[H]ardOCP: Alphabet Launches Chrome Extension That Filters Comments With AI

Remigio Civitarese
Marzo 14, 2019

To make your online life slightly stress-free, a Google Chrome extension is taking the help of artificial intelligence.

The internet is a merciless place and dealing with trolls is a pain. Not everyone is a name worthy enough for such backend solutions, so, for the rest, Tune might bring some peace of mind. Tune works for comments on a variety of platforms, including Reddit, Twitter, Facebook, YouTube, and Disqus. "It still misses some toxic comments and incorrectly hides some non-toxic comments".

Jigsaw acknowledges that Tune's machine learning algorithms are still experimental, meaning that it will not hide all toxic comments and may sometimes block non-toxic comments.

Users can keep Tune on for a certain website and off for another, it is entirely up to them.

Alphabet's Jigsaw launches tool to hide noxious comments
Google's AI-Powered Tune Chrome Extension Hides Toxic Comments

With the Tune Chrome extension, you can choose to turn all toxicity off (called "zen mode") or experiment with different settings to find one that works for you.

Most users will want to set their volume somewhere in the middle, and Tune can filter out attacks, insults, profanity, and more. The tech that came in to being in the form of an API back in 2017, now essentially uses the same machine learning tech with the Tune Chrome extension.

"We hope that Tune can inspire platforms and developers to explore viewership controls for readers and enable communities to join discussions without relying exclusively on comment moderation", wrote CJ Adams, Product Manager, Jigsaw, in the blog post.

Needless to say, all the comments that the Tune blocks does not disappear from the world wide web, it's just that you don't get to see it because you chose not to.

Altre relazioniGrafFiotech

Discuti questo articolo

Segui i nostri GIORNALE