That was the strangest part. Most games about history get to wait until the history is over. This one was written in the middle of it, by someone with no inside access, from the same headlines everyone else was reading.
Here are a few things I learned along the way.
Half of it is history, half of it is a guess
The game starts in 2020 and runs to the end of 2031. The first half is a stretch most of us remember: models that could suddenly write a decent essay, the scramble for chips, open letters, hearings, people quitting in public. The second half hasn't happened yet.
Writing across that line was harder than I expected. The years we've lived through are easy to get wrong in the details but right in the feeling, because everyone shares the feeling. The future works the other way round. Every specific prediction I tried felt dated within weeks of writing it.
What worked was keeping the future vague and the mood specific. Instead of guessing what happens in 2029, I tried to guess what it might feel like to work in the industry then. Busier, probably. A bit off, in ways people don't say out loud. That kind of guess holds up much longer than a forecast.
There's no villain
The easy version of this game has a cartoon villain: a CEO who knows exactly what harm they're doing and does it anyway. I wrote a few of those early on and deleted them.
The more honest version is that nobody in the story has to be evil. Everyone makes decisions that make sense on the day: raise the money, ship before the competitor does, hire the brilliant person who wants to move faster. The problem is what those decisions add up to. That's harder to dramatise, because there's no single moment where someone crosses a line. Things just drift, and the drift is what the game is about.
It also meant the satire had to point at incentives rather than at individuals. Every company and person in the game is made up. Partly that's for legal reasons, but mostly it's because singling out one real company would let everyone else off the hook.
I tried not to preach
I have opinions about AI. The game is better when it keeps them to itself.
The temptation never really goes away. You write a scene and you want the player to know what you think about it. But once a game tells you the right answer, you stop thinking and start hunting for the approved button. That's when it gets boring.
So the game shows you what your choices led to and leaves the judging to you. Two people can play it very differently and both feel it had something to say to them. At least, that's what I was aiming for. Players will tell me whether it worked.
The jokes ended up doing a lot of the work here. As things get darker, a silly headline or a ridiculous job applicant gives you a moment to breathe, without the game pretending everything's fine.
Deciding what to leave out
Some cuts were easy. No real names. Nothing built on a real tragedy. And no jokes about people losing their jobs, even though that's part of the story. Those moments are played straight, because the people it happens to aren't a punchline.
Other cuts were harder. Most of the obvious AI jokes, the ones everyone has already made, went too. A nod to a famous fictional AI is a nice wink once; a game full of them is just a list of references. I kept a few, because spotting one is a small pleasure, and dropped the rest.
The hardest cut was anything that claimed to know how this ends. I don't know, and I don't think anyone does. A game that pretended to would come across as either smug or scared.
Keeping the dread small
The game starts out cheerful and slowly wears down over the decade. My first pass at the later years overdid it. There were too many warnings, and it started to play like a horror game, which wasn't what I was going for.
What I ended up with is quieter: small things that are a little off and never explained. The way people talk shifts slightly. The office feels emptier than it should. Nothing is confirmed. If it works, a player can't point to the moment it got unsettling. They just notice at some point that it has.
I think that's closer to how big changes feel when you're living through them. There's rarely an alarm. You look back and realise things are different.
Making entertainment out of something serious
I'm still not entirely comfortable with having made a game about this. It's an actual industry with actual people in it, and their lives are changing while I make jokes about it.
Where I've landed is that games are one of the few things that put you inside a set of incentives instead of telling you about them. Reading that a race pushes everyone to cut corners is one thing. Being the one who decides whether to cut them, again and again, with a rival close behind, is another. If the game does anything useful, I hope it's that: a better sense of why the question is hard, even without an answer.
The game is LLM Wars, for iPhone, if you're curious.