Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Why don't we make AI have no access to skin color. I feel like high crime rates …
ytc_UgxMYmJwY…
G
3.7 billion years ago, that little bastard crawled out of the soup and now I hav…
ytc_Ugyvl_zqj…
G
Some people have very obviously started using ChatGPT to generate answers to r/a…
rdc_j8b7b5t
G
Points are 1. Cameras are not good enough for autonomous driving. 2. Accident…
ytc_UgyXM95yD…
G
This shit is as hilarious as it is stupid, meta-data exists for nearly every pie…
ytc_Ugzn9I5sT…
G
Can you fix my regex, write this in bash for me, tell me about posix, and then i…
rdc_munmkuc
G
The video is interesting but, as someone who is working on developing AI tools, …
ytc_Ugy1onCVK…
G
Honestly, the only thing ChatGPT was good for was porn. Then they put in all the…
ytc_UgwxLJU7K…
Comment
To jump on a point about "things that large language models do that we can't code by hand"... I wrote a chatbot a long time ago that could make "jokes" and seemed to "understand parts of humour". It could also do haikus Didn't take a whole datacenter either. Iunno, it's really tempting to attribute any LLM joke as the delusions of pattern seeking animals projecting humour onto what amounts to rolling the dice on a big lookup table of words... Sure somtimes it says something surprising and people will perceive that as "funny"...
It's also really tempting to comment on the efficiency question. It takes so much electricity and space and microchips to make a computer that *kind of* seems like it's talking, and comparatively little fuel and resources to, you know, create and maintain human brains that currently do the job *better*. Not to say that strong AI or superintelligence is completely impossible, but I do wonder if a bunch of digital switches is maybe not the wrong design for the machine that could actually do it... Especially as the richest men in the world seem to have no interest in keeping up the infrastructure and international cooperation and all the rest that will be required to actually acheive this by brute forcing it with digital electronics.
youtube
AI Moral Status
2025-11-01T19:1…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | developer |
| Reasoning | mixed |
| Policy | none |
| Emotion | mixed |
| Coded at | 2026-04-26T23:09:12.988011 |
Raw LLM Response
[
{"id":"ytc_UgwqDZPwS0sJhzustSl4AaABAg","responsibility":"government","reasoning":"consequentialist","policy":"regulate","emotion":"approval"},
{"id":"ytc_UgwfVIgjc9RUVbtK2Yx4AaABAg","responsibility":"ai_itself","reasoning":"mixed","policy":"none","emotion":"outrage"},
{"id":"ytc_UgwBkZ0RB2dzvKO0Wc54AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"ytc_UgyRx8kIRspv6bsRE4J4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"indifference"},
{"id":"ytc_UgxDgzcIUZXZuAzgHSR4AaABAg","responsibility":"developer","reasoning":"mixed","policy":"none","emotion":"mixed"},
{"id":"ytc_UgzYdePcFg5OXhfaaV14AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"fear"},
{"id":"ytc_Ugz8uz5IUjT5JCw33wF4AaABAg","responsibility":"none","reasoning":"none","policy":"none","emotion":"approval"},
{"id":"ytc_Ugy3sz23nrUfIxdlCFJ4AaABAg","responsibility":"ai_itself","reasoning":"mixed","policy":"none","emotion":"indifference"},
{"id":"ytc_UgwnojViKzl0G8CMj794AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"resignation"},
{"id":"ytc_UgyH4hSorqWq8zxU7AN4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"}
]