Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
Ahhh! Interesting. I vividly mind reading or hearing about something similar; this is a little bit Chinese Whispers but I *think* there was an old story about a famous old actor (might have been the director, I'm hazy on it) being asked about a particular scene/movie they were in, for which they may have won an award. The gist was something along the lines that the director had asked the actor several times to try to express a particular emotion during the scene (something about stairs or walking up to a railing?) and wasn't satisfied with how it was looking on camera. Eventually, after multiple takes, the director just asked the actor to completely wipe their mind, have no specific emotion or motivation, and just walk up to the railing while looking off into the middle distance... which ended up being the shot the director used in the final cut & deemed to be perfectly capturing the character in the moment! With regard to other comments making the argument that "the emotion is happiness in both cases"; that does sound like a fair point, but I think maybe the counterpoint is that any human emotion is too nuanced to be able to boil down as simply as that; especially for the purposes of companies trying to judge your emotional state for advertising. The happiness of a grandfather looking at his family isn't quite the same as the happiness (horniness?) of a lech eyeing up a stranger on the beach, no? Even at that, the first clip could still arguably be, with a sufficiently malicious context, actually more malevolent than it first appears. The second could even be the opposite; what if he's smiling because the attractive woman just has a striking resemblance to his wife in her younger days? The wholesome Hitchcock clip would maybe be better targetting adverts for gifts for a small child or a female relative; the lecherous Hitchcock clip would perhaps be more into LoveBunny, condom adverts or movies that star attractive actresses prominently (and no, not those types of movies, you dirty sod you. Ha.) If they combine facial recognition with being able to look at your search history over the last, say, 24 hours? That might be enough context & they can target adverts if you're in, say, a happier mood and more likely to buy a particular product (or in a bad mood and want something to comfort eat/cheer you up...)
youtube AI Moral Status 2024-08-29T18:3…
Coding Result
DimensionValue
Responsibilitynone
Reasoningunclear
Policynone
Emotionindifference
Coded at2026-04-27T06:24:59.937377
Raw LLM Response
[ {"id":"ytc_Ugzweb9kQDSVnDg2bYN4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgyXFlmOQDupiZzRy_J4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_UgwjjQ9bFzjHRzGwxBR4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_UgwXIvCCGRhilT3foB54AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"none","emotion":"fear"}, {"id":"ytc_UgwDGJG6hwnaxz2H6Rd4AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"none","emotion":"outrage"}, {"id":"ytc_UgwfIyc8DpmK19hjYDB4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"indifference"}, {"id":"ytc_Ugzg-LKU9Wgd0IukAQN4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"fear"}, {"id":"ytc_UgxbQFeMUe2WTGdTCmJ4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_UgxCjF-MP0uelJx5yD54AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"indifference"}, {"id":"ytc_UgxbEQSUwrYPlq1clCV4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"} ]