Blog Ipsa Loquitur

Ed Sheeran’s having a lousy summer. He’s a pop star and defendant in a couple of copyright infringement lawsuits filed in June and August. The lawsuits claim he stole the melody for two of his hit songs, “Thinking Out Loud” and “Photograph.” One of these lawsuits is well-written, and the other is about to get tossed out of court.

Eriq Gardner at the Hollywood Reporter does some great legal reporting and lawsplains the situation:

Get past the lawyers involved. (The plaintiff in the “Photograph” case is represented by the victorious attorney who convinced a jury that “Blurred Lines” was lifted from Marvin Gaye’s “Got to Give It Up,” while confusingly, the one now fighting to protect Marvin Gaye’s “Let’s Get It On” is a Florida attorney without much of a reputation in entertainment circles.)

Just focus on what was actually filed in court. The contrast is pretty stark. Here, for example, is the way that Sheeran allegedly infringed “Let’s Get It On.” It comes from the plaintiff’s complaint:

That’s pretty much it. How is the melody, harmony and rhythm similar? No word. That will likely become a problem for plaintiffs as copyright law only protects expression — not generic ideas. A failure to specify the striking similarity leaves the plaintiff open to a challenge for not pleading a plausible claim of infringement. Emphatic underlining hardly cures this fault. (The case is being brought in New York federal court, which isn’t particularly friendly to copyright plaintiffs.)

Gardner includes screenshots of the actual complaints, and compares and contrasts the draftsmanship behind the two lawsuits. I wish everyone who reported on lawsuits was so thorough.

Published on under to Your Loving Arms

Charlie Warzel, for Buzzfeed News, outlines the continual and systematic failure of Twitter to address abuse on its platform in“A Honeypot For Assholes”:

Part of what makes Twitter so powerful is its ability to level the communication field; tweets from a non-famous, little-followed user, are, in theory, just as easy to surface as a celebrity’s. Simply, Twitter is built differently than most social networks. Two accounts don’t have to follow each other to interact, and once a tweet is out in the world, the original tweeter doesn’t have the ability to moderate responses, like they might on an Instagram or Facebook comment. This unique design is responsible for some of Twitter’s most revolutionary and serendipitous moments; it’s also perfect for abuse.

“For years, it allowed this equal footing, where a troll you didn’t follow and your best friend who you follow and interact with all the time were given equal weight, and that’s crazy,” a former senior employee said. “Seriously, if you were an alien and you came down to look at this thing, you’d say, ‘Oh, the product was basically built for maximum ease of trolling.’ Like, they must have built this for trolls.

Warzel talked to ten former high-level employees to try to make sense of Twitter’s inability to keep its platform troll-free. He notes that while executives keep saying they want to combat abuse on their platform, Twitter has a philosophical commitment to free speech that’s just fundamentally incompatible with tapping certain users on the shoulder and saying “hey, knock that off.”

Part of this internal conflict stems from the fact that most of the people involved in these values judgments at Twitter aren’t subject to the kind of sustained harassment that makes their platform unusable for so many people. That’s not necessarily fatal unless you also ignore your users when they say your platform is infested with trolls, because you’re in love with the idea of completely unrestricted speech.

Unless you post a six-second video of an Olympic event. For God’s sake, show a little humanity.

Published on under We Can't Have Nice Things

It’s been a weird, weird election cycle so far, and I’m not certain we’re going to get back to “normal” after November. While we’ve got a few months to go before the election itself, I can’t stop thinking about what Donald Trump means for the Republican Party. Clare Malone in FiveThirtyEight has a compelling theory: this is The End Of A Republican Party.

Many have assumed that adherence to a certain conservative purity was the engine of the GOP, and given the party’s demographic homogeneity, this made sense. But re-evaluating recent history in light of Trump, and looking a bit closer at this year’s numbers, something else seems to be the primary motivator of GOP voters, something closer to the neighborhood of cultural conservatism and racial and economic grievance rather than a passion for small government.

Her whole piece is fantastic, and it’s FiveThirtyEight, so you know the data science and chart games are solid. For my money, one of the most telling moments is when Malone discusses the Republican Party’s future with Ben Howe, a contributing editor at

Howe’s theory for the racial animus of Trump supporters boils down to simple attrition: “Everybody who was reasonable seems to have gone home in 2012,” he said. Romney’s loss in 2012 discouraged many of the once-energized fiscal conservative activists. “This isn’t the most artful way to say it, but it’s like, where do you go when the only people who seem to agree with you on taxes hate black people?” Howe laughed ruefully. “I think what you do is you say, ‘Well, I may lose but I can’t align myself with them.’”

But instead, Howe said, he made moral compromises he regrets. “There are some things that I don’t have core values about, that I can be negotiable on, compromise on. But then there are other things that I can’t budge on,” he said. “I think I thought I had to budge on some things: ‘Yeah, this guy talking to me right now just said he agrees with my taxes and also we need to get that Kenyan out of office.’ Why did I stand there and say, ‘Yeah’? You know? I shouldn’t have done that. I should’ve said, ‘Wait, what? No, that’s stupid. You’re stupid. Don’t be stupid.’”

