Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
The look on her face... I think this is the first recorded instance of an AI bei…
ytc_Ugzn6K52J…
G
I don't understand the point of language, conversation and any sort of expressio…
ytc_UgxEL6LJr…
G
ai is a thesaurus, it's not a lexical thesaurus though. It is a semantic thesaur…
ytc_UgyZ6jC9i…
G
I always say please and thank you when asking it anything, and when I finished a…
ytc_UgyshGENL…
G
@lizzzbenoist Do you realise how stupid you actually sound? That's 2 grade leve…
ytr_Ugz2C1Nli…
G
This is a great example! It went wrong in so many ways. We also did a video wher…
ytc_UgyugRxPr…
G
I agree with you that AI art is bad. On the fence if it's nessisarilly "stealing…
ytc_UgyLHTxA1…
G
AI will make humans insignificant. When AI evolves, Humans will depend on AI on …
ytc_UgxQTcUP8…
Comment
you do realise that you are actually helping to train the AIs right? while they might give worse immediete results upon encountering your drawing, it is exactly this that makes them learn how to make and perceive something in a more human way. 'poisoning' art is not some magic that only AI falls for, it exploits the specific way in which most models recognize images.
if you purposefully give tricky questions to a child they will probably answer wrong, but they will eventually learn them and become 'smarter' overall. the same is true with AI models.
by feeding them things that initially confuse them (but that are for humans easy to understand) you only help future AI models get better at drawing and understanding art how a human would
PS. mare than anything this just shows you misunderstand what nightshade is and how it works, not to mention what AI training is. while I do like the pushback against AI art, this is not the way. as you may have realised, AI is pretty good at marking nightshade as AI. during training there is a lot of effort put into filtering out bad data, which includes other AI things, because, as one might expect, without that filtering, AI would perform worse and worse as more of the internet is AI generated slop.
youtube
Viral AI Reaction
2025-12-29T03:1…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | user |
| Reasoning | consequentialist |
| Policy | none |
| Emotion | mixed |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytc_UgztSfi-QPsJaDMdj494AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"ban","emotion":"outrage"},
{"id":"ytc_UgyAKL6I8QdVfPnYniV4AaABAg","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytc_UgwFU5NWiX1OkxLdubt4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"},
{"id":"ytc_Ugwz3rgH1BsuVVn1xOp4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"ban","emotion":"outrage"},
{"id":"ytc_UgwI1gZx6l2cktbS1At4AaABAg","responsibility":"user","reasoning":"mixed","policy":"industry_self","emotion":"indifference"},
{"id":"ytc_Ugz3epb1tCTfrOzdTi14AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"mixed"},
{"id":"ytc_UgycAxEspB0U4aH0EQh4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"},
{"id":"ytc_UgxkB9fYqP-ApoXVlIt4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"ban","emotion":"outrage"},
{"id":"ytc_UgxzB7rlNIvGaw194PV4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"},
{"id":"ytc_UgxvyXhTkSDXmt2Ozwp4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"mixed"}
]