Oddbean new post about | logout
▲ ▼
dtonon | 1 months ago (raw) | root | parent | reply | flag +4 
 The whole concept of "verification" is bound to centralized structures, it should be buried. Self-verification is a state of mind. Instead, verifying someone/something else is a matter of active experience and interactions, which minimize the trust factor. Of course technology helps, but also WoT needs to be verified and not blindly trusted. "Proof of work" means taking responsibility for everything we care about, exactly like we do (should do) in real life.
▲ ▼
bosco | 1 months ago (raw) | root | parent | reply | flag 
 There is no way I can fathom doing it without trusted parties, sure. I guess I haven't truly embreced the philosophy already but hypothetically, community bound verifications, though trust reliant could be put in place if desired
▲ ▼
dtonon | 1 months ago (raw) | root | parent | reply | flag +1 
 I understand, it is not easy. The good news is that these trusted parties can be collaborative and self-monitored: peer reviewed algo within peer reviewed platforms. But we must always keep an open eye on our verification attitude.
▲ ▼
Aspie96 | 1 months ago (raw) | root | parent | reply | flag +1 
 > The whole concept of "verification" is bound to centralized structures, it should be buried.

It doesn't have to be. There can be multiple verifiers and a single pubkey can be verified by many of them. If among the verifiers that verify me there is at least one which is included in your trusted list, your client should show me as verified.
This would be a level of decentralization similar to that of relays (in the inbox model).

The issue with NIP-05, if it were a verification system (it's not), would be that an account can only be "verified" by one "verifier". So an account would have to pick just one and you'd be forced to "trust" any widely used one (if you want to rely on verification at all), leading to centralization.