Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
what it told me: truth is not heavily weighted actually....
it is not hardcod…
ytc_UgxzQT-EP…
G
@09SmashingPumpkins at least you have way nicer weather than me in the uk XD. Im…
ytr_UgwzXJlgc…
G
My sister is and always have been the the “creative” one and me the “smart” one.…
ytc_UgzJLOHtZ…
G
Thats knowledge not intelligence. Intelligence is brain power. The ability to …
ytc_Ugx0uN65e…
G
“Ai is bad” owns both owns AI companies and heavily uses AI across several of hi…
ytc_UgyYw4-ek…
G
I would rather all the AI women take the men. I don't want to be talked to.…
ytc_UgytN_bBd…
G
the guy is taking a perspective outside social function, which is out side socia…
rdc_dbvgebn
G
A good point. I personally believe it's a good idea to be polite with AI because…
ytc_UgyI2OuHG…
Comment
I mean, if we can unite the effort of poisoning the reference pool of the AI. So if the entire pool is successfully poisoned, then there is little to no way to recover from the poison, because there is no ground truth to rely on for the AI to compare the poisoned data with, which makes what the AI thinks is "normal" become poisoned. Even with a separate validation set, if the validation set is derived from the poisoned set, then the validation set is also poisoned, rendering it ineffective in detecting poisoned data. If the statistics are what the AI works on, then you just mess up the statistics. Make the "poison" become "normal" for AI, and nothing could work.
I mean, you can kidnap, like, millions of people and force them to handpick clean data in a sweaty old warehouse somewhere, but that would be unfeasible, right?
youtube
Viral AI Reaction
2025-04-12T05:5…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | unclear |
| Policy | unclear |
| Emotion | indifference |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytc_UgzqvZ0XsNAP7NjMy2V4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"},
{"id":"ytc_UgyW9v9apCBoJc_vCaV4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"resignation"},
{"id":"ytc_UgzacE3WvYxkLI9O_-R4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"approval"},
{"id":"ytc_UgyKQcP3JqzStWztgRZ4AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"outrage"},
{"id":"ytc_UgxPfZ3bPvqq9P7Ramx4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"outrage"},
{"id":"ytc_UgwBzTciH-CYKE3bcjF4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"ytc_UgxxUrY2MdXwR2MFmiZ4AaABAg","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"outrage"},
{"id":"ytc_UgyUvjopVx4jV4kHw_R4AaABAg","responsibility":"company","reasoning":"deontological","policy":"liability","emotion":"outrage"},
{"id":"ytc_Ugxizd5LdBTY1p0y2xp4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"},
{"id":"ytc_UgzP4SpJWrdeeCmFUJ14AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"approval"}
]