Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
AI is not conscious yet. It's still just software running on silicon-based hardw…
ytc_UgzxUJ9HZ…
G
Another liar AI can be trusted to guardrail?
This tech is more dangerous than a…
ytc_UgzYAJDzR…
G
Hoje a tecnologia está tão avançada silicone perfeito como se fosse um rosto hum…
ytc_UgxTs1zzd…
G
From what I know about child development and how the human brain works and grows…
ytc_UgxackYgs…
G
AI absolutely and unavoidably will replace all software engineers. It already ca…
ytc_UgytYNN60…
G
I think there are a lot of problems here. Gen AI makes non-consensual deepfakes …
ytc_Ugzdy4D13…
G
9:47 What a jackass. This is horrifying. It just goes on and on, gets worse and…
ytc_UgxObLe_P…
G
So they no AI is going to take over eventually but they still proceed to advance…
ytc_UgxZg9Xhr…
Comment
I see a lot of “this is your job” comments, but don’t see any real explanation as to why this won’t be ideal to repeat in the future.
4 year old children are in a peak social development phase. They are learning how to interact with people - how to converse with them, how to listen and actually talk WITH someone, not AT someone. Keyword - learning. Obviously these are earlier stages, but these are also some of the most important.
Your child is not old enough to truly comprehend that the human voice talking to him is not like an actor on tv or someone on the phone, and not actually a person at all. Is this one off situation going to permanently alter your son’s social skills? Highly unlikely. But if you allow him to return to this - he’s going to become accustomed to having conversations that only he leads, that are only about his interests, and that don’t require him to have any actual social skill at all. Children are incredibly impressionable, and it takes MUCH less time for a child to become hyper fixated on something - such as realizing “when I talk to other people they don’t always like to talk about what I want to talk about, but when I talk to this guy, he ALWAYS talks about exactly what I want to” - and that idea being reinforced even a few times at his age can have a surprisingly profound impact longterm.
In the bigger picture of this situation though, I think the actual f up wasn’t letting your son talk to ChatGPT - it’s that this went on for two hours.
reddit
AI Harm Incident
1748841769.0
♥ 158
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | developer |
| Reasoning | virtue |
| Policy | unclear |
| Emotion | mixed |
| Coded at | 2026-04-25T08:13:13.233606 |
Raw LLM Response
[
{"id":"rdc_muq3jxx","responsibility":"unclear","reasoning":"consequentialist","policy":"regulate","emotion":"fear"},
{"id":"rdc_mva4xct","responsibility":"government","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"rdc_mva720b","responsibility":"government","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"rdc_mvjhdnu","responsibility":"developer","reasoning":"virtue","policy":"unclear","emotion":"mixed"},
{"id":"rdc_mw0h0mr","responsibility":"none","reasoning":"virtue","policy":"none","emotion":"approval"}
]