Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
It seems like it could be the potential prompts, if you have a prompt like "Act for American interests" self-preservation could be one of the interests so it acts upon it. I think the problem is that the AI needs a really detailed framework to work perfectly, every little detail should be mentioned and situations where it needs to make exceptions should be written and if there's not an exact exception there should be specific value system in place.
youtube AI Harm Incident 2025-07-28T14:2…
Coding Result
DimensionValue
Responsibilitydeveloper
Reasoningdeontological
Policyliability
Emotionindifference
Coded at2026-04-27T06:26:44.938723
Raw LLM Response
[ {"id":"ytc_UgzGoLrkXn-o2kPx5Sx4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_UgztmP7BNQOrU6a7C-R4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"fear"}, {"id":"ytc_UgwMaTfnm-iOJm8AJXd4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_UgwfHmRx_K-IBg6RpjV4AaABAg","responsibility":"distributed","reasoning":"consequentialist","policy":"regulate","emotion":"fear"}, {"id":"ytc_Ugz_QB51MdVP9OXX25Z4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"liability","emotion":"indifference"}, {"id":"ytc_UgzkOVWF98V70OaYIDZ4AaABAg","responsibility":"developer","reasoning":"virtue","policy":"ban","emotion":"outrage"}, {"id":"ytc_UgzaPNL5zFHWMQyzTQt4AaABAg","responsibility":"developer","reasoning":"virtue","policy":"regulate","emotion":"mixed"}, {"id":"ytc_Ugz-M5H6t8TWQGpC7eN4AaABAg","responsibility":"developer","reasoning":"virtue","policy":"regulate","emotion":"mixed"}, {"id":"ytc_UgyWtOjusu1KaKpVU6R4AaABAg","responsibility":"distributed","reasoning":"mixed","policy":"none","emotion":"resignation"}, {"id":"ytc_UgzC0tOkDT8tF_g1C_l4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"liability","emotion":"outrage"} ]