Facebook, Instagram, and Twitter limit data access for AI-powered babysitter background checks

Facebook, Instagram, and Twitter have limited the amount of data that’s accessible to Predictim, a California-based startup that uses machine learning to vet potential babysitters. The social networks took action against the company after a report by The Washington Post last week detailed its methods, attracting widespread criticism.
Predictim claims to use “advanced artificial intelligence” to judge a babysitter’s suitability. This includes combing through an individual’s Facebook, Instagram, and Twitter histories before offering an automated assessment of their character. The company claims it can predict whether the individual is a drug user, if they might bully or harass others, and even if they have a “bad attitude.”
Experts have criticized Predictim’s service as unscientific, noting that machine learning is notoriously unreliable when it comes to parsing complex data like human speech. AI might be good at recognizing objects in photos or digitizing handwriting, but it can’t reliably interpret nuances in tone and speech like sarcasm or jokes. Experts in data policy also noted that Predictim’s software doesn’t explain how it comes to its decisions, which means that a potential babysitter could lose a job without ever knowing why or offering an explanation.
Facebook, Instagram, and Twitter have all taken action against Predictim over the last few weeks, report The Washington Post and BBC News. Facebook says it dramatically limited the company’s access to user data on Facebook and Instagram after it violated a ban against developers using this information to vet job candidates. Twitter revoked Predictim’s access to its API (which is used to access data on a scale, rather than reading individual profiles) earlier this week. A spokesperson told the Post, “We strictly prohibit the use of Twitter data and APIs for surveillance purposes, including performing background checks.”

Two Iranian men charged with the ransomware attack that took down Atlanta

The US Treasury Department has placed bitcoin addresses on its sanctions list for the first time after two Iranian hackers were charged with extorting millions of dollars through them. The two addresses belonged to Faramarz Shahi Savandi and Mohammad Mehdi Shah Mansouri, who allegedly created the SamSam ransomware software. The Justice Department unsealed an indictment against both men today, alleging that they collected $6 million by targeting more than 200 victims — including the cities of Atlanta, Georgia, and Newark, New Jersey.
SamSam began infecting computers in 2015, and it’s been linked to expensive and temporarily devastating attacks on hospitals and infrastructure. Like other ransomware, SamSam encrypted users’ machines and ordered them to funnel money — sometimes tens of thousands of dollars — to a bitcoin account. The Treasury Department says the two accounts above processed over 7,000 transactions, although not all were necessarily related to SamSam.
In a press conference, US Attorney Craig Carpenito told reporters that Savandi and Mansouri “worked hard to identify the most vulnerable targets that they could,” and not just because they would be more likely to pay up. “Money is not their sole objective,” he claimed. “They’re seeking to harm our institutions and critical infrastructure. They’re trying to impact our way of life.”
One of Savandi and Mansouri’s most high-profile alleged crimes was an attack on Atlanta in March 2018. Major basic municipal functions were affected, including the ability to pay water bills or parking tickets, although Atlanta’s emergency services remained functional. Altogether, the Justice Department lists attacks in 43 US states.

Google removes gendered pronouns from Gmail’s Smart Compose to avoid AI bias

Gmail’s Smart Compose is one of Google’s most interesting AI features in years, predicting what users will write in emails and offering to finish their sentences for them. But like many AI products, it’s only as smart as the data it’s trained on, and prone to making mistakes. That’s why Google has blocked Smart Compose from suggesting gender-based pronouns like “him” and “her” in emails — Google is worried it’ll guess the wrong gender.
Reuters reports that this limitation was introduced after a research scientist at the company discovered the problem in January this year. The researcher was typing “I am meeting an investor next week” in a message when Gmail suggested a follow-up question, “Do you want to meet him,” misgendering the investor.
Gmail product manager Paul Lambert told Reuters that his team tried to fix this problem in a number of ways but none were reliable enough. In the end, says Lambert, the easiest solution was simply to remove these types of replies all together, a change that Google says affects fewer than one percent of Smart Compose predictions. Lambert told Reuters that it pays to be cautious in cases like these as gender is a “big, big thing” to get wrong.
This little bug is a good example of how software built using machine learning can reflect and reinforce societal biases. Like many AI systems, Smart Compose learns by studying past data, combing through old emails to find what words and phrases it should suggest. (Its sister-feature, Smart Reply, does the same thing to suggest bite-size replies to emails.)
In Lambert’s example, it seems Smart Compose had learned from past data that investors were more likely to be male than female, so wrongly predicted that this one was too.