Howe’s onto something, but I’m not sure if ‘reasonable people going home’ is really what’s happening here. The Republican Party’s Southern Strategy was a cynical but effective maneuver to appeal to people with very specific ideologies. Republicans have had decades of opportunities to tell these people “what? That’s stupid. Don’t be stupid.” They dragged their feet. Don’t take it from me or Howe: an NBC poll this week found that 72% of registered Republicans still aren’t sure if President Obama was born in America.

Now the inmates run the asylum, and it’s dawning on Republican leaders that tax policy isn’t nearly as important to millions of their voters as “getting that Kenyan out of office.” I guess we’ll just have to wait and see if Obama builds his presidential library in Chicago or Nairobi.

Published on under The News

T. Christian Miller in ProPublica: The FBI Built a Database That Can Catch Rapists — Almost Nobody Uses It. The system is called the Violent Criminal Apprehension Program, or ViCAP. While knowing that The Feds have a centralized database of crime might set your Police State Spidey Sense a-tingling, the truth is more like the year 1984 than the book 1984:

That’s what’s striking about ViCAP today: the paucity of information it contains. Only about 1,400 police agencies in the U.S., out of roughly 18,000, participate in the system. The database receives reports from far less than 1 percent of the violent crimes committed annually. It’s not even clear how many crimes the database has helped solve. The FBI does not release any figures. A review in the 1990s found it had linked only 33 crimes in 12 years.

Now, your honor, I’m not some fancy big-city lawyer. And I’m certainly no data scientist. But I do know that when you’re talking about big data, the first part of that term is “big,” and ‘less than 1% of violent crimes’ is not “big.” In fact, you might want to call this an Anecdotabase. You can’t crunch the numbers and catch the bad guys with such lousy data.

The good news is that this isn’t costing us a lot of money.

In an agency with an $8.2 billion yearly budget, ViCAP receives around $800,000 a year to keep the system going. The ViCAP program has a staff of 12. Travel and training have been cut back in recent years. Last year, the program provided analytical assistance to local cops just 220 times.

220 times! If there are 18,000 police agencies in America, it would take 81 years to provide “analytical assistance” to each local agency exactly once. Even Time Warner’s customer service picks up faster than that.

You know, maybe the FBI should just scrap this thing. If it hasn’t worked by now, this just might not be a system that can ever work. Who knows how many computer systems the various local police agencies use? You can’t expect anyone to interface between all of those.

The U.S. need only look north for an example of how such a system can play an important role in solving crimes. Not long after ViCAP was developed in the United States, Canadian law enforcement officials used it as a model to build their own tool, known as the Violent Criminal Linkage Analysis System, or ViCLAS. Today, the Royal Canadian Mounted Police maintains a database containing more than 500,000 criminal case profiles. The agency credits it with linking together some 7,000 unsolved crimes since 1995 – though not all of those linkages resulted in an arrest. If the FBI collected information as consistently as the Mounties, its database would contain more than 4.4 million cases, based on the greater U.S. population.

Instead, the FBI has about 89,000 cases on file.

Our Northern Nemeses strike again! The most surprising thing here is that the Canadian version doesn’t have an extraneous U in it somewhere. Showoffs.

Seriously, though. You get it by now. ViCAP has underwhelmed. It doesn’t get results because it doesn’t have enough police records to analyze, and it doesn’t have enough police records in it because… the police don’t put their records in it.

As the money made its way to the bomb shelter, the FBI conducted a “business review.” It found that local cops were sending the agency only 3 to 7 percent of homicides nationwide. The miniscule staff — about 10 people — could not even handle that load, and was not entering the cases on a timely basis. Cops on the street saw the system as a “black hole,” according to “Cold Case Homicide,” a criminal investigation handbook.

To get information into the database, local cops and deputies had to fill out by hand a form with 189 questions. The booklet was then sent to Quantico, where analysts hand-coded the information into the computer. It was a laborious process that flummoxed even Brooks. He had a hard time filling out the booklet, according to one account — as did officers in the field. Only a few hundred cases a year were being entered.

I bet I can tell you exactly how those 189 questions were determined. The FBI sat down every analyst they could find, and asked them what they’d like to know when solving crimes. And instead of considering local police departments as the users of ViCAP, and anticipating that ain’t nobody got time for 189 questions, the FBI neglected to trim down that list to the top 50 or so.

The craziest part is that this process wasn’t updated when it became obvious that the FBI couldn’t keep up with the few police departments that participated in ViCAP. As Miller notes above, the FBI had to manually re-enter 189 answers into their computers every time a crime was reported, and they became overwhelmed almost instantly.

Still, ViCAP is a laudable attempt to eliminate data silos (previously seen here on BL), despite falling way short of its goal. Here’s hoping ViCAP 2.0 is a little more successful.

Published on under Procurement Hell

As a resident of Brooklyn for the last decade, affordable housing is more than a passing interest to me. The number of apartments in New York City is significantly outstripped by the number of people who want to rent apartments here. As you’d imagine, this leads to a slightly crazy real estate market; tiny apartments sell for five or six times the cost of the home I grew up in in the suburbs, and the average monthly rent for a Brooklyn apartment is $2,800.

