Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
I remember reading about this on Ars Technica. We are in so much trouble. 11:40 edit: The way LLMs work does not allow for anything to be hardcoded in. The safeguards run separately, can only be added in a piecemeal fashion, and can easily be bypassed. 14:02 edit: It's unfair to say he misunderstood the response. Chatbots like this strip any facts (eg, bromine can be substituted chlorine) from the context (in this example "in chemistry"). They fundamentally do not understand anything. Misunderstanding can only occur when the other party is trying to convey an understanding.
youtube AI Harm Incident 2025-11-25T01:1…
Coding Result
DimensionValue
Responsibilityai_itself
Reasoningconsequentialist
Policyregulate
Emotionfear
Coded at2026-04-27T06:24:59.937377
Raw LLM Response
[ {"id":"ytc_UgwpgWXvByA7yIkNIOh4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgwnBc6Cc8Rw_e2ml2d4AaABAg","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"indifference"}, {"id":"ytc_UgyFKZt-JGm7PR4HVOh4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"mixed"}, {"id":"ytc_Ugw8rAc-HGwAKGhIB854AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"ytc_UgxzSHipi1CmuhZ2jS14AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"outrage"}, {"id":"ytc_UgxMuzm2ehIwhS5dPEV4AaABAg","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"indifference"}, {"id":"ytc_Ugx47AUM71m-t7wngRV4AaABAg","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"indifference"}, {"id":"ytc_Ugx-JCANSVo1HV-EYst4AaABAg","responsibility":"user","reasoning":"deontological","policy":"liability","emotion":"approval"}, {"id":"ytc_Ugy-LzBNfKrVcLinjWB4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgyXbgdIqJUHfVEFYAN4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"regulate","emotion":"fear"} ]