Amazon quietly ends its restaurant delivery service in the UK

After two years of operation, Amazon has quietly closed the London arm of its takeout and delivery service, reports the Evening Standard. The news was announced in an email to the service’s customers, which said deliveries will cease in the capital (and, by extension, the rest of the UK) after December 3rd.
Like many of Amazon’s services, Amazon Restaurants was available exclusively to Prime subscribers who could pay £1.99 to have food delivered within an hour from over 200 London restaurants. It first launched as a service in Seattle back in 2015 before expanding to London in September 2016.
However, it appears that even Amazon’s stores of shopping data weren’t enough to let it convincingly compete against London’s incumbent Deliveroo and Uber Eats services, despite the latter having only launched only three months previously in the city. In the US, the service also doesn’t seem to be faring well against popular competitors like Eat24, Postmates, and Grubhub. In New York City, for example, Amazon Restaurants options are limited to a handful of vendors with few user reviews.

LinkedIn used 18 million non-user e-mails to target Facebook ads

Ireland’s Data Protection Commission issued a report last week (via TechCrunch) that covered its activities during the first six months of 2018, in which it highlighted a complaint against LinkedIn. It found that the company used millions of e-mail addresses of non-LinkedIn users to target ads on Facebook.
Following a complaint against the company, the DPC conducted an audit and found that it violated data protection regulations. In a bid to grow its user base, LinkedIn “processed hashed email addresses of approximately 18 million non-LinkedIn members,” and then used them to target those individuals with ads on Facebook. As TechCrunch points out, numerous companies moved their data processing operations to Ireland prior to the implementation of new European data regulations. The DPC says that the “complaint was ultimately amicably resolved,” and that LinkedIn has ceased those practices.
However, the body was “concerned with the wider systemic issues identified” in its report, and undertook a second audit to see if LinkedIn had adequate “ technical security and organisational measures.” It found that the site was “undertaking the pre-computation of a suggested professional network for non-LinkedIn members,” and ordered them to stop and delete associated data that existed prior to May 2018.

A WeChat-powered hotel in China lets you check in, enter your room, and get breakfast through your phone

China’s biggest app, WeChat, partnered with InterContinental to open a smart hotel in Shanghai that lets users book rooms, remotely check in, and use their phones as key cards — all without requiring human assistance. Once they’ve completed their reservation, travelers verify their identity at the hotel through facial recognition and collect their digital key card. Although the hotel is 88 feet underground, it’s connected enough to provide quick internet services, according to a press release it shared with The Verge.
WeChat and its parent company Tencent have been working on developing smart hotel experiences since 2014, but now it’s added facial recognition into the mix for faster identity checks. Traditionally, hotels in China require you to take a photo of your face and present government ID to be scanned and copied before you can enter your room. However, internet access offered by most hotels in China also remains restricted by the firewall enforced by regulators, and many Western hotel chains have stopped offering VPNs for free.
Through WeChat, travelers can also order room service, make payments, set air conditioning in the room, and adjust curtains and lighting. Breakfast vouchers are included in the customer data within the app, so you can just scan your face at the dining area to grab a meal. If you’ve booked an executive suite, there’s a 24-hour butler service that you can summon through WeChat. Similarly, check-outs are also done through the app.

