14-07-2021, 06:45 AM
Just as a postscript to my post yesterday about the online racist abuse, it has now emerged that much of the abuse took the form of banana and monkey emojis and phrases like 'go home' which circumvented Instagram and Facebook's policing system - computer bots.
This from today's Time magazine:
"The problem is, those algorithms don’t always work very well. After the England players missed their penalties, several users shared images on social media of Instagram’s moderation system responding to reports of posts containing racial slurs and emojis saying “our technology has found that this comment probably doesn’t go against our Community Guidelines."
What would be obvious to a human moderator, computer AIs are blind to. The sheer number of posts, particularly for high-profile figures and high-profile events, would require a huge army of human moderators to police.
Maybe the best solution to to allow other users to put the abusers in their place. But as Woody Allen once said about how to respond to anti-semitism "It's difficult to satirise someone with shiny boots."
This from today's Time magazine:
"The problem is, those algorithms don’t always work very well. After the England players missed their penalties, several users shared images on social media of Instagram’s moderation system responding to reports of posts containing racial slurs and emojis saying “our technology has found that this comment probably doesn’t go against our Community Guidelines."
What would be obvious to a human moderator, computer AIs are blind to. The sheer number of posts, particularly for high-profile figures and high-profile events, would require a huge army of human moderators to police.
Maybe the best solution to to allow other users to put the abusers in their place. But as Woody Allen once said about how to respond to anti-semitism "It's difficult to satirise someone with shiny boots."