This is a problem that affects lots of cities in America and around the world. If lots of peple are moving to a city, and there aren’t enough places to live for everyone, prices are going to go up in a hurry. Timothy B. Lee writes in Vox about Tokyo’s One Weird Trick to avoiding runaway rent: building new places to live.

Tokyo does a better job of allowing housing supply to keep up with housing demand. In 2014, Tokyo issued permits for 142,417 new housing units. In contrast, the entire state of California — which has three times the population of Tokyo — issued permits for only 83,657 new housing units. Little wonder that demand for housing has outstripped supply in the Bay Area.

In the United States, local housing markets are plagued by grassroots “Not In My Back Yard” (NIMBY) activists who organize to stop efforts to build town homes and apartment buildings in their local neighborhoods. Because every construction project is located near somebody, the result tends to be that little housing gets built anywhere.

In contrast, Harding writes, Japan sets housing regulations at the national level. As a result, if a Tokyo landowner wants to knock down his single-family home and replace it with a six-unit condo building, there’s little that his neighbors can do to stop it. That can be annoying to individual homeowners, of course. But it also has the huge upside of keeping housing costs under control.

Note of course that homeowners in any neighborhood/city have a strong economic incentive to limit the construction of new housing. If housing is difficult to come by, you can sell or rent your home for a higher price than if housing is plentiful. While this may not explicitly motivate any individual activist attempting to block new construction, collectively, homeowners all benefit from this. (That’s that invisible hand we’re all familar with.)

Over at the Daily Beast, Clive Irving has another great example of a system unintentionally incentivizing lousy behavior. His article is aboutVolkswagen’s decision to spend a decade cheating at emissions tests:

Executives even carefully evaluated what the cost would be to the company if they were caught. Reviewing previous cases of violations of environmental regulations by auto manufacturers in the U.S. they predicted that the likely fines posed “only a moderate cost risk.” They cited the highest fine, imposed against Hyundai/Kia as amounting to “barely $91 per vehicle” and added “fines in this amount are not even remotely capable of influencing the share price of a globally operative company such as Volkswagen.”

VW is going to settle the various lawsuits for $15.3 billion, which covers about 600,000 cars. That’s a fine of about $25,500 per vehicle, so VW’s executives were only off by about twenty-eight thousand percent. They had no way of knowing the federal government was about to make an example out of them, so VW were acting as though the pressure to not cheat was barely a rounding error. That’s why they did some pretty sleazy stuff:

And the New York case provides a whole lot more insight into the lengths VW went to cheat the system. It details six successive versions of defeat devices used across the range of VW, Audi, and Porsche diesels. Moreover, it reveals that VW went a further step to make sure it was not caught. In New York and other states the annual inspections do not actually measure the exhaust emissions. They rely on the car’s onboard diagnostics (OBD) to detect whether the car is running clean.

“To allow its defeat device equipped vehicles to pass New York’s inspection and maintenance tests,” says the lawsuit, “Volkswagen therefore needed to, and in fact did, implement a further cheat: It programmed the OBD systems to falsely report at inspection time that the automobile emissions systems were working properly.”

Got that? VW cheated because federal and state regulators asked VW to police itself, and had historically provided a slap on the wrist for cheaters. All while VW was subject to increasingly stringent emissions requirements. It doesn’t take a genius to see that’s a system with a strong incentive for “creative” (read: cheaty) solutions.

All is not lost, though. This $15.3 billion settlement is a great way to change the calculus for people who would cheat in the future: it’s not just a slap on the wrist anymore. The guys responsible for this are going to think long and hard before pulling this kind of stunt agai-

VW’s Management Board, the nine men who had presided over the perpetration of fraud, the cover-up and then a public relations debacle that followed its exposure, were awarded $70 million in executive compensation for 2015 alone.

Oh, come on.

Published on under We Can't Have Nice Things

Madeleine Clare Elish, on the future of machine intelligence, says future designers of autonomous systems will consult ethnographers. She opens her piece by recounting a taxi driver driving to the (nonexistant) back entrance of an airport because Google Maps told him to ignore the front gate. Elish’s piece is all about the expectations society places on the human operators of these systems:

In a previously published case study of the history of aviation autopilot litigation, Tim Hwang and I documented a steadfast focus on human responsibility in the arenas of law and popular culture, even while human tasks in the cockpit have been increasingly replaced and structured by automation. Our analysis led us to thinking about the incongruities between control and responsibility and the implications for future regulation and legal liability in intelligent systems. The dilemma, as we saw it, was that as control has become distributed across multiple actors (human and nonhuman), our social and legal conceptions of responsibility have remained generally about an individual.

We developed the term moral crumple zone to describe the result of this ambiguity within systems of distributed control, particularly automated and autonomous systems. Just as the crumple zone in a car is designed to absorb the force of impact in a crash, the human in a highly complex and automated system may become simply a component—accidentally or intentionally—that bears the brunt of the moral and legal responsibilities when the overall system malfunctions.

Published on under The Digital Age