Month: December 2018

(credit: Jeremy Brooks / Flickr)

Almost two dozen apps with more than 2 million downloads have been removed from the Google Play market after researchers found they contained a device-draining backdoor that allowed them to surreptitiously download files from an attacker-controlled server.

The 22 rogue titles included Sparkle Flashlight, a flashlight app that had been downloaded more than 1 million times since it entered Google Play sometime in 2016 or 2017, antivirus provider Sophos said in a blog post published Thursday. Beginning around March of this year, Sparkle Flashlight and two other apps were updated to add the secret downloader. The remaining 19 apps became available after June and contained the downloader from the start.

“Serious harm”

By the time Google removed the apps in late November, they were being used to click endlessly on fraudulent ads. "Andr/Clickr-ad," as Sophos has dubbed the family of apps, automatically started and ran even after a user force-closed them, functions that caused the apps to consume huge amounts of bandwidth and drain batteries. In Thursday's post, Sophos researcher Chen Yu wrote:

Enlarge (credit: Aurich Lawson)

How long does it take someone to have a YouTube impersonation claim reviewed, confirmed, and enforced? That kind of data is hard to piece together across such a giant video-sharing platform. But in the case of one user, Penguin Books author Celeste Ng, the process took a little over one full day—and required a big pile of public shaming in the process.

Ng's ordeal began when she discovered someone had created an account with her first and last name that primarily posted racially and culturally insensitive videos—including apparent dog-whistling mentions of mixed-race marriages. This appeared to be a targeted impersonation attempt, as Ng has previously been targeted by online communities for marrying a non-Asian man.

She took to Twitter to ask her followers how to report impersonation claims on YouTube. After filing a Wednesday impersonation report, which included proof of identify (her published book's jacket sleeve, complete with author photo) and screenshots of the offending, fake YouTube account, Ng received a Thursday response from YouTube: her report did "not meet our impersonation reporting guidelines." A quick scan of YouTube's reporting page includes a request for "a clear, readable copy of your valid driver's license, national ID card, or other photo ID" as an attached image. Ng's posts did not confirm whether her book's jacket photo was YouTube's point of contention.

Enlarge / Employees fufill online orders at the Amazon.com Inc. fulfillment center in Robbinsville, New Jersey, on Thursday, June 7, 2018. (credit: Bess Adler/Bloomberg via Getty Images)

On Wednesday, 24 Amazon employees working in a Robbinsville Township, New Jersey, warehouse were sent to five different area hospitals after a 9oz canister of bear spray was deployed.

Early reports that the canister was punctured by an Amazon robot have not been confirmed by either Amazon or the Township's communication official, John Nalbone. Nalbone told Ars that Robbinsville first responders only reported that an "automated machine" was to blame, which could mean anything from a programmed robotic arm to an automated conveyor belt. (Of course, may readers of Ars would agree that defining the meaning of "robot" is more than just an exercise in trivia. This Wired article explores the topic more deeply.)

On Wednesday evening, one of the 24 people hospitalized was in critical condition, while another 30 were treated on the scene, ABC News says.

Enlarge / The Mars 2020 rover will likely carry artificial intelligence software to help manage the science workload. (credit: NASA)

Welcome to Ars UNITE, our week-long virtual conference on the ways that innovation brings unusual pairings together. Each day this week from Wednesday through Friday, we're bringing you a pair of stories about facing the future. Today's focus is on AI in manufacturing and space—stand by to blast off!

NASA can't yet put a scientist on Mars. But in its next rover mission to the red planet, NASA's Jet Propulsion Laboratory is hoping to use artificial intelligence to at least put the equivalent of a talented research assistant there. Steve Chien, head of the AI Group at NASA JPL, envisions working with the Mars 2020 Rover "much more like [how] you would interact with a graduate student instead of a rover that you typically have to micromanage."

The 13-minute delay in communications between Earth and Mars means that the movements and experiments conducted by past and current Martian rovers have had to be meticulously planned. While more recent rovers have had the capability of recognizing hazards and performing some tasks autonomously, they've still placed great demands on their support teams.

Enlarge / Starting from random play and knowing just the game rules, AlphaZero defeated a world champion program in the games of Go, chess, and shoji (Japanese chess). (credit: DeepMind Technologies, Ltd.)

Google's DeepMind—the group that brought you the champion game-playing AIs AlphaGo and AlphaGoZero—is back with a new, improved, and more-generalized version. Dubbed AlphaZero, this program taught itself to play three different board games (chess, Go, and shoji, a Japanese form of chess) in just three days, with no human intervention.

A paper describing the achievement was just published in Science. "Starting from totally random play, AlphaZero gradually learns what good play looks like and forms its own evaluations about the game," said Demis Hassabis, CEO and co-founder of DeepMind. "In that sense, it is free from the constraints of the way humans think about the game."

Chess has long been an ideal testing ground for game-playing computers and the development of AI. The very first chess computer program was written in the 1950s at Los Alamos National Laboratory, and in the late 1960s, Richard D. Greenblatt's Mac Hack IV program was the first to play in a human chess tournament—and to win against a human in tournament play. Many other computer chess programs followed, each a little better than the last, until IBM's Deep Blue computer defeated chess grand master Garry Kasparov in May 1997.