Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
At this rate, GTA six won’t be able to come out before the robot apocalypse…
ytc_UgwWrzlbW…
G
i had no idea that it was even possible to poison ai generators. i wanna start d…
ytc_Ugy_AhurR…
G
I've also seen a lot of Youtube videos, that use AI voices. Some people can't re…
ytc_UgwTDChzZ…
G
"ChatGPT, you started to sound like Jordan Peterson there defining truth".
"I u…
ytc_UgwR--dBh…
G
I partly agree, AI is just a tool for artists. It won't replace artists but you …
ytc_Ugw--_ywd…
G
I see a short future with a blend of the two movies i Robot and the termanator f…
ytc_UgwUGYrkY…
G
Another screwed up part of this is, if you know how to actually use ChatGPT, it’…
ytc_UgxNNPl8-…
G
My wife followed behind me in our '99 Toyota Corolla, and I drove a 2002 Ford E4…
ytc_UgzsGRPZ5…
Comment
It seems that the framework of the solution is easy.
We set up many AI pairs that are designed to act badly (in a simulation), and with the other AI watching and attempting to predict bad actions. They will both get better... use them both to help design measures to stop a breakout by looking for early signs of growth in the wrong direction.
Look for early signs of an AI's attempts to investigate places that it could use to break out.
When you find early signs of problems apply training that reinforces preferred behavior patterns.
Regular training that moderates their behavior toward a norm. Based on feedback from observers that have a goal to keep the AI contained.
youtube
AI Moral Status
2023-08-25T23:5…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | developer |
| Reasoning | consequentialist |
| Policy | regulate |
| Emotion | mixed |
| Coded at | 2026-04-26T23:09:12.988011 |
Raw LLM Response
[
{"id":"ytc_UgwviVWNo4VSsADOgrN4AaABAg","responsibility":"distributed","reasoning":"virtue","policy":"none","emotion":"fear"},
{"id":"ytc_UgyfKyHhZM3QVMDwDzZ4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"mixed"},
{"id":"ytc_UgyKSe3m-7-aXilb5Uh4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugwym_-WI7mM9mzp8294AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"indifference"},
{"id":"ytc_UgyOyG0yPCz2DX5Npy54AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"},
{"id":"ytc_UgwQ8APOlvSug49V9ZJ4AaABAg","responsibility":"distributed","reasoning":"deontological","policy":"none","emotion":"mixed"},
{"id":"ytc_UgzOuOpAioCL3g4D3Bd4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugz0NSzYdbOunFS_DEB4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"resignation"},
{"id":"ytc_Ugwo5xG5jauU-DUfmwd4AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"liability","emotion":"fear"},
{"id":"ytc_UgxzgYl2-Q_0qXP8VJd4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"regulate","emotion":"mixed"}
]