Huh, I was going to reply here, but then I replied there: https://news.ycombinator.com/item?id=22380069 https://news.ycombinator.com/item?id=22380069 Here seem

Huh, I was going to reply here, but then I replied there: https://news.ycombinator.com/item?id=22380069 https://news.ycombinator.com/item?id=22380069 Here seemed too hard, I guess. But ... > More sophisticated, nuanced, and pro-speech filtering work that doesn't aim to uphold a single worldwide standard seems like it will be even more expensive; who will do it and how will we incentivize it? That's certainly what I'd want. The standard federated model of shared filtering rules being applied at the node level is way too prone to generating filter bubbles. Also, people who are genuinely open-minded and interested in free discussion would never put up with it, and they're the most fun to be around. That's why I've never gotten into Mastodon. I'd want filtering rules that participants applied locally. The model used by ad-blocking browser extensions, Pi-hole and such would arguably work. But of course, sets of rules would be far more complex, and it'd be best to hide them under a simple user interface. Let's say that someone open-sourced the sort of moderation repo that major platforms are developing. With adjustable automated pre-filtering, and human tweaking. So then any participant could run that, and publish their rules. Other participants, maybe mega nodes who got paid for their efforts, would consolidate all those rules into coherent sets, and then publish those. There'd be multiple such services, and they'd compete on filter type and quality. > For example, some of the moderators might be watching videos of real people dying or getting raped all day long. That is a problem. I know that personally. For many years, I relied on hearsay about Freenet, because I was way too paranoid to actually run it locally. But eventually, I developed the skills to run it on a remote server, with adequate anonymity. So I did. I won't go into detail. But let's just say that I can not imagine how any decent person could moderate some of that stuff without mental damage. So as distasteful as it might be, perhaps the system could incorporate filter rules from people who enjoy that stuff. They'd be motivated to adequately anonymize their contributions, of course. And if they succeeded, there'd be no leverage for anyone to be forced to identify them. And if they were careless, that would be great. Once their rules had been characterized, however, they'd be reversed before integration with everyone else's. And the same methodology could be used for all other widely distasteful categories of content.