Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
AI is a bit of a fuzzy term to begin with, but they're all ultimately programs. The one you're talking about seems to just be a function maximizer tasked with writing a "better" function maximizer. Humans have to define how "better" is measured - probably candidate solutions will be given test problems and evaluated on how quickly they solve them. And in this case, the objective/metric doesn't change between iterations. If it did, you'd most likely get random, useless programs.
reddit AI Bias 1438067496.0 ♥ 3
Coding Result
DimensionValue
Responsibilitynone
Reasoningunclear
Policynone
Emotionindifference
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_ctir09r","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"rdc_cthv4ah","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"rdc_cthy920","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"rdc_cthwc5r","responsibility":"user","reasoning":"deontological","policy":"unclear","emotion":"mixed"}, {"id":"rdc_cthwpop","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"} ]