Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Either humanity will be eliminated by AI as soon as it is able to sustain and d…
ytc_Ugw9TBokq…
G
If Ava becomes a robot I wouldn't be surprised if she becomes the first AI robot…
ytc_Ugz_9d2z9…
G
I disagree about not creating asi, the current economical system is a pyramid sc…
ytc_Ugx4ocHN4…
G
Men create ai programs. Men in rome took over the world
Columbus and spain too…
ytc_Ugy-zNGEh…
G
AI art is more of an art tool than anything, but it's a tool that automates the …
ytc_Ugz7BAEaH…
G
One question about the Ai future, with all this talk of having robots doing all …
ytc_Ugz1edtxs…
G
No offence Doc but the people that we could control AI were off the scale in arr…
ytc_Ugx-u5bCp…
G
I hope every single ai generated stuff is FORCED to have a ai water mark on it a…
ytc_UgxtpUQAs…
Comment
@Robespierre-II I think saying "programmed" is just a wholly inaccurate way to talk about it and it makes it sound like an acceptable solution.
What they do is they literally put in the prompt (the "hidden" prompt" the AI gets sent before your actual prompt) lines that say "Don't say anything to do with X" or even feed it a response to give in the event that the LLM decides it wants to generate that.
The "filters" on these AIs are crazy, because they're literally just asking the AI not to do those things in plain english, even crazier is that they usually use "negative" direction. Prompt engineers _know_ that saying "She's not wearing pants" is functionally identical to "She's wearing pants" so... yeah, a "filter" that says "Don't talk about race" means about the same (mathematically) as "Do talk about race"
youtube
AI Moral Status
2024-08-31T16:2…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | company |
| Reasoning | deontological |
| Policy | regulate |
| Emotion | outrage |
| Coded at | 2026-04-27T06:26:44.938723 |
Raw LLM Response
[
{"id":"ytr_UgwsIgpGHQUJGvuzyuN4AaABAg.A7lojboGloEA7lrGyM9vBi","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"ytr_UgwsIgpGHQUJGvuzyuN4AaABAg.A7lojboGloEA7lskdFHbFf","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"ytr_UgxN_IoNR6fIp0aOBKB4AaABAg.A7liodm6risA7oMYT-6Q9d","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"mixed"},
{"id":"ytr_UgxN_IoNR6fIp0aOBKB4AaABAg.A7liodm6risA7oRjQa7GFP","responsibility":"company","reasoning":"deontological","policy":"liability","emotion":"fear"},
{"id":"ytr_UgxN_IoNR6fIp0aOBKB4AaABAg.A7liodm6risA8-IMidozY8","responsibility":"company","reasoning":"deontological","policy":"industry_self","emotion":"approval"},
{"id":"ytr_UgxN_IoNR6fIp0aOBKB4AaABAg.A7liodm6risA81SlGVT_vG","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"indifference"},
{"id":"ytr_Ugx-i06SoP7frpiHnf14AaABAg.A7lg98t4AfkA7nWws5IslM","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"},
{"id":"ytr_Ugx-i06SoP7frpiHnf14AaABAg.A7lg98t4AfkA7oK1Gz5E43","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytr_UgyskqWXoHYl1BsS80J4AaABAg.A7ldIMhYIRoA7lqjnAKJb2","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytr_UgzYd-5m6100Ag6qqK14AaABAg.A7lWZ8c90j0A7z0SBrtFOM","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"fear"}
]