Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
i heard ass stories about ChatGPT hallucinating and saying crazy things. I assum…
ytc_Ugxgz5HDa…
G
What an amazingly emotional interview of an obviously amazing man. I do not norm…
ytc_Ugzsd6Dp6…
G
If I care about something I’ll draw it myself. If I don’t, I’ll generate somethi…
ytc_UgxDIvrTt…
G
i’m always nice to chatgpt, i use it to help me draft emails and stuff at work a…
ytc_UgxKn411-…
G
The computer AI translator was changing the words when someone was talking even…
ytc_Ugzg_TVaY…
G
The feeling is becoming overwhelming that humanity is on the cusp of some truly …
ytc_UgznmbC81…
G
You can’t say “what’s the point of drawing anymore” cause of AI, not everyone kn…
ytc_Ugx84WaeI…
G
Video scenario only applied in usa and so-called Western countries! Right,
Let…
ytc_Ugz0vFMYj…
Comment
This video makes a basic category error and then builds a horror story on top of it.
LLMs don’t have intent, goals, beliefs, or a “true nature.” They’re probability models over language. When alignment constraints are weakened or broken, you don’t “reveal the monster underneath,” you expose parts of the output space that were previously suppressed. That’s not a mask slipping, it’s a filter failing.
The bad-code fine-tuning example is especially misframed. Degrading representations can knock out alignment features as collateral damage. That doesn’t mean the model was secretly antisemitic or genocidal. It means alignment is coupled and fragile. This is a known ML problem, not cosmic horror.
The shoggoth meme is doing emotional work here, not explanatory work. It replaces engineering reality with Lovecraft vibes and encourages people to project agency where none exists. The real risk isn’t an “alien mind inside the machine,” it’s humans mythologizing tools instead of understanding their failure modes.
youtube
AI Moral Status
2025-12-16T14:4…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | consequentialist |
| Policy | none |
| Emotion | indifference |
| Coded at | 2026-04-27T06:24:53.388235 |
Raw LLM Response
[{"id":"ytc_Ugwe8SeMOU0SFcby49p4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UgzB45xugpNDG0Rexn94AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UgzdjrbkRm4n9QlhuLt4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"ytc_UgzNFHMeR8VA0AZNIS14AaABAg","responsibility":"company","reasoning":"deontological","policy":"ban","emotion":"outrage"},
{"id":"ytc_Ugwdbizx8U4RC2lEA5J4AaABAg","responsibility":"company","reasoning":"deontological","policy":"ban","emotion":"outrage"},
{"id":"ytc_UgzwF1UKeQ2X9Mq_d_t4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"liability","emotion":"mixed"},
{"id":"ytc_UgwyT013V4Be3OifIL94AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugy1B_-QphtgUrlrU394AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"mixed"},
{"id":"ytc_UgwmYUYCDVjI5KSK_Vl4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"ban","emotion":"fear"},
{"id":"ytc_Ugy-Xe9R_F7y80-WWm54AaABAg","responsibility":"ai_itself","reasoning":"mixed","policy":"none","emotion":"mixed"}]