Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
DAMNNNN AI really gettin me to feel nothing ubt it will get me emotional in 2-3 …
ytc_UgzgQQu9T…
G
Funniest thing is that deepfakes of male streamers have existed for years even s…
ytr_UgxC7Njl-…
G
You art community losers really need to get new material. I swear every video is…
ytc_Ugy8oVTOz…
G
Yeah, i was trying to use Chatgpt for research for my thesis, it's decent with g…
ytc_UgxwcUnLX…
G
I think most people run into it when trying to ask social questions/topics.
The…
rdc_jhgjr2n
G
Is there an AI app to recognize what AI generated and what's Real ? If not, than…
ytc_UgxR5JK27…
G
We have to remake education. LLMs and all of the technological innovations of th…
ytc_UgwCza2vt…
G
One of the biggest current problems with AI-generated art is IP - if an AI creat…
ytc_UgyRrbpK7…
Comment
It very definitively is not storing it, for example, an image in an image model represents less data than two characters of text in terms of data. The reason it can reproduce, sometimes verbatim, is because the text prediction model predicts the shape of the article enough to just reproduce it
There's also something called "overfitting" but thats not whats happening with a times article - its just that language can only be expressed so many ways and news articles are written similarly enough that if it knows the shape of your article it has a chance of reproducing your text. So while that entire article it scanned might be two bytes of mathematical data, if it's scanned that writers entire history of work it can probably - from that tiny datapoint - reproduce the text *very closely*
Which is why if you run articles through AI detectors they will be detected as AI even if they werent
youtube
AI Responsibility
2026-04-11T19:4…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | unclear |
| Policy | unclear |
| Emotion | indifference |
| Coded at | 2026-04-27T06:26:44.938723 |
Raw LLM Response
[
{"id":"ytc_UgxR6pgJiXK68YCg5G54AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UgzSLqlT0uFkccswhZB4AaABAg","responsibility":"distributed","reasoning":"deontological","policy":"liability","emotion":"mixed"},
{"id":"ytc_Ugwnr-US85qnT5RRmzV4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"},
{"id":"ytc_Ugw1g_kQ7yfuanUJjdJ4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"ytc_UgwkKAwwSUCYQHach3R4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"},
{"id":"ytc_UgzKX0pdAKQSbszDkzt4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"mixed"},
{"id":"ytc_UgylHoLsXZKdsV-KfGd4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"unclear","emotion":"fear"},
{"id":"ytc_UgxrxsN8OyoTHprHA-B4AaABAg","responsibility":"distributed","reasoning":"mixed","policy":"regulate","emotion":"mixed"},
{"id":"ytc_Ugzah7ADs2OjAvDH95R4AaABAg","responsibility":"unclear","reasoning":"deontological","policy":"unclear","emotion":"mixed"},
{"id":"ytc_Ugx8I5FZY_SvdxmQlL94AaABAg","responsibility":"company","reasoning":"deontological","policy":"liability","emotion":"outrage"}
]