Twitter bans ‘Impostor Buster’ bot that ID’d Nazi trolls

It’s been a rocky road to Twitter’s newfound pledge to kick members of hate groups off its platform. For those on the end of racist abuse, reporting trolls to Twitter has been a frustrating process. But, earlier this year, one journalist decided to take matters into his own hands. After suffering a record amount of harassment, reporter Yair Rozenberg teamed up with dev Neal Chandra to build a bot that unmasked impersonator accounts manned by bigots. It got off to a great start, explains Rozenberg in a New York Times op-ed, before Twitter shut it down at the behest of Nazis.

The bot, dubbed “Impostor Buster,” went after trolls with fake profiles of real individuals from ethnic minorities. To fool unwitting users, they’d adorn these accounts with clear identifying markers “like a yarmulke-clad Hasid or a woman in hijab,” along with descriptors like “jewish,” “muslim” or “enemy of the alt-right” in their bio, writes Rozenberg. Armed with the fake identity, they’d interject their racist bile in Twitter conversations by high-profile users, in the hopes of defaming an entire community.

Tapping into a crowdsourced database of impersonator accounts, which were curated to avoid false positives, Impostor Buster leapt upon the hijacked conversations to publicly expose the fakers. In a matter of weeks, the bot accrued thousands of followers, while its creators received thank-yous from victims, making it a success story for our troubled times. Then the Nazis mounted a fightback.

Inundated with reports of harassment from bigots, Twitter briefly suspended the bot in April. Despite fine-tuning it to evade Twitter’s alarms, the bot was permanently banned this month. It seems Twitter gave into the reports of spam and unsolicited replies from the neo-Nazis the bot had been targeting.

The move is just another sad chapter in Twitter’s paradoxical fight against hate on its platform. The moral of the story, according to Rozenberg, is that the company’s top-down enforcements are unfeasible. Instead of censorship, he posits a bottom-up approach using bots like his own to nurture a “healthier culture below.”

Source: The New York Times

Scientists made a virtual supernova you can walk around

At 11,000 light-years away, it’s safe to say you won’t be able to visit Cassiopeia A even if private space corporations start selling seats to amateur spacefarers. Thanks to a team of scientists, though, you might be able to see the 300-year-old supernova remnant up close — or even walk inside it. The team used data from NASA’s Chandra X-ray Observatory, the Spitzer Space Telescope and ground-based facilities to create a 3D model of Cassiopeia A with VR and AR elements. You’ll be able to walk into the model, engage with parts of the remnant and conjure up short captions on what you’re seeing and “touching.”

To build the 3D model, they looked at the elements inside a supernova (such as iron, silicon and argon) and the light they emit when heated. That gave them a way to figure out how fast the supernova’s debris are moving and in what direction. The result is a visualization showing the spherical component making up the outer parts of the remnant and the flat, disk-like component making up the inner part. You can also see high-velocity plumes or jets shooting out of the structure:

Since the team created the model specifically for other astronomers building supernova models, it’s unfortunately not that easy to get your hands on it. You’ll have to contact Kimberly Arcand, the Visualization Lead for NASA’s Chandra X-ray Observatory, to get the open access files for VR caves (immersive VR rooms) and for the Oculus Rift. The team is also working on a Google Cardboard version, so you may want to wait a bit before sending in a request if that’s your preferred device.

Source: Chandra X-ray Observatory

Snapchat Stories May Soon Be Shared Online Like Embedded Tweets

Snap is reportedly working on a new Snapchat feature that will let users share stories outside of the mobile app, in an effort to boost Snapchat’s presence in the social media space (via The Verge).

Called “Stories Everywhere”, the feature should see Snapchat’s premium video and editorial content appear on news sites and elsewhere online. The move aims to bring greater visibility to the brand, in the face of increasing competition from the likes of Facebook and Instagram.

Live streaming news network Cheddar reports that former Storyful CEO and News Corp executive Rahul Chopra is leading the “Stories Everywhere” project. Chopra’s job is to attract more active users to the social platform – similar to how, in 2011, Twitter began letting people embed tweets off-platform for the first time, for example.

The strategy comes at the end of a bad year for Snap, which saw third quarter revenue fall after its March initial public offering failed to excite investors, many of which struggled to understand how the app works. In response, the company earlier this month released a redesigned Snapchat app that aims to offer a more personalized user experience that’s easier to navigate, especially for new users, with brand and friends’ content kept separate.

Tag: Snapchat
Discuss this article in our forums

WhatsApp to Pull Support for Older iPhones From Next Year

WhatsApp announced on Thursday that its mobile app will stop working on some older smartphones from next year, as it concentrates on developing new features for the popular messaging platform.

The decision means that from January 1, 2018, WhatsApp will pull support for the app on iPhone 3GS and older models, as well as iPhones running iOS 6 and earlier. The list of unsupported phones also includes Android 2.1 and 2.2, BlackBerry OS, and Windows Phone 7.

