[Cthulhu_]

> I've been brewing up strategies of letting the community itself moderate because a machine really cannot "see" what content is good or bad, re: context.

There's a ton of material on that subject, thankfully; look at news groups, HN itself (flagging), Stack Overflow, Joel Spolsky's blogs on Discourse, etc etc etc. My girlfriend is active on Twitter and frequently joins in mass reporting of certain content, which is both a strong signal, and easily influenced by mobs.

[halfmatthalfcat]

Do you have any references to those materials? Would love to read up on some more academic takes.