#(i say as if i didn't mostly trace these over the 3d models i made in daz lmao)

[wishbonemotel]

More OC references

[pizzaronipasta]

I'm sorry to drag back in the ai discourse, but this one thing has been bothering me: I think that accessibility by making art programs more user friendly would be the big impact here. Finding ways to work with people and make the process overall much easier to navigate and not as physically demanding would be huge for all users. It would also allow for new mediums for people to choose to engage with at a pace and focus right for them.

I also will say, I think AI would work best as a tool, something to maybe help generate an idea, or to guide in the right direction. My anatomy sucks, so I often use a 3D modeling program to angle the camera after a pose a blank base so I can achieve dynamic movement. Tracing as an early artist is sometimes encouraged, because you have to know what exactly you're drawing. All thats required is the acknowledgement of what into that work, and that it's a tool being used.

If you have put keywords into a computer and it produces an image, you can't trace what made that exact image, unless you put art in for it replicate. In which case, it's based on someone's work heavily, and credit is due. If you write an essay about the dangers of milk consumption, and even cite sources and quotes some of your information... but most of your paragraphs are paraphrased from other people's articles on the same subject, you didn't really share your own thoughts. There isn't a discussion on what you think, the way you see the world, because the bulk of that product is someone else's words and the ideas that are yours aren't allowed to shine.

And that's why I can't really consider typical AI generated images, just by themselves, as art. It's a picture, and it can look nice enough, but I would rather see what innovations someone can add to that picture to make it their own. You don't have to be limited to traditional artwork, you can make through whatever medium works best for you.

So, at this point, I'm wondering if I'm just poorly reading and misinterpreting what you meant, and want to clarify:

Do you mean AI art, as in AI generated images then edited significantly by humans with the AI image and base product, or AI just by itself?

Because the thing about art is that it's definetly building on previous ideas, there's a reason "steal like an artist" is a popular phrase. But the difference is that AI doesn't tell you where it's sources are: therefore you can never credit them. Generative AI is based on an algorithm, but it doesn't tell you what that algorithm is coming from. When you learn how to draw over time, you're influenced by your memory of what things sort of resemble, but it's almost never exact, and thus even when you make something mimicking that piece, you default to your own movements you've learned. If you try to draw the Mona Lisa, you'll end up drawing your own stylization, with the way you've taught yourself to draw eyes, the way you've learned to render color, and backgrounds, ect.

I don't think people should have to put themselves through physical stress or harm to create art. What I want is for art to be more accessible so more people can express how they feel. Generative AI can only express a few sentences of an idea, but not the deeper thought process and emotions and experiences, and because of that I feel it fails as art.

Thank you for your time.

I'm having a little trouble understanding your perspective on a couple of points, so I don't think I can adequately respond to the whole thing. Feel free to elaborate in further asks if there's anything you want me to understand better. However, I do feel that I can provide my thoughts on a particular detail you brought up. You say that some of your problems with generative AI come from the fact that it doesn't cite sources. I would argue that an AI's sources are too diluted to warrant or even enable conventional citation, considering that it's mostly unknowable what specific influence any given source would have on the program's output. Barring exceptional circumstances, AI doesn't really draw from existing works in any remotely straightforward way unless specifically prompted to, in which case the user is responsible for that decision, not the computer.

More importantly, however, I must stress that AI training is comparable to human learning. No human is able nor expected to credit every source that contributed to their every habit. When making art, we are drawing inspiration from literally everything we have ever experienced in our lives in at least some small way. Yet we only consider credit to be due for the most direct influences. I don't see why AI should be held to a different standard.

To answer your question, yes, I do consider both edited and unedited AI-generated images to be fully capable of being art. In general, I have a maximalist view of what counts as art.

I hope I haven't misunderstood anything you were trying to say. If I have, please correct me. Thank you.

[oekaki-chan]

hi clef!! i first stumbled upon your art on twitter via a comic tutorial you posted. i finally decided play AA a month ago and remembered about your art and thankfully found your account again. After also finding your tumblr,, its just. wow. going through people's art tags is always a journey, especially when theres stuff in it that are older than a decade, haha. i just wanna say that your improvement over time motivates me to keep drawing, as ive been recently studying forms and shapes in other people's art, and yours is definitely one of the best examples. As im taking a gap year to work on a portfolio for college, i finally had the proper opportunity to start studying art as i wanted.

i mainly wanted to ask- when drawing, what types of references do you use for character poses? they always seem really natural but also have that "homemade feeling"- as in applied really well into an art style. The way you draw expressions is also really immaculate! I was wondering (if its alright with you), whether if you have tips on that too?

thank you regardless, and i hope you have a great day!

I hope this doesn't come off as bragging because honestly when I'm drawing poses I almost never use references 😭💦 During my art journey I conditioned myself to draw without references because my parents didn't allow me to have internet access until I went to high school, it was hard to get references for drawing so it was mostly just "frick around and see what happens", I just tried to project whatever I have in mind into the canvas 😂 Instead of training my eyeballing skill, I ended up training my muscle memory and the way I imagine poses in my head (that's why I'm a bit bad with realism drawing because I have bad eyeballing and real-time observation skills).

I do use references sometimes when the pose is a bit hard to get because it has many foreshortenings and hidden parts like this Diavolo doodle I made, but I studied the reference instead of tracing it so I can get a better gist of how things work in case I'll need to draw the same pose again in the future. (Also I hate that whenever I try to find some references for a specific pose google suggested me porn sites instead zzzz)

But I'm not saying using references is wrong, not at all 😭!! It's just my way of drawing! I'm lazy and get distracted easily once I'm scrolling the internet pages, it's also why I don't usually use 3D models in CSP because trying to get the right pose with that may take a while and the outcome isn't really worth the time because it often looks stiff, weightless, and unnatural.

And same goes for expressions! I think being a (somewhat) avid manga reader extends my visual library for expression (and um, I think drawing steamy stuff helps too because you can practice drawing exaggerated expressions and poses, just make sure you're not drawing the same type of expression and pose/angle over and over 😳💦)

I hope this helps and I'm sorry it got a bit long 😭😭!!