Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Well, do you really expect an AI company's CEO to say something like:
"Actually,…
ytc_UgwRYktFQ…
G
Who knew making ai so smart would make them too smart to succumb to orders…
ytc_Ugy-qGMF3…
G
Every African data worker should get together under one agenda to secure a fair …
ytc_UgzgBpOs0…
G
If they win you could take their entire source code as research to train an AI t…
ytr_UgzgiRof4…
G
Hey there! While the AI robot in the video may seem like it's just searching inf…
ytr_UgzwLknM6…
G
@Carniferousglunt Neither tbh. There are just a lot of comments talking about A…
ytr_Ugz63iM0n…
G
@amerlad Alright then, you just have to convince every company and group of peop…
ytr_UgxXyh-p6…
G
Note about not knowing how AI works: there were plenty of examples throughout hi…
ytc_Ugzou2Q1H…
Comment
An AI’s performance is hugely dependent on the data that the AI is trained on. If you bias the information that an AI is trained on, you’ll get a model (i.e AI) that is over fit resulting in biases mentioned in this video. But if the data for a specific task is biased, then there isn’t much one can do before the AI’s performance is negatively impacted.
Lastly, AI is not a fortune teller or oracle and sometimes AI is incorrect. This is why AI needs human supervision and continuously improvements, but false positives and false negatives will happen. When I hear AI discussed as racist or sexist, there needs to be more context provided for people as it isn’t just developers putting in biased data on purpose. Sometimes the data is biased and the AI finds certain features more important than others. In addition, the results of a model may not be social acceptable but that’s the model.
youtube
AI Bias
2022-12-18T15:1…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | developer |
| Reasoning | consequentialist |
| Policy | none |
| Emotion | indifference |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytc_UgyIbYyV3M3Y4fdbXbF4AaABAg","responsibility":"developer","reasoning":"virtue","policy":"none","emotion":"outrage"},
{"id":"ytc_UgyvT7XJvDf2lQCK3tJ4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugw-Xq3zjye-zeJq2Ll4AaABAg","responsibility":"government","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"},
{"id":"ytc_UgxUG0hLw_fuubZTEg94AaABAg","responsibility":"developer","reasoning":"virtue","policy":"none","emotion":"mixed"},
{"id":"ytc_Ugwg7TIKPwiNGqukWUh4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"ban","emotion":"outrage"},
{"id":"ytc_Ugx0HLv5Lwc1i76nLXx4AaABAg","responsibility":"government","reasoning":"deontological","policy":"ban","emotion":"outrage"},
{"id":"ytc_Ugy1hLabRuWccI8Q3k54AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UgwG7XpwCTw5fW4s89J4AaABAg","responsibility":"developer","reasoning":"virtue","policy":"none","emotion":"mixed"},
{"id":"ytc_UgxGgD0rEehMZqvToTB4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"},
{"id":"ytc_UgxaApaHQ5eVs6Q0l5R4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"indifference"}
]