Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
Claude doesn’t hallucinate nearly as aggressively as GPT. It got to where GPT confidently made stuff up and I couldn’t trust it anymore, about nearly anything. ClaudeCode is LIGHT YEARS better too. I’ve only kept my GPT Pro account because I got lazy and didn’t want to download all my old keeper contexts. I initiated the data download this afternoon and either they’re getting slammed, or they’re slow walking it. Shutting my account down as soon as it arrives.
reddit AI Responsibility 1772330952.0 ♥ 2
Coding Result
DimensionValue
Responsibilitynone
Reasoningconsequentialist
Policynone
Emotionapproval
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_o7wxmyt","responsibility":"developer","reasoning":"virtue","policy":"none","emotion":"outrage"}, {"id":"rdc_o7wz6tj","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"rdc_o7zfldi","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"rdc_o7zgd82","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"outrage"}, {"id":"rdc_o80qvva","responsibility":"none","reasoning":"unclear","policy":"industry_self","emotion":"approval"} ]