I wonder if the solution involves partitioning the social graph to allow accounts to coexist?
Instead of trying to censor accounts, because I’m going to assume accounts aren’t used purely for offensive content — that’s the easy case — but rather the account is generating ‘mixed’ content.
Bans are a primitive form of isolating a part of the graph. Particularly if they extend to commenting/replying to that account’s posts.
The false abuse reports similarly should carry an extremely high cost to the submitter. If an abuse report is flagged maybe the account is no longer trusted ever to report a post again. Maybe an abuse report should actually have to carry some monetary value (like hashcash).
Instead of trying to censor accounts, because I’m going to assume accounts aren’t used purely for offensive content — that’s the easy case — but rather the account is generating ‘mixed’ content.
Bans are a primitive form of isolating a part of the graph. Particularly if they extend to commenting/replying to that account’s posts.
The false abuse reports similarly should carry an extremely high cost to the submitter. If an abuse report is flagged maybe the account is no longer trusted ever to report a post again. Maybe an abuse report should actually have to carry some monetary value (like hashcash).