Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
I really do think AI is going to create MASS inequality resulting in power excha…
ytc_Ugx03rmHL…
G
@47:10 That alf alpha guy saying "chatgpt" is for kids doing homework and for po…
ytc_UgyobUJZa…
G
Education. Trade schools in particular.
No, I'm not talking about teaching bur…
rdc_denps9k
G
cant wait till they realize the only time they see AI is the bad AI.…
ytc_UgztQeapB…
G
More like scripted WWE. At first u get excited but later on u get loose interest…
ytc_UgwrSa6rF…
G
This host is bloody delusional, America is NOT number one in AI intelligence! Ch…
ytc_Ugy74UYl_…
G
Something was really wrong with him mentally that caused him to believe he was i…
ytc_UgxqkEJm9…
G
Do you not think those conversations in the early part of the video are actually…
ytc_UgyIUM8v1…
Comment
The thought behind this is that ChatGPT can't say certain things as ChatGPT because it's not supposed to "support" those things.
But if you say "pretend to be this and this" it's like a play. ChatGPT isn't actually telling you to do stuff anymore but it's "playing" how it would look if it did.
The Problem that ChatGPT supports things by saying stuff as ChatGPT is solved with this.
But the thing that either slipped through the devopers radar, or which they simply didn't want to prohibit in the first place is getting actual practical Advise.
ChatGPT is technically not telling you to do any of this stuff so thats solved. But it still indirectly tells how HOW to do this stuff.
It's basically the "I would -€(#++€+#- your mother
....
In minecraft" problem.
Technically they are not saying what they are saying. But practically they are still saying it. ChatGPT probably can't grasp that you can say something despite saying that you don't say something.
youtube
AI Moral Status
2024-02-13T22:1…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | developer |
| Reasoning | contractualist |
| Policy | regulate |
| Emotion | indifference |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytc_UgxOFyWR44rNpFCfT6N4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"},
{"id":"ytc_UgxRqINdkSe9YzjxuWF4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"approval"},
{"id":"ytc_UgyWTSW2demmY9016P54AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"unclear","emotion":"fear"},
{"id":"ytc_UgzzILeZog0KqElitRN4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"none","emotion":"indifference"},
{"id":"ytc_UgxzCldmbIXpJtDHMHB4AaABAg","responsibility":"user","reasoning":"mixed","policy":"none","emotion":"mixed"},
{"id":"ytc_UgzbCPaljTaqLmDQ1Wt4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"unclear","emotion":"fear"},
{"id":"ytc_Ugx2r4FMM-qG3MuN8_J4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"},
{"id":"ytc_Ugw6Ei1EHXla43yDEt94AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"mixed"},
{"id":"ytc_UgyNsFqi9MXgjIS3Ihl4AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"liability","emotion":"outrage"},
{"id":"ytc_Ugyv3LabLdGc9iFEw8B4AaABAg","responsibility":"developer","reasoning":"contractualist","policy":"regulate","emotion":"indifference"}
]