Raw LLM — Corpus Dashboard

Look up by comment ID

Random samples — click to inspect

G I HATE when people say they are artists or put that in their bio but use AI. Its… ytc_UgyndTt4a… G All these kids in the comments cheering on unregulated AI replacement of human l… ytc_UgwsDGF07… G @Frosty708 Ok forget the word racist for two seconds, it's blinding you for a p… ytr_Ugw2BlAeM… G The Basilisk isn't real....have you seen 2 AI talk, they don't say please and th… ytc_UgyCDPL-4… G A thing about AI artists that i atleast interacted with is that they are really … ytc_UgwnJB5ku… G There is no intent behind AI "art" so in my mind there is no point to it existin… ytc_Ugw_SjtHb… G No one does that. They really should just anonymize the names upon receipt and… rdc_luztjx2 G I don't understand what he means by *"properly equip the American workforce."* W… rdc_denkltl

Comment

The idea of training AIs is hilarious. At the macro level, LLMs trained on millions of data sets, yes, but they're not training on your usage at the micro level. Not at all. Have you ever tried to create a custom GPT? LOL. It’s hilariously impossible. No matter how well you try to write the instructions, it simply does NOT obey you. It randomly ignores all your instructions. For instance, I’m trying to get my copywriting GPT to stop contrast framing (which is “It’s more than an A. It’s a B.” or “Our mission is not just to A. It’s to B.”). It does this like five times per 200 words. But it just can’t stop. Same with the em dashes. It uses these long hybrids that are long like an American em dash but with spaces on the side like a UK en dash. I keep instructing it to not do this, but it can’t remember. Also, in terms of frequency, I tell it to not use the en/em dash more than once per 100 words. You think it listens? LOL. It simply CANNOT help itself. It does what it wants. I realize that some agencies and brands are using LLMs to write their own copy now, but god help them. If they don’t have an experienced writer working with their LLMs, then they’re producing junk.

youtube Cross-Cultural 2025-10-14T12:5…

Coding Result

Dimension	Value
Responsibility	ai_itself
Reasoning	mixed
Policy	unclear
Emotion	mixed
Coded at	2026-04-27T06:24:53.388235

Raw LLM Response

[
{"id":"ytc_UgzU0hSOO3RN49FRyRx4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"mixed"},
{"id":"ytc_Ugx8k1rAIDGOx0rBiYh4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"indifference"},
{"id":"ytc_Ugy2M2W-ktP-VAYSi1N4AaABAg","responsibility":"ai_itself","reasoning":"mixed","policy":"unclear","emotion":"mixed"},
{"id":"ytc_Ugx6GZs6w05A4TtzPjZ4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"resignation"},
{"id":"ytc_UgwQh2zPYv6RzNRkTQ54AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"outrage"},
{"id":"ytc_UgxgQURbOzz2KrOInJh4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"mixed"},
{"id":"ytc_UgzFVXR-m2HAw-CJOn94AaABAg","responsibility":"company","reasoning":"virtue","policy":"unclear","emotion":"outrage"},
{"id":"ytc_UgxwGWmyLlErIqFuPMF4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"unclear","emotion":"approval"},
{"id":"ytc_UgwXmy4UjPVAFL13KrZ4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"liability","emotion":"outrage"},
{"id":"ytc_UgzVL2JFkumRqkN5Dil4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"regulate","emotion":"fear"}
]

Raw LLM Responses