Google to stamp out internet trolls with new software

The internet may have made communication easier and more widespread, but it’s also heralded an era of online confrontation and harassment. Google is planning to use artificial intelligence to automatically tackle the online trolls and make the internet a more harmonious place for all of us.

The company has launched its new Perspective software, which monitors comments and automatically flags up any that are potentially abusive.

You can see the software in action on Perspective’s website ( perspectiveapi.com). Simply type a phrase into the Writing Experiment box, and you’ll see how ‘toxic’ your words are. For example, typing “You are a complete idiot!” told us that our words were 97 per cent similar to comments that people had considered toxic.

Perspective, developed by Google subsidiary Jigsaw, uses machine learning to automatically update its sense of what’s acceptable. It was trained originally by being fed millions of comments from the New York Times website, Wikipedia and other sources.

Critics have warned that Perspective could threaten free speech and become a form of automatic censorship. But the software’s developers say it’s a step towards better online communication.

“It’s a milestone, not a solution,” Jigsaw’s founder, Jared Cohen, told Wired. “We’re not claiming to have created a panacea for the toxicity problem.”