Start Up No.934: Facebook blames spammers for hack, how Saudi Arabia hacks dissidents, Brazil’s WhatsApp problem, iPad alert!, and more

Apple’s podcasting charts got gamed: now we know how. Photo by Nicolas Solop on Flickr.

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 9 links for you. See? Another week done. I’m @charlesarthur on Twitter. Observations and links welcome.

As Facebook shows off its “election war room,” a massive WhatsApp scandal hits Brazil • Buzzfeed News

Ryan Broderick in Sao Paolo:


Brazil’s biggest newspaper, Folha, released a bombshell report on Thursday that local marketing firms have been buying bundles of phone numbers and using them to mass-WhatsApp voters anti-leftist propaganda. The report was released the same day that WhatsApp’s new CEO, Chris Daniels, published a piece in Folha, writing, “We have a responsibility to amplify the good and mitigate the bad.”

Thursday morning, also, appears to have been the time when Facebook allowed access stories from American journalists such as CNN covering Facebook’s new “election war room” to publish. The timing of the embargo — an agreement between news organizations to publish news provided by a source at the same time — the investigation by Folha, and Daniels’ op-ed throw into question exactly how Facebook intends to monitor fake news and hyperpartisan misinformation, especially in a WhatsApp-dominated country like Brazil.

“We know when it comes to an [election], every moment counts,” said Samidh Chakrabarti, head of civic engagement at Facebook, who oversees the war room, told the Verge during their tour of the facility. “So if there are late-breaking issues we see on the platform, we need to be able to detect and respond to them in real time, as quickly as possible.”

Misinformation on WhatsApp has been a huge concern for Brazilian journalists and fact-checkers. About 40% of the country’s 207 million people are using the app. Its messages are encrypted, which means it’s virtually impossible to monitor exactly how political actors are using the app.


The problem with WhatsApp is that it can spread information, and misinformation, virally, far faster than text messages could. It’s like weaponised Ebola when it comes to viral spread.
link to this extract

Report: only 1% of exchange location data useful for offline attribution • MarTech Today


The debate about the relative accuracy and value of location data derived from the exchange bid-stream and that derived from first-party apps has been raging for about three years, with partisans on each side. First-party data is more accurate but less plentiful; third-party location data is much more available but often very polluted or inaccurate.

The latest missive in this debate comes from Placed, a location analytics company recently acquired by Snap. The company just issued a report (registration required) on location accuracy.

Exchange-derived location data usable for in-store attribution

Source: Placed (2017)

The often-technical report asserts that “the average accuracy of exchange-derived locations is over 4 New York City blocks.” It also finds that “only 1% of locations from bid requests are useful for in-store measurement (based on a location accuracy < 50 meters)." Bid-stream location data comes from multiple sources including GPS, cell towers, WiFi and IP addresses, but it rarely comes from the device itself. The report goes on to critique location data coming from exchanges on multiple fronts. Among the criticisms, which all go to the accuracy and utility of the bid-stream data, are the following:

• 80% of bid requests are made while people are between visits — and most of the rest are made at home (so of limited value for attribution).
• Bid stream data overindexes on location data from certain categories (e.g., Lodging, and Gyms & Fitness Centers), likely due to readily available WiFi combined with extended time spent at a given business.
• Key retail categories such as Fashion, Sporting Goods and Computers & Electronics are under-represented in bid data.


That’s for offline, but of course for online it’s going to be a lot bigger.
link to this extract

Facebook finds hack was done by spammers, not foreign state • WSJ

Robert McMillan and Deepa Seetharaman:


Internal researchers now believe that the people behind the attack are a group of Facebook and Instagram spammers that present themselves as a digital marketing company, and whose activities were previously known to Facebook’s security team, said the people familiar with the investigation.

Facebook has previously said it was working closely with the Federal Bureau of Investigation on a criminal probe into the incident.

The incident immediately raised questions about the hackers’ motivation, in part because Russian and Iranian operatives have in the past used social media, including Facebook, to cause mischief in the U.S. Other countries, including North Korea and China, have in the past been accused of cyberattacks for various purposes.

The stolen tokens are digital keys that allowed the hackers to access any part of a user’s Facebook account, and would be of great use to state-sponsored attackers looking to conduct espionage, according to security researchers.

However, the Facebook internal probe suggests the goal of the hackers was financial, not ideological, the people said.

The hackers accessed only a limited subset of the data they could have taken, Facebook said last week. Instead of accessing personal messages, they accessed contact details—including phone numbers and email addresses—gender, relationship status, and search and check-in data belonging to 14 million users. For another 15 million users, only names and contacts were accessed; and the attackers didn’t obtain personal information from 1 million people affected by the breach.


Lot of effort to go to for some customer data.
link to this extract

Supreme Court case could decide Facebook, Twitter power to regulate speech • CNBC

Tucker Higgins:


The Supreme Court has agreed to hear a case that could determine whether users can challenge social media companies on free speech grounds.