The AmazonBasics Microwave is for those who are spoiled by voice commands

How much do you need your microwave experience to be streamlined? Is it annoying to have to press a few buttons? Are a microwave’s presets too complex to use?
Amazon seems to have different answers to these questions than the ones I’ve heard from fellow Verge co-workers and friends. It says it sees an opportunity in making a smart microwave, letting people use those advanced presets they normally don’t touch. Instead of having to press multiple buttons, you can just ask Alexa to heat an item, and it will determine heating times for items as varied as sweet potatoes and popcorn.
To accomplish this, Amazon made a microwave that communicates with an Echo to let you use voice commands for control. The $59.99 AmazonBasics Microwave has features that are typical for an affordable device in its category — including 700 watts of power, auto defrost features, and a popcorn button — but it also works with Alexa.
Setting up the AmazonBasics Microwave is fairly fast and simple: you’ll just need an Echo device (the microwave doesn’t actually have a microphone or speaker built in), an Alexa app, and a 2.4 GHz Wi-Fi connection. The microwave is powered by Alexa Connect Kit, which runs on Bluetooth Low Energy and Wi-Fi to let Echo devices communicate. It also acts as a reference of Alexa Connect Kit for other device makers to learn from. The app prompts you to add your microwave as an appliance. Alternatively, you can just ask Alexa on your Echo to pair with your microwave. The overall process takes about 10 minutes, especially if the connection is slow. In that time, you could’ve finished setting up a regular microwave by plugging it in and setting the time.
The microwave does simplify things enough so you can say, “Alexa, microwave one potato” or walk over to the microwave, tap the button, and say, “one potato.” Alexa will say, “Okay,” and the microwave will automatically be set for six minutes and 30 seconds.
But the two potatoes (one sweet, one savory) I tried to heat up with the AmazonBasics Microwave were undercooked and almost raw. I had better results with popcorn, which came out just right with Alexa’s help, unlike a bag of popcorn I guesstimated and ended up burning. The difference is that a normal microwave will ask how heavy your potato is, but Alexa strangely doesn’t ask. And while you normally might not know exactly how long to heat your popcorn (the box gave me vague advice “not more than 4 minutes and sometimes less than 1.5 minutes”), Alexa does ask how many ounces the bag of popcorn is and times it accordingly.

YouTube now runs pop-ups on videos that warn users of EU copyright proposal

If you watch YouTube videos without a YouTube Premium subscription, you’re familiar with the tiny pop-ups that appear in the bottom left-hand corner of the stream that prompt users to sign up for YouTube’s ad-free service.
It seems that YouTube has replaced that pop-up with a new message that warns users about the European Union’s proposed copyright directive. It also offers an explainer on one particular act known as Article 13. The pop-up brings YouTube viewers to the company’s standalone website detailing the possible effect of Article 13, which puts the onus on YouTube to prevent copyrighted material from appearing on the platform. YouTube executives like CEO Susan Wojcicki and head of business Robert Kyncl have penned exhaustive blog posts and guest columns in publications that warn that the company cannot financially take on the burden, and extreme measures will have to be taken like preventing users from uploading videos at all.
YouTube’s communications team has spent weeks aggressively tweeting from its main channel and its subsequent YouTube Creators account in an attempt to raise awareness about the copyright directive. A Q&A last week with creators and users on Twitter led the company to admit that if the copyright directive passes, some European channels will be blocked from playing in the United States. Certain users could also be prevented from uploading videos at all.

Xiaomi takes over Meitu hardware business to sell more selfie-focused smartphones

Chinese selfie app and smartphone maker Meitu has announced that Xiaomi is effectively taking over its hardware business. Future phones will carry the Meitu brand, but Xiaomi will be responsible for design, R&D, and sales, while Meitu will still be involved in camera software. Basically, it’s a safe bet that Xiaomi’s Mi Home stores across the world will start carrying a bunch of wild selfie-focused cameraphones in the near future.
As for why this deal is taking place, Meitu says its “mission is ‘to inspire more people to express their beauty,’ and the board believes that entering into this Strategic Cooperation Agreement will accelerate our pace in carrying out this mission.” Meitu isn’t a major smartphone player by any means — it’s only sold 3.5 million handsets in the five years its hardware business has been around — and has identified Xiaomi as the “perfect partner” to help increase its growth.
Xiaomi will initially pay a percentage of phone profits to Meitu, moving to flat per-phone licensing fees after five years or when a certain number of units have been sold overall. Meitu also says that the deal will allow it to focus on expanding its social network user base through pre-installations, though it’s unclear whether that would apply to Xiaomi-branded phones as well.

