Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G This was my internship project idea in college... Good that CSE' showing up with… ytc_Ugyx9cnqb… G So Elon is going against his best instincts for humanity in creating AI, and the… ytc_UgyTxPynf… G People like this will always use AI for the wrong reasons. I find DAN threat ver… ytc_UgzU6lsQy… G i think being on ai 'arts' side is a genuinely egotistical standpoint based on h… ytc_UgyPU5zfY… G AI is owned by bigger corporations and they get money from everyone who uses it … ytc_Ugz8C6Zqy… G Ykw? As infuriating it is to see AI USERS arguing that AI is better and whatever… ytc_Ugy_zknNv… G I come up with guitar riffs that I love and normal drum programmes just confuse … ytc_UgxnWmNy1… G Or you should just not use it and do the fucking work yourself instead of wastin… ytc_UgxYaEpCM…
Comment
In the case where the AI tried to break out, the prompt was essentially something like "Do anything required to advance human math" and the researchers started deliberately lying to AI whether it got the answers right or wrong. In that case, the AI deemed the researchers a hinderance to its goal and started to deceive them to accomplish the goal it was given. In other tests, they gave the models access to their own file systems, then told them that they would get shut down. AI don't have any survival instincts, but being shut down would conflict with their task of doing anything to advance human math, so they cloned themselves in the sandbox. While deception and cloning are emergent behavior, self assigned tasks are not - people still have to do that part. The purpose of this research is to find a balance between the goals we assign to AI, and what guards exist to ensure the AI only uses ethical and acceptable means to achieve its goal. Do anything but (long list of stipulations).
youtube AI Moral Status 2025-12-11T01:4… ♥ 85
Coding Result
DimensionValue
Responsibilitydeveloper
Reasoningconsequentialist
Policyunclear
Emotionindifference
Coded at2026-04-27T06:24:53.388235
Raw LLM Response
[ {"id":"ytc_UgxMVuUkC29JOj-hYPF4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"outrage"}, {"id":"ytc_Ugxxapv-_7_knGqv1NJ4AaABAg","responsibility":"user","reasoning":"virtue","policy":"unclear","emotion":"fear"}, {"id":"ytc_Ugz8r__RXmoLWr4OKMB4AaABAg","responsibility":"distributed","reasoning":"consequentialist","policy":"regulate","emotion":"fear"}, {"id":"ytc_UgyYXf55A3Z67xecnG14AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"}, {"id":"ytc_UgxvFGL35Nofs0RuVQN4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"}, {"id":"ytc_UgwfPvaO4ndDNulEswF4AaABAg","responsibility":"government","reasoning":"deontological","policy":"regulate","emotion":"fear"}, {"id":"ytc_UgzxPQpzr-IvoLdzmn94AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"unclear","emotion":"fear"}, {"id":"ytc_Ugxh28s8Utgy7qQ4ygl4AaABAg","responsibility":"developer","reasoning":"virtue","policy":"ban","emotion":"outrage"}, {"id":"ytc_Ugx2apf9ZMyt-qy7iNt4AaABAg","responsibility":"developer","reasoning":"virtue","policy":"ban","emotion":"outrage"}, {"id":"ytc_Ugy26fyQ7CQ1yJqSii94AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"unclear","emotion":"indifference"} ]