Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
It's all about context. The same AI model will say completely different answers to the same question, if you first ask it a related question that steers it or whatever. It's like it activates parts of the simulated mind and primes it. So if you ask it something about medical text or something then ask it for advice it might sound like a doc and be relatively good substitute for a doc's advice (compared to aunt Doris at least), but if you ask it about government giving pregnant mothers plutonium and then give it the medical text and ask for advice it might be very critical of the text. The trick is to know it's just predicting what you want to hear * what it knows, and how to give it proper context and tools so it can give the *proper* response.
youtube AI Harm Incident 2025-11-26T21:2…
Coding Result
DimensionValue
Responsibilitynone
Reasoningunclear
Policyunclear
Emotionindifference
Coded at2026-04-27T06:24:53.388235
Raw LLM Response
[ {"id":"ytc_UgwXj31gZXjyfnR-8up4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_Ugz95s8kc4TgnNNi_IB4AaABAg","responsibility":"ai_itself","reasoning":"unclear","policy":"unclear","emotion":"fear"}, {"id":"ytc_UgyjlcVDkQq3j3BRrE54AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"industry_self","emotion":"approval"}, {"id":"ytc_Ugx-hZdCuHMWwxGDycV4AaABAg","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"}, {"id":"ytc_UgxNfC7qVMPQcwSB0jN4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_Ugysiw5QjG6QDLAznrV4AaABAg","responsibility":"ai_itself","reasoning":"virtue","policy":"unclear","emotion":"mixed"}, {"id":"ytc_Ugx2ANk3EIvvzbvkY5B4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"approval"}, {"id":"ytc_Ugw0vPJtRa0pcmTh0lF4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"ytc_Ugwb35NSCRp1OZc1CsV4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"regulate","emotion":"outrage"}, {"id":"ytc_Ugxayw4_NQ-AG2hDevR4AaABAg","responsibility":"distributed","reasoning":"consequentialist","policy":"none","emotion":"resignation"} ]