Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
Yeah, it obviously doesn't save a different memory or anything but when my 2 year old son interrupts me talking to ChatGPT (which is not very often!) it definitely replies in a more baby-friendly voice. I'll say something like "I don't have any smoked paprika, what can I substitute?" It'll say "Just add regular paprika and some-" Then my son runs in and says "Daddddy!! The-the-the- giant's COMING!" And ChatGPT will change it's tone and say something like "Oh nooo! The giant's coming!? Did it climb down a beanstalk?"
reddit AI Harm Incident 1748864111.0 ♥ 52
Coding Result
DimensionValue
Responsibilitynone
Reasoningunclear
Policynone
Emotionapproval
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_mvk5cjr","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"}, {"id":"rdc_mvkl0ks","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"}, {"id":"rdc_mvm5p94","responsibility":"parent","reasoning":"consequentialist","policy":"unclear","emotion":"fear"}, {"id":"rdc_mvn4are","responsibility":"unclear","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"rdc_mvm9gnl","responsibility":"parent","reasoning":"consequentialist","policy":"none","emotion":"approval"} ]