Conversations can become vicious on Social Media. Instagram is painfully aware of that, even on its platform, that’s supposed to be filled with beautiful imagery. So, it has appealed to machine learning to sift through comments and eliminate those offensive #securemagic
Until now, Instagram had blocked comments that it deemed offensive by first cross-referencing them with a limited list of words and phrases it considered insulting. But as we all know, people can get creative when trying to do some damage and the system definitely didn’t cover those.
That’s when Instagram decided to employ the help of AI to understand the context of some comments and replies and by doing so, really get rid of all forms of unacceptable behavior online. When the machine learning algorithm discovers a negative comment, it’s blocked for everybody, except the user who left it. So, he/she won’t know if their reply was accepted or not.
Unlike the previous time, Instagram is turning this AI filter on by default now, but gives users the option to turn if off from settings. So far, the company managed to implement this filtering system for comments in English and said they were working to expand the service in other markets.
Instagram took advantage of this opportunity to officially announce a smarter spam filtering system, also based on machine learning. Unlike the comment-blocking one, this is already active in nine languages.