Our stance on AI Art

I’ve been thinking a lot about the AI art discussion. I joined AI art groups as soon as I saw Midjourney results, I used numerous generators early on to try and understand this new tool. Even from its early stages, it was clear that we were seeing a total world shift in art. AI tools (for images, audio, code, and more soon in sure) will continue to get better at an exponentially rapid rate, I feel that’s just the reality at this point that we’ll need to adjust to.

I’ll start by saying that I believe removing barriers to entry for creative endeavors is a good thing, and AI Art is now part of that. I believe that AI Art could be an accessible and beautiful medium that allows millions of people to finally get a creative project off the ground.

As it stands I have two issues with AI Art and image training models:

1. these models were trained on a huge database containing protected works, medical records, and even explicit revenge images, without artists’ or owners’ knowledge or consent. Monetizing access to learning based on those images without compensation doesn’t sit well with me, and granting general commercial license (like Midjourney does for example) creates a secondary revenue stream yet again built on the backs of artists without their consent.

2. Artists do not currently have a reasonable way to opt out of the training model. I understand fair use regulations and the comparison to fanart, but this isn’t the same, because it’s not a person making the images (the promoter could be compared to an art director, not the artist), and the threshold of deviation allowed by seed images makes it incredibly easy to abuse. Artists deserve the opportunity to say “no” and opt out of the learning models.

Musicians aren’t being treated the same way as visual artists, those ai models are only being trained on royalty free/public domain works, and in the words of Stability “because diffusion models are prone to memorization and overfitting, releasing a model trained on copyrighted data could potentially lead to legal issues.”

I feel that visual artists aren’t being afforded the same courtesy because they don’t have the same sort of powerful centralized industry that musicians do. This tells me that Stability knows that it’s probably unethical, but that they just aren’t afraid of the repercussions.

The more I’ve learned about AI art the more I’ve come to accept that it’s here to stay, but until there are ethically sourced databases that can compensate artists, and a reasonable means of opting out of training models, I will be saying no to AI Art. We will not be using AI Art in any of our projects here at Witchway Games, I will not use AI Art in my freelance work, and we will prioritize supporting projects that celebrate human creators.

Stay gritty, this feels like it will be a difficult road.

-Eric Streed (Creative Lead at Witchway Games)

#createdontacrape #notoaiart #noaiart

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this:
search previous next tag category expand menu location phone mail time cart zoom edit close