Start Up: pricing iPhones, see humans evolve!, why credit systems are broken, Manc-y Oyster, and more


In 2011, Facebook compared political ads on its site to – guess what? Photo by vijay chennupati on Flickr.

Hey web readers: you can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

A selection of 10 links for you. You can choose not to pay $1,000 for them. I’m @charlesarthur on Twitter. Observations and links welcome.

AI will soon identify protesters with their faces partly concealed • Motherboard

Louise Matsakis:

»

A new paper to be presented at the IEEE International Conference on Computer Vision Workshops (ICCVW) introduces a deep-learning algorithm—a subset of machine learning used to detect and model patterns in large heaps of data—that can identify an individual even when part of their face is obscured. The system was able to correctly identify a person concealed by a scarf 67% of the time when they were photographed against a “complex” background, which better resembles real-world conditions.

The deep-learning algorithm works in a novel way. The researchers, from Cambridge University, India’s National Institute of Technology, and the Indian Institute of Science, first outlined 14 key areas of the face, and then trained a deep-learning model to identify them. The algorithm connects the points into a “star-net structure,” and uses the angles between the points to identify a face. The algorithm can still identify those angles even when part of a person’s mug is obscured, by disguises including caps, scarves, and glasses.

The research has troubling implications for protestors and other dissidents, who often work to make sure they aren’t ID’d at protests and other demonstrations by covering their faces with scarves or by wearing sunglasses. “To be honest when I was trying to come up with this method, I was just trying to focus on criminals,” Amarjot Singh, one of the researchers behind the paper and a Ph.D student at Cambridge University, told me on a phone call.

Singh said he isn’t sure how to prevent the technology from being used by authoritarian regimes in the future.

«

But note that this is a long way from reliability, or real-time, or anything that would stand up in court. 67% accuracy sounds a lot, but it leaves gigantic holes for doubt. That won’t stop authoritarian regimes using it, of course.

link to this extract


Your next phone will probably cost you $1,000 • Bloomberg

Mark Gurman:

»

On Tuesday, Apple will introduce its latest top-of-the-line iPhone, and even the cheapest model is expected to cost about $1,000. A few days later, Samsung’s Galaxy Note 8 goes on sale for a comparable amount. The iPhone is expected to be made from glass and stainless steel, while the Note has an exceptionally large, bright screen with a metal-and-glass case. New features for the iPhone will include upgraded cameras and the ability to unlock your phone with a 3D scan of your face. All that stuff has pushed up prices, and there’s a risk that even many longtime early adopters will balk at laying out four figures, including tax.

“A thousand dollars is a line in the sand,” says Ramon Llamas, an analyst at researcher IDC. “There’s going to be a comparison of what $1,000 is to people’s everyday lives, and whether or not that purchase is justified. For some people, $1,000 represents a single paycheck. For others, it represents several weeks of groceries.”

«

That’s the cheapest model of the top-of-the-line phone, and nobody is forcing you to buy that one. These articles are written as though people were being lined up at the point of bayonets and made to purchase them.

Nice graphic though. The reason why prices keep moving up: it’s where the profit margin is.
link to this extract


Autonomous cars: the level 5 fallacy • Monday Note

Jean-Louis Gassée on the idea that cars will be completely self-driving (“Level 5”):

»

In prior Monday Notes that discussed electric and autonomous cars, a subject of endless fascination, I evoked scenarios where SD cars can’t cope with circumstances that require human intervention. Today, I’ll offer the pedestrian crossing at the intersection of Hayes and Octavia in San Francisco:

Understandably, the Google Street View picture was taken in the early morning. Now, imagine the 1 pm Sunday scene with crowded sidewalks and sticky car traffic. In today’s world, pedestrians and drivers manage a peaceful if hiccuping coexistence. Through eye contact, nods, hand signals, and, yes, courteous restraint, pedestrians decide to sometimes forfeit their right-of-way and let a few cars come through. On the whole, drivers are equally patient and polite (an unceasing subject of amazement for Parisians walking the streets of San Francisco).

Can we “algorithmicize” eye contact and stuttering restraint? Can an SD car acknowledge a pedestrian’s nod, or negotiate “turning rights” with a conventional vehicle?

No, we can’t. And we don’t appear to have a path to overcome such “mundane” challenges.
But you don’t have to believe me, or think I’m not “with it”. We can listen to Chris Urmson, Google’s Director of Self-Driving Cars from 2013 to late 2016 (he had joined the team in 2009). In a SXSW talk in early 2016, Urmson gives a sobering yet helpful vision of the project’s future, summarized by Lee Gomes in an IEEE Spectrum article [as always, edits and emphasis mine]:

»

“Not only might it take much longer to arrive than the company has ever indicated — as long as 30 years, said Urmson — but the early commercial versions might well be limited to certain geographies and weather conditions. Self-driving cars are much easier to engineer for sunny weather and wide-open roads, and Urmson suggested the cars might be sold for those markets first.”

