mix-up, Francis said he had no hard feelings.
"I think anything we can do to stop the terrorists and the funding of terrorists, let's do it," he told KTVU. "And if it means an inconvenience to me and my dogwalker, then that's a price I'm totally willing to pay."
Maybe it's time we just pack this whole America thing up and try again with something new.
Maybe you've been to a store with the self-check-out lanes where you have to scan and bag your own groceries. Some stores have had success with this kind of self-serve technology, but others haven't. Still, brick and mortar stores need to try out different ways to be more convenient for shoppers in order to compete with same-day shipping from various online retailers.
Maybe this past weekend was a bit snowy for you if you live on the East Coast. Shoveling your driveway, and then shoveling it again, and then digging through the big mound of snow where the street meets your driveway -- fun times, for sure. If you were wondering if anyone had a robot snowblower, wonder no more.
Robots are taking away valuable jobs from more and more humans everyday. People used to connect phone calls. People used to categorize links on the internet (and some still do). But those jobs have been largely replaced by more efficient algorithms. Jobs that require some human creativity are supposed to be immune from an attack of automation, but it really depends on what kind of creativity.
Lots of jobs these days couldn't be done as efficiently or as well (or at all) without a little computer help. Computerized assistants (eg. Siri/Cortana/Alexa/GoogleNow) are even helping us with more and more everyday tasks. Increasingly skilled machines are able to perform jobs -- and may be destroying some tedious segments of the labor market. But it's not just manual labor that might be replaced by robots. Automation can help almost anyone become more productive in their work, and as the costs of machines come down with time, we might want to prepare for semi-automated jobs to become fully automated.
Let me start out by saying that I think online harassment and bullying is a significant problem -- though also one that is often misrepresented and distorted. I worry about the very real consequences of those who are bullied, harassed and threatened online, in that it can often lead to silencing voices that need to be heard, or even causing some to not even bother to participate for fear of the resulting bullying. That said, way too frequently, it seems that those who are speaking out about online bullying assume that the best way to deal with this is to move to push for censorship as the solution. This rarely works. Too frequently we see "cyberbullying" being used as a catchall for attacking speech people simply do not like. Even here at Techdirt, people who dislike our viewpoint will frequently claim that we "bullied" someone, merely for pointing out and discussing statements or arguments that we find questionable.
There are no easy answers to the question of how do we create spaces where people feel safer to speak their minds -- though I think it's an important goal to strive for. But I fear the seemingly simple idea of "silence those accused of bullying" will have incredibly negative consequences (with almost none of the expected benefits). We already see many attempts to censor speech that people dislike online, with frequent cases of abusive copyright takedown notices or bogus claims of defamation. Giving people an additional tool to silence such speech will be abused widely, creating tremendous damage.
So, imagine what a total mess it would be if we had a ContentID for online bullying. And yet, it appears that the good folks at SRI are trying to build exactly that. Now, SRI certainly has led the way with many computing advancements, but it's not clear to me how this solution could possibly do anything other than create new headaches:
But what if you didn’t need humans to identify when online abuse was happening? If a computer was smart enough to spot cyberbullying as it happened, maybe it could be halted faster, without the emotional and financial costs that come with humans doing the job. At SRI International, the Silicon Valley incubator where Apple’s Siri digital assistant was born, researchers believe they’ve developed algorithms that come close to doing just that.
“Social networks are overwhelmed with these kinds of problems, and human curators can’t manage the load,” says Normal Winarsky, president of SRI Ventures. But SRI is developing an artificial intelligence with a deep understanding of how people communicate online that he says can help.
This is certainly going to sound quite appealing to those who push for anti-cyberbullying campaigns. But, at what cost? Again, there are legitimate concerns about people who are being harassed. But one person's cyberbullying could just be another person's aggressive debate tactics. Hell, I'd argue that abusing tools like contentID or false defamation claims are a form of "cyberbullying" as well. Thus, it's quite possible that the same would be true of this new tool, which can be used to "bully" those the algorithm decides is bullying as well.
Determining copyright infringement is already much more difficult than people imagine -- which is why ContentID makes so many errors. You have to take into account context, fair use, de minimis use, parody, etc. That's not easy for a machine. But at least there are some direct rules about what truly is "copyright infringement." With "bullying" or "harassment," there is no clear legal definition to match up to and it's often very much in the eye of the beholder. As such, any tool that is used to "deal" with cyberbullying is going to create tremendous problems, often just from misunderstandings between multiple people. And that could create a real chilling effect on speech.
Perhaps instead of focusing so much technical know-how on "detecting" and trying to "block" cyberbullying, we should be spending more time looking for ways to positively reinforce good behavior online. We've built up this belief that the only way to encourage good behavior online is to punish bad behavior. But we've got enough evidence at this point showing how rarely this actually works, that it seems like perhaps it's time for a different approach. And a "ContentID for harassment" seems unlikely to help.