Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Doing multiple lane changes at once and cutting people off? Wow, self driving ca…
ytc_Ugx6lSDO0…
G
I think ChatGPT has gotten so good and advanced that is now able to troll humans…
ytc_UgyCgWsgS…
G
For this boss, don't ghost.
Day-1: Call and say you're running late but sho…
rdc_ohqqojm
G
The more you do videos actually calling out this kind of comment and waste time …
ytc_UgyF861Me…
G
Ok *trained* models to literally do somebody’s style is TERRIBLE please don’t us…
ytc_UgzCqEm9_…
G
because of AI, every unwanted thing is happened only in India
In India, every di…
ytc_UgwuKx7vw…
G
I heard that this doesn’t work anymore and that you have to put it through some …
ytc_UgwTHACWY…
G
They may only need 1 person instead of 5, but they sure are not paying that 1 pe…
ytc_UgxVk-nvD…
Comment
@wardm4 are you saying we can’t put guardrails at the point between the user submitting the prompt and LLM inference taking place? Seems to me that’s something we could be doing if we’re really concerned about LLMs going off the rails. In fact, I’m pretty sure most of these companies have versions of this already, but it does seem they’re trying to limit doing that for now. But as more people harm themselves as a result of talking with LLMs, I’m pretty sure we’re gonna see more of this. Granted, users can find ways around these types of guardrails as they’re implemented today. But if OpenAI and the others wanted to they could spend plenty of time and resources to reduce the chance of circumvention.
youtube
AI Governance
2025-10-17T23:3…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | company |
| Reasoning | consequentialist |
| Policy | liability |
| Emotion | fear |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytr_UgyOjzmTIoLZXJQ8_614AaABAg.AOLq_Q-3U1jAOMOxQavCEm","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytr_UgyOjzmTIoLZXJQ8_614AaABAg.AOLq_Q-3U1jAOMZQfG2nIu","responsibility":"ai_itself","reasoning":"mixed","policy":"unclear","emotion":"mixed"},
{"id":"ytr_UgxHMLXRr5iWJK8TPC14AaABAg.AOKcxEEl8x_AOL8M15bXkM","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"ytr_UgxBIFZaYUxl9uHUwgR4AaABAg.AOKakxbuc3SAOMoJcYULyg","responsibility":"developer","reasoning":"deontological","policy":"industry_self","emotion":"approval"},
{"id":"ytr_UgxBIFZaYUxl9uHUwgR4AaABAg.AOKakxbuc3SAOO6NOe7PSi","responsibility":"user","reasoning":"mixed","policy":"unclear","emotion":"outrage"},
{"id":"ytr_UgxBIFZaYUxl9uHUwgR4AaABAg.AOKakxbuc3SAOOyD59Zh8i","responsibility":"company","reasoning":"consequentialist","policy":"liability","emotion":"fear"},
{"id":"ytr_UgwnFgN90LTWyWMpQ1x4AaABAg.AOJlHYQMm7EAOLpvvvwYb0","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"resignation"},
{"id":"ytr_Ugx0eO84iCVdGa-cKip4AaABAg.AOJau-ynNbIAOKRp7QKg86","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytr_Ugx0eO84iCVdGa-cKip4AaABAg.AOJau-ynNbIAOR8w2Bjz6b","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytr_Ugzt0860SlM1kjlnlQV4AaABAg.AOJZLkvQuUxAOLpkVsaMkt","responsibility":"ai_itself","reasoning":"mixed","policy":"unclear","emotion":"fear"}
]