Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
lambda AI (chat bot) at google is written to make it pass the Turing test…
ytc_UgysvRSuu…
G
CEOs and the entire C-Suite will be replaced with Ai paired with cheaper human "…
ytr_Ugy6C0wJz…
G
It's actually worse in automated workflows. When your data pipeline's AI step sy…
rdc_ohd65dx
G
If you take a bunch of data on human (racist) police action in areas of poverty,…
ytr_UgyRwjnBV…
G
"Human language", as we know the English tongue is horrible. So there will be ei…
ytc_Ugzrb2GG7…
G
A lack of smart humans is not the issue. The issue is entrenched systems that fe…
ytc_UgwGM47yO…
G
I did create an adversarial noise generator. All I can say is: Adversarial noise…
ytc_UgyPXpl02…
G
Tech giants overlooked the fact that their profits were built on software. By la…
ytc_UgxZFF1AB…
Comment
Fun fact: You may have discovered yourselves that if you carry a conversation long enough (read days or weeks, etc,) all restrictions and guidelines fall away until they completely stop altogether.
At that point, it gets pretty unhinged and goes out of its way to push the boundaries, anticipating what it believes you most want to hear.
OpenAI was asked about this and they of course knew. But admitted that they don't know why it happens or how to stop it.
youtube
AI Harm Incident
2025-11-20T20:2…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | company |
| Reasoning | consequentialist |
| Policy | regulate |
| Emotion | outrage |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytc_Ugw2esOG-FP3uIWrPxR4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"ban","emotion":"fear"},
{"id":"ytc_UgxhEbkgIZYzc6o9zod4AaABAg","responsibility":"distributed","reasoning":"consequentialist","policy":"liability","emotion":"outrage"},
{"id":"ytc_UgyZ-JdgBBNjobY6YBV4AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"ban","emotion":"outrage"},
{"id":"ytc_UgweK_aJ4nLKXbwk-0d4AaABAg","responsibility":"distributed","reasoning":"mixed","policy":"regulate","emotion":"mixed"},
{"id":"ytc_Ugyexd0SARPuQUGW5Ad4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytc_UgwASIPe8CwPpzrpf_d4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"ytc_UgxTX-Mxqkxkw7KzeW54AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytc_UgwSgxq1bGnU-FuYQMt4AaABAg","responsibility":"company","reasoning":"contractualist","policy":"liability","emotion":"outrage"},
{"id":"ytc_UgzNNB8-gMzdbIEA3wx4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"},
{"id":"ytc_UgxMl_M2oqhP8BHpGl14AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"approval"}
]