[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/ic/ - Artwork/Critique

Search:


View post   

>> No.7000987 [View]
File: 165 KB, 1098x710, Screenshot 2023-10-04 173415a_proca.jpg [View same] [iqdb] [saucenao] [google]
7000987

>>7000944
>you are using this as an argument that you are learning by using AI
lol......no. i don't know why you would read my post that way.
i'm telling you the simple truth that AI is learning from its training data. you're so misinformed and confused when it comes to AI that you can't even understand a simple english sentence.

when the AI takes in a thousand images of circles, it doesn't "learn" a thousand circles. instead it learns what circle GENERALLY is like. the concept of it. and that representation is what it then uses to create its images. and the same happens for every concept the AI learns, going from "cat" to more complex ones like color relationships and reflections or lighting.
the AI is not human, but the AI is learning despite it. this is what makes it work in the first place.

i find it funny how even now, you people cannot read this simple sentence as it is without adding some random baggage on top of it due to your own lack of understanding.
when i say that the "AI is learning from its training data", i mean literally just that. not that it is human, not that it learns exactly like a human, not that i am learning, none of that.

this is not some cute approximation, or some justification, it is just how the technology works.

>>7000975
>It's more like googling for an image that perfectly matches your lines in a universe where every image that could exist has already been made, and you can find more of those images by modifying the input(the search).
>in a universe where every image that could exist has already been made
but they haven't been made, so your analogy isn't quite the same.
those images couldn't even exist at all without said line input.
and lines weren't my only input. you can use your whole image as input, down to the % of it that you allow to change.

>> No.6930719 [View]
File: 165 KB, 1098x710, 1697046150114374.jpg [View same] [iqdb] [saucenao] [google]
6930719

>>6930492
>>6930323
do you ask people for consent when using their images for reference?
do you ask people for permission when studying their art?
no?
then know that what the AI is doing is more vegue than even that.


>>6930492
>If EVERY non-consensually scraped artwork were removed, it would cease to function in an impressive manner where art is concerned.
i already adressed this in >>6929707
TLDR
maybe you think people are being facetious when they say AI is "learning".
but i'm telling you, we're not. we literally mean it.
i don't think there is any point at all in arguing until you at least understand your opposing viewpoints.

>> No.6904095 [View]
File: 165 KB, 1098x710, 1676291769223440.jpg [View same] [iqdb] [saucenao] [google]
6904095

>>6904061
that video is made by a person who has no real understanding of AI either. only someone who pretends to read a paper and then puts all his biases on top of it anyway.

and tbf none of it is easy to understand. even i am only slowly gaining a better understanding, partly because i'm trying to get better at explaining it. my earlier analogies were way worse than what i tend to use now.

>As is the human intelligence tasks that are being performed by economically destitute third world slaves.
they only label images. if they are actually underpaid then that's an issue. but not one directly related to whether AI art is ethical...
(and i read that they aren't underpaid relatively to where they live anyway.)

the AI sees the patterns on its own. without human intervention.
it's like you and me seeing a bunch of human faces and realizing that all of them are oval from the front. these kind of patterns, among many many others and much more complicated ones.

>It is disingenuous when you intentionally and repeatedly ignore facts that are being provided to you and make analogies that are based on misinformation.
sorry? like what? any examples?
it's very rich when i hear my arguments thrown back at me from people who can't explain a single thing about how SD works.

Navigation
View posts[+24][+48][+96]