Dec 2022
1 / 3
Dec 2022
Jan 2023
Mallory
Dec '22
1

Hey friends, I know this is a hot topic right now, so I wanted to jump in and see if anyone was interested in a quick collab to address the most common argument being used right now to discount the art theft of AI image generators. The argument is that AI image generators scraping art is identical to artists using references, and I feel like pretty much every artist knows that it simply isn't true. Referencing is the act of collecting information through limited means (sensory input), filtering them through your world-view, perspectives and knowledge, and then creating an output that is further constrained by your technical expertise, physical ability, and also artistic goals and preferences. Image generation, on the other hand, relies on distilling art to a mathematical equation, then averaging the outputs of that equation to create the final result. Referencing is interpretive, while AI image training is derivative. The test for this, of course, is that you can still make art without references, but AI image generation would not exist without original art.

I wrote a rough script that is meant to be 10 square pages long (with a cover image, and then 9 "story" images containing some text and about 2-4 diagrams), which we can of course alter. It's using the humble smiley face as the reference art, for simplicity's sake.

I'm hoping that we can make it together in a way that would be easy to share and disseminate (online or even in person via printed pamphlets!). Please note that I'm not really trying to debate the philosophical and existential problems with AI art, I think that's a whole different topic, and I also personally think that ethically sourced and implemented AI image generation can have legitimate uses in the creative process. I just want to address the ethical and intellectual property theft concerns in a way that is hopefully illustrative and not completely confrontational.

The script in question: https://docs.google.com/document/d/1JGcbfKJW3r1hYwXmjT68hZGva730dG9iMkFPr4JtQp4/edit?usp=sharing4

I was hoping to make something that looked like a sketchy infographic (something like the bottom images here1), so I'm not expecting this to be a huge time commitment from anyone involved. Ideally we get one artist (doesn't matter what your medium is) to commit to each of the 10 pages so that we can each spend at most about 30-45 minutes on this. If we don't have enough artists, we may need to double-up (and if we have too many, we can create different versions of this comic, or have people collaborate on the same image... um, let's cross this bridge if we get to it, haha)

Let me know if you're interested, and of course, invite any interested friends! You can claim one of the pages or just say you're interested and I'll assign the remaining pages to the people who don't have a preference. I wrote down some visual suggestions in the script, but they can be completely re-imagined if you have an idea that works better.

(For anyone who doesn't know me, I'm a digital art educator/technician with the public library, kidlit illustrator and also the artist/writer for a webcomic here on tapas called Engram. I'm also someone dealing with a complex neck injury that prevents me from being able to draw some days).

TL;DR:
- Looking for peeps to collab in an informational comic about referencing in art vs AI scraping art
- Hoping to collab on a total of 10 sketchy infographic pages with text and a few images (let's say, 3”x3” at 600 DPI to be safe) - whatever time you wish to commit to this
- Art medium and skill doesn't matter as long as the final art is clear in what its trying to communicate
- I have written a script (see link4) and will also contribute artistically wherever necessary (portfolio here1).
- Looking for any number of artists to collab!
- Deadlines - I don't really want to impose any, but I feel like the sooner we get this going, the more fresh the discourse will be.

Last thing - if you wish to contribute to this project, but want to debate (or simply alter) some of the points, I'm all ears. But I'm not really trying to make this into a discussion on AI image generation, so if you don't want to contribute to the project but rather want to debate the topic, I ask that you find or create a separate thread for that. Thank you!

  • created

    Dec '22

  • last reply

    Jan '23
  • 2

    replies

  • 519

    views

  • 2

    users

  • 4

    likes

  • 3

    links

1 month later

closed Jan 16, '23

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.