From the Art Department
First off, for anyone who needs a primer on AI Generated Art here's a concise video on the subject to get you up to speed: LINK
I used Midjourney I started out with these prompts to see what I would get: Astronaut with a skull head +skull + skeleton + red spacesuit + character design + full body + red + sci-fi + star wars + Ralph McQuarrie + Jake Parker, cartoon, cartoon network, adventure time style:
Thought I'd mix it up and see what prompts like octane render + 3d would give me:
I didn't like what I was getting so I went back to my original prompts but added rubberhose animation + pixar + disney + 3d:
The results were better, so I asked for more variations. When I got these I decided I was getting diminishing returns and called it a day. Lots of interesting ideas to put into my design, but no one design really felt like it nailed the vibe I got from my original design.
After this experiment here's what I think:
1) AI Art Generators will only become more powerful/capable. So wishing it away is a waste of time. The only path forward is figuring it out how to implement it and how to regulate it.
2) With revolutionary technology comes a reorganization of status and power. The status quo HATES this and will do almost anything to stop it from happening.
AI Generated Art shifts the power (and wealth) of creating images from people who have training, to people who don't.
This reorganization happened in the 2000s when programs like Maya, 3D Studio Max, and Photoshop made art creation a lot more accessible to people who couldn't paint traditionally, or sculpt clay.
It allowed places like animation studios to be havens for creative people to make art who might not have been able to draw really well. Which really upset people who had trained to animate in 2D on paper, and who studied classical painting techniques. Which leads me to 3:
3) Not everyone who is creative can make art, and not everyone who can make art is creative. The creative people who could also adapt and learn new tools absolutely thrived in the new digital art world.
A lot of the art I've seen generated from AI is a lot like hearing someone impersonate English but who doesn't know the language. It sounds right, but they aren't actually saying anything.
4) I see these AI Art generators as tools. Another resource for creative people to add to their toolbox to make them even more creative. Or at the very least, make their job easier.
5) AI isn't an end to end problem solver for productions. There's still a needs to be an artist to translate it into something usable. Someone needs to interpret AI art into something a modeler can model, or set designer can build.
Example: After a producer plugs a bunch of prompts from a script into Midjourney they take it to the art dept. The crew gets a brief from an art director and instead of a lot of back and forth, the art director points at a page of AI art and says "Make it look like this"
6) Questions I’m still thinking about:
- Do these AI Art Generators actually undermine illustrators, photographers, concept artists? Or does it actually elevate these industries?
- Is it bad to democratize something like art creation?
- Who truly benefits from this shift in power? Where is the money flowing to?
- Should artists have the option of their artwork being removed from the AI generator's databases? Or is any art you post online free game? Does the AI generator do anything different than what an artist does who has strong stylistic influences from other artists?
7) I'm still learning about this, and still reading up on all the pros and cons.
I would love to know your thoughts. We've been discussing it over on the Discord for a couple weeks now: LINK
I also posted this on IG and it blew up. I could not keep up with the comments. Over 750 of them! If you want to get a vibe check on what the broader art community thinks of this check it out here: LINK
-Jake