Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
The problem is that robots don't buy stuff. They don't consume anything other th…
ytr_UgzcymAQm…
G
Do you use AI but not to generate photos? Also, if you go to the comment section…
ytc_UgwI_2Z2f…
G
The trend is a good one, but as Sam said, the fast food analogy makes a lot of s…
ytc_UgyaW8WRD…
G
Yeh...don't lift samples from the AI generative stuff, bro. This is the basis o…
ytr_UgyKuewO2…
G
If AI ever turns on us.. Alex will be the first to go! They won’t want to go thr…
ytc_Ugyicm8j3…
G
It seems to me that a case could be made - as a class action suit, I'd imagine -…
ytc_UgxbvYwAF…
G
@stephenr85 Well OpenAI recently got rid of important evidence in New York Time'…
ytr_UgzFqF5Dc…
G
What's the difference if AI kills us or climate change? We're extinct either way…
ytc_UgwkPjlTS…
Comment
Is it just me or do AI videos stick out like a sore thumb? I can't point to the direct reason, but I can instantly tell when something is AI.
My friends feel this way too. One night me and two of my friends decided to test out thoughts. We each made 4 groups of videos with 5 videos in each group. Each person got 4 groups; two from each person.
We uploaded the videos to a friends server to make sure our links were as uniformed as possible. Each person picked a random letter so they would name their "1at" with "a, b, c, d" representing the groups.
Then we all watched them and guess if they were AI, wouldn't trust, or real life.
Every single AI video was labeled either "AI" or "wouldn't trust". Of the 60 videos shared 37 of them were AI and 23 were real life. 7 AI videos were labeled "wouldn't trust" with 30 videos being absolutely "AI".
The interesting part came on our 27 "real life videos". 8 of them were mislabelled as "wouldn't trust" with 4 being labeled as "AI" with only 11 videos being labeled real life.
We tried to limit any bias so we took random steps to avoid this. We used a selection box instead of typing and once we labelled each video we couldn't see the previous results. We also took a 15 minute break in-between without using technology.
It was a really fun process/game we did. We plan on doing it again, but now we want to add videos that are AI and real life.
While I do think AI videos will always set of our internal " Spidey tingles" I think just AI muddling the waters will work to make us distrust real videos. To me being able to tell if a video is AI or not is equivalent to not being able to tell if a video is real life or not.
If we aren't able to tell if a video is real life that will end up having the same affects of not being able to tell if AI is real life or not.
reddit
AI Jobs
1771267784.0
♥ 2
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | unclear |
| Reasoning | unclear |
| Policy | unclear |
| Emotion | mixed |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_o5sed36","responsibility":"company","reasoning":"mixed","policy":"unclear","emotion":"outrage"},
{"id":"rdc_o5q1l1m","responsibility":"ai_itself","reasoning":"consequentialist","policy":"unclear","emotion":"fear"},
{"id":"rdc_o5qdavf","responsibility":"company","reasoning":"mixed","policy":"unclear","emotion":"indifference"},
{"id":"rdc_o5qiqvc","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"rdc_o5qaveu","responsibility":"unclear","reasoning":"unclear","policy":"unclear","emotion":"mixed"}
]