In a recent post, I discussed the social media monster that is cyberbullying. While we take most issues regarding online harassment into our own hands, it seems that Internet giants such as Instagram want in on the war against trolls and spammers.
Instagram is using AI to try and understand the context around offensive speech and get rid of abuse even if it doesn’t trigger specific keywords. That should help mitigate the cat-and-mouse nature of fighting hateful comments, although it currently only works in English.
Instagram’s keyword filter feature has been a hit with users who prefer not to see certain types of content–but to know that these platforms continue to battle hate speech is reassuring.
The AI system, called DeepText, is able to identify offensive language. For example, it can understand that the word “white” may not be offensive as a color or title (White Sox, white snow), but that “white power” could be offensive in many uses.
Accurately discerning what is offensive remains to be somewhat sticky for AIs, but there is no doubting that technologies like these will see constant improvement.
Are you using this new feature?