Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
IA ne remplacera jamais l'inspiration qui peut se connecter au Futur proche ou l…
ytc_Ugzf9llTn…
G
Grenades. Like in call of duty. If you hit me my bike automatically drops a gr…
ytc_Ugys9JRs1…
G
speaking as a robotics researcher (who started at MIT in the 90's) I really like…
ytc_UgyD_vVgK…
G
Dawg this is such a stupid idea. What's the point of making an apology video if …
ytc_UgzFM46oo…
G
@casperaaron5530 true. They actively need to work on it as soon as possible and…
ytr_UgzoB34e3…
G
Look I hate Elon, and am unsure about full self driving, but I won't lie, whenev…
ytc_UgzHwNntf…
G
@laurenpinschannels for what it's worth as someone who has done some research in…
ytr_UgxcjR158…
G
Let's see if Aurora's in-charge can put his money where his mouth is. Have his f…
ytc_UgwuptTBH…
Comment
>I am definitely unintentionally condescending and toxic to the AI
I have caught myself doing this too. One of my recent prompts ended up being something to the effect of:
"Please DO NOT attempt to search for a sequential series of hex identifiers for <objects> the tool used to generate these objects generated them sequentially during the development process, and so they are not arranged in blocks of similar <object_types>. You keep trying to add that to the code, despite being instructed not to in every bit of reference documentation I have given you, every instruction I have given you, and the markdown best practices instructions I have given you. STOP DOING THAT. Do not add it as a fall back, do not add it in the notes, do not add it to a separate script! Just DONT DO IT."
This was after hours of more reasonable prompts. It would generally forget that it was not supposed to add that code every 15 mintues. I quickly devolved into thinking of it in a similar way to some teenagers I once had to supervise while working in food service. The need to always tell them exactly how to do the exact same task, the exact same way, every single time they need to do it is remarkably parallel.
reddit
AI Governance
1757794603.0
♥ 3
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | user |
| Reasoning | virtue |
| Policy | none |
| Emotion | resignation |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_ne1vqjp","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"resignation"},
{"id":"rdc_ne1xfhd","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"rdc_ne20l2n","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"mixed"},
{"id":"rdc_ne2eedh","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"approval"},
{"id":"rdc_ndzqour","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}
]