Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Can you imagine these are probably their oldest AI's.. I'm sure they have more a…
ytc_UgwaRt5X3…
G
He left out the most important part. He was shot because the police were surveil…
ytc_Ugx-uSAm2…
G
As someone who uses AI within the context of a stage performance, I think AI is …
ytc_UgyLk3Q6w…
G
Yeah, I was thinking the same. AI writes the script, then a human dots the i’s a…
rdc_jwv654s
G
It is not just that, but also because they know as well as everyone else that th…
rdc_e2wh52e
G
Remember to ask your AI to be "brutally honest" and take "a very critical look" …
ytc_UgzLn1T_s…
G
I hope it's swarms of thousands of AI drones that wipe out all the little green …
rdc_oi1kb3t
G
It doesn't matter what your emotions think. It matters what will help the childr…
ytr_UgzoJUe8_…
Comment
Chat GPT and other AI's have shown racial and other biases--COMPASS's purely algorithmic approach is at least equally biased. All of these amount to the algorithm used in the old South in slavery and Jim Crows days: are you black? if so guilty. With computers, we can be much more sophisticated in our prejudices. Even prior criminal history (the best predictor we have) is biased. People of color and poor people of any color are more likely to have had wrongful convictions as part of their record. Likewise, a sufficiently rich person who is criminally inclined may have a series of wrongful acquittals.
youtube
2024-02-26T15:1…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | company |
| Reasoning | consequentialist |
| Policy | regulate |
| Emotion | outrage |
| Coded at | 2026-04-26T23:09:12.988011 |
Raw LLM Response
[
{"id":"ytc_UgxL-foTm4hFoJl3FBt4AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"outrage"},
{"id":"ytc_UgyzHrz-WNYeqCj7sjp4AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"outrage"},
{"id":"ytc_UgyTvBJ-5-tFaU0wd5R4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"outrage"},
{"id":"ytc_Ugxr84xnJe6mFa6dRuN4AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"outrage"},
{"id":"ytc_Ugz5ang0cNbk8TlfvUl4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"mixed"},
{"id":"ytc_UgxNRQRZCRwuZs1IdzZ4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"},
{"id":"ytc_Ugz8sPkNGiv6g5oOMt54AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UgxdWZuiADGDNXyMPIZ4AaABAg","responsibility":"government","reasoning":"deontological","policy":"liability","emotion":"fear"},
{"id":"ytc_UgzZ1zdzDnTs0XuzVtx4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"},
{"id":"ytc_UgzNXywCG8irU835pLZ4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"liability","emotion":"mixed"}
]