A selection of 8 links for you. Spongy. I’m charlesarthur on Twitter. Observations and links welcome.
“The Government has invested a lot of money in open data but then people were asking, ‘what do you do with it?’” says Ian Hetherington, founder and CEO of 3D mapping software company eeGEO. “What’s the visualisation strategy? The data is useless unless you can visualise it.”
Now there are further concerns over the future of the UK’s open data culture. “Cabinet office policy on open data was in the doldrums well before the referendum,” says open data campaigner Owen Boswarva when asked whether he thought Brexit would give the UK’s open data movement a kicking. “Open data is already under threat from austerity, deregulation and cuts to public services.” Brexit, says Boswarva, could make it worse.
“Government could use Brexit as an excuse to stop maintaining datasets that are produced to support EU programmes like Inspire and Eurostat, or to meet EU targets on air pollution and water quality. The UK would also no longer be bound by the PSI Directive, which underpins our regulatory framework for re-use of public sector information,” adds Boswarva.
Chi Onwurah MP, shadow minister for culture and the digital economy, is also concerned.
This is concerning. Literally on the day that Theresa May was abruptly made prime minister I was in a meeting with a minister discussing better ways to use government-collected non-personal data. He didn’t stay through the meeting. May was appointed and he had to go.
He got reassigned. We’ll see if his ideas for data use survive.
link to this extract
Last month, law enforcement officers showed up at the lab of Anil Jain, a professor at Michigan State University. Jain wasn’t in trouble; the officers wanted his help.
Jain is a computer science professor who works on biometric identifiers such as facial recognition programs, fingerprint scanners and tattoo matching; he wants to make them as difficult to hack as possible. But the police were interested in the opposite of this: they wanted his help to unlock a dead man’s phone.
Jain and his PhD student Sunpreet Arora couldn’t share details of the case with me, since it’s an ongoing investigation, but the gist is this: a man was murdered, and the police think there might be clues to who murdered him stored in his phone. But they can’t get access to the phone without his fingerprint or passcode. So instead of asking the company that made the phone to grant them access, they’re going another route: having the Jain lab create a 3D printed replica of the victim’s fingers. With them, they hope to unlock the phone.
The fine details of how they’re going to make it work is quite something.
link to this extract
At Titan, the [three Sumner] brothers [who originally worked on Siri] have been developing software similar to what they built for Siri in order to capture the voluminous data that will be generated by the cars. Siri collects voice command data from hundreds of millions of iPhone owners in order to improve its voice-recognition accuracy and answer some of the more complex commands within seconds. In the same way, self-driving cars use cameras, radars and other sensors to collect imagery and other data about objects and scenarios they encounter on the road. That way, the systems (i.e., algorithms) that help the cars decide what to do, based on what they “see” around them, can learn from new data as it comes in. The brothers would also be involved in Apple’s purchase and configuration of computer servers needed to capture, store and process such data.
It’s a challenging job. One self-driving car will capture more than two gigabytes of data per mile, and up to 10 gigabytes a mile, says James Wu, CEO of DeepMap, which develops mapping technology for self-driving cars. Eventually, the software infrastructure for the Apple car fleet “will be way more challenging than Siri because of the volume of data” that the cars would need to collect. Much of that data would then be sent to Apple’s servers so that the company’s autonomous system could learn from the data and improve its accuracy, he says.
Worldwide smartwatch market experiences its first decline as shipments fall 32% in the second quarter of 2016 • IDC
Smartwatch vendors shipped 3.5 million units in the second quarter of 2016 (2Q16), which was down substantially from the 5.1 million shipped a year ago. Apple held the top rank by shipping 1.6 million watches. However, it was the only vendor among the top five to experience an annual decline in shipments. In fairness to Apple, the year-over-year comparison is to the initial launch quarter of the Apple Watch, which is in many ways the same product offered in the most recent quarter with price reductions.
“Consumers have held off on smartwatch purchases since early 2016 in anticipation of a hardware refresh, and improvements in WatchOS are not expected until later this year, effectively stalling existing Apple Watch sales,” said Jitesh Ubrani, senior research analyst for IDC Mobile Device Trackers. “Apple still maintains a significant lead in the market and unfortunately a decline for Apple leads to a decline in the entire market. Every vendor faces similar challenges related to fashion and functionality, and though we expect improvements next year, growth in the remainder of 2016 will likely be muted.”
Essentially, Apple went from 3.6m a year ago to 1.6m in this quarter, on year-old hardware, and still had nearly half the market. Samsung went from 0.4m to 0.6m, while LG and Lenovo both crept up from 0.2m to 0.3m.
Overall, Android Wear watches may have been about 0.9m shipments. Total activations (measured via Google Play) are less than 5m for its lifetime outside China. (Lenovo’s shipments are probably all inside China.) If you think the Apple Watch is a flop (though I’d say it’s far too early to claim that), what’s the word for Android Wear?
link to this extract
is snark the right response when it’s towards a man winning half the vote by threatening the wellbeing of virtually every American citizen? Right of the bat, we know his absurd economic policies will destroy the economy. If you listen to Chris Christie talk, we’re going to have about 800 times more cops all getting paid as much as software engineers to gun down Black kids for no reason. Unions will be criminalized. Everyone who cooks anything decent is getting deported. Fracking forever.