The case, Manhattan Community Access Corp. v. Halleck, No. 17-702, centers on whether a private operator of a public access television network is considered a state actor, which can be sued for First Amendment violations.

The case could have broader implications for social media and other media outlets. In particular, a broad ruling from the high court could open the country’s largest technology companies up to First Amendment lawsuits.

That could shape the ability of companies like Facebook, Twitter and Alphabet’s Google to control the content on their platforms as lawmakers clamor for more regulation and activists on the left and right spar over issues related to censorship and harassment.

The Supreme Court accepted the case on Friday. It is the first case taken by a reconstituted high court after Justice Brett Kavanaugh’s confirmation earlier this month.

On its face, the case has nothing to do with social media at all. Rather, the facts of the case concern public access television, and two producers who claim they were punished for expressing their political views. The producers, DeeDee Halleck and Jesus Melendez, say that Manhattan Neighborhood Network suspended them for expressing views that were critical of the network.

In making the argument to the justices that the case was worthy of review, attorneys for MNN said the court could use the case to resolve a lingering dispute over the power of social media companies to regulate the content on their platforms.

While the First Amendment is meant to protect citizens against government attempts to limit speech, there are certain situations in which private companies can be subject to First Amendment liability.


We’re betting on Kavanaugh ruling in favour of it being a “state actor”, yes?
link to this extract

Apple launches special event page for October event with dynamic set of Apple logos • Mac Rumors

Juli Clover:


Apple today sent out invites for an upcoming October 30th event set to be held in Brooklyn. Apple did something special for its invites this time around, and each one features unique artwork with a different Apple logo.

Apple also designed a new event page for the event, and each time you reload the page, you can see a new Apple logo that Apple created.

It’s not clear how many different Apple logos Apple designed for the event, but it appears to be at least several dozen. You can see a selection of approximately 10 of them by refreshing the event page, but not all of the artwork that showed up on the invites appears to be on the page.

Apple’s event, which will focus on the iPad Pro and its Mac lineup, is set to take place on Tuesday, October 30 in New York City at the Howard Gilman Opera House in Brooklyn. It is Apple’s first NYC event in several years.


Bound to be speculation about why New York rather than the custom-built place in California? I suspect it’s about time zones – 10am EDT is 3pm in the UK, 7am PST, so it’s a little easier to get the word around.

Very much looking forward to seeing the new iPad Pros. (That’s surely part of it, right?) AirPods? Mac minis? …AirPower…?
link to this extract

Saudis tried to silence associate of Jamal Khashoggi, recordings show • The Washington Post

Loveday Morris and Zakaria Zakaria:


As he criticized the Saudi leadership as a contributing columnist to The Post, [Jamal] Khashoggi had encountered the pro-government Twitter accounts that Saudi activists refer to as “the flies.”

“Jamal was insulted so much by the Saudi bots,” [exiled Saudi, Omar] Abdulaziz said. “They were focusing on Jamal as he was the voice in the Western media.”

Abdulaziz said he suggested an online countermovement. He just needed some cash to get it off the ground. “We call them ‘the fly army,’ ” he said. “We call ourselves ‘the bee army.’ ”

The plan, he recounted, was to buy SIM cards with Canadian and American numbers that Saudis inside the kingdom could use. Twitter accounts must be verified with a phone number, and activists in Saudi Arabia are scared of linking their Saudi numbers to their Twitter accounts, fearful they could be traced and arrested for being critical of the government, he said. They’d already allocated 200 SIM cards to people.  

Khashoggi had also asked Abdulaziz to help on a short film showing how the Saudi leadership was dividing the country, he said. And Khashoggi had asked for help designing a logo for a new foundation he was forming — Democracy for Arab World Now. Abdulaziz was also helping him design a website to track human rights issues.

But Khashoggi was particularly apprehensive about the SIM card project. “He told me this project is too dangerous,” Abdulaziz said. “He told me to be careful. . . . Twitter is the only platform we have, we don’t have a parliament.” 

In a June 21 message, Khashoggi wrote to Abdulaziz: “I will try to get the money. . . . We should do something. You know sometimes I’m [affected] by their attacks.”

Two days later, Abdulaziz placed an order on Amazon. He clicked a link sent to his phone to track a parcel delivery. He suspects that the action infected his phone. 

The Citizen Lab, a University of Toronto project that investigates digital espionage against civil society, warned him in August that his phone may have been hacked. Two weeks ago, the group concluded with a “high degree of confidence” that his cellphone had been targeted. The group said it believed the operator is linked to “Saudi Arabia’s government and security services.”


