AI as a tool

Blog

Per the earlier post, this was step 3. I opted to make it its own post, because of the controversial nature.

As a disclaimer, I have not, nor do I ever plan to use AI art in a commercial product. Other people have opinions on this subject, but that is my line in the sand. For development, I think it is a valuable tool.

Comic scramble, by its very design, needs a lot of art. To date I’ve created over 400 images as I experiment with different styles, concepts, and iterations of tiles to figure out what does and doesn’t mesh. That’s how I got the learnings listed in the last post. But I’m not done. I would guess that I’ll have generated well over a thousand by the time I’m done, to have close to 100 that I would want for the game proper.

So, let’s say that an image by an artist will cost $40, which I believe is in the ballpark of reasonable. That would mean that I would’ve needed to spend $16k for test art to date, and ultimately $40k to get art for the final project.

“Big deal!” Some may say. “That’s the cost of development,” others may push. And that’s a fair case for companies that can afford it. My development budget is closer to $2k a year, as this is all bootstrapping to get things off the ground. That would mean 20 years of development on this project to the exclusion of everything else to get final art, for a project that may not even be viable.

Could I raise the funds on Kickstarter? Maybe, but $40k is a big ask, and that’s just for the art (admittedly, one of the biggest development expenses of any game). My last game was 1/10th that, and few projects successfully raise that much.

So, that’s the economics of it, but I believe there is a very real benefit to a designer using a tool like this. It helps me to figure out what I don’t want! Everything is amorphous and intangible until you see something for the first time, and very rarely is the response, “Yes! It’s perfect!” As a designer of games and products in general, it’s best to get a concept in front of people as quickly as possible so that they can say, “No, not that, because of X, Y, and especially Z!” It’s not a critique on you as a designer (it may feel like it, but criticism is a part of any design process).

I would much rather go to an artist with the lessons I’ve already learned on what works well with the game stylistically, and examples of what to avoid. If I’m paying by commission, that will reduce the number of iterations they need to spend on my project to get what I ultimately need, freeing them up to move on to even more projects!

Leave a Reply

Your email address will not be published. Required fields are marked *