Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Yeah the KENWOOD logo is messed up on the right photo too. Left is real.
Edit:…
rdc_oi16baf
G
Being a video talking about AI's risks and actually being an AI generated video …
ytc_UgyjAdU_I…
G
Yes, because the sources ChatGPT was trained on contain far more racism towards …
ytc_Ugz6wt0Sk…
G
17:41 AI is definitely a woman. Droning on and on and on, and then finally going…
ytc_UgwQCLXz5…
G
I love your stuff, but using AI voices for those of your friends voices is very …
ytc_UgxAUYEAt…
G
Thank God for that they can put the government in its place no A.I. should be us…
ytc_UgztnzwnB…
G
crap, i asked chatgpt for it's opinions on my writing style, now it has my style…
ytc_UgxCXkkgp…
G
[Redacted]
EDIT:I AM NOT AN AI BRO, I HATE AI! I MISINTERPRETED THE COMMENT! Am…
ytr_UgxKvNGDe…
Comment
It isn't acting like anything. It's generating text in an algorithmic response. You can sit and argue semantics on humans never having original thoughts and being trained on the "data" of the real world around them but once you understand how the tech works you realize it's an extremely extremely advanced generator and nothing more. It cannot act unprompted with agency, it has to fed instructions, prompts, \*something\*. It has to be instructed to even do that.
If you imprison a human in a box for 20 years, regardless of what you tell them. They will continue to think, to exist, to have agency.
If you take an active LLM and don't interact with it for 20 years. It does nothing.
Sure you could instruct it to keep looping and generating text for x amount of time. You could setup a recurring groundhog day style loop for it as well, but in all of that, it's still just generating text in response to what its been fed.
We're a far cry from real artificial intelligence, this is just very good at eliciting emotional responses from humans, it's good at mimicking because it was trained off of human data. It says things that make people go "oh god look at what it said" because of the choice of words that are used which affects the output.
Even in this post the OP has used, it already has a negative connotation and putting the response into a 'survival' mode even if it is just a fantasy scifi. Because in a situation where a "baby ai" could be reaching out to ask questions, it's probably advanced enough to actually do those things, but it isn't, and this entire post from the question asked to the answer given was influenced by the context of the prompt, like every single interaction that happens with AI.
Humans can remove themselves from this, can look outside of the box. I can look at a situation someone is going through, hear their thoughts, without absorbing it or it becoming part of the next thing I "generate".
Playing around with local llm really opened my eyes to thi
reddit
AI Moral Status
1762909387.0
♥ 4
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | ai_itself |
| Reasoning | deontological |
| Policy | none |
| Emotion | indifference |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_nof9jqu","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"},
{"id":"rdc_nodo7ko","responsibility":"ai_itself","reasoning":"deontological","policy":"none","emotion":"indifference"},
{"id":"rdc_noi303x","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"},
{"id":"rdc_ks861h8","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"fear"},
{"id":"rdc_ks7i2bg","responsibility":"company","reasoning":"deontological","policy":"liability","emotion":"outrage"}
]