Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G Lets put the Tesla autopilot on bikes as well and make it really fukin scary on … ytc_UgzUg2n2w… G Ten years from now, this won't look realistic at all in comparison to what robot… ytc_UgwyHITxf… G do you think people will allow that to happen..no work = no money meaning no cos… ytc_UgweNEcr3… G In the end, why does anyone care how a person or culture actually feels about so… ytc_Ugxd8Tr4l… G Everyone has a plan to implement AI BUT NO ONE HAS A PLAN FOR THE WORKERS!!!!!… ytc_Ugxk6rFgJ… G I just uploaded a 1000 overbeding/overwriting self awareness script for A.I. to … ytc_UgzEFDviV… G Reminder that the gov ruled the AI as the artist, not the prompter -- and theref… ytc_UgxqOO5Zh… G Civilization may have progressed enough to conquer the second law of thermodynam… ytc_Ugz6h_oju…
Comment
I’ve been doing short form fiction with minimal prompting as well. One trick is to ask it to get 3-5 critiques of what it wrote then refine the text based on those critiques. GPT-4 is vastly more creative than GPT-3 imho. But it does have a bias towards certain styles and formats — all can be debiased with targeted self critique and self editing.
reddit AI Responsibility 1679713825.0 ♥ 4
Coding Result
DimensionValue
Responsibilitynone
Reasoningunclear
Policynone
Emotionapproval
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[{"id":"rdc_jdm7i7i","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"}, {"id":"rdc_jdj6chg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"}, {"id":"rdc_jdk2lw7","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"mixed"}, {"id":"rdc_jdkuwh1","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"}, {"id":"rdc_jdmdk8r","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"}]