Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
My big solution to any AI too powerful dilemma is just to use it. Try developing a software with those supposedly frontier models. Even, if you pick the best of the best, according to the most strict benchmarks possible, that "PhD" level intelligence on paper, fails very spectacurarily on medium difficulty problems. Once you stray away from anything that is in the learning set, they behave like a drunk idiot.
reddit AI Moral Status 1765372389.0 ♥ 1
Coding Result
DimensionValue
Responsibilitydeveloper
Reasoningconsequentialist
Policyindustry_self
Emotionindifference
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_nta66jp","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"}, {"id":"rdc_nta7iql","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"}, {"id":"rdc_nta8jb4","responsibility":"developer","reasoning":"consequentialist","policy":"industry_self","emotion":"indifference"}, {"id":"rdc_ntaf3ph","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"resignation"}, {"id":"rdc_ntax468","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"} ]