AI Art and Indie Game Development

Posted Aug 10, 2024

There has been some concern about the use of AI art in the gaming industry. This blog post will clarify how Orbs CCG uses AI art and why it is important for indie gaming.

I first made Orbs CCG back in 2012. AI art did not exist back then, so I hired artists to make the art for the game. This was a very expensive process, and I was not able to make the game effectively in part due to the costs associated with indie gaming. I raised a small amount of funding, less than $100,000, and was looking at the prospect of spending a significant portion of that on art alone. That was a large part of the reason I was only able to release one set of cards.

Let's do some math to put hard numbers on this. A new set of cards could be 350 cards. If I paid just $100 per card, that would be $35,000 for a set. That's a lot of money for an indie game for just one set of cards. And that's not even considering the other costs of making a game. Assuming one set of cards per quarter, that's $140,000 for the year, more than I had raised.

There's a lot of work that goes into making a CCG game. You have to design the cards, playtest them, code the cards, make the game UI, code the fundamental game logic and game server, do marketing, and more. It's a lot of work. And art is significantly less important than the other parts of the game. People will play a good CCG with weak art, but wouldn't play a weak CCG with good art.

From a business standpoint, when you're making a new game, it's a bit of a gamble. You're not sure how much revenue you're going to make. I was unsure back in 2012 whether I'd make $140,000 in revenue in a year (I didn't). It was a big risk to spend that much on art alone, especially when so many other more important aspects of the business need funding as well and are more critical to the game's success.

The fact is that, prior to AI art, it was pretty much infeasible to make an indie CCG game from scratch without a significant budget going towards art. I'm familiar with games that used stick figure art or other cheap things, as a way to work around this reality, but I never really liked that idea.

Fast-forward to today. AI art is now good enough to use in real products. I happen to think that technological development is an overall good thing, although I am concerned with what it means for the future of human work. My career is in software engineering, where I believe AI will make significant impact in the near future too. In fact, it's likely that most human jobs will be supplanted by AI in our lifetimes. So while it is unfortunate for artists to have to compete with AI, they weren't the first profession to be impacted by technology, and in a relatively short amount of time I suspect we all will be in a similar position.

AI art actually isn't good enough on its own, yet, that I'd want to use it if I had unlimited resources. The art can be weird and have strange artifacts. It can be difficult to get it to do exactly what you want. I often find myself editing the art manually after the fact, to fix some of the artifacts or change things about it. I'd prefer to work with humans for various reasons, if budget were not a factor. However, AI art is good enough that it fundamentally changes the cost structure of making new sets for indie CCG games, and that's a big deal for indie games like Orbs CCG.

For the first set, about half the cards were made by human artists, and half were AI art. Going forward, I expect to continue to use a mix of AI and human art.