Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
Even in writing boilerplate unit tests, which is one of AI's strengths, I've found you have to be very careful, because AI is really good at writing proper-seeming tests that don't actually test the thing you want them to. It's really easy to miss something like that when it's buried in a thousand lines of code that were all written at once, and it feels like the extra scrutiny you have to put into the code review largely cancels out the time savings.
reddit AI Jobs 1754617986.0 ♥ 15
Coding Result
DimensionValue
Responsibilityuser
Reasoningconsequentialist
Policynone
Emotionfear
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_n7i33az","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"rdc_n7j0cdt","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"fear"}, {"id":"rdc_n7hkrj2","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"mixed"}, {"id":"rdc_n7j0xqq","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"rdc_n7h9uuu","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"} ]