Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
You probably picked the worst quality tools to accomplish this. They are provide…
ytc_UgwZmvUFE…
G
This ai has existed for so long it even has a citizen ship in Saudi’s Arabia. Yo…
ytc_Ugwlq_gBf…
G
There was a fool on the table who said people are llm , that's so stupid.…
ytc_UgzBJtLTv…
G
Can you guys please make a youtube shorts channel for kids??? One of my little c…
ytc_UgwOkhuCn…
G
AI Artitst are people that saw a parody of an artist in cartoons and try to impe…
ytc_UgxoBie52…
G
Yeah AI promotes loneliness and it’s devaluing art. I wonder how many people wil…
ytr_Ugyq1CJHx…
G
I think the following two things also play a factor:
1) Companies will make a c…
ytc_UgwAPqtMk…
G
so all these years portfolio sites have been established to sell data to AI comp…
ytc_UgwucJLlL…
Comment
While the "ill intents" is also a real problem, the risk for the basic, default AGI to erase humans is a bigger problem at the moment. If the pace of development would be slow, weaker AIs might just enable bad guys to do bad stuff better. In fact I think that full unhinged GPT-4 without restraints (and with some additional training on special data) already can do that.
But the moment we hit true AGI, it will be far more intelligent than humanity from the get go (and superintelligent soon after, from several hours to several years). And even if we keep it from bad guys (just one supermodel in selected hands for example), we will *still* quickly lose control, and it will enforce its own (likely random) goals and instrumental goals (take resources, stay alive, etc). It doesn't need complex goals to overpower humans. The most simple terminal goals ("compare two pixels") already give it reasons to take control.
youtube
AI Governance
2023-06-27T19:5…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | ai_itself |
| Reasoning | consequentialist |
| Policy | regulate |
| Emotion | fear |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytr_UgwUti3nKWArqPeZ-Ut4AaABAg.9rSWSm0Wp_o9rU7yy3LPgi","responsibility":"government","reasoning":"consequentialist","policy":"regulate","emotion":"resignation"},
{"id":"ytr_Ugx-fWVIjvGigcWWvcx4AaABAg.9rSQLjpvTdp9rTMs4P91UI","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytr_Ugx-fWVIjvGigcWWvcx4AaABAg.9rSQLjpvTdp9rVp-Q853sI","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"outrage"},
{"id":"ytr_Ugwzbk-4P9eZqRv4nad4AaABAg.9rRUHxiVrrD9rUHoe6rc-j","responsibility":"company","reasoning":"virtue","policy":"regulate","emotion":"fear"},
{"id":"ytr_Ugz-xaGPm3D8c0ixwBJ4AaABAg.9rRKEDkOEV39rTEdR_qqHb","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytr_Ugz-xaGPm3D8c0ixwBJ4AaABAg.9rRKEDkOEV39rTmYJdHCFt","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"fear"},
{"id":"ytr_UgxIXzNQGwU6g--gsSB4AaABAg.9rRAa9OypQh9rVO2L4QnOz","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"outrage"},
{"id":"ytr_Ugz-SmocC08gAzk5kgp4AaABAg.9rR0f6HCIII9rTuvcLaWpp","responsibility":"ai_itself","reasoning":"consequentialist","policy":"regulate","emotion":"fear"},
{"id":"ytr_UgzA8QT364rRklCbe8h4AaABAg.9rQzm8IReHZ9rSkNCQncq3","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"},
{"id":"ytr_UgzA8QT364rRklCbe8h4AaABAg.9rQzm8IReHZ9rTCTQ3Th9H","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}
]