2 ms·

There are schemes where you can prove authenticity without revealing identity.

• amelius 6y ago
  That's not the point. What matters here is if you can still link messages to the same person, even if you don't know who that person is.
  • cosmojg 6y ago
    Right, again, get rid of public keys with zero-knowledge proofs, and you've solved this problem. The only way to screw up is by publicly disclosing which messages are yours. Although, I guess a large part of the problem is that some users might accidentally do this implicitly, and many may do it just by writing in some unique way. Any thoughts on a solution?
    • DennisP 6y ago
      Train your own GPT-3 on some other users' writings, and post the output from that instead of writing it yourself. You could have your public persona that you write yourself in your own unique style, but use GPT-3 for things you don't want attributed to you.