Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Companies severely overestimated AI. Today’s knowledge fundamentally binds them.…
ytc_UgxwbabK3…
G
The AI takes photos from the internet and compiles it all into one photo, so by …
ytc_UgzgLClNo…
G
nothing can stop it just imagine what china does. And ohhh the pour artist like …
ytc_UgyM3dQC0…
G
This is the video that the AI is going to show the other AI is to convince them …
ytc_Ugyry2Bdk…
G
Wow that robot does everything except for the 1 thing you would want it to do. S…
ytc_UgxTG0Nyl…
G
Heyyy something I'm a little bit knowledgeable about!
Eliezer is kind of a weir…
ytc_UgwAR-miK…
G
This is clearly a Marxist media firm. Great interview and insights by Karen.
F…
ytc_UgwDrWGaz…
G
@ColbyAzimuthNot surprising. It's like expecting a blind man to describe red t…
ytr_Ugzxbaz6M…
Comment
Exactly. ChatGPT will, if he ever hears the word „suicide“, provide you with emergency sites and numbers. It’s nearly impossible that he said that. Maybe the guy asked him to rate his plan in a fictional scenario or something.
reddit
AI Harm Incident
1756221853.0
♥ 125
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | user |
| Reasoning | consequentialist |
| Policy | none |
| Emotion | approval |
| Coded at | 2026-04-25T08:13:13.233606 |
Raw LLM Response
[
{"id":"rdc_n8n5jww","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"},
{"id":"rdc_nas2d5i","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"rdc_nb83qbc","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"fear"},
{"id":"rdc_nbs2b4q","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"resignation"},
{"id":"rdc_nc22lrk","responsibility":"company","reasoning":"mixed","policy":"industry_self","emotion":"approval"}
]