Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
How do we stop AI noticing these patterns?
I guess blacks could stop committing …
ytc_UgzIYDiQI…
G
say we find a way to make ai sentient, why, just why? it's fine the way it is. w…
ytc_UgzDzLQCH…
G
Sir Roger is correct. The world intellect does not apply to machine learning, bu…
ytc_UgzTaMbbl…
G
What is art? Baby don't hurt me. Don't hurt me. No more.
Go watch something rec…
ytc_UgxoqZRPz…
G
AI probably will replace one day most of professional artist in industry, but it…
ytc_Ugz6F4MZO…
G
The solution is not forcing AI and robotics companies to subsidize the welfare s…
ytc_Ugz6GKtBZ…
G
@Kinshu-ep1jc well sports is different then art 😐well kind of…😐for one my icon i…
ytr_UgzkAebuO…
G
There’s also all the facial recognition things that have a problem with POC’s fa…
ytc_UgxH2HnU9…
Comment
Coincidentally it's actually the "solving new problems" capability that makes neural network generative AI so interesting. They're not specifically just repeating content they find on the internet (like Google does) - it learns the rules and relationships that holds the content together and then uses those rules and the current context to generate a response to a prompt. This is why it's so good at being confidently wrong - it is inventing something novel on the spot which can be incorrect.
I'm a software engineer and asked chatGPT to write some code to use an API - it didn't effectively just paste the content from the API's documentation; it instead completely invented a non-existent website and wrote the code to use that. It understands what, at the fundamental level, an API is and the general pattern of how to use one.
The difference is a human brain can store vastly more relationships than current models can but that is something that is rapidly changing as models have been experiencing an expontential growth in size lately.
reddit
AI Moral Status
1674094639.0
♥ 6
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | unclear |
| Policy | none |
| Emotion | approval |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[{"id":"rdc_j4xug0a","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"rdc_j4y86t4","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"rdc_j4y6d02","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"rdc_j4xjr38","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"rdc_j4xxgig","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"fear"}]