Where I'm from, minimum wage is around 15k per year. So hiring 5k people would cost 75M in salaries, and let's just double it for other operation costs, 150M. People don't work 24/7, so we must quadruple it for triple shifts + weekends. That leaves us at around 600M: yes, that's terribly expensive. 1
But maybe we don't need to review all the reports! What if we rank them? Correct reports grand a user positive 'report karma', and they climb higher. Posts that are 'trending' in reports should also climb higher. New posts take priority over old posts, ie let's police new content, not everything that has been ever written. And of course, we can also use modern techniques, ie AI to score the reports.
If Twitter did that and could review the top 10% of reports, at 60M per year, would that be worth it? If successful, I think so.
1: this number is biased because it assumes all reports are independent, but I suspect a significant amount of reports would concern a low number of posts, eg controversial or political posts from accounts with a large follow count.
But maybe we don't need to review all the reports! What if we rank them? Correct reports grand a user positive 'report karma', and they climb higher. Posts that are 'trending' in reports should also climb higher. New posts take priority over old posts, ie let's police new content, not everything that has been ever written. And of course, we can also use modern techniques, ie AI to score the reports.
If Twitter did that and could review the top 10% of reports, at 60M per year, would that be worth it? If successful, I think so.
1: this number is biased because it assumes all reports are independent, but I suspect a significant amount of reports would concern a low number of posts, eg controversial or political posts from accounts with a large follow count.