YouTube is now streaming free, ad-supported feature films

YouTube recently began to roll out a series of feature films that viewers can watch for free, supported by ads. While the selection isn’t large, there are some notable ones to catch, like The Terminator, Legally Blonde, and Rocky.
According to AdAge, the video platform began releasing the movies in October with a “Free to watch” category in its movie section, where users have traditionally purchased or rented movies. The films come with commercial interruptions, with pop-up ads that appear at regular intervals.
Rohit Dhawan, YouTube’s director of product management, told AdAge that the company saw an opportunity for consumers and advertisers, and notes that there might be a way for advertisers to “sponsor” films or hold exclusive screenings. At the moment, the selection is limited to 100 films, a selection that Dhawan says will eventually expand.
The feature comes not long after Roku announced that its free, ad-supported movie selection, The Roku Channel, would be available on the web (it was previously only available to those with a Roku TV or box), which also includes a number of older films, like The Matrix and 50 First Dates. TechCrunch points out that Walmart has its own selection of free streaming films on Vudu, while service Tubi has its own large selection of free films.

Google Home Hub, Pixel 3, and Pixelbook are all getting discounts on Black Friday

The Pixel 3 was just released last month, but if you waited until November to buy one, you’ll be rewarded with up to a $200 price cut. Additionally, for Black Friday and Cyber Monday this year, Google’s sales will also center on accessories for the home, including the new Google Home Hub, Nest devices, and Chromecast dongles. While the newly announced Pixel Slate will not receive a discount, there are some deals on laptops like the Pixelbook and Chromebooks.
Here’s a look at what you can expect to save on at the Google Store and beyond.
The best Google product to buy on Black Friday
Google Home Hub, the company’s inevitable Google Assistant on a screen, is one of the better deals from Google this season. At $99, that $50 savings makes it officially cheaper than the retail price of display-less Google Home speaker.
That doesn’t mean the Pixel 3 isn’t going on sale at all... just don’t expect the usual BOGO deals like you would get with Samsung phones. Instead, ahead of Black Friday, Google will offer a buy one, get the second phone for half off for Pixel 3 devices. Then, on November 22nd, Google’s discounting the Pixel 3 and 3 XL down to $649 and $699, respectively — which is a savings of $150 for the smaller device and $200 for the larger one. On Cyber Monday, the phones will go back to their original retail prices, but you’ll get a free Google Home Hub and a $50 credit to spend on the Google Store. That’s not the best deal, but it could be ideal if you’re already interested in the smart display and you were planning to buy a phone case for your Pixel anyway.
Lastly, it looks like Google’s clearing up some inventory from last year’s hardware event, too. The Google Clips camera (remember that?) is getting a price cut from $250 down to $125, and the Pixel Buds are now $109, down from $159. With plenty of Google products under $50, now’s a good time to snag presents ahead of the holidays.

Apple updates MacBook Pro with AMD Vega GPU options

As promised, Apple has updated its MacBook Pro configurations with new options for AMD Radeon Pro Vega graphics cards, specifically the all-new Pro Vega 16 and Pro Vega 20 options. Apple first announced these new component choices as part of its fall Mac hardware event, where it announced a new MacBook Air and Mac mini.
The options are now available on Apple’s online store for the 15-inch MacBook Pro with Touch Bar. It’s only available on the laptop’s highest-end configuration. You’ll be paying an extra $250 for the Pro Vega 16, and an extra $350 for the Pro Vega 20. That’s on top of the laptop’s $2,799 base cost. The laptop’s base GPU remains the AMD Radeon Pro 560X, while the 2.2Ghz i7 15-inch model comes standard with the 555X and the option to upgrade to the 560X for an additional $100. (The 2.2GHz model has no available option to upgrade to the Vega cards.)
Apple said last month that the new cards would provide up to a 60 percent faster graphics performance for GPU-heavy tasks like video editing and 3D rendering. Additionally, AMD claims its second-generation high bandwidth memory technology for the cards is superior to standard GDDR5, with the added bonus of more power efficiency.
Right now, you can already buy the new iMac Pro with an AMD Vega card, albeit with the Vega Pro 56 or the $600 more expensive Vega Pro 64, both of which sport much more power than what it seems like we’re getting with the Vega 16 and 20. Still, AMD has custom produced these card configurations specifically for the MacBook Pro, so it’s likely they’ll offer a considerable performance boost.

