Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
Definitely, dude was using discord bots to make this and quality control the illustrations. That's the wrong way of doing it because the artistic style will be all over the place. The real way would be to train a model of the artistic style you want and then quality control your prompts with that. Unfortunately Midjourney is not open source so I dont think you can do that.
reddit AI Harm Incident 1671071122.0 ♥ 39
Coding Result
DimensionValue
Responsibilityuser
Reasoningmixed
Policynone
Emotionindifference
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_j0b5ipo","responsibility":"developer","reasoning":"mixed","policy":"none","emotion":"indifference"}, {"id":"rdc_j0cxmjk","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"outrage"}, {"id":"rdc_j09lin2","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"}, {"id":"rdc_j09tam1","responsibility":"user","reasoning":"mixed","policy":"none","emotion":"indifference"}, {"id":"rdc_j0aen8g","responsibility":"company","reasoning":"consequentialist","policy":"liability","emotion":"fear"} ]