I try to avoid writing about AI when I can because I know all of us are already drowning in it. Stick with me here, because there is a point to this and something I want you to take away that’s not just AI doomerism.

If you want one thing to blame for this post of mine, I suppose it would be Epic Games going “all in” on AI. But really, this is bigger than that and something I’ve been feeling for weeks.

I don’t like preaching like this, but I think you need to understand my thought process for my point to land. Depending on AI in any nontrivial capacity is short-sighted and unsustainable. You’re switching off your critical thought and learning processes and replacing them with a hollow imitation rented to you by a megacorp. In doing this, you give up the opportunity to invest in your own abilities instead.

But let’s say you don’t care about that. You don’t care about developing your own skills, you just want the thing made and you’re using AI to write the code for you. At the end of the day the question of whether the code works is a simple yes or no. It might be an inscrutable mess, but it either works or it doesn’t. If it doesn’t, you can probably get to a version that does if you have good tests and a big enough compute budget.

A game is different. There are no tests an AI can run to find out if a game is “good”. There is no correct way to design a game. It takes lots of testing and tweaking, and it all depends on what it is exactly that you’re trying to say and what you want the player to feel.

Games, like any other art, are created with intent. They are created by people with tastes that have been developed over time from lived experience. These are things which an AI model, however large, will never have. Anyone claiming that AI is going to solve or streamline game development is either ignorant or trying to sell you something.

But does a game even need to be “good” to be good? Or any art for that matter? For the past while, I’ve found myself developing a new appreciation for things that are clearly made by a real person. Even the really amateur looking stuff.

In fact, I love noticing the little human imperfections now. They’re evidence that someone cared enough about what they were doing to do it themselves. The world we live in is one where they could have easily typed a few words into a computer to generate some version of what they were going for, more or less. That they didn’t is respectable. It takes some courage to put something you made on display while knowing all of its imperfections. Even if what they made isn’t “good”, they weren’t afraid to put themselves into it and they respected the craft enough to learn it. Even just doing that much is an important statement to be making these days.

Like I said, I’ve been having these kinds of thoughts for weeks. They’re a big part of what’s given me the drive to start work on a new personal game project. It’s not a big project, or anywhere close to that. It might not even be fun to play. Someone else could probably make the same game and do it much better.

But I want to make a thing that someone will notice the rough edges on and maybe get inspired by to make their own thing.

I guess the takeaway I’m going for is that AI is helping me appreciate that things are still worth doing yourself, even when you know they’ll be flawed. You leave a mark on the things that you do and create which an AI could never replace. They leave a mark on you too, and everything you do and create after.

— JP