Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
@IgnisKhanit's not sentient. How does it roleplay hacking nuclear arsenals or …
ytr_UgzsFSPFi…
G
If all the employees are fired, what will the corporates do when they have no cu…
ytc_Ugx64yTma…
G
@heeyy3currently large language models without any fine-tuning prefer unicorns …
ytr_UgzO9Sqep…
G
Tax AI like we tax humans. One metric might be to tax by the token of AI used. …
ytc_UgyG2Eo2N…
G
I'm from California and I gotta say it's cool to see how the Europeans do things…
ytc_UgwpZ9lAb…
G
Love you guys but I think you vastly underestimate what AI is capable. I hate it…
ytc_UgzIMOXrQ…
G
Brilliant questions. There are a few basic solutions that should prevent these s…
ytc_UgibjtNUD…
G
Sometimes higher has prices are complicated. It has to do with geopolitics, so o…
rdc_oi0sg42
Comment
I really don't think people understand the full extent to which facial recognition software is and will screw people over.
First of all, it is created with a heavy bias towards identifying white and southeast Asian people, due to heavier samples used for testing. Darker skintones that have obvious differences to the human eye look completely identical to this technology. And even if you meet the best match criteria, you are still affected by things such as camera quality, lighting, hairstyle changes, posture, and expression.
[This technology is being used for job applications](https://www.washingtonpost.com/technology/2019/10/22/ai-hiring-face-scanning-algorithm-increasingly-decides-whether-you-deserve-job/). My boyfriend applied to hundreds of jobs a couple of years ago and had multiple "interviews" where he recorded himself answering questions in video form. These videos are analyzed and screened by a facial recognition software before being passed off to real people. These services claim that they can check for minute movements and facial changes that somehow make one candidate better than others.
But as you can see from this article, this kind of software can't even tell two totally different dudes apart from each other, let alone perform whatever kind of phrenology garbage they claim to be able to do.
reddit
AI Harm Incident
1609201197.0
♥ 22
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | developer |
| Reasoning | consequentialist |
| Policy | regulate |
| Emotion | outrage |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_ghc3axh","responsibility":"none","reasoning":"deontological","policy":"liability","emotion":"outrage"},
{"id":"rdc_ghc1eb4","responsibility":"unclear","reasoning":"unclear","policy":"unclear","emotion":"unclear"},
{"id":"rdc_ghbvo88","responsibility":"none","reasoning":"deontological","policy":"regulate","emotion":"approval"},
{"id":"rdc_ghc6ies","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"rdc_ghcay0o","responsibility":"developer","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"}
]