Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
8:35 First you ask an AI to give solutions without moral boundaries,
and then ur…
ytc_UgyXxO9lg…
G
Thanks Silicon Valley, you're turning the Terminator series into a DOCUMENTARY o…
ytc_UgzRNOqlG…
G
2:50 Specifically Waymos cannot be controlled remotely, so they are always drive…
ytc_UgyUT6TZR…
G
Talk about me... I made illustration and design my career and AI is killing some…
ytr_Ugw-sI8Yq…
G
But if someones work SOOOO good, that nobody can make it that good just like him…
ytc_Ugx1Kj_kc…
G
Fun fact, people ended up in hospitals after trying online recipes that turned o…
ytr_Ugw-kSrKl…
G
Which government? With AI in control I doubt the government will be anything lik…
ytr_UgxvltSQz…
G
Did you have a stroke? Sorry, that was ableist - Listen bud, people in remote pa…
ytr_UgxkefnLh…
Comment
Oh hey, I know what echo chambers are and their dangers, but I didn't realise LLMs would give people an opportunity to make a small echo chamber with themselves and polarise their psychosis. It doesn't help that the short term goal is to make these LLMs as pleasant to interact with, which means that they will never challenge your views but instead will reinforce them.
There may be a solution; present an unpalatable idea and arguments in a humorous way and trigger curiosity. Of course, there's always the risk that you're associating the unpalatable idea as a joke, which discredits it and if you're actually trying to influence opinions, by having the LLM have an opinion, people will catch on. Honestly, the only solution is trying to educate people to be more critical and open-minded, so maybe there's no hope.
But oh well, I still hope that somewhere, some flat Earth believer tries to interact with an LLM on the subject, and it replies with a thick Chinese accent: "Are you stooooopid?!" and then proceeds to profusely insult them while delivering arguments.
youtube
AI Moral Status
2025-07-09T16:0…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | company |
| Reasoning | consequentialist |
| Policy | regulate |
| Emotion | fear |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytc_UgwiEWNctlQGcmI5v4F4AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"outrage"},
{"id":"ytc_UgzZHZMoBHdqrAYuN5F4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"},
{"id":"ytc_UgwOlxeDtogNV5rtI-l4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"indifference"},
{"id":"ytc_UgwVjTkEXnHizg3vTZ14AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"mixed"},
{"id":"ytc_Ugx_TvYWK3AUfW5D5Kx4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"ytc_Ugx3vewk2Koz79HGhUl4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UgyEsEcuG23f9oCgGAd4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"fear"},
{"id":"ytc_Ugx9oN9oM8Kpeo-uyX14AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"fear"},
{"id":"ytc_UgxefB2nRluc-Rr2QVF4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"ytc_Ugxr3f9pW8RKGEhvqFV4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"}
]