playing ai games

a $632 space rts, and why i won't play it.

This was on my feed this morning:

a post on x: a homeworld-style space rts built with opus 5 for $632.65 in tokens, generated models, no assets. the screenshot shows a large capital ship over a ringed planet with a full rts interface.

It isn't the only one. Today I've also seen a Halo 1 clone with matchmaking and leaderboards, a one-prompt game somebody called "like a top game on steam", and another clone built in Unity. So what am I supposed to take from this, as someone who plays games and as someone who makes them?

As a player, I don't want to play it. It isn't anything to me. It's code and systems and algorithms running into each other. That's all the explanation it needs, too, because that has always been true. Every game you've ever played is code and systems and algorithms running into each other.

As a developer, it mostly points back at something games have always shown us: people care about the experience, not the structure. They want to get dropped into another world, another story, somewhere they've never been, and believe it for a while. That's suspension of disbelief, and getting a player there is most of the job.

It's also how a game stands out in a market that's already packed with clones. Think of a game you loved, or one that stuck with you for years. It wasn't the feature list. It was what the game did with those features, how they added up to something you could feel. Even Tetris is more than blocks moving down a well. It's the feeling of solving the puzzle with the next piece already coming.

None of this bothers me, as a gamer or as a developer. I know I'll never play that RTS, and that's my choice; it has nothing to do with whether it ever comes out. It's also clearly better looking, and probably better built, than anything I've made or will make any time soon. So why doesn't it bother me?

Because the part I actually care about, every time I start something new or go back to one of the projects I've been dragging along for years, is the experience and the story. LLM tools can build pieces and stick them together. I use them too. What they can't do is put a person in it. That's where the person playing and the person who made it meet, and it's why a game stays with you after you put it down.

So my advice to other players is to cut these games out of what you look at. They won't satisfy that itch to explore and find something new. I don't care that AI made them. What's missing is anybody guiding you through them.

Go find the games, and the books and movies and songs, that open a door to somewhere else and then take you through it, even if the way they guide you is handing you a sandbox and getting out of the way.