Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
LANGUAGE CHANGE AI BHOT PHALA HE AH GAYA THA.
POV: TV MAH DISCOVERY CHANNEL MAH…
ytc_UgyVfTjl-…
G
This is really just a commercial for Khan Academy. Most of what is presented wi…
ytc_UgwCU_XWS…
G
My personal experience in AI. I am 61 years old, and worked my way up the ranks …
ytc_UgzRbUZlk…
G
For me, the most breathtakingly stupid part here is "I thought it was telling th…
ytc_UgzYQ6B-E…
G
There's so many things wrong with this but a couple of things got my attention. …
ytc_UgzyRgUKe…
G
I do have one question for you!
I 100% AI “art” is not art.
However, I had come…
ytc_UgwZtkD4o…
G
Hey Lavender? I wanted to ask if there’s mobile support for the AI poisoning I w…
ytc_UgwQbm3jl…
G
Why are more people not scared of this? Am I catastrophizing by thinking this is…
ytc_UgwErsxup…
Comment
This is such a mistake about AI, it doesn't take the average, it takes your input, and then it runs it through cascading probability tables. These cascading probability tables attempt to come up with the best result at each level, so it becomes more like a decernment table. were it selects and rejects previous conclusions until it comes to the best fit it can to produce the closest to the most relevant single elements in the training data.
This means that with a carefully developed prompt with proper instructions and a large enough token base you will get good code. A great example is to add instruction prompts at the start that define what good code is before requesting it to develop code. This will get you marginally better results because in this case, it might target a training sample that was provided from a text book (as the text book might have used words or phrases that described good code)
youtube
AI Jobs
2024-07-06T05:3…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | unclear |
| Policy | none |
| Emotion | indifference |
| Coded at | 2026-04-26T23:09:12.988011 |
Raw LLM Response
[
{"id":"ytc_UgzRjMtjPvCyJ3bh-5V4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"outrage"},
{"id":"ytc_UgzqDDAYiOOGTMPPNKB4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"},
{"id":"ytc_UgxJaLZR3BzBG65I-Kh4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"indifference"},
{"id":"ytc_UgxK5tzu3S_UgpRzD0p4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"outrage"},
{"id":"ytc_Ugyw_9JoiD_V7BqNtfJ4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugz-52TgPxidp_O5G6N4AaABAg","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"mixed"},
{"id":"ytc_Ugz0DB-T_LchQL-hhn54AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"indifference"},
{"id":"ytc_UgyH7Fr0Ws1XGsnEbSZ4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"ytc_UgzzKn-Op2CxL92Cwyp4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"industry_self","emotion":"approval"},
{"id":"ytc_UgwvmFRUm6nZjVkweQB4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"}
]