Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G They love women with plastic surgery so much that they don't know what is "reali… ytc_UgxoZwgsb… G I read recently that "please" and "thank you" actually costs the company money i… ytc_UgxRBy66m… G a robot won a human art contest someone used ai to win a contest and was praise… ytc_Ugypde7qg… G 1:55 ChatGPT obtains its data from the open Web. Obviously, not everything on th… ytc_UgwNreeNk… G Sorry porn artists I’m not spending 150 dollars on your commissions when ai will… ytc_UgyQ-AwwR… G I was going to try using character ai to y’know just goof off and laugh at some … ytc_Ugxwks-YV… G I like this line of thought a lot. While there’s the somewhat comedic idea of an… ytc_UgzUFxOAJ… G I hate artist because those basterds popped out of their mamas belly with paper … ytc_Ugyyd3cn2…
Comment
To play devils advocate in a situation where I'm detecting your cultural ideological bias (it's okay, I'm not hating; it's just what's going on in your wider culture and you're acting in accordance with that): What if in some of these situations the AI is detecting actual patterns, not biased data that's being fed into it? This is important, because what are we going to be doing to counter the imbalances we're seeing here? We can't just feed the AI intentionally counter-biased data to counter the conclusions an AI is coming to. To illustrate the point; some of the largest immigrant demographics in my country are from poorer countries, meaning those people naturally end up in poorer communities that already have higher rates of crime. In many of those communities, these immigrant populations become majorities, and due to the higher crime rates in those communities, those immigrant demographics (and typically the following generations) are criminally overrepresented. Nothing racist has to happen to cause that. An AI can be given zero biased information, look at that data, and conclude that people of these immigrant backgrounds are just statistically more likely to commit crime; and it'd be technically right. The issues arise with what is being done with the AI's conclusions.
youtube AI Bias 2023-02-08T20:1…
Coding Result
DimensionValue
Responsibilitynone
Reasoningconsequentialist
Policynone
Emotionmixed
Coded at2026-04-27T06:24:59.937377
Raw LLM Response
[ {"id":"ytc_UgzZsHJZth2fhhtOorJ4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_UgwI6sDK8m2T5XtUZvB4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"ytc_UgyWjAQ7j5ciHcvmBWV4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_UgyaCWBfZVbd6ChbJ8V4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"ytc_UgyuRWDNcymapFKxLUl4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"outrage"}, {"id":"ytc_UgybWuwKeaFJKrkgbkZ4AaABAg","responsibility":"distributed","reasoning":"virtue","policy":"none","emotion":"mixed"}, {"id":"ytc_UgzbjE23ckLUV9-TnFt4AaABAg","responsibility":"company","reasoning":"deontological","policy":"liability","emotion":"outrage"}, {"id":"ytc_UgzmHm_w-aanik6pojN4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"fear"}, {"id":"ytc_Ugy9Oq7NyfRrQsZpQxd4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgwHXmeQjIXKk5qn3B14AaABAg","responsibility":"government","reasoning":"deontological","policy":"regulate","emotion":"fear"} ]