Explaining its reason for the move, WhatsApp said that when it was founded back in 2009, the mobile landscape was very different.

“About 70 percent of smartphones sold at the time had operating systems offered by BlackBerry and Nokia,” the company said. “Mobile operating systems offered by Google, Apple and Microsoft – which account for 99.5 percent of sales today – were on less than 25 percent of mobile devices sold at the time.”

Apple’s iPhone 3GS was released in June 2009, while iOS 6 was released on September 19, 2012, followed by iOS 7 in 2013. According to Apteligent Research, less than 0.04 percent of iPhone users are currently using iOS 6. Apple’s most recent mobile OS, iOS 11, was released on September 19, 2017, and accounts for over 75 percent of active iOS users.

(Via Sky News.)

Tag: WhatsApp
Discuss this article in our forums

iMac Pro Base Model Teardown Reveals 2x SSD RAID Configuration and Four 8GB DIMM Modules

While we await iFixit’s inevitable comprehensive teardown of Apple’s new iMac Pro, third party Mac component supplier OWC has just published its own teardown video, providing some interesting tidbits on the internal configuration of the non-user upgradeable machine.

Whereas standard 27-inch iMacs have a small hatch in the back that allows the RAM in the machine to be upgraded after purchase, the iMac Pro does not. Fortunately, an Apple Store or an Apple Authorized Service Provider is able to open up the iMac Pro and swap out the RAM, and here’s what they can expect to find upon doing so.

The teardown reveals that in the 32GB base model, there are four 8GB DIMM modules, a configuration type that appears to be mirrored in the 64GB (4 x 16GB) and 128GB (4 x 32GB) models. The good news is that this means the iMac Pro supports quad-channel memory, but the bad news is that it also means users looking to upgrade from, say, 32GB to 64GB will have to replace all four modules to do so.

The teardown also reveals that in the iMac Pro 1TB base model, Apple has chosen to use two 512GB SSDs in a RAID configuration. Rather than soldering the flash storage on the main board, both drives exist as separate modules that are attached via screws, so replacing them is at least technically feasible, even if Apple does not make it easy.

OWC says that in the near future it will be offering a DIY memory upgrade kit for the iMac Pro, although most users are likely to hand over such an undertaking to a qualified service provider. For more details on the iMac Pro’s internals, including the Intel Xeon W eight-core processor, be sure to watch the video embedded above.

Related Roundup: iMac ProTag: OWCBuyer’s Guide: iMac Pro (Buy Now)
Discuss this article in our forums

This self-driving delivery robot uses facial recognition to unlock the goodies

Small, wheel-based delivery robots continue to be of interest to many startups around the world, though considering the tragedy that befell this hitchhiking robot a few years ago, it’s hard to know if the human race is actually ready to fully embrace the idea of autonomous robots trundling along the street by themselves.

One outfit that isn’t giving up on delivery robots is E-Novia. Based in Milan, Italy, E-Novia is currently testing YAPE — short for Your Autonomous Pony Express — a compact self-driving robot that has an impressive range of some 50 miles (80 kilometers).

Bringing to mind Piaggio Fast Forward’s Gita robot, YAPE moves along on two-wheels and self-balances while reaching speeds of up to 12.4 mph (20 kmh), according to New Atlas.

Able to transport packages weighing up to a hefty 154 pounds (70 kilograms), YAPE uses a combination of sensors, cameras, and GPS to navigate sidewalks and other off-road routes such as cycle lanes to reach its destination. It also communicates with sensors fitted to traffic lights and other road equipment so it can respond to fast-changing street conditions.

When it reaches the delivery address, the recipient simply shows their face to YAPE’s facial recognition camera to unlock the compartment containing the package.

“YAPE is simple to use thanks to an app connected to a control platform,” E-Novia explains on its website. “To send packages via YAPE, all you have to do is input the destination address in the app, or simply to whom the package is addressed, leaving the task of finding the correct address to YAPE’s GPS system. On delivery, the recipient will only be able to open the package compartment using a facial recognition system.”

It all sounds rather cool, though regulatory hurdles will need to be overcome if such technology is ever to make it into the mainstream. And we still worry about ne’er-do-wells who believe it’d be hilarious to throw a blanket over YAPE as it goes on its merry way, or simply lift it up and toss it into a canal. Remember poor ol’ hitchBOT.

If a secure delivery system can be worked out, then we’re right behind YAPE and other similar designs offered by the likes of Starship Technologies and DHL. Heck, robots like these don’t even have to go out by themselves — they could simply act as a personal helper or robotic companion that follows you around and carries your stuff.

E-Novia will be showing off YAPE at CES 2018, and DT will be there to check it out.

