Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
AI sucks like, why should a lack of yours be taken from me and shared with the v…
ytc_Ugz2Kh79e…
G
Remember, with self-driving cars, you will always be asking permission no matter…
rdc_ebuiz8u
G
Ok i think china would be a really bad choice to start payments through facial r…
ytc_UgykAmHRV…
G
Here's an idea. Ban self driving if a human can drive it. If an accident happens…
ytc_UgwRdvXW3…
G
@Sydney88383 I did and he does not want that AI takes the useless jobs that nobo…
ytr_UgwICWyqj…
G
My ai keeps on giving me incorrect answers after some abuses it gives me the cor…
ytc_Ugz9bQXWx…
G
It’s odd growing up I really disliked art bc I sucked at it. I also thought the …
ytc_UgzBie3mN…
G
i'm a begginer artist and i learn how to draw rn
in my perspective ai is a tool …
ytc_UgwF4ComO…
Comment
**Where the model runs** ***is*** **the workflow risk.**
When you're running a model locally, *everything* stays on your computer.
No cloud providers possibly rifling through your chats, no servers in the middle of your traffic to intercept anything, etc.
Not sure on your threat model, but locally hosted LLMs prevent *any* data getting out unless you specify it to.
Just *don't* use something like OpenClaw or some shit.
Stick to ollama if you don't know what you're doing.
Unsloth studio is "better", but a bit more wonky to set up.
llamacpp is the "best" (since everything is just running llamacpp under the hood) but you need a frontend too technically.
**But yeah, just grab ollama and try something like Qwen3.5\_9B.**
See if it's something that even works for you.
reddit
AI Surveillance
1777003826.0
♥ 2
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | user |
| Reasoning | consequentialist |
| Policy | none |
| Emotion | approval |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_d6t2tli","responsibility":"unclear","reasoning":"unclear","policy":"unclear","emotion":"unclear"},
{"id":"rdc_ohyau3x","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"rdc_oi0tpi5","responsibility":"none","reasoning":"deontological","policy":"unclear","emotion":"outrage"},
{"id":"rdc_f8so31j","responsibility":"unclear","reasoning":"unclear","policy":"unclear","emotion":"fear"},
{"id":"rdc_f8tluf4","responsibility":"unclear","reasoning":"consequentialist","policy":"unclear","emotion":"approval"}
]