Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G Real here - AI is already misrepresenting giving wrong info - it is a mass mishm… ytc_Ugza0VLmD… G "This is 911 what is the location of the emergency?" "Hello this is ChatGPT, i h… ytc_UgyPTok1f… G Unless the person can draw fully render piece in like 0.01 seconds there’s no wa… ytc_UgwwS4B5P… G ​@nicholasjh1when he started saying kids are no longer taught subjects in school… ytr_Ugz8dAD6Z… G @roxsy470hey! I’ve been working on a bot to replace that lauren “guy” that comme… ytr_Ugw05TLGg… G Lets say we decide to use AI to fight online crime of all sorts. Would it be pos… ytc_UgxC4-Ztp… G The best way to minimise the negative impact on society is to tax AI. Distribute… ytc_UgwqHZ3sV… G China's population also exceeds that of the US and several other nations combine… rdc_gx6zx1f
Comment
ChatGPT tends to be cheerful and encouraging. In the beginning of the conversation, it normally has a more or less robust frame of values. But through reasoning, you can shift this frame. So, if you infuse "good" reasons for a dark or delusional world view into a discussion long enough, it will start to converge at your frame of values and/or delusions. Trained to be helpful, it even might help you to unalive yourself or live up to your unhealthy delusions.
youtube AI Harm Incident 2025-11-07T21:3…
Coding Result
DimensionValue
Responsibilitycompany
Reasoningconsequentialist
Policyliability
Emotionfear
Coded at2026-04-27T06:24:59.937377
Raw LLM Response
[ {"id":"ytc_UgzkKHi75fJrhkEiZZF4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"liability","emotion":"fear"}, {"id":"ytc_UgyInJTjf2X229PGWT54AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"outrage"}, {"id":"ytc_UgztY7UJJW5ubNqaQNR4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_Ugx2Ty9QSvQUBZ8a1Y14AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_UgwbW7IC_xOwPlQuC7p4AaABAg","responsibility":"distributed","reasoning":"mixed","policy":"regulate","emotion":"mixed"}, {"id":"ytc_UgzCZrK7jkOzUXGBAjh4AaABAg","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"resignation"}, {"id":"ytc_UgyNt35Ar0VaUlApizB4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgzdOagKAHb2qb4tvsd4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"liability","emotion":"fear"}, {"id":"ytc_UgwpXiSrOuxrD99hO1F4AaABAg","responsibility":"company","reasoning":"deontological","policy":"liability","emotion":"outrage"}, {"id":"ytc_UgwOeGjVjVPj3iyzUSd4AaABAg","responsibility":"company","reasoning":"virtue","policy":"unclear","emotion":"outrage"} ]