RebelOfBabylon on Nostr: Interesting talk, thanks for sharing. Some thoughts: - I think web of trust solves a ...
Interesting talk, thanks for sharing. Some thoughts:
- I think web of trust solves a lot of the problems of spam and trustworthiness on the internet but not in the way web of trust is implemented today. I don't like that the current implementation of web of trust is based on follows. A follow is not necessarily an endorsement of any kind. Also, I may follow you but not necessarily trust the people you follow by default.
- I think a proper web of trust should be built on vouching. A new, explicit action separate from following or friending that could be based on cryptographic signatues. In nostr, this could be a new event kind where you include a pubkey (the pubkey that you want to vouch for) in the p tag field. This message would essentially be a formal declaration from you that you vouch for this person aka you trust them.
- now all of a sudden we have a verifiable way to prove you vouched for someone and this unlocks many possibilities. You can filter content based on degrees of separation. You trust the people you vouch for so their notes appear in your feed, and you trust whoever they vouch for but maybe no further (call that 2 degrees of separation). You can also as an admin ban based on degree of separation. Or refuse pubkeys that have not been vouched for. Also, if someone does join your community and they have been vouched for and they're exhibiting bad behaviour, you can see who vouched for them and you can potentially take action accordingly. Maybe on your relay, everyone gets three strikes and their out and a strike could be someone you vouched for was banned.
- this kind of web of trust I think is way more practical in many situations as well. Like for example there is currently a crisis in science where scientific papers are being published in increasing numbers to such a degree that journals cannot keep up. Some fraudulent papers are slipping through the cracks. A potential solution would be to create a scientific journal on nostr where different scientists are allowed to publish papers based on this model of web of trust where a vouch would hold a lot of weight and responsibility. And this could effectively decrease the fraud in spam in your journal.
- because a vouch is just a nostr event, its easy to find them. They could also be replaceable events where you could disown/condemn instead of vouch so that a previously good actor turned bad could be removed from your web of trust.
- I think web of trust solves a lot of the problems of spam and trustworthiness on the internet but not in the way web of trust is implemented today. I don't like that the current implementation of web of trust is based on follows. A follow is not necessarily an endorsement of any kind. Also, I may follow you but not necessarily trust the people you follow by default.
- I think a proper web of trust should be built on vouching. A new, explicit action separate from following or friending that could be based on cryptographic signatues. In nostr, this could be a new event kind where you include a pubkey (the pubkey that you want to vouch for) in the p tag field. This message would essentially be a formal declaration from you that you vouch for this person aka you trust them.
- now all of a sudden we have a verifiable way to prove you vouched for someone and this unlocks many possibilities. You can filter content based on degrees of separation. You trust the people you vouch for so their notes appear in your feed, and you trust whoever they vouch for but maybe no further (call that 2 degrees of separation). You can also as an admin ban based on degree of separation. Or refuse pubkeys that have not been vouched for. Also, if someone does join your community and they have been vouched for and they're exhibiting bad behaviour, you can see who vouched for them and you can potentially take action accordingly. Maybe on your relay, everyone gets three strikes and their out and a strike could be someone you vouched for was banned.
- this kind of web of trust I think is way more practical in many situations as well. Like for example there is currently a crisis in science where scientific papers are being published in increasing numbers to such a degree that journals cannot keep up. Some fraudulent papers are slipping through the cracks. A potential solution would be to create a scientific journal on nostr where different scientists are allowed to publish papers based on this model of web of trust where a vouch would hold a lot of weight and responsibility. And this could effectively decrease the fraud in spam in your journal.
- because a vouch is just a nostr event, its easy to find them. They could also be replaceable events where you could disown/condemn instead of vouch so that a previously good actor turned bad could be removed from your web of trust.