That is my take on AI/auto content moderation and the enforcement - they can't have it both ways; centralized platform, with no responsibilities.
So - hire people. Actual human beings - train them and hire large numbers from many demographic and geographic locations.
But - this would cut into profits.
Because the userbase is so ludicrously big "Massive" doesn't really convey the level of staffing in relative terms.
How many posts does the network have per moderator-workday?
What is the number for integrity, that you think would solve the problem and why? how so this compares to other companies and facebook. Luckily we have so many social networks we can do statistics?
And then, vefore you know this, and the problem at hand, maybe you should suspend your harsh judgment?
https://www.bbc.co.uk/news/technology-57088382
https://www.nytimes.com/2021/08/31/technology/facebook-accen...
>...15,000 people whom Facebook has said it has hired to inspect its posts.
How many similar contractors are there, say what about India?
Another thing, the moderation is a combination of ml technique and human, the next question is how much load ML takes a way from human operator. How efficient are their tools?
What is recall and precision of removing violent content?
How does this compare with other social networks?
How much moderation is done organically by users?
its about money, not the lack of ability.
I'm gonna assume the latter. And forget about 'massive', they clearly don't even want to hire enough to make it decently large.
Tech companies are fine hiring people when they can't just push the costs (social and monetary) onto others. But Facebook thinks it can - so it is.