AI art will not steal artist’s commissions (yet)
Art is art (is art is art is art).
In the most recent wake of AI generators not only getting better at what they are doing but also being available to millions of people for free, I was curious enough to take a deep dive.
I have always been a creative person, you see. I started dabbling in art at a very young age (meaning intentionally, not just doodling like most kids do anyway) and kept going. Given the cost of tutoring, I had to teach myself. Without the several free tutorials that are currently available, it was a tedious and arduous process with only books as resources.
It was painful. I hated myself more often than I thought “this was a good attempt”. Not having a lot of money didn’t really help because I never knew how to improve, I had no one in the family to ask.
Purchasing any art supplies was difficult and frequently led to the purchase of materials that were of lower quality. Which wasn’t too much of a problem, but it limited the way how I could experience art.
When PCs became more capable of processing, I was able to obtain a cheap but still decent graphic tablet. It was a no-name product from Aldi, and I felt like the world would open up to me. Finally, I was able to create art without wasting anything.
Still, digital art was still in its infancy, nothing we see nowadays. And something I heard very early on was “Digital art is not real art”. If I had gotten a penny every time I heard that sentence, I would probably drown in art supplies now.
For me, art was nothing more than a hobby, I never really wanted it to be a career because I learned early on that I am not really suited for a career as an artist. There were many reasons. One was that I didn’t really have the time for it. Nor the patience. One other was that at times I just lose interest. Sure, it comes back, but basing a living off of it?
For a project, I tested out Canva’s AI generator and was interested enough to take that deep dive into what is currently possible.
There are several tools out there, several services that allow you to experiment with AI-generated art — at least for a bit until you have to pay. I tried to evaluate my options and found out that there’s even an open source project — and man, I love open source projects!
Sure enough, you need a graphics card that can handle it, but else from that, there’s not much hindering you from using Stable Diffusion.
And I feel like I have come full circle!
I am back to where I discovered a technology in its infancy and am extremely invested! I just love it. From any kind of perspective.
But it sure has its haters. Now, instead of hearing that digital art is not real art, I hear “AI art is not real art” and saying these voices are entirely wrong is actually wrong. But we have to look at the bigger picture.
So, what actually is AI-generated art? To describe it in my own words and to the best of my knowledge, it is a picture generated by words that define a direction. It is diffusing words into an image. It is an algorithm that was trained on millions of pictures on the Internet and can create images that may resemble someone else’s artwork but are not the same. They are ideas having taken shape but not form.
To explain the process I added the images to this post to show how such a picture can come into existence.
It’s by no means a “bad” picture, meaning the quality is good, you can recognize what I intended, but does it also have artistic intention?
I would say no.
Since I have to enter text (which is called a prompt), this is what I came up with in the end:
tree with colorful chinese lanterns, detailed swirls of colour by huang guangjian and greg rutkowski and ghibli, cyberpunk city, insanely detailed, volumetric light, digital art, intricate, ornate, octane render, unreal engine, photorealistic
Don’t worry, I am not going to take you through the entire process, but I posted my prompt for you to understand the difference between what I described and what I came up with.
Especially in the versions that I posted here, you can see that there’s a tree standing on a tree. In one of the first versions it looks more like the tree just grew thicker again, but especially in the later stages, it looks like there are two trees. That was definitely not my intention and I had to go in and correct that step. Something else the AI decided for me was the lighting, the coloring, the positioning… something an artist may intentionally choose to convey emotions, atmosphere, or even opinions or political meanings.
Intent. That is what AI-generated art can not do. As an artist that does more than just paint pretty pictures (which AI does at the moment) you learn why you insert dramatic lighting, when to use the rule of thirds or the golden spiral, or all the other things. That is by what we define art or should in future because that is where AI generation will take us. It may not yet be a medium which we take into our fingers and sculpt, but I think we should more see it like a camera through which we observe something and tell our own story.
