Perhaps I am naive. It seems they do benefit from hate, in that negative emotions drive more engagement.
IF they truly didnt want to benefit from hate, they would find a way to not always rate negative engagement (arguing) as "engaging".
IF they truly didnt want to benefit from hate, they would find a way to not always rate negative engagement (arguing) as "engaging".
It's really hard to tell, because nudity or illegal items for sale is removed within minutes of posting. Sometimes Facebook clearly overreacts in these cases, but it seems they're not reacting to hate speech or racism at all. At least they should be able to flag and hide posts by simple keyword searches and then sent them of to human moderators, who can then unhide false positives and delete offending posts.
Honestly I think the truth is that Facebook simply isn't able to moderate their platform, the volume of data is to high and their A.I/filters/keywords search technology isn't nearly as good as customers, media, politicians and investors believe, but Facebook won't admit it.