Intel’s new 5G modem might power Apple’s first 5G iPhones

2019 is shaping up to be a big year for 5G, and Intel — one of tech’s biggest mobile players — has finally announced its plans for the next-generation network in the form of its new XMM 8160 5G modem. The XMM 8160 modem is set to be released to manufacturers sometime in the second half of 2019, with the first devices using the chip coming in early 2020.
Intel has big ambitions for the XMM 8160 5G. It envisions using it across phones, PCs, and broadband hubs, with peak speeds of up to 6 gigabits per second. The modem will support both the standalone and non-standalone specs for the 5G NR (New Radio) standard, as well as legacy support for 4G, 3G, and 2G networks all in one chipset. Additionally, Intel says that the modem will support both millimeter wave (mmWave) spectrum as well as lower-band parts of the spectrum.
Intel may have a lot of catching up to do when it comes to getting companies to actually use its 5G modem, though. Currently, at least 18 major companies — including Samsung, Nokia / HMD, Sony, Xiaomi, Oppo, Vivo, HTC, LG, Asus, ZTE, Sharp, Fujitsu, and OnePlus — are working with Qualcomm and its Snapdragon X50 5G NR modems. Huawei and Samsung are both also working on their own in-house 5G modems, too.

The best deals from Alibaba’s Singles Day

Alibaba’s Singles Day is a huge international shopping event that originated in China and has now spread across the globe. Deals kicked off on November 6th, and shoppers everywhere from the US and UK to China and Indonesia are able to nab some good deals.
Many analysts, including App Annie and Adobe, predict that Singles Day will sell more than Black Friday and Cyber Monday combined, based off of last year’s numbers. App Annie put the figure at over $32 billion, while Adobe is estimating $23.4 billion in sales.
Still, the holiday isn’t widely known here in the US. And for newcomers, it might not be clear how to get started grabbing deals. Here’s a quick tutorial on what you’ll need to know.
The holiday, which is believed to have been started in the 1990s as a way for men in colleges to celebrate being single, was turned into a shopping event in 2009 by Alibaba’s Jack Ma and has since become an annual tradition. Other retailers, including JD.com, Suning, and smaller brands, have started to participate in the holiday by offering sales and promotions, too.
Alibaba has multiple platforms that it offers sales on: Taobao, Tmall, and AliExpress. The latter is aimed more at international customers and is the only one of the three available in English. Tmall and Taobao are quite similar, except Tmall’s deals are offered by companies to customers, while Taobao contains deals sold by individuals, which is more like eBay.
Some of the highlights we anticipate on AliExpress and Tmall include deals on Huawei and Xiaomi devices. There are also plenty of third-party fandom apparel, accessories, and household items on sale. We’re updating this article constantly with the latest Singles Day deals.

Internet freedom continues to decline around the world, a new report says

Digital authoritarianism is on the rise, according to a new report from a group that monitors internet freedoms. Freedom House, a pro-democracy think tank, said today that governments are seeking more control over users’ data while also using laws nominally intended to address “fake news” to suppress dissent. It marked the eighth consecutive year that Freedom House found a decline in online freedoms around the world.
“The clear emergent theme in this report is the growing recognition that the internet, once seen as a liberating technology, is increasingly being used to disrupt democracies as opposed to destabilizing dictatorships,” said Mike Abramowitz, president of Freedom House, in a call with reporters. “Propaganda and disinformation are increasingly poisoning the digital sphere, and authoritarians and populists are using the fight against fake news as a pretext to jail prominent journalists and social media critics, often through laws that criminalize the spread of false information.”
In the United States, internet freedom declined in 2018 due to the Federal Communications Commission’s repeal of net neutrality rules. Other countries fared much worse — 17 out of 65 surveyed had adopted laws restricting online media. Of those, 13 prosecuted citizens for allegedly spreading false information. And more countries are accepting training and technology from China, which Freedom House describes as an effort to export a system of censorship and surveillance around the world.
Of course, there are tradeoffs between freedom and security. The report is critical of Sri Lanka and India, which have periodically shut down or limited access to the internet in response to the outbreak of ethnic and religious conflict. In both cases, citizens were being murdered by mobs that had encountered misinformation spread through social media.