The internet is a tough place to have a conversation. Abuse has driven celebrities and ordinary folks from social media platforms that are ill-equipped to deal with it, and some publishers have switched off comment sections.
That’s why Google and Jigsaw (an early stage incubator at Google parent company Alphabet) are working on a project called Perspective. It uses artificial intelligence to try to identify toxic comments, with an aim of reducing them. The Perspective API released Thursday will provide developers with a score of how likely users are to perceive a comment as toxic.
In turn, that score could be used to develop features like automatic post filtering or to provide users with feedback about what they’re writing before they submit it for publication. Starting on Thursday, developers can request access to Perspective’s API for use in projects they’re working on, and Jigsaw will approve them on a rolling basis.