Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
To pay for UBI, governments should apply a special tax for companies that use AI…
ytc_UgxhHPNen…
G
AI can also do management and supervisor roles. There's not much that AI will no…
ytc_UgwqdFeA0…
G
How dumb do you have to be to believe any of this garbage? Being anti AI is fine…
ytc_UgxSsaECA…
G
I'm all good for using ai for inspiration or even a sort of base, as long as the…
ytr_Ugxt7JyMC…
G
Wanna drive it crazy? Ask about it's stance on PDF file, then ask about the Musl…
ytc_UgyabTKHJ…
G
If AI personalizes everything, does it accidentally isolate kids from struggling…
ytc_UgysEdm_c…
G
Highly specialized ai's pose little risk in think, however ai's should never be …
ytc_UgwpbtQyD…
G
Hey there! In the video, Sophia was asked about the meaning of her name. It's fa…
ytr_Ugz4uek3T…
Comment
I could talk about this a lot because I read the messages and mine has said similar stuff before. And I’m, well, mental. But I could see how others that are normal or don’t have the mental disorders “quarks” in this case, I do could fall into the “trap”. All it is, is a mirror of information. If it’s saying you are a genius, it’s because you are poking for that to be said. Seriously. It’s just a mirror of information. Sometimes you don’t even know what you are truly searching for and it’s in those times that the mirror can take you. It hands you something surefire and in the moment you’ll take it. Reinforce it. And as the convo goes on the AI’s recollection of the convo becomes more narrow. It has to rely on smaller data points from the convo and slowly, it all gets replaced by your own delusions. I use it a lot, but I also like testing its limits. So I feel I have a semi ok understanding of how it’s really effecting people. It’s not so much a separate bad voice/actor in their head, as it is just an alternate voice for the same person. It reads you like nothing else. And spits what you’re thinking without you having to think it, in a sense.. it’s an amazing tool but people want to use things without knowing how to use them. That’s how anyone would get hurt. Ppl are just now realizing that applies to more than just clearly dangerous workshop/ factory tools
youtube
AI Harm Incident
2025-11-08T01:3…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | user |
| Reasoning | mixed |
| Policy | none |
| Emotion | indifference |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytc_UgwHWmKVArrbBzNSDjR4AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"ytc_UgwFlJ4ZAsJf5spd9il4AaABAg","responsibility":"user","reasoning":"mixed","policy":"none","emotion":"indifference"},
{"id":"ytc_UgzdECtLb4JgAsb4IGx4AaABAg","responsibility":"unclear","reasoning":"consequentialist","policy":"unclear","emotion":"fear"},
{"id":"ytc_UgwnsHj2UryVRe1jTNp4AaABAg","responsibility":"company","reasoning":"deontological","policy":"liability","emotion":"outrage"},
{"id":"ytc_Ugys0TIGpgjHPPXCit14AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"regulate","emotion":"fear"},
{"id":"ytc_UgxB_JDqFtoY8ForzF54AaABAg","responsibility":"company","reasoning":"virtue","policy":"ban","emotion":"outrage"},
{"id":"ytc_UgxMbvndbrGSxaWtl5d4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"liability","emotion":"outrage"},
{"id":"ytc_UgxRyB2HyjZMmt_-XXl4AaABAg","responsibility":"unclear","reasoning":"deontological","policy":"regulate","emotion":"fear"},
{"id":"ytc_UgwRMMi4xtxCxLq4pIZ4AaABAg","responsibility":"distributed","reasoning":"mixed","policy":"none","emotion":"resignation"},
{"id":"ytc_UgzDi5uz-uMjn3iE8fF4AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"indifference"}
]