Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
I have AI's do guided research tasks and math tasks for me pretty regularly and then I review the results, sometimes they make mistakes but less often than people do at that kind of task in my experience. I would consider that close to AGI. I think a lot of models are sycophantic and overly RLHF trained but I don't doubt that modern transformers have some capacity for general intelligence with the right data and architecture.
reddit AI Moral Status 1752780418.0 ♥ 5
Coding Result
DimensionValue
Responsibilitynone
Reasoningmixed
Policynone
Emotionapproval
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[{"id":"rdc_n3pzcao","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"indifference"},{"id":"rdc_n3osekd","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},{"id":"rdc_n3oqer5","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"},{"id":"rdc_n3os95o","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"mixed"},{"id":"rdc_n3or9gj","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"outrage"}]