Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
I have a hard time calling that "altered by AI" because that is a a term used for situations where you change the content. This is more or less just a bad looking codec which tries to optimize the available data rate for the important stuff. Showing details where it makes most sense. Things codes do for decades. I wouldn't be surprised at all if they use AI to find those important areas and to enhance those details with as little data as possible while reducing detail in other areas squeezing it in fewer bytes. But that's not what AI altering means in my eyes. That's crappy encoding with AI. Pretty sure it is to save bandwith. But yes, it is a problem if genuine content looks more like AI. If it makes you trust youtube less, finally, that's a good thing.
youtube 2025-09-18T11:5…
Coding Result
DimensionValue
Responsibilitynone
Reasoningconsequentialist
Policyunclear
Emotionapproval
Coded at2026-04-26T19:39:26.816318
Raw LLM Response
[ {"id":"ytc_Ugyu2KDam2KLoZ8btvh4AaABAg","responsibility":"company","reasoning":"deontological","policy":"unclear","emotion":"fear"}, {"id":"ytc_UgwLaWY3Z9eQqpeXqX54AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"ytc_Ugyzt52uHJeKQb9CUQF4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"approval"}, {"id":"ytc_Ugx04bcVfVaSkuRR7hx4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"indifference"}, {"id":"ytc_Ugz2edC4inESQh88qyB4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"approval"} ]