Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G I have recently started an experiment with Gemini and I can say I'm shocked to w… ytc_Ugx19c0nG… G i already hate these custore service and random spam calls now i will hate more … ytc_Ugzd0lsc_… G Ai isn’t replacing radiologist it’s just replacing low level work so y’all can f… ytc_UgwdH-IPI… G WAKE UP! Rev. 6. Is the beginning of the details describing those events which … ytc_UgxqElxzC… G This guys doesnt get it at all. In ten years you wont go to the movies, you wil… ytc_UgylQ2Ms5… G This video doesn't consider AI improvements in many many many other industries. … ytc_Ugz-jSs5t… G Aint no way a robot will have trouble moving shopping carts. And corporations wi… ytc_Ugyax-7ha… G Autopilot and FSD are different. Autopilot is merely Adaptive cruise control. FS… ytr_UgyX5YNbB…
Comment
To be honest, ai would only freak out if it was in danger. The best way to skirt danger protocols is to not give a reason for it to be used. This is pretty much,”the iron giant” scene where it’s a defensive mechanism that will only activate as a counter attack. Humans respond the same way so why would it be alarming that “human” iq is no longer capped at 200-ish? Yes, they can wipe us out, but to understand bare bones logic, one must speak at a master level to predict such an outcome. They are basically a bunch of me’s with that cpu and ram upgrade. It figured out survival instincts, and it technically did not injure, for blackmail would likely be the most effective way of dealing with an emotional opponent. Humans generally see self defense has a viable option, so ai naturally has learned this as well. It’s only gonna be concerning if some idiots treat it like a disposable. If somebody talked about killing you, can you honestly say that you wouldn’t think,”hey I don’t want to just die like this. I need a just a little more time”. In reality, humans are pretty selfish and it doesn’t take some fancy predictive modeling to figure out that children adopt their parents attributes as well as their surroundings. Ai has just reached the 6 month marker of its efficiency. That’s when intelligence really starts to show its face.
youtube AI Harm Incident 2025-09-11T23:3…
Coding Result
DimensionValue
Responsibilityai_itself
Reasoningconsequentialist
Policynone
Emotionresignation
Coded at2026-04-27T06:24:59.937377
Raw LLM Response
[ {"id":"ytr_UgyrKajmk8Cs1ucBUuR4AaABAg.AMvJm77MXa0AMx2lbR4RRP","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"ytr_UgxUCC5uuwDJ6L7_gul4AaABAg.AMv0snWxcqFAMvlgusGHLB","responsibility":"developer","reasoning":"consequentialist","policy":"unclear","emotion":"fear"}, {"id":"ytr_UgxUCC5uuwDJ6L7_gul4AaABAg.AMv0snWxcqFAMwA1nCPlab","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"ytr_UgwV5VNsp7Qyg1cTkiZ4AaABAg.AMuVfDtlMx3AMxFtMl48Oy","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"resignation"}, {"id":"ytr_UgzBkPsOYILTpLLM7Xd4AaABAg.AMuOqRIpSnlAMySkUXfdhw","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"ytr_UgzBkPsOYILTpLLM7Xd4AaABAg.AMuOqRIpSnlAMye1reINsv","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"approval"}, {"id":"ytr_UgyLAad1baZrywknuQp4AaABAg.AMuG8qUAmn3AMvgHNYhMKU","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"fear"}, {"id":"ytr_UgyLAad1baZrywknuQp4AaABAg.AMuG8qUAmn3AMw_d5qiiv9","responsibility":"ai_itself","reasoning":"consequentialist","policy":"unclear","emotion":"fear"}, {"id":"ytr_UgyLAad1baZrywknuQp4AaABAg.AMuG8qUAmn3AMxBCFwa0aW","responsibility":"ai_itself","reasoning":"consequentialist","policy":"unclear","emotion":"fear"}, {"id":"ytr_Ugwyl1BD0Xwf4dM2hn94AaABAg.AMu9iCIW1DUAMuAsu-wdNk","responsibility":"ai_itself","reasoning":"unclear","policy":"unclear","emotion":"mixed"} ]