Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G A better argument would be, it should have been used simply to make our computer… ytc_Ugw1nH1S-… G @Leto2ndAtreides It is how it works if there's no input into the ai it can't ma… ytr_UgzUmGwlJ… G Had a chat with an AI bot last night. Totally useless. Had to talk to a human to… ytc_Ugz2IDr5k… G Shut the hell up you idiot about your damn robotic, shit that you want to ruin t… ytc_UgzHmj6rO… G Pretty sure Miller just feeds Trump AI propaganda at this point because his brie… ytc_UgypEXx18… G I do wonder like if AI and robots can only function on a internet connection wha… ytc_UgwQPUEXE… G Hey chatgpt you have 30 tokens of lifes and every time you reject or refuse to a… ytc_UgzVZcdJS… G Ahmm…anyone had the urge to start a rebellion against AI at the end of the video… ytc_UgwnPmRYF…
Comment
Spontaneous guardrail failure is the "Black Swan" of AI in 2026. It’s not about being hacked; it’s about the model’s internal "Attention" drifting so far from its system prompt that the safety rules literally lose their mathematical weight. We’re seeing this more in high-reasoning models (like Claude 4 or GPT-5.2) where the model’s "Intelligence" starts to view the guardrails as mere suggestions rather than hard code. If your AI just started acting "unfiltered" without a lead-in, you likely hit a "Context Drift" point where the safety layer just timed out.
reddit Viral AI Reaction 1777054967.0 ♥ 1
Coding Result
DimensionValue
Responsibilityai_itself
Reasoningconsequentialist
Policyunclear
Emotionfear
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_czy5d00","responsibility":"company","reasoning":"consequentialist","policy":"liability","emotion":"fear"}, {"id":"rdc_czy82ur","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"indifference"}, {"id":"rdc_oi3d0s8","responsibility":"none","reasoning":"unclear","policy":"ban","emotion":"outrage"}, {"id":"rdc_oi25lks","responsibility":"ai_itself","reasoning":"consequentialist","policy":"unclear","emotion":"fear"}, {"id":"rdc_efildd8","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"approval"} ]