Editors’ Recommendations

• Bigger isn’t always better: Five of the smallest smartphones worth buying
• Rylo hands-on review
• Lyft riders can now take a driverless-car trip in Boston
• Facebook facial recognition prevents someone from stealing your photo
• Max your mobile: How to use the Game Booster app on the Razer Phone

Olympic organizers may use facial recognition to manage guests

Japan is looking for a different way to confirm the identity of the hundreds of thousands of athletes, officials and journalists who will be attending the 2020 Olympics. The Japan Times reports that sources close to the Tokyo 2020 Olympic organizing committee say organizers are planning to utilize facial recognition technology. Not including spectators, attendees are expected to number in the 300,000 to 400,000 range and officials are concerned that ID sharing or theft will reduce the Games’ safety. Facial recognition technology could get around those issues while also reducing entrance waiting times.

The technology wouldn’t be used on spectators, say the sources, and it was tested at the Japan House during the 2016 Rio Olympics where it was used to manage journalists coming and going from the visitor center. NEC Corp is said to be the developer of the facial recognition program. Other NEC customers include the Department of Homeland Security, CaliBurger, Chicago police, and Georgia.

The program hasn’t been officially announced and therefore none of the concerns surrounding the use of this technology have been addressed yet. Along with the typical privacy worries that go along with facial recognition software, it’s currently unclear whether attendees could opt out of the system or even if the organizing committee has the authority to make it a requirement.

The Tokyo Olympics will take place between July 24th and August 9th of 2020.

Via: Gizmodo

Source: The Japan Times

Amazon hopes for major expansion of its online ad business

You probably don’t associate Amazon with online ads outside of sales pitches for Echo speakers, but you might think differently in the near future. CNBC sources claim that Amazon is aiming to expand its online ad business well beyond its own sites and products. It’s partnering with mobile ad houses to link promos on mobile and TV, for instance — you’d see a commercial on the big screen followed by a related ad on your phone. It reportedly wants to expand its video ads, as well, by creating “branded” videos with celebrity hosts.

The internet retailer would entice advertisers by giving them the one thing they want the most: user data. Two contacts told CNBC that Amazon was willing to share more info than rivals like Facebook and Google if an advertiser was willing to spend enough. And when Amazon already has both extensive knowledge of buying habits and a huge volume of online traffic, its offers could be tempting.

Amazon hasn’t commented on the apparent scoop. It has every reason to want to grow its ad efforts, mind you. Facebook and Google dominate the online ad world (more than 70 percent between the two, Pivotal estimates), and Amazon could seize on that by becoming a third major option that increases competition and forces prices down. Would-be advertisers might flock to Amazon simply to keep its ad rivals humble.

We’d add that Amazon may want to diversify its sources of income. Even with its growing cloud services, Amazon is still heavily dependent on its store (including Prime subscriptions) for revenue. A large ad business could give Amazon a cushion if the store struggles, and could help fund some of Amazon’s experiments, such as delivery drones.

Source: CNBC

LG unveils its first smart speaker with Google Assistant

LG is acting on its promise to unveil a slew of AI-powered gadgets at CES this January. The tech giant has unveiled a 2018 speaker lineup headlined by the ThinQ Speaker, a Google Assistant-powered device that promises “premium” sound alongside its smarts. It supports lossless high-resolution audio like LG’s more advanced phones, and uses Meridian Audio’s know-how to pump out “more natural and warm” music. The launch also hints that many of LG’s newer appliances will be Assistant-savvy — the company envisions turning on your air purifier with a voice command.

The rest of the lineup is led by the SK10Y soundbar, a 5.1.2-channel companion to your TV with Chromecast support, Dolby Atmos processing, lossless audio and Meridian-made Height Elevation to put the sound at the level of the screen, theoretically providing extra clarity. There’s also the PK-series Bluetooth speakers that use Meridian to produce “powerful and undistorted bass,” and party speakers that kick out as much as 1,800W of power while supporting features like DJ and karaoke modes.

As is usually the case with these pre-CES teases, LG hasn’t provided availability details for the ThinQ Speaker and other devices. And that may be important. We don’t know for sure whether it’s aimed at relatively low-cost speakers like the Sonos One or pricier models like Apple’s upcoming HomePod. However, LG is clearly trying to avoid directly competing with the Amazon Echo, Google Home and other speakers where audio fidelity clearly isn’t the top priority.

Source: LG Newsroom

WeChat accounts could double as state IDs in China

Some WeChat users in China will soon be able to use the chat app as a state-issued ID card. Financial Times reports that those living in Guangdong’s capital city Guangzhou will be able to use facial recognition to link their identities to the app in the Nansha district this week. The feature will apparently roll out across the country in January.

FT notes that the ID-linking system is a bit limited as far as on-device registrations go. You’ll be able to log into an internet cafe with the limited-function, on-device registration. For anything more involved that’d require state ID, however, you’ll need to hit a registration hard point in the city. After that, your phone will cover “most functions” you’d need a typical ID card for. Like registering a business, for example.

While this sounds like the type of future we’ve expected, the feature’s privacy implications can’t be understated. Linking someone’s face to what’s essentially the de facto way to communicate in China is a bit scary considering how heavily the government polices the app.

Source: Financial Times