«

«

link to this extract


How to generate FiveThirtyEight graphs in Python • Dataquest

Alexandru Olteanu:

»

If you read data science articles, you may have already stumbled upon FiveThirtyEight’s content. Naturally, you were impressed by their awesome visualizations. You wanted to make your own awesome visualizations and so asked Quora and Reddit how to do it. You received some answers, but they were rather vague. You still can’t get the graphs done yourself.

In this post, we’ll help you. Using Python’s matplotlib and pandas, we’ll see that it’s rather easy to replicate the core parts of any FiveThirtyEight (FTE) visualization.

We’ll start here:

And, at the end of the tutorial, arrive here:

To follow along, you’ll need at least some basic knowledge of Python. If you know what’s the difference between methods and attributes, then you’re good to go.

«

If you’re into Python and graphs, this is what you want.
link to this extract


Taxi medallions, once a safe investment, now drag owners into debt • The New York Times

Winnie Hu:

»

Owning a yellow cab has left Issa Isac in deep debt and facing a precarious future.

It was not supposed to turn out this way when Mr. Isac slid behind the wheel in 2005. Soon he was earning $200 a night driving. Three years later, he borrowed $335,000 to buy a New York City taxi medallion, which gave him the right to operate his own cab.

But now Mr. Isac earns half of what he did when he started, as riders have defected to Uber and other competitors. He stopped making the $2,700-a-month loan payment on his medallion in February because he was broke. Last month, it was sold to help pay his debts.

“I see my future crashing down,” said Mr. Isac, 46, an immigrant from Burkina Faso. “I worry every day. Sometimes, I can’t sleep thinking about it. Everything changed overnight.”

Taxi ownership once seemed a guaranteed route to financial security, something that was more tangible and reliable than the stock market since people hailed cabs in good times and bad. Generations of new immigrants toiled away for years to earn enough to buy a coveted medallion. Those who had them took pride in them, and viewed them as their retirement fund.

Uber and other ride-hail apps have upended all that.

«

The New York taxi medallion business is crashing, hard. Difficult not to see this as people who happened to be looking in the wrong direction when the articulated lorry of technological change came down the road.
link to this extract


Our entire credit bureau system is broken • The Verge

Russell Brandom:

»

It’s easy to point to Equifax [the credit reference agency which was thoroughly hacked] as the problem, and its poor handling of the breach (and possible insider trading) certainly doesn’t help. But the problem is bigger than any single company. In a world flooded with information, we’re still relying on a tiny set of sensitive data to protect us from fraud, and putting the burden on the average consumer when that data leaks out. We treat data as private when it’s already been exposed in breach after breach. This system has reached its breaking point. It’s time to burn it all down and start over.

In the most basic terms, credit bureaus work as a reputation service. You submit someone’s name and get back a report on all the money they’ve borrowed over the years and how it’s been repaid. That’s valuable information if you’re deciding whether to lend someone money, so businesses (or their customers) are often willing to pay for it. In that situation, the biggest risk to the lender is an impostor who runs up someone else’s tab and then skips town. So along the way, credit bureaus have become an identity service, too. Along with the potential client’s name, they ask for a Social Security number, and if those things don’t match, they know they’re dealing with fraud.

This is a terrible way to manage identity. From afar, a Social Security number looks kind of like a password. But you can change a password, and you shouldn’t use the same one with every service. To get slightly more technical, you can hash passwords, which lets services verify your identity without keeping your exact password easily available. Right now, I could count the number of places my Gmail password exists anywhere on one hand, whereas I’ve been writing my Social Security number on forms since I was 12. By now, hundreds of organizations have it, from old jobs to old dentists. That number was never going to be safe from scammers. The system was set up for failure from the very beginning.

«

Powerful, and spot-on, piece.
link to this extract


Massive genetic study shows how humans are evolving • Nature News & Comment

Bruno Martin:

»

A huge genetic study that sought to pinpoint how the human genome is evolving suggests that natural selection is getting rid of harmful genetic mutations that shorten people’s lives. The work, published in PLoS Biology1, analysed DNA from 215,000 people and is one of the first attempts to probe directly how humans are evolving over one or two generations.

To identify which bits of the human genome might be evolving, researchers scoured large US and UK genetic databases for mutations whose prevalence changed across different age groups. For each person, the parents’ age of death was recorded as a measure of longevity, or their own age in some cases.

“If a genetic variant influences survival, its frequency should change with the age of the surviving individuals,” says Hakhamanesh Mostafavi, an evolutionary biologist at Columbia University in New York City who led the study. People who carry a harmful genetic variant die at a higher rate, so the variant becomes rarer in the older portion of the population.

Mostafavi and his colleagues tested more than 8 million common mutations, and found two that seemed to become less prevalent with age. A variant of the APOE gene, which is strongly linked to Alzheimer’s disease, was rarely found in women over 70. And a mutation in the CHRNA3 gene associated with heavy smoking in men petered out in the population starting in middle age. People without these mutations have a survival edge and are more likely to live longer, the researchers suggest.

