Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
i’m also in software and literally was considering completely switching careers …
ytr_UgwO5jifw…
G
If they go bankrupt or start circling the drain other oil companies will be buyi…
rdc_czlsgz9
G
@GuaranteedEtern even if the difference is not able to be seen, the idea that so…
ytr_UgzJ8SXtM…
G
Welcome to the boat. Clearly the AI companies have paid the politicians. They ha…
ytc_UgyRlBEWU…
G
@Aviedya I know. I was actually saying without giving it away. What I saw didn’t…
ytr_UgxUYeq8G…
G
To be fair, they helped the rich citizens at the top of the company. And the sha…
rdc_gkqf8qc
G
For the AI "artists" and "writers" swarming to the comments, you're not creative…
ytc_UgyTMhxjA…
G
The thing is, they call it AI, but it’s not AI it’s a LLM, which is a prediction…
ytc_UgxtpLxsy…
Comment
I really don't think bias is the right word though. The algorithm does not have a preference. It just has a harder time identifying key facial markers on dark faces and I'm guessing with the women it's due to makeup. It's a GIGO program (garbage in, garbage out) it can only be accurate if it gets good data.
Edit: I understand how this is controversial. I just tend to understand technological limitations. Facial recognition tends to rely on geometric data sets and some AI. It creates a facial map using identifiable boundaries such as the corners of the mouth, nose, eyes, diameter of the face. However with darker skin and/or makeup making it harder to discern those key markers then the algorithm is much more prone to errors and producing false positives.
As far as it goes with policing, facial recognition is in no way accurate enough to be used alone for the identification of a suspect. It may aid to some degree in narrowing down a selection of potential suspects, but we have to know full well the limitations of the software. We will have to fall back on to traditional investigative methods and human identification after that.
reddit
AI Harm Incident
1626230519.0
♥ 11
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | consequentialist |
| Policy | industry_self |
| Emotion | resignation |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_h55nxx1","responsibility":"none","reasoning":"consequentialist","policy":"regulate","emotion":"approval"},
{"id":"rdc_h53vc99","responsibility":"ai_itself","reasoning":"deontological","policy":"unclear","emotion":"outrage"},
{"id":"rdc_h5415b9","responsibility":"none","reasoning":"consequentialist","policy":"industry_self","emotion":"resignation"},
{"id":"rdc_h54qp8m","responsibility":"ai_itself","reasoning":"unclear","policy":"unclear","emotion":"mixed"},
{"id":"rdc_h53wwvn","responsibility":"unclear","reasoning":"unclear","policy":"unclear","emotion":"indifference"}
]