During an election year, one might expect a lot of mud-slinging between candidates and political parties. Politics as usual, right. However, for years,social media platforms have been battling abusive (political and non-political)language aimed at users. According toSearch Engine Land, many different strategies have been tried to curtail these obnoxious posts and the users that put “toxic” ideas into what often began as civil conversations.
Google and its companion site, Alphabet, are trying a different approach, using an AI called Perspective, which was released this week.Perspective has been designed to identify and filter abusive and offensive comments before they can be posted on social media.
They haven’t worked out all the kinks as yet because Perspective sometimes takes out comments that uses certain “offensive words” to emphasize in a positive manner, but the wording patterns are misidentify as inflammatory remarks.
The abusive language becomes so volatile that some websites shut down their commentaries feature rather than deal with the hassle of monitoring them. As a result, the abusers go to open platforms to vent their incendiary passions.
Some of Google’s subsidiary companies like YouTube have had trouble with abusive posts in their comment sections as have other social media platforms. Perspective could be a forerunner to other AI moderators that will be a solution (hopefully, long-term) to this kind of problem.
Perspective could also be made available to users as a self-monitoring device. It has been suggested that the AI could be used as a pre-posting check, a tool users could utilize to find out if a questionable comment might be considered offensive. Users would be able to voluntarily censor themselves.