Raw LLM — Corpus Dashboard

Look up by comment ID

Random samples — click to inspect

G AI should be better. Writers rewrite old Storys to fit a certain Narrative. To … ytc_UgzSCFGCL… G If AI can really replace most jobs, a system where people sell their time for mo… ytc_UgwiobJb9… G AI artists are so embarrassing xD. thank you so much for continuously speaking a… ytc_UgyfTo1l-… G Also not only that but things like git history aren’t always publicly available … rdc_nm8x7yl G First of all, I want to say that your art is really inspiring, it's hypnotic, th… ytc_UgzWQAVLH… G i don't care how benevolent or malevolent ai might seem, it WILL take over the w… ytc_UgxDJMIcZ… G This video has finally convinced me to cancel my subscription to midjourney. As … ytc_UgyLVD1bH… G Ok... The robot held that gun still, in one steady position, os how did the bull… ytc_UgzUyT6Dr…

Comment

Hank, in the bit on AI-induced psychosis, when you say it seems like the AI is roleplaying - I think you're touching on something really important there. *Literally everything an LLM does* can be understood as roleplaying. Try thinking of an LLM as nothing more than a roleplaying machine, and all the weird or surprising things it does suddenly make perfect sense. Take "hallucinations." Hypothesis: AI doesn't actually hallucinate. If you ask it to write, say, a legal brief, it'll write as plausibly as it can what a lawyer would write. That includes referencing real cases it's seen in its training data - the same ones a real lawyer would reference - but it also includes inserting references that look indistinguishable from the sorts of references it has seen in its training data. (For that matter, it doesn't know that all the legal cases it's read *aren't* just someone roleplaying.)

youtube AI Moral Status 2025-12-16T17:5… ♥ 32

Coding Result

Dimension	Value
Responsibility	none
Reasoning	mixed
Policy	none
Emotion	approval
Coded at	2026-04-26T23:09:12.988011

Raw LLM Response

[
{"id":"ytc_UgyE3RhkarsXglEKbel4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"ytc_UgwNedXKXHpm55QxMKN4AaABAg","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"fear"},
{"id":"ytc_UgxeGSODXXNQ_a7cN5d4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugy1t_U_DgBnORUuUZ54AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"mixed"},
{"id":"ytc_UgzwJs-yWxL2Zw8kGr54AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"outrage"},
{"id":"ytc_Ugw1YrapuCCq5OnagPh4AaABAg","responsibility":"industry","reasoning":"deontological","policy":"regulate","emotion":"fear"},
{"id":"ytc_Ugz3ZD8TXDmd2iQNSRt4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"ytc_UgwGjcwuCAXJBOSJhFF4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"},
{"id":"ytc_UgzEEG3LPOv5PP-a6Fd4AaABAg","responsibility":"none","reasoning":"none","policy":"none","emotion":"approval"},
{"id":"ytc_UgyPYtYsY7Y8ajDzra14AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"fear"}
]

Raw LLM Responses