This is not, by itself, evidence of evolution at work. In evolutionary terms, having a long life isn’t as important as having a reproductively fruitful one, with many children who survive into adulthood and birth their own offspring. So harmful mutations that exert their effects after reproductive age could be expected to be ‘neutral’ in the eyes of evolution, and not selected against.

But if that were the case, there would be plenty of such mutations still kicking around in the genome, the authors argue.

«

link to this extract


Google appeals against EU’s €2.4bn fine over search engine results • The Guardian

Daniel Boffey:

»

Google is appealing against the record €2.4bn (£2.2bn) fine imposed by the European Union for its abuse of its dominance of the search engine market in building its shopping comparison service.

The world’s most popular internet search engine has launched its appeal after it was fined by the European commission for what was described as an “old school” form of illegality.

The Luxembourg-based general court, Europe’s second-highest, is expected to take several years before ruling on Google’s appeal, which had been widely expected. The Silicon Valley giant had responded to the fine at the time of its announcement by saying that it “respectfully” disagreed with the legal argument being pursued.

«

But still has to stop boosting its shopping service in contravention of EC rules; has until 28 September to comply. The EC is looking at its proposal on this, apparently.
link to this extract


May 2011: Facebook: exempt us from federal election commission rules • POLITICO

Jennifer Epstein, in May 2011:

»

Facebook, the company that has helped put so much of what was once private out in open on the web, is looking for a sort of corporate privacy setting of its own — the company is looking to ensure that it is exempt from federal election rules requiring campaign advertisements to include disclosures of who paid for them.

In a request to the Federal Election Commission made late last month, lawyers for the social networking powerhouse argued that the small ads on Facebook’s website should not have to include disclosures because of the limited amounts of room for text.

While it’s easy to include disclosures on a television ad, billboard or email, Facebook argues, it’s more difficult with the tiny ads posted along the side of its webpages. “With some mediums … – e.g. bumper stickers, buttons, pens, T-shirts, concert tickets, and text messages – it is inconvenient or impracticable to include a disclaimer,” three lawyers from the Washington office of the firm Perkins Coie write in their request for an advisory opinion from the FEC.

The company says it has made a conscious decision to keep the ads on its site small and less obtrusive to the user experience, and does not want to take away from that experience or penalize campaign advertisers. “Facebook gives a wide range of candidates and causes a voice where they would otherwise not be able to afford one through more traditional political advertising,” spokesman Andrew Noyes said in a statement to POLITICO. “We encourage the FEC to consider these benefits and other fundamental differences between some online ad formats and newspaper and TV advertising.”

«

Facebook was in effect claiming that its ads – including the political ones – are the equivalent of skywriting (which doesn’t need disclosure about who paid for it). The FEC agreed. This, of course, turns out to have been a significant turning point, even though nobody saw it at the time.

Imagine if all the political ads on Facebook in the 2016 election had had to declare who bought them. The discourse around the company would be very different. (Twitter too have used this get-out, I believe.)
link to this extract


A beginner’s guide to using My Get Me There • Medium

Susil Nash on “Manchester’s hilarious attempt at reinventing London’s Oyster” (the latter, for Americans, is an RFID system which can be used to pay contactlessly for trips on buses and underground; it’s worked, pretty much perfectly, since 2004:

»

The first of the new system’s fun quirks is that My Get Me There isn’t just a card. It’s an app too. Now, you might think that’s to be expected — it’s a convenient way to manage your card, right? The two work together in harmony, right? Wrong. The app and the (presumably ironically-named) ‘Smart Card’ are two completely separate systems that work entirely independently of one another.

Your first decision is therefore whether to opt for the app, the Smart Card or, as will be the case for most travellers, both. The app is certainly less tricky to get hold of (more on that in a moment) but the significant downside is that it can only be used on Metrolink — Manchester’s tram network. Which means no smartphone fun for Team Bus or the vehicle-agnostics, but app-tastic news for all tram devotees.
Having said that, there are a couple of things that even you dedicated Metrolinkers should watch out for before ditching the paper tix. Firstly, know that you’ll need to make sure you’re not low on battery when heading out the door because, if your phone gives up mid-travels, you could be hit with a £100 fine. Secondly, you’ll need to remain online… ish. The reason for the ‘ish’ is that you don’t actually need web access to use the app once you’ve bought your ticket. However, any tickets on your phone will expire if that device “has not been connected to the internet for a long period” (that’s literally the timescale specified on their website).

So do make sure your phone has a plenty of juice and has been connected to the internet at least once in the most recent ‘long period’.

«

Even worse: it’s not a top-up scheme. It’s a “specific tickets for specific journeys” system. And you have to be over 16. It’s as if they wanted to keep cash forever.
link to this extract


Errata, corrigenda and ai no corrida: none notified

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s