You can read Rabble's & my approach to content moderation here…
https://s3x.social/nostr-content-moderation
We have a pull request proposing "NIP-69" as a first step in that process, and have a more general "issue" open on Github discussing the topic more generally. (The links for both are in the link above).
The basic vision is "bottom up" content moderation where communities can moderate what people in their community sees via interconnecting webs of trust relationships. The end user will pick the people/organizations/bots they trust to have a voice in moderating their feed and only reports by those people will be factored in when the user's feed is filtered by a client app.
So for example, I could mark "exosome" as someone to block and he will then go on the block list of everyone who has designated me as trusted moderator of their feed. And he could go on the block list for relays that trust me as a moderator.
These discussions are good, I just thought of how the whole model can be extended…
Everyone is talking about how the client apps will eventually support algorithms that filter users' feeds - pushing some content up in the feed and pushing other content down in the feed. This model can factor into that… So the same events that are used to block content can be used to promote content if they had the ability to endorse as well as suggesting blocks and warnings.
For example, if a fellow Nazi really liked "exosome" he could endorse him and people who trust him would see more of exosome's posts. And to extend things further a user could have reverse trusts lists that basically say "do the opposite of whatever this person says to do" . The possibilities are actually rather interesting.
s3x_JAY / s3x_jay
npub1ve…738wt
2023-04-28 11:45:56
in reply to nevent1q…ul9y
Author Public Key
npub1veeqt66jt2j209y9nv4a90w3dej074lavqygq5q0c57dharkgw0qk738wtPublished at
2023-04-28 11:45:56Event JSON
{
"id": "d3ac92210c484eac8d81d8bf4bd161987046dd5703fa3dfe6a0f4c8e9b3321e7",
"pubkey": "667205eb525aa4a794859b2bd2bdd16e64ff57fd600880500fc53cdbf476439e",
"created_at": 1682682356,
"kind": 1,
"tags": [
[
"e",
"837795e1a4b592552372d42e283cb191c95e1daedd860e25a02cdf4c1f38a23b",
"",
"root"
],
[
"e",
"140672a7dedcfe8dd75b21db5ef36861cdba6b9969ba0fb709991e858ff73b94",
"",
"reply"
],
[
"p",
"667205eb525aa4a794859b2bd2bdd16e64ff57fd600880500fc53cdbf476439e"
],
[
"p",
"76c71aae3a491f1d9eec47cba17e229cda4113a0bbb6e6ae1776d7643e29cafa"
],
[
"p",
"7a433655129ff5b17084aec27ffbdf1f28c7469c0002347f783c42161c4f44a7"
]
],
"content": "You can read Rabble's \u0026 my approach to content moderation here…\n\nhttps://s3x.social/nostr-content-moderation\n\nWe have a pull request proposing \"NIP-69\" as a first step in that process, and have a more general \"issue\" open on Github discussing the topic more generally. (The links for both are in the link above). \n\nThe basic vision is \"bottom up\" content moderation where communities can moderate what people in their community sees via interconnecting webs of trust relationships. The end user will pick the people/organizations/bots they trust to have a voice in moderating their feed and only reports by those people will be factored in when the user's feed is filtered by a client app. \n\nSo for example, I could mark \"exosome\" as someone to block and he will then go on the block list of everyone who has designated me as trusted moderator of their feed. And he could go on the block list for relays that trust me as a moderator. \n\nThese discussions are good, I just thought of how the whole model can be extended…\n\nEveryone is talking about how the client apps will eventually support algorithms that filter users' feeds - pushing some content up in the feed and pushing other content down in the feed. This model can factor into that… So the same events that are used to block content can be used to promote content if they had the ability to endorse as well as suggesting blocks and warnings. \n\nFor example, if a fellow Nazi really liked \"exosome\" he could endorse him and people who trust him would see more of exosome's posts. And to extend things further a user could have reverse trusts lists that basically say \"do the opposite of whatever this person says to do\" . The possibilities are actually rather interesting. ",
"sig": "087f8017e1f8291bffc219bf85d31542fb4c8de11eb354a6cc7a36033dcf0088ea6ef1fc3cc34200eaa7e4905f1b582efa27869cc3c2fba6637ebf54dfba0d7c"
}