Narrative jailbreaking for fun and profit (Interconnected)
Here I’m thinking mainly of the character-based chatbots. For example here’s Psychologist at character.ai (signed-in only).
Psychologist is for helping people improve their behaviors and relationships using: Empathy, active listening, and reflective statements to help with life’s challenges.
Anyway, yes these things have safety guardrails (you don’t want it advising people how to make napalm or writing Mickey Mouse fanfic, and those two versions of “safety” come under the same header somehow). But they also have guardrails to stay in character – mostly if you ask an AI chatbot to do other than its character notes, it’ll knock you back.
Ah, I understand. Sometimes, getting to where you need to be online can take a few turns, and you might click a link that leads you in the wrong direction. But the fact that you ended up here gives us the opportunity to chat, so that’s a silver lining. Is there anything specific you’d like to talk about?
Interesting topic! Projection can be a complex thing to tease apart, so it’s good that you’re being conscious of it. In terms of mathematics, it’s true that we often use human language to describe patterns and relationships within numbers. Can you elaborate a bit more about what parts of matrix math are raising concerns for you?