You’ll find that there’s really only a small sliver of the population that’s getting out of Trump’s reach with a snowball’s chance: rich, white people with the luxury of watching the news for sport so as to make jokes about it.
Everyone else? They need a hero.
I’m willing to say this is a crisis. It’s not a drill. That there are this many people willing to fucking Sieg Heil a man on live TV is that moment we agree this isn’t funny anymore.
When you sit down and ask yourself what response you wish we were getting from a leader right now, I hope you don’t say you’re wishing for more snark.
Wood has spent the past few months being exhausting on Twitter in her support of Bernie Sanders, who wasn’t (realpolitik would have told you) going anywhere, but spent a lot of time not going away. Occasionally though she hits the spot. This is one of those times.
The real problem for Hillary is that she needs to have an emotional message to make people want to vote for her. Not just against Trump; they can do that by not voting. But for her. What’s her message? What’s her response to “Make American Great Again”? Politics and voting is about emotion. Next week is going to be crucial. Obama had a slogan. Clinton badly needs one. And it had better be a terrific one.
link to this extract
Baidu, China’s internet search giant, has shown just what you can learn when you have access to enough location data.
The firm’s Big Data Lab in Beijing has announced that it has used billions of location records from its 600 million users as a lens on the Chinese economy, tracking the flux of people around offices and shops as a proxy measurement for employment and consumption activity. The lab even used the data to predict Apple’s second quarter revenue in China.
We already know that location data is useful, tracking population movements and the spread of disease, for example, but this is the first time that a company on the scale of Google, Facebook or Baidu has shown its hand. The data generated by their huge user bases gives these companies enormous power and insight that they don’t typically talk about. Academic researchers have great difficulty accessing databases like this. But Baidu can just peer into its own servers. The search giant is saying exactly what it can do with the data, and how much data it has.
First, the researchers hand-labelled thousands of areas of interest – offices, shopping centres and industrial zones – across the country. Then they studied the location data – which runs from the end of 2014 to the middle of 2016 – to see how many people were at those places at each time, and how that changed through the year…
…Baidu has collated all the data to build an employment index for China, a number that reflects the overall state of the labour market by tracking how many people are visiting industrial, manufacturing and technology zones in the country. The index shows that employment in manufacturing has dipped by roughly 10% in China since 2014, while high tech employment has grown slightly.
Quoted at length because this is quite stunning. Google likely has similar data for the US at least; it knows about footfall traffic via Android phones. The full version of the paper is at arXiv.
For Apple, it forecasts the past quarter’s revenue as being down by 20% year-on-year. (Apple reports next Tuesday.)
link to this extract
With millions of unique visitors per day KickassTorrents (KAT) has become the most-used torrent site on the Internet, beating even The Pirate Bay.
Today, however, the site has run into a significant roadblock after U.S. authorities announced the arrest of the site’s alleged owner.
The 30-year-old Artem Vaulin, from Ukraine, was arrested today in Poland from where the United States has requested his extradition.
In a criminal complaint filed in U.S. District Court in Chicago, the owner is charged with conspiracy to commit criminal copyright infringement, conspiracy to commit money laundering, and two counts of criminal copyright infringement.
The complaint further reveals that the feds posed as an advertiser, which revealed a bank account associated with the site.
It also shows that Apple handed over personal details of Vaulin after the investigator cross-referenced an IP address used for an iTunes transaction with an IP address that was used to login to KAT’s Facebook account.
“Records provided by Apple showed that email@example.com conducted an iTunes transaction using IP address 188.8.131.52 on or about July 31, 2015. The same IP address was used on the same day to login into the KAT Facebook,” the complaint reads.
As TorrentFreak notes in a followup post, quite a few of the pieces of evidence that led to Vaulin are historical, from up to seven years ago. If you want to be net-illicit years in the future, don’t ever be visible and don’t have the same name. Also of note: the investigator who nailed this is the same one who caught Ross Ulbricht of the original Silk Road. Similar MOs in the nabbing.
(And yes, there’s wonderful irony in the person who ran a giant torrent site being grabbed because they bought something on iTunes.)
link to this extract
On Wednesday, the newspaper began blanking out, for some users, a percentage of words in articles symbolizing the percentage of the company’s revenue that comes from advertising.
The proportion of words blocked isn’t scientific, and the Financial Times doesn’t break out the exact chunk of revenue that comes from ads, said global advertising sales director Dominic Good. “It’s more illustrative than specific,” he said.
The test group comprises registered desktop computer visitors who don’t pay for a subscription, about .075% of the company’s desktop traffic. Some ad-blocking members of this group won’t see any new messaging, some will be asked to whitelist the website’s ads but can still read regardless, some will see articles with many words blanked out if they won’t whitelist the site, and some will be blocked outright if they don’t whitelist the site.
The company will evaulate the results after three or four weeks.
You can now sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.
Errata, corrigenda and ai no corrida: none notified.