Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
I find AI allows me to work more at "architectural" scale - I feel I can refactor easily - as long as I am clear with my requirements, keep the updates limited, have testability for things. But that's with keeping a close eye over what it does. I've tried (as an experiment) "vibe coding" and my God the amount of garbage I got. I had to go through and re-write most of it by hand I hear people just giving AI unfettered access to their code bases... and I will not be doing that. At least not with current gens I find it best for refactoring, searching code bases, translating code from one paradigm to another (say I have some code that works in a NextJS code base, and want to convert to a python service or something)
reddit AI Jobs 1754603536.0 ♥ 1
Coding Result
DimensionValue
Responsibilityuser
Reasoningconsequentialist
Policynone
Emotionmixed
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_n7hu5cg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"rdc_n7hug3o","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"rdc_n7hw47o","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"outrage"}, {"id":"rdc_n7hziye","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"rdc_n7i523z","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"mixed"} ]