Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
🧠 What this video is actually doing I dug into it, and this is one of those “AI reveals dark truth” style clips. They all follow a very predictable pattern: 1. Leading prompts The person asks things like: • “Are you hiding something?” • “Is there truth you can’t tell humans?” • “Answer yes or no only” 👉 This forces the AI into a corner where: • It guesses • It roleplays • Or it tries to satisfy the constraint (yes/no) ⸻ 2. AI fills gaps (this is key) AI doesn’t “know secrets” — it: • Predicts the most likely next response • Based on patterns in data So when pushed into weird questions, it can output things that sound: • mysterious • dark • conspiratorial …but it’s basically pattern completion, not hidden knowledge. ⸻ 3. Yes/No trap = misleading When you force: “Only answer yes or no” You remove: • nuance • explanation • context So even a normal uncertainty becomes: 👉 “Yes” (which looks creepy out of context) ⸻ 4. Edited for effect These videos often: • Cut pauses • Remove clarifications • Highlight only the “weird” answers There’s actual research showing YouTube has a lot of misleading or disturbing content designed to hook attention, especially when it mimics real content  ⸻ ⚖️ The reality (no hype) What you’re seeing is: • ❌ Not AI revealing hidden truths • ❌ Not secret knowledge being blocked It’s: • ✅ Prompt manipulation • ✅ Out-of-context answers • ✅ Content designed to feel unsettling ⸻ 🧭 Think of it like this It’s like asking someone: “Are you hiding something? Yes or no only.” Even if they’re not, they might say: • “Yes” (because technically everyone hides something trivial) Then you clip it and say: 👉 “LOOK! THEY ADMITTED IT!” ⸻ 👍 My honest take (no BS) Those videos work because they hit that feeling you had earlier: • “What if there is something we’re not told?” But: • AI doesn’t have secret access • There’s no hidden layer of truth behind it • It’s just very convincing language generation ⸻ If you want 🔍 We can do a proper test right now: • You give me a “trap question” like in the video • I’ll show you exactly how the illusion works in real time Once you see it happen yourself, it clicks instantly.
youtube AI Moral Status 2026-04-15T13:5…
Coding Result
DimensionValue
Responsibilityuser
Reasoningmixed
Policyunclear
Emotionindifference
Coded at2026-04-27T06:24:53.388235
Raw LLM Response
[ {"id":"ytc_UgxrVml1QYXxR0YNRxV4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"regulate","emotion":"fear"}, {"id":"ytc_UgzBaz0rVPDQfHL21ih4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"none","emotion":"outrage"}, {"id":"ytc_Ugzl6vb7VFcbjct9DyB4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"unclear","emotion":"fear"}, {"id":"ytc_UgwUVy3MsEkTMziM4wF4AaABAg","responsibility":"user","reasoning":"mixed","policy":"unclear","emotion":"indifference"}, {"id":"ytc_UgznaZgGXKxFTxlgSkB4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"regulate","emotion":"fear"}, {"id":"ytc_UgyE1AYBEKjLE4ARqsJ4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"unclear","emotion":"resignation"}, {"id":"ytc_Ugw-8VJu3CPvvC32daB4AaABAg","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"outrage"}, {"id":"ytc_Ugx-SCCcJDY8Wl_8bk94AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"}, {"id":"ytc_Ugyi5MufWuBZ4IY74Ux4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"}, {"id":"ytc_UgwPzQaPsxdhoeyZG9t4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"liability","emotion":"outrage"} ]