Facebook’s New Anti-Violence Plan Will Sometimes Leave Up Disturbing Videos To Help Authorities

Justin Sullivan / Getty Images

Facebook is adding 3,000 workers to fight violence on its platform, and their mandate isn’t simply to remove violent content.

Facebook may not immediately remove all violent live videos, Facebook CEO Mark Zuckerberg said during his company’s first quarter earnings call Wednesday. The company will also work with law enforcement to prevent violence, he said, even if it means risking airing the footage.

“A lot of what we’re trying to do is not just about taking the content down, but also about helping people when they’re in need on the platform,” Zuckerberg said.

Law enforcement recently prevented someone from killing themselves after Facebook alerted them to a suicidal broadcast, Zuckerberg said. In that case, authorities communicated with the person through the live video itself, and prevented them from going through with the act. For Facebook, a social giant of 1.94 billion users, using its power to simply remove violent content is not enough. It wants to prevent the violence itself.

Facebook’s 3,000 new content moderators, combined with the 4,500 moderators it already has, won’t be fighting violence alone. They will be enlisting Facebook’s artificial intelligence as well. Zuckerberg noted that even with 7,500 people scanning Facebook’s posts for possible violence, “we’ll never be able to look at everything.” Artificial intelligence, he suggested, could fill in the gap, though its presence may not be felt for years.

“Right now there are certain things that AI can do in terms of understanding text and understanding what’s in a photo and what’s in a video,” Zuckerberg said, noting that the technology was likely years away from working effectively. “Over time, for sure, more AI will do this.”

AI, in many cases, is still far from the romanticized treatment it gets in the press and in film. “This is still very far from where we want it to be,” Facebook’s head of AI research Yann LeCun told BuzzFeed News in a recent interview.

The highly-publicized violence on Facebook Live did little to scare Facebook’s advertising customers away in the first quarter of 2017. The company brought in more than $8 billion in revenue, up from $5.3 billion in the same quarter last year.

Quelle: <a href="Facebook’s New Anti-Violence Plan Will Sometimes Leave Up Disturbing Videos To Help Authorities“>BuzzFeed

Here Are The Funniest Tweets About The Google Docs Scam

Today, a massive, fake Google Docs phishing scheme spread across the internet.

Giphy

The attack worked by sending people emails that imitated how Google emails look when people share real Google Documents with each other. But in this case, the links were from a replica of Google Docs — and if you clicked on them and granted the app permission to access your account, the attack then spammed everyone on your contact list with requests to share more fake Google Docs.

If you've already clicked on a spam link, go to your Google account permissions and revoke access for “Google Docs.” The real Google Docs doesn't appear individually in that permissions list since it's part of Google Drive, so you won't be interfering with any of your real work.

After the initial panic subsided, people came up with some jokes.

If this kind of email hack sounds familiar, may I present to you several jokes at the Democratic National Committee's expense:

And if you've seen that one episode of Black Mirror, this tweet has you shook:

No one's yet claimed responsibility for the hack, but at the end of the day, it's obvious who's really behind it.

Quelle: <a href="Here Are The Funniest Tweets About The Google Docs Scam“>BuzzFeed

Google Technology Is Helping Exterminators Kill Bugs Faster

Sergeytoronto / Getty Images

Imagine opening your favorite breakfast cereal, only to find half a dozen feasting weevils inside the box. Panicked, you call an exterminator — but what if instead of a dime-store Ghostbuster, the person that shows up at your house is a smartphone-wielding, AI-using, Google-powered, bug-destroying 21st century pest technician?

If you live in the general vicinity of Reading, PA, that scenario could become a reality sooner than you might think.

Using an app called PestID, exterminators who are part of the pilot program can take photos of bugs, upload them, and — using Google’s Vision API — instantly get information about what kind of insect it is and how to kill it.

Google’s Vision API tells you in words what an image depicts. PestID is trained on proprietary visual data owned by an extermination company called Rentokil, which is currently testing the technology in the field in Pennsylvania. The app, which is the result of a partnership between Google and Accenture, uses machine learning to constantly improve its accuracy.

“If I took a picture of a confused flour beetle, Google’s out of the box capability would be able to tell me it was a beetle,” said Nisha Sharma, an Accenture Mobility executive who worked on PestID. “But with Rentokil, it can actually tell you it’s a confused flour beetle.”

Pest extermination doesn’t exactly scream out as the obvious first workplace use case for advanced artificial intelligence. But apparently, it’s not as easy as you’d think. While most exterminators have plenty of experience and most extermination jobs are fairly routine, says Rentokil’s Keith Chisolm, a hard to recognize pest can really slow a pest technician down.

“If you do come across something that's particularly unusual, what would typically happen is they'd get in touch with a more experienced technician, or ask one of the support team to come and meet them on site,” he said. “ [PestID] helps them to deal with it themselves, and not have to lean on technical teams.”

Chisolm, who first had the idea of using AI to identify bugs when he saw that Facebook was automatically identifying people in photos, said exterminators who have used the app are blown away by how quickly it can identify a bug.

PestID can help Rentokil technicians with the more banal parts of pest extermination, too, like prompting the technician with the right treatment, and auto-populating paperwork; it can even use GPS to guess whether the pests are located inside or outside a facility. “There's a lot of stuff that these guys have to do manually,” said Accenture’s Sharma. “Using this app, we can automate it, and make it quicker.”

Employing faster technicians who can do more jobs in a single day is obviously attractive to a company like Rentokil, but it’s not the only perk of partnering with Accenture and Google. While the big companies own the underlying technology, Rentokil owns the database, which means as the system continues to learn about new bugs, Rentokil will own the most accurate and easy-to-use pest identification system on the market, which it can market directly to consumers.

“If you found an insect in your house, in the larder, munching away on your cereals, and you wanted to know, What is it? Do I need to be worried? What do I do about it? You could use the Rentokil app to identify the pest, we could give you advice, link you to a local pest controller, who could come and help you with that problem,” said Chisolm.

Even in a world where homeowners and companies pay for an app that identifies pests, Chisolm says a professional would still have to come by and actually kill the things. And that may be true, for now — but pest-killing drones and termite-destroying robots already exist.

“Could you replace that with a robot?” Chisolm asked. “I guess you could.”

Quelle: <a href="Google Technology Is Helping Exterminators Kill Bugs Faster“>BuzzFeed