There's a "history rhymes" kind of dynamic to this. The first recognizable, critical-mass online social media(broadcast, topic driven, message posting) came fr

There's a "history rhymes" kind of dynamic to this. The first recognizable, critical-mass online social media(broadcast, topic driven, message posting) came from BBS operators, mailing lists, Usenet and proprietary variants. The moderated ones gradually disappeared as users migrated; the unmoderated ones drowned in noise over time. Pieces of it still survive as relics. The next generation was website driven: home pages, forums and weblogging. Again, a similar pattern: moderated places disappeared, unmoderated ones drowned. But, this time the scale was larger, with different dynamics. Self-publishing was getting easier: at first there were web rings, later a "blogosphere" emerged. Forums were more varied. But again, some relics were left over. And from there, social media platforms started emerging, aggregating the space. And each one in turn has followed this life cycle, on slower or faster timelines. Moderating keeps the signal alive, but after a few years people tire of the status quo and look for the next thing, somewhere with fewer gatekeepers, perhaps, or more structure, or stronger or weaker identity models. The signal decays in amplitude rather than drowning in noise. What tends to be overlooked in social networking and virtual worlds discussions is that there's always a quality-of-service tradeoff dynamic to them. If you engineer the rules to improve one area you may lose some others. This is true even when looking at search systems: they retrieve correlated info that often has an intentionally wrong meaning.