Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Manipulated And Corrupted Data As Source(s) Also Makes AI Corrupt And Manipulati…
ytc_Ugz4iT3tT…
G
Not the same at all, not even close. Its Simple, You (Generated) a prompt but th…
ytr_UgyvHnGic…
G
The problem is that if the people who create these conversations have these cons…
ytc_UgxQwzuox…
G
Why does that robot in the middle look like doja cat with no makeup and if she w…
ytc_UgwGh6SdL…
G
If technicians program responses to important topics a certain way, then surely …
ytc_Ugx2bpDKw…
G
What’s funny is that in a podcast with Joe Rogan he said AI needs to be heavily …
ytr_UgxG06_8_…
G
The controversy of AI use fades away very quickly when you think of vs whom that…
ytc_Ugw6luSqh…
G
Your video contains some information that might be misleading to people who don'…
ytc_Ugzr9d3r-…
Comment
Because "AI" is often just a short-hand for using huge amounts of computational power to replicate complicated patterns in the real-world training data. It turns out racism is also complicated pattern of behavior in the real-world. Sometimes, the correct real-world prediction for things like arrests is "the cops will act like racists." Likewise, sometimes the correct prediction for real-world promotion behavior is "they will promote a bunch of rich, white frat bros." Good AI developers check their predictions to make sure that the results reflect their hopes, instead of blindly laundering the status-quo with recommendations based on their predictions.
reddit
AI Harm Incident
1576169661.0
♥ 62
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | consequentialist |
| Policy | none |
| Emotion | resignation |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_fakmnp3","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"rdc_fakqrdm","responsibility":"developer","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"rdc_fanio3o","responsibility":"developer","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"rdc_fal9bvn","responsibility":"developer","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"rdc_fal61p2","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"approval"}
]