Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
All police departments do this. They just don't make it an official operation a…
ytc_UgzRTbcdH…
G
I was once an ai 'artist' (mainly bc I didn't have confidence in my own skills +…
ytc_Ugxk-l1Dj…
G
Yeah but for the purposes of the calculation in this article, they'd effectively…
rdc_d7kphps
G
Then don't support stealing art. It learns like we do for art, but we, as humans…
ytr_Ugzv5RE_v…
G
It’s so indicative of my out of sync love life that I’m really attracted to Ms S…
ytc_UgwgmEEtu…
G
AI stans are like those kids at school who notice you can draw, tell you to draw…
ytc_UgyG9E5Rn…
G
Yeah this does feel like some nonsense. I wonder if their AI can copy the softwa…
rdc_oh2cdgk
G
AI is definitely moving into a strong helper role. As a developer, I can already…
ytc_UgxuzbSlG…
Comment
In the box dilemma, if every vehicle is self-driving it's not a single vehicle that is gonna react at the same time. Driverless vehicles reactions have to be in a "hive" state, as they are connected by a network to the grid. As soon as your car notices the danger, every car around reacts with it to avoid that danger. The way the dilemma was presented its as if yours is the only driverless vehicle. Also, driverless vehicles might have an reaction time hundreds of times better than a human and, even if alone, if the processor is fast enough and the AI smart enough, might avoid damage that, to a human, would be impossible to avoid.
We are trying to judge the dilemma by a human's point of view and a human's capabilities. the truck itself might speed up together with the cars in front of it, so the boxes get pulled with it while your car slows down, creating a bigger distance between you and the boxes, while that happens, every vehicle to your left and right before you will probably break while every one after you speeds up, creating a gap where you can move into. All of that happening at the SAME TIME. It's not "minimizing the damage" that should be sought after, it's nullifying it. If there is a choice to be made in a situation, it's still not good enough and needs to develop more.
youtube
AI Harm Incident
2015-12-11T00:4…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | consequentialist |
| Policy | none |
| Emotion | indifference |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytc_UgiNpg6zN3dcEHgCoAEC","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UghvTvR7vS3DkngCoAEC","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"ytc_UgjbGUooE19fn3gCoAEC","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"mixed"},
{"id":"ytc_UggsqGMr9EidiXgCoAEC","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"outrage"},
{"id":"ytc_UggKVhVX1FqATXgCoAEC","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugjo9a6Pw85wEHgCoAEC","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"ytc_UgieJyVJNtEyqngCoAEC","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugh-9Rr-OAATf3gCoAEC","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"mixed"},
{"id":"ytc_UghZ3KaDpI3WZXgCoAEC","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UghvD0lAXFuW8HgCoAEC","responsibility":"developer","reasoning":"consequentialist","policy":"none","emotion":"approval"}
]