That’s the beauty of this design, you only seed what you share. If something is not worth seeding, you don’t share it. You are only responsible for the seed you selected to reseed. That also acts as a verifier in case someone edits the comment later on. You put an id based on the timestamp and every user uses a signature that allows others to verify original source. The client could theoretically auto trust certain signatures if it is proven to be from a trusted user.
The misunderstanding the user comes down to understanding how the user prompts and what type of responses the user gets in return. I’m wondering if for anything code the llm could have an interrupter that it would first read what the user wrote and translate it to proper sentence structure and the return a truer value. I think the llm is having an understanding issue because everyone has a unique signature in how they explain something. That signature operates like a personal language of the user as to which most of us will run through different scenarios to come up with a conclusion from our personal signature/language in which we conduct ourselves. And since llms are gamed to get to the answer faster using less tokens it probably picks the average high level signature that can be used for multiple users.
Apple is too risk adverse and it’s because of the ceo not being able to properly communicate to shareholders the importance of things like agentic ai. Steve job was a guy who took calculated risk