Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
That's an interesting thought! The future of AI and robots like Sophia is defini…
ytr_UgwchRPUI…
G
@toomylight2311Perturbation too.
There is no justification, no education, no we…
ytr_UgyRbkajd…
G
At the end of the article it mentioned that Facebook suspended research into Cha…
rdc_dlhan4o
G
It doesn’t matter whether or not we agree that developing AI is ethical. Someon…
ytc_UggA6P1aB…
G
I agree with the Doctor when he says we "don't need super intelligence." I perso…
ytc_UgznlOnMI…
G
When ai art got popular, I tried to make some ai art with a bunch of different a…
ytc_UgwQdYMDK…
G
AI 'artists' can go rot in a hole for all i care. Not only do they ruin the envi…
ytc_Ugy0BsJCf…
G
I still don't get why so many people want a robot to sound/act human it just see…
ytc_UgzSdM6ql…
Comment
Interesting video. Three points stick out for me.
1. Correction at 6:40, the AI does not directly take items from the data set, and actually cannot recreate existing images save for items with over-defined boundaries - the overfitting mentioned. Imagine showing a person who had never seen a giraffe a single cartoon giraffe every time they asked what it was, and then asked them to draw the giraffe. It would probably look like that cartoon giraffe. Even then, there are a host of different factors which make exact replication nigh impossible in all but the most tightly defined concepts ( famous pieces, celebrities, pop culture icons, etc.)
Remember, it is not using the art from the training set at creation. It is a common misconception among artists that AI collages existing work, but what the training data is doing is refining boundaries around concepts within a multidimensional latent space reflecting all possible images.
It is easier to think of it like a 2d graph of all possible pixel configurations for an image, and on that graph there's a lot of random junk from random pixel placement, but also every image that ever was and will be. The AI uses the training data to draw circles around parts of the graph where certain words can be found.
2. An artist's style is not protected by copyright, and for good reason. We want people to learn from other artists, and we want them to create without fear that someone already worked in a style. Hell, we want people to copy their favorite artists, because that is how they learn and eventually experiment with their own styles.
The problem is we're not the only ones learning now, and knee-jerk reactions are leading to some artists calling for art styles to be covered by copyright...which probably will kill art, as we'll see corporate interests consolidate styles like they did intellectual property. Ironically, that would lead to a future where only a few corporations would own most styles, and would probably generate art anyway because they own the styles and it's cheaper.
3. AI wants to take your job in art. AI wants to take your job in music. AI wants to take your driving, piloting, and construction jobs. AI wants to take all your jobs. This is a good thing!
The only way we move away from people requiring jobs to survive is to make it so most jobs don't need people. Instead of creating art for money, artists could create art just to create. If no one needs a job, people can be free to pursue any job they want.
youtube
Viral AI Reaction
2022-12-24T19:1…
♥ 2
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | unclear |
| Policy | none |
| Emotion | mixed |
| Coded at | 2026-04-27T06:24:53.388235 |
Raw LLM Response
[
{"id":"ytc_UgxDLNG3AfNxSwTwd2V4AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"ban","emotion":"outrage"},
{"id":"ytc_Ugy64dVdg4WJ9o7NGNl4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"resignation"},
{"id":"ytc_UgxYb5YK7ggnlb-mAd94AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"liability","emotion":"fear"},
{"id":"ytc_Ugw4IA2HxdADWudPHep4AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"none","emotion":"fear"},
{"id":"ytc_UgxFlE3XIIertr7pRed4AaABAg","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytc_UgwGpo5InXBYBbRlZYZ4AaABAg","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"approval"},
{"id":"ytc_UgwToVINBlnxxs6kFn94AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"fear"},
{"id":"ytc_UgzUp5VDb3fkBTB-lo94AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UgxrhcdxTZ9bM6ZppG54AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"mixed"},
{"id":"ytc_UgwVr5BKCot6gJEgaex4AaABAg","responsibility":"user","reasoning":"deontological","policy":"liability","emotion":"indifference"}
]