Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
American drivers with their texting, applying make up, arguing and not paying a…
ytc_UgxQ935jU…
G
But, those 17 different doctors got their money for trying to treat a symptom. M…
ytc_Ugz4Pt8ns…
G
I had to watch the video 4... Or was it 5 times before understanding the situati…
ytc_UgyL3-7YW…
G
9:47 What a jackass. This is horrifying. It just goes on and on, gets worse and…
ytc_UgxObLe_P…
G
Ive seen motors on huge tunnel washers spitting out water like a fire hose becau…
ytc_UgztYbhD7…
G
Things change when there's MASS unemployment and all the wealth is being concent…
ytc_UgwyjdDsg…
G
@soakedpasta (and it's an ai generated drawing, not a crime, murder scene and a …
ytr_UgwOPoQu_…
G
I don't think a non conscious AI has any motivations. So yes, if they can become…
ytc_UgzlFdKnJ…
Comment
It's not news because people are bad at statistics, you just don't understand why people are upset.
You are correct that this is an *extremely* accurate system from a statistical and technological perspective. 99.5% accuracy is quite good.
But you're still wrong. The fact remains - the overwhelming majority of people flagged were false positives. This *isn't* an argument that the system is flawed - it's doing what it's designed to do and doing that pretty effectively. It's an argument against sweeping facial recognition based mass surveillance entirely. You're mistaking a moral argument for a technological/statistical one.
In fact, what you're saying drives the point home even more: the system is working quite well and doing what it's supposed to do and what the police wanted it to do. Yet in spite of that, 81% of results are false positives. Those are real human beings with rights too.
It's a little depressing that you've posted a convincing argument for why any sort of large scale automated mass surveillance is inherently repugnant, only to completely miss the point.
reddit
AI Harm Incident
1562213397.0
♥ 55
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | developer |
| Reasoning | consequentialist |
| Policy | regulate |
| Emotion | mixed |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_fvx67z5","responsibility":"company","reasoning":"deontological","policy":"ban","emotion":"outrage"},
{"id":"rdc_fvx40us","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"rdc_esqiowe","responsibility":"developer","reasoning":"consequentialist","policy":"regulate","emotion":"fear"},
{"id":"rdc_esqzw30","responsibility":"developer","reasoning":"consequentialist","policy":"regulate","emotion":"mixed"},
{"id":"rdc_estsssr","responsibility":"developer","reasoning":"deontological","policy":"liability","emotion":"fear"}
]