Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
You had a chance to share some truely helpful information about the hidden motiv…
ytc_Ugw-nxZtf…
G
Yeah... click the cookie button on every single page first and then tell me abou…
ytc_Ugz7rVMdu…
G
32:44 the Guy ist Wrong! He hast No Idea how Dangerous erotic AI really is.…
ytc_UgzZ95CgS…
G
I thought that I've had is from book of Revelations AI can be the first person b…
ytc_UgyAYVXLJ…
G
Look, i dont give a shit what stranger goes through the shit ive said, i use it …
ytc_UgxprMh3a…
G
I have multiple projects in ChatGPT. Are you saying that Claude isn't designed w…
rdc_o7xeo6o
G
HUMANS SLOW TF DOWN WE DOING TO DAMM MUCH LIKE ALWAYS. Go watch I robot these pe…
ytc_Ugyf_k4I1…
G
The Shoggots were protoplasmic, half-intelligent biomechs, grown for constructio…
ytc_Ugyzk5RcK…
Comment
@dauchande LLMs are very clearly shown to deal better with confabulations the better they perform, similar to how better-educated folks usually are less prone to believing and spreading nonsense. There is nothing that hints at it being remotely ingrained, there is no aspect about their underlying architectures that makes it susceptible to making up information. Never mind that it is mostly a user error to begin with, if you force someone at gunpoint to tell you about your paranoid schizophrenic framework for interfacing with your third eye, they too would make up complete horseshit to keep you rambling rather than turn you into their stabby nightmare.
If humanity managed to build upon all our knowledge despite the vast archives of hallucinated poppycock and pseudo-scientific literature diluting our achievements, current-day LLMs are pretty much fine already and definitely will be going forward. It just doesn't matter, really.
youtube
2025-03-12T18:1…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | user |
| Reasoning | consequentialist |
| Policy | none |
| Emotion | indifference |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytr_UgxkublFe5_VUWRgNXF4AaABAg.AF_I9HDsjBZAF_LsqXHIwN","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"},
{"id":"ytr_Ugy-OOqcuAbtaWOSWF94AaABAg.AF_Hd9jv-CDAF_U0aRQmna","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytr_UgzrCZwRnJEGfCCH92F4AaABAg.AF_EMt0DzUFAF_MSNWnqhd","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"ytr_UgzH6ki7mw8m6Mh17a54AaABAg.AF_CxfwSfW0AF_Eo7w3g0f","responsibility":"company","reasoning":"consequentialist","policy":"liability","emotion":"fear"},
{"id":"ytr_UgzH6ki7mw8m6Mh17a54AaABAg.AF_CxfwSfW0AFa9NQ0G0oG","responsibility":"none","reasoning":"virtue","policy":"none","emotion":"approval"},
{"id":"ytr_UgzH6ki7mw8m6Mh17a54AaABAg.AF_CxfwSfW0AFbb__5jg8B","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"ytr_UgzH6ki7mw8m6Mh17a54AaABAg.AF_CxfwSfW0AFeEFZej5dA","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"indifference"},
{"id":"ytr_UgwsaGrh8vkMvM8q5LB4AaABAg.AF_CFz2nkogANdtkF2elVz","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"ytr_Ugyy0sarhE1EAW79Vfx4AaABAg.AF_B63K6nUdAFaHJoaFjaM","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"ytr_Ugyy0sarhE1EAW79Vfx4AaABAg.AF_B63K6nUdAFb1AlW51yn","responsibility":"distributed","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"}
]