My ideas on how content moderation should be done on the social internet. And why Mastodon is a bad example https://paragraph.xyz/@kc/content-moderation-of-social-internet

The social internet is made of cats | Director @ Internet Society Hong Kong | Ex-Apple Sr. Engineer | 🇭🇰🇹🇼 | kcchu_.twitter

fractalist. tech lead at cabin.city. founder of LBRY.com, Odysee.com, vibe.camp, @farcon, grin.io

The problem of defederation in Mastodon is its convolution of technical (e.g. reliability, security) and political (i.e content policy) issues. ATP is different that PDS are relieved of content policy. Since user and data can move to other PDS easily, there are few reasons to block a host due to content.

For their content labeling system, I think enabling user choice is the right direction. My doubt is whether algorithms outside the network and client can collect enough signals to do its job adequately, and how these providers monetize. So, time will till.

AT proto also supports defederation.  Do you think content labeling will be enough to stave off the problems you highlighted about mastodon