GenAIArt Rules of Thumb
- Aug 16, 2025
- 3 min read
Why we need Rules of Thumb
It's easiest to start all of this by thinking of Art as a form of Communication that just so happens to involve pictures.
Like any other form of communication, there are right ways and wrong ways to say what you want to say. Just as spoken/written language has things like adjectives, nouns, and verbs - the visual language of Art also has fundamental building blocks that determine how to construct something comprehensible.
The visual language of Art is organized into Elements, Principals, and Techniques based on things like the biology of the eyes, neurology of visual processing, the psychology of how we g r o u p information, and the sociology of symbols <3 combined with media literacy <3 <3 <3
In Fine Art and Social Media Posts you can use those fundamental building blocks to communicate anything. But in Commercial Art you have to communicate what suits the business objectives of your employer using the technological and financial restrictions they operate under.
Commercial Artists work with the understanding a piece of Art can fail to communicate what it was supposed to with the resources it was given. The eternal struggle is that oversight on a failure of resources is easy, but oversight on a failure to communicate is hard. In videogame production, it's challenging to make it systemic and scalable. I can write tools for days that can determine if an asset is using too many resources, but there's no script in the world that can determine if an asset failed to communicate.
Furthermore, on a individual level Artists really struggle to quantify to themselves and to upper management when their work is failing and by how much. It's hard to know for sure if you're successfully communicating without asking the people you're communicating with, and that's expensive Market Testing. So on a practical level, it falls to the Art Director to be the Arbiter of Failure.
This is unlike Engineers, who can verify if code runs and integrates successfully on an individual contributor level in concrete 'pass' / 'fail' terms. It's closer to the kind of challenge Game Designers face when trying to verify if gameplay is Fun, except they wouldn't be able use their own playtesting to see if they're going in the right direction.
If Artists don't have an easy accessible way to measure and verify their own work, they definitely don't have a way to do that for GenAIArt. So working with GenAIArt on a case-by-case basis the way Engineers do with GenAIText, can be unrealistic at best, and at worst a bottleneck.
Instead, I'm proposing it's wiser having some Rules of Thumb to guide when and where to use GenAIArt in the first place.
I'm going to organize my suggestions based on the communication goals Art Directors I've worked with over the years have had. There may be more kinds of Art Directors out there that I haven't worked with yet, but Rules of Thumb are experience-based and this is the experience I've got.
Part3: If Game Art is Storytelling; -------
Most Art Directors I've worked with are some combination of these, but do tend to prioritize one when push comes to shove. The Rules of Thumb I'm suggesting aren't mutually exclusive, so they can be used together. This is written for Video Game Development professionals of all experience levels who want to learn more about Generative AI when used to create Art.