The way to avoid having a shared blocklist "get polluted by the politics of its maintainers" is for it to have transparent policies, including how to get on and how to get off. The reason people or instances are listed needs to be clear as well. There needs to be a clear appeals process.
This all seems obvious to me. It also seems obvious to me that the reason it seems non-obvious to others is that they don't actually care much about solving the problem. They'd rather just throw up their hands and talk about how complicated and hard it is.
It's not complicated and hard. You just have to stop comparing real solutions to these non-existent "perfect" solutions that exist only in your head, like "heuristics". I have done this for a living, from being the abuse administrator for a nationwide ISP to Second Life to Facebook. It's waaaay easier to moderate when you are just tweaking heuristics and shadow-banning people in ways they can barely even detect, can't prove, and can't appeal. And you'll probably get away with it if you have a billion users. But it can never work here. And honestly it doesn't actually "work" for Facebook, because it's unethical.
Real moderation doesn't scale very well. It requires a moderator per 100 users or something around that. We have that. We just need to accept that the Fediverse runs on trust and that that's a good thing, not a bad thing.
Charles U. Farley /
npub1a3…f07t5
2024-08-01 16:05:41
in reply to nevent1q…t39z
Author Public Key
npub1a36n5u2z95sv8jnmdmeusrf7t9nrvlweferv0xf2at0fpmaflggqef07t5Published at
2024-08-01 16:05:41Event JSON
{
"id": "b625770130eb8435c5d9d31b5edc31cd96147134ea36a24de7af6abee4d69e11",
"pubkey": "ec753a71422d20c3ca7b6ef3c80d3e5966367dd94e46c7992aeade90efa9fa10",
"created_at": 1722528341,
"kind": 1,
"tags": [
[
"p",
"ec753a71422d20c3ca7b6ef3c80d3e5966367dd94e46c7992aeade90efa9fa10"
],
[
"proxy",
"https://retro.social/@freakazoid/112887617403843788",
"web"
],
[
"e",
"d627c1ea47bbd8b16592bf0ca2a2e14c0de8e53d63826162673baa2a5805e69d",
"",
"root",
"ec753a71422d20c3ca7b6ef3c80d3e5966367dd94e46c7992aeade90efa9fa10"
],
[
"proxy",
"https://retro.social/users/freakazoid/statuses/112887617403843788",
"activitypub"
],
[
"L",
"pink.momostr"
],
[
"l",
"pink.momostr.activitypub:https://retro.social/users/freakazoid/statuses/112887617403843788",
"pink.momostr"
],
[
"-"
]
],
"content": "The way to avoid having a shared blocklist \"get polluted by the politics of its maintainers\" is for it to have transparent policies, including how to get on and how to get off. The reason people or instances are listed needs to be clear as well. There needs to be a clear appeals process.\n\nThis all seems obvious to me. It also seems obvious to me that the reason it seems non-obvious to others is that they don't actually care much about solving the problem. They'd rather just throw up their hands and talk about how complicated and hard it is.\n\nIt's not complicated and hard. You just have to stop comparing real solutions to these non-existent \"perfect\" solutions that exist only in your head, like \"heuristics\". I have done this for a living, from being the abuse administrator for a nationwide ISP to Second Life to Facebook. It's waaaay easier to moderate when you are just tweaking heuristics and shadow-banning people in ways they can barely even detect, can't prove, and can't appeal. And you'll probably get away with it if you have a billion users. But it can never work here. And honestly it doesn't actually \"work\" for Facebook, because it's unethical.\n\nReal moderation doesn't scale very well. It requires a moderator per 100 users or something around that. We have that. We just need to accept that the Fediverse runs on trust and that that's a good thing, not a bad thing.",
"sig": "03871da2dc27f8ff83e314cdf7104a6d2cbc941235d0358db183955e5822dd3ce98a37dd9b570670d8b8b370db6217c779ea29d58d1206e222234eb65f7a48bd"
}