Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Did no one watch minority report? Jesus Christ why the actual hell is predictive…
ytc_UgyuIC4Mk…
G
Trash union propaganda. Screaming “Stop!” In the face of progress. No one is s…
ytc_Ugx2gQN_c…
G
I'm not really sure what the relationship is between bias and variance. Sometime…
ytr_UgzyEE3xv…
G
Why do we even need AI cameras? Remember that one guy who dresses up like a poli…
ytc_UgyyLFWbR…
G
The difference with digital art tools is that if you have an adobe licence, your…
ytr_UgwXGmMoz…
G
if god exists then we are robot built by him. if god doesnt exist then the diffe…
ytc_Uggut43AX…
G
I hate change and have so many routines with ChatGPT. Started working with Claud…
rdc_o7w9uo5
G
i want ai that can keep filth like porn and gore off facebook and youtube. I wan…
ytc_Ugw6S6ic3…
Comment
Note: Just wanna clarify, people faking that one prompt they did is real art they went and made is infuriating and stupid. Wanna put that first. Buut...
The tool section here kinda ignores mass amounts of control and work that *can* actually go into images. Stuff like ControlNet that lets you have more control over how the AI does things. Which lets you give it an image and ask it to say, texture some walls in the background for you. The best example I can think of is people adding different textures to CSGO maps with a prompt. That's madness.
Or something like img2img where you feed the AI an image, ask it to tweak it and go from there. You might draw a layout and ask the AI to colour it in and add something. Or you might ask it for a concept of something you're doing. But want it to tweak it so you bring it to an inage editor and mess around.
A lot of really cool "easy" AI art can take a lot more than just a prompt. You might do a layout and put it through img2img, you might edit that img and put it back through and then go back over it in full adding your own changes. Or you might take a drawing you made and want to add some details with ControlNet to specific areas.
Is it infuriating that people do a single prompt, get one of the pretty good random results and say that they made it entirely themselves? Yea.
To only focus on that disregards everything that you can REALLY do with it though. Cause there's a lot.
youtube
Viral AI Reaction
2024-09-16T18:3…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | user |
| Reasoning | mixed |
| Policy | none |
| Emotion | mixed |
| Coded at | 2026-04-27T06:24:53.388235 |
Raw LLM Response
[
{"id":"ytc_UgwB1WHEP8kdiembi094AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"resignation"},
{"id":"ytc_UgyUZnXgQ4AJCaMipOV4AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugw3Xoun1QJaunVtHrp4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"},
{"id":"ytc_UgwevkqJcGwLwhm-jYZ4AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"disapproval"},
{"id":"ytc_UgxHChU_buzrQZzORcV4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"mixed"},
{"id":"ytc_UgxXolyuyJuEGphB-Wx4AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"outrage"},
{"id":"ytc_UgwaaxLqXwKd9VXnvol4AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"indifference"},
{"id":"ytc_UgwoV17TK_wtrd2XK0p4AaABAg","responsibility":"user","reasoning":"mixed","policy":"none","emotion":"mixed"},
{"id":"ytc_UgwqQHIx-4eCCZfyV1Z4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"approval"},
{"id":"ytc_Ugzjq8xvvZfc3EmfYsV4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"}
]