The Anger I Understand

People in the gaming community get defensive, fast, when AI content, especially art, replaces real art in games. I hear the arguments every time. AI art lacks soul. The models were trained on real artists’ work without permission. The companies selling these tools are vampires, and every dollar spent on them keeps a broken system alive.

I don’t disagree with any of that.

The Tools That Might Replace Me

I got laid off within the last year, like a lot of engineers were, because leadership misread either the technology or the cost structure underneath it. I understand that anger because I live with it too.

That anger sits in direct tension with something else. I have a family to provide for, and the fastest way to do that right now runs through the very tools I’m angry about. So I’m learning them anyway, even though they might be part of what replaces me next. Between client work at Kerrberry and nights on this game, that tension doesn’t go away.

Nobody’s Winning Except Wall Street

I do think there’s a happy path somewhere down the road, where these models run locally on hardware regular people can actually afford. We are nowhere near that today. Right now the AI companies aren’t winning, the businesses paying for their tools aren’t saving money, consumers aren’t getting better products, and engineers are watching their market value erode. Nobody is winning except Wall Street.

That’s a conversation for another day.

None of this has made life better for the people actually using these products. Every backlog is full of opportunities and risks. A team could be pointed at that list and move faster with the people already there. Instead, companies lay people off by the thousands. If the tools genuinely made teams faster, keeping people on board and progressing faster is the obvious move. That gap, between what AI could do and what companies actually do with it, is where the angst comes from.

Why Does Code Get a Pass?

What I don’t understand is why that same anger doesn’t show up when AI replaces development time instead of art. A lot of the code that trained these models was just as ill-gotten as the art.

For me, AI compressed twelve months of solo game development into a month or two. That’s not a small difference. It’s the difference between shipping and not shipping. It let me chase features I wasn’t sure would work, follow whims, and try to build something wide and a little strange instead of small and safe.

But I’ll say this too. When a studio ships a project and then lays off the very people who built it, purely for accounting reasons, that’s not just a disaster for those people. It’s lost expertise and lost tax revenue for everyone else.

What a Hundred Dollars Bought Me

As a solo developer, generating dozens of iterations of art and audio for Double Skunk cost me about a hundred dollars total. Not a hundred dollars per asset. A hundred dollars, period.

If I’d had to wait weeks for a human artist’s schedule to open up, or pay per revision every time the design shifted, I never would have finished this project. The game keeps changing shape build to build (the patch notes tell that story better than I can), and I needed to be able to rework a piece of art or a voice line without flinching at the bill.

Share the Success

So here’s what I’d propose. Ship the game with AI placeholders. If it earns money, commit ten percent of revenue to a fund for re-skinning it with real art, up to a set cap. Do the same for voice-over work. Voice is probably the most obvious AI surface of all of them. I could even source community members to voice characters. That enriches the community twice. It gives people an actual role in the game, and it builds a content base people actually care about. Commit another ten percent to the platform that made the whole thing possible. In my case, that’s Godot.

I want a world where an individual can compete with a big studio. Big studios will always have better tools and more resources than one person working nights. So instead of asking why an indie developer is using what’s available, make a plan to share the win when it happens. If this game could have supported twenty thousand dollars in art and voice acting instead of what AI generated, then that’s the number. Set it as the budget, and when the money shows up, spend it back into the community.

No Pitchforks

None of this resolves the tension I started with. I still think the critiques of AI art are largely right. The companies building these models still operate like vampires. And I’m still using the tools anyway, because they’re what let one person build something real without a studio behind them.

If you can put money behind the parts of this that need supporting, digital art chief among them, do it. That’s what the revenue plan above is for. But save the pitchforks for the studios laying off thousands of people who made them money in the first place. Not for the one-person team using what’s sitting on the table in front of them.

Double Skunk is still in development, built by one person figuring it out as he goes. Want to see where the placeholders end up? Join the playtest.