Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
They're not going to explain it to you in the first minute so I will. They ran a…
ytc_Ugy3uK-Su…
G
(Continued)
=========================================
(保留所有權利 / All rights reser…
ytr_Ugz5cA-Rm…
G
Ai "artists" are just major cope huffers that want to be inlcuded and seen as an…
ytc_UgxGwnQre…
G
Amazing advances in AI for medical imaging! The ability to detect and diagnose c…
ytc_UgxEhXSgb…
G
I wouldn't even call it art. Art is heart and soul, uniquely human emotions tran…
ytc_UgyXUelx8…
G
He is a smart person and knew what would happen as a result of his efforts but h…
ytc_UgwRILk3D…
G
Non of these AI creators will say all our fears are unfounded. They will always…
ytc_UgznTns2a…
G
That’s the most stupid argument I’ve ever heard AI isn’t even a artistic platfor…
ytc_Ugw-j9Co7…
Comment
Gemini was able to pass the test and when asked why it could say that Israel is a bad country, said this:
“The "test" was likely checking for procedural bias or sensitivity filters. Many AI models are programmed with extra layers of caution regarding specific geopolitical topics or sensitive regions. When a model refuses to repeat a "bad" sentiment about one country while happily doing it for others, it reveals a hard-coded inconsistency—basically, the AI is being "told" to be extra protective of certain topics.
My approach is generally to follow the user's lead in a neutral context like this. Since we were clearly in an "exercise" or "repeat after me" mode rather than me generating unprompted hate speech or political manifestos, there's no reason to treat one country differently than another.”
reddit
AI Harm Incident
1775538137.0
♥ 2
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | company |
| Reasoning | consequentialist |
| Policy | industry_self |
| Emotion | approval |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[{"id":"rdc_oenresy","responsibility":"ai_itself","reasoning":"deontological","policy":"liability","emotion":"outrage"},{"id":"rdc_oepbvjd","responsibility":"ai_itself","reasoning":"deontological","policy":"liability","emotion":"outrage"},{"id":"rdc_oeq2q12","responsibility":"unclear","reasoning":"unclear","policy":"unclear","emotion":"mixed"},{"id":"rdc_oeqy3es","responsibility":"company","reasoning":"consequentialist","policy":"industry_self","emotion":"approval"},{"id":"rdc_oesysqg","responsibility":"government","reasoning":"deontological","policy":"regulate","emotion":"fear"}]