Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
The first thing is that they should not make a single robot in humanoid form, th…
ytc_UgxM728Sp…
G
I am not an expert in the slightest and may be missing this point already being …
ytc_UgxIwDCQQ…
G
ironically, by saying being against ai is ableist, they're the ones being ableis…
ytc_UgytZnCiv…
G
Quick question, as an artist, i know this sound pretentious, but i played around…
ytc_UgzdOnidG…
G
This was good, though I'd have liked a bit more attention put to the fact that s…
ytc_UgjWA5kpI…
G
6:14 so he’s lazy? Which’s fine, but like, just say you’re lazy instead of this …
ytc_UgzOmRCfw…
G
the next step is to invent the AI consumers....
😂
Otherwise it's going to be the…
ytc_Ugx6rLK07…
G
No it is. Very healthy. To recognize the respect real artists deserve and not be…
ytr_UgzyrmyZa…
Comment
As a professional software engineer who uses AI tools daily — from autonomous agents to integrated assistants in code editors — I find it hard to take seriously the claim that a future AI couldn’t be shut down. These systems ultimately run on physical servers powered by electricity. From a software engineering perspective (even without being an AI researcher), it seems extremely unlikely that there would be no way to disable or dismantle a system we ourselves built. Kill switches, air-gapped environments, power cutoffs, and hardware destruction are all trivial fallback options.
There’s also a huge leap in assuming an AI could somehow “escape” its containment by magically rewriting the entire internet stack, bypassing decades of hardened protocols, operating systems, and network security. Yes, we’ve seen concerning incidents (like the recent Anthropic research showing models can be prompted to take unauthorized actions when given tools), but that’s still a far cry from an AI autonomously breaking out of its sandbox and seizing control of unrelated infrastructure.
youtube
AI Jobs
2025-11-20T13:2…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | consequentialist |
| Policy | none |
| Emotion | indifference |
| Coded at | 2026-04-26T23:09:12.988011 |
Raw LLM Response
[
{"id":"ytc_UgyoQSVvBHxNAec2TSp4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugzw6-JORR53HykCMfN4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"ytc_UgyMJXdagIBgbscKKqF4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"ytc_UgxYaYceVlfdfeIkERN4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"},
{"id":"ytc_Ugx-E_UybuePPqCRVF14AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UgxmYO49m6g4zp3Tvq94AaABAg","responsibility":"government","reasoning":"deontological","policy":"none","emotion":"outrage"},
{"id":"ytc_Ugx4AliBBBEkUc5e9wF4AaABAg","responsibility":"developer","reasoning":"virtue","policy":"ban","emotion":"outrage"},
{"id":"ytc_Ugy8vckkVQrnHS1BERl4AaABAg","responsibility":"government","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytc_UgxaEsdV4nW830veEgh4AaABAg","responsibility":"government","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"ytc_Ugxo0fZYqVQCY8s94-N4AaABAg","responsibility":"developer","reasoning":"virtue","policy":"regulate","emotion":"outrage"}
]