Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
GPT-5 is not much better than GPT-4, but that's not the problem. The problem is that GPT-5 is no longer blowing smoke up its users asses. It used to be ridiculously and obnoxiously affirming. I actually *hated* that part about it, as it felt like it was spending more tokens buttering me up than it was giving me the data or analysis I wanted. But some people responded very positively to it to the point of actual psychosis. That was actually unintended and undesired (I think). So they had to dial the glazing down. And when it doesn't glaze them, they don't trust it as much.
reddit AI Jobs 1754759511.0 ♥ 15
Coding Result
DimensionValue
Responsibilitynone
Reasoningmixed
Policynone
Emotionindifference
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_n7sz371","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"indifference"}, {"id":"rdc_n7s06n6","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"rdc_n7taoz4","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"disapproval"}, {"id":"rdc_n7tn60n","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"rdc_n7srops","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"outrage"} ]