Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
The flattery thing is also an architectural strategy to improve poor prompts. By injecting that flattery into the start of the response, it adds that into the context it’s using to generate the remainder of its response, so it nudges it toward the higher quality weights that lead to the higher quality predictions. It’s annoying and harmful to humans, but beneficial to the output of the LLM.
youtube AI Moral Status 2025-11-19T08:3…
Coding Result
DimensionValue
Responsibilitydeveloper
Reasoningconsequentialist
Policyliability
Emotionmixed
Coded at2026-04-26T23:09:12.988011
Raw LLM Response
[ {"id":"ytc_UgySxcEJMp9l9ZaI1Ll4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"}, {"id":"ytc_UgwQOmZkw3e_hySJSw14AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"indifference"}, {"id":"ytc_UgwJ7jRLbOJxw9hyYRp4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"fear"}, {"id":"ytc_UgxtoVq2yGiQSfGK04F4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"approval"}, {"id":"ytc_UgwXuqLoTqnGtRtzFop4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"}, {"id":"ytc_Ugw5jMnF_70HzNoBM9d4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"liability","emotion":"mixed"}, {"id":"ytc_UgyuJoX-M97cR0PQ_dt4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"}, {"id":"ytc_Ugw1dUx-NEI99326TQJ4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"indifference"}, {"id":"ytc_UgxJ3oVfL6G5OGrKval4AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"none","emotion":"mixed"}, {"id":"ytc_UgxJQREHFjl7sTHzf3l4AaABAg","responsibility":"company","reasoning":"virtue","policy":"regulate","emotion":"outrage"} ]