As a reminder, Apple in September 2016 issued an urgent security update to address spyware that Saudi Arabia bought from an Israeli company for about $1m to infect the phone of another dissident, Ahmed Mansoor. Mohammad bin Salman, the current ruler of Saudi Arabia, didn’t take over until June 2017. So this isn’t new.
link to this extract

What damage control looks like in Saudi Arabia • Bloomberg

Donna Abu-Nasr and Vivian Nereim:


“I’m shaking now, literally,” says a Saudi businessman vacillating between fear and disbelief that his country might have resorted to the methods of late dictators such as Iraq’s Saddam Hussein and Libya’s Moammar Qaddafi. He spoke on condition of anonymity, a usual request nowadays in a country where the prince has been willing to detain even royals and billionaires to get his way.

Repression is key to damage control at home. A young Saudi who recently returned to the kingdom after studying abroad wrestled with how to react to the Khashoggi news before concluding he had to defend his country above all. Saudis have to side with the government no matter what, he says. As the prince consolidated power in the past two years, many in Riyadh became increasingly cautious about what they say in public. “Talking costs you dearly now,” one Saudi academic said in August after declining to meet with a Bloomberg News reporter. Those still willing to talk suggest rendezvous in secluded settings. They leave their phones behind or seal them in containers in other rooms, hoping to prevent the microphones from being used as listening devices. Sometimes they whisper in the privacy of their own homes.


link to this extract

Chartbreakers: how spammers are gaming the podcast charts • Chartable

Dave Zohrob on how a podcast called “Bulletproof Real Estate” abruptly zoomed to the top of the iTunes charts:


I wanted to see how this cluster of podcasts [in the top of the Apple podcast charts] related to other top shows on the charts, like Serial and Joe Rogan. I grabbed them all for the top 50 podcasts and made another network graph:

Again, every box on the graph represents a podcast, and every arrow represents a recommendation. The chart easily breaks into four clusters, and we can draw some quick conclusions from them.

First, there’s one “main cluster” that includes most popular shows. You can see some natural sub-clusters—for example, one sub-cluster around Joe Rogan includes similar talk shows; another around Someone Knows Something includes true crime shows.

Clusters 1, 2, and 3 are completely disconnected from the main cluster. There are zero recommendations in common between them. Bulletproof Real Estate lives in Cluster 1. You can see by the density of connections that the isolated clusters also have many more connections between the shows than even the most popular sections of the main cluster.

The isolated clusters are highly interconnected, but with very different subject matter. For example, Breaking the Underdog Curse for Chiropractors is related via subscriptions to many podcasts from both Clusters 1 & 2, but has little in common with them in terms of subject matter. The same goes for shows like Winning with Shopify, an ecommerce podcast, and This is Hot Bowga, “home of THE greatest hunting podcast ever created,” in Cluster 3.

So, what can we conclude from this network graph? Here’s my take:

If the podcast charts are based on subscription velocity, it’s highly likely that some or all of the podcasts in the isolated clusters have artificial subscriptions.


Spammers, basically.
link to this extract

Exclusive: Amazon has shut down Liquavista • The Digital Reader

Nate Hoffelder:


Launched in 2006 as a spin off from Philips, Liquavista had been developing a unique type of screen tech that was based on running an electric current through a liquid. This is called electrowetting technology, which is a fancy way of saying that each pixel in a Liquavista screen contained 3 liquids (red, green, blue), and that the color shown by a pixel depended on the amount of power fed into each liquid.

Here’s a demo of a Liquavista screen from 2013. Recorded shortly before the Amazon acquisition, this was the last time Liquavista showed off their screen tech.

The screens were originally being developed as a solution to the battery life issue. Mobile battery life was terrible back in the pre-iPad, pre-iPhone, and pre-netbook era, and people were willing to pay a premium for a screen which used less power than typical LCD screens.

That was why the company was launched, and why Samsung bought it in 2011, but by the time Amazon bought Liquavista in 2013, it was pretty clear that there was no broader market for this tech. The problem of mobile battery life had been solved and battery capacity was already improving year by year, and screens were getting more and more energy efficient.

Coincidentally, I was the first to report that Samsung bought Liquavista in 2011, and the first to report that it had been sold to Amazon in 2013, and now I am the first to officially report Liquavista’s demise.


So that’s that? Until someone finds a better use for electrowetting.
link to this extract

Errata, corrigenda and ai no corrida: none notified

1 thought on “Start Up No.934: Facebook blames spammers for hack, how Saudi Arabia hacks dissidents, Brazil’s WhatsApp problem, iPad alert!, and more

  1. Re. Apple event. Also, Xiaomi is soon announcing the v3 of their Mi Mix. A very interesting $600-800 “experimental flagship” for which Xiaomi doesn’t play it safe (or, ape Apple). In the past they’ve featured bone-conducted sound, ceramic body, bezel-less on 3 side with a very small chin (that held a…nostril cam), Pie beta, camera that made it to DxO’s top ten…

    Expected this year: 10GB RAM (more than my Windows PC), full screen that slides to hide the selfie cam. Meh vintage. One of the few phones Xiaomi releases globally.

Leave a Reply

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

You are commenting using your 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 )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.