Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
And we’re going to get restricted pretty soon ! fake is becoming real and real i…
ytc_UgzJyZ69n…
G
They already are choking.
But it's kind of great news to me, a data engineer.
…
ytr_UgxCjjHml…
G
Third world countries with no internet and minimal tech who still rely on manual…
ytc_UgyfW___-…
G
are the 5 percent of profitable AI when they tried replacing CEOs to say hyped u…
ytc_UgxzIOtDp…
G
I tested AI right and it said if people were right behind the scenes l…
ytc_UgyHNl79s…
G
Let me guess, chat gpt wrote this for you?
AI can achieve great things, but this…
ytr_UgxECCn4e…
G
What you said about conservatives mocking a formal training in the visual arts h…
ytc_UgxlChlie…
G
Years ago it was tictactoe, then chess players
In 2016 (around then, I was too y…
ytc_Ugyb9tGnM…
Comment
The technology actually is racist, but not necessarily intentionally. As someone who actually works on these systems there’s two main technical reasons off the top of my head why facial recognition objectively performs worse on minorities. The most obvious one is training data. These systems are trained on millions of images and it’s natural that if there are less images of minorities (because they are minorities), then the system will bias towards non minority populations. The second reason is cameras themselves. Cameras were historically designed (and continue to be built) for optimal capture of white faces. This results in pictures of minorities being generally worse quality and therefore harder for machines to successfully identify.
reddit
AI Harm Incident
1593051542.0
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | developer |
| Reasoning | unclear |
| Policy | unclear |
| Emotion | mixed |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_fvy9ex6","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"rdc_fvz2b0i","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"approval"},
{"id":"rdc_fvz4sxq","responsibility":"developer","reasoning":"unclear","policy":"unclear","emotion":"indifference"},
{"id":"rdc_fvyoiqf","responsibility":"government","reasoning":"deontological","policy":"ban","emotion":"outrage"},
{"id":"rdc_fvwzrtz","responsibility":"developer","reasoning":"consequentialist","policy":"unclear","emotion":"mixed"}
]