Serendipity35

As part of Twitter's efforts to limit the visibility of abusive posts, the company is using more automated methods. They (and other social media and news sites) have gotten a lot of bad press the past year for allowing the spread of harmful content. This is certainly a worthy goal, but a difficult one to implement. At one time, many sites relied on human monitors and community managers to respond to abuses, but the volume of posts makes that impossible.

Twitter has expanded its mute and filter features. Now, they will also notify users when they take action on harassment they reported, even if it's not directed at them.

Perhaps most pwerfully - and therefore most frighteningly - Twitter is implementing tools that can automatically identify patterns in the way harmful tweets are made and distributed. CNN reports that "If it identifies a harmful account, it will truncate the user's visibility. For instance, if a user routinely replies to accounts that don't follow it, or repeatedly violates Twitter's rules, Twitter might make it so that the tweets are invisible to everyone except the user's followers for a set amount of time."

That means if you routinely reply to a favorite celebrity (even with kind words) and the celebrity doesn't follow you (and most will not) that invisibility function kicks in. The algorithm thinks you are stalking/harassing the person. This is what someTwitter users have called "Twitter jail."

What behaviors might trigger the algorithm and put you in jail or ban you from Twitter? The company won't say because they don't want people to find workarounds.

Big Brother isn't watching you - an algorithm is...