How a person makes a game today — and how AI will do it tomorrow
This morning I was working on a castle defence game. Not on graphics, not on music — on a single button. A round, gold-bordered button that appears on Android and stays invisible on Windows.
This morning I was working on a castle defence game. Not on graphics, not on music — on a single button. A round, gold-bordered button that appears on Android and stays invisible on Windows. This button upgrades the selected tower. Touchscreens have no right-click, so something had to be done about that.
These small details are what real game development looks like. Not the spectacular parts — but the parts that go unnoticed, that simply work, and no one pauses to think about who built them.
But today I no longer work alone. An AI sits beside me, reads my code, understands the logic, and completes in minutes what used to take hours. This is not frightening. It is a tool — like a pen for handwriting, or a DAW for composing.
AI does not take away game development. It gives back the creative part.
I remember spending two days on a simple mobile side project just to figure out why a button wasn't appearing on the right device. CSS specificity, touch event bubbling, different browsers behaving differently — these are not creative challenges. They are technical obstacles in the path of creation.
AI reduces those obstacles. Not eliminates — because the decision of what the button should be, where it should sit, what it should do, what feeling it should give — that is still mine. But the implementation is increasingly becoming a dialogue. I describe what I want. I receive the code. I test it. I ask for a change.
This workflow also changes who makes games. You no longer need to be a programmer to create one. But you do need something else: vision, taste, patience — and the ability to know precisely what you want and say it clearly.
Much can be said about the future of game development. What I know for certain: more people will be making games, with fewer technical barriers. That means more unique voices, unique worlds, unique stories on the screen.
In the castle defence game, runners race along roads, towers grow level by level, the enemy thinks and attacks. I designed all of it by hand — every level, every difficulty curve, every balance point. The AI only helped the player see my fingerprints more clearly.
But if we are talking about the future of game development, it is worth looking at where the genre as a whole is heading. Because today it is no longer enough to make a good strategy game. The player expects to stumble upon something unexpected along the way.
The best example I can give is 7IS — Army of the Seven Gods, which I am currently working on. In terms of scale, this is no small undertaking: it is not just a strategy game, but a world in which the player moves from event to event and discovers curious little minigames, hidden mechanics, unexpected turns. Each of these small things represents days of work on its own — a polish here, a fine-tune there, a balance adjustment — and you need thousands of them before the whole thing feels alive.
This principle is now almost a baseline expectation: a game within the game, a program within the program. An RPG inside a strategy game. A video game expansion inside a larger RPG. The creator builds layers upon layers, and behind every layer sits the same question: what will the person who digs a little deeper find here?
Thousands of small refinements. Days spent on a single scene, a single sound effect, a single button. And yet — today, the person who works alone is not truly working alone. The AI is there beside them, or we, or whatever one prefers to call it. It replaces nothing. But it sits in the workshop, and when your hands tire, it takes over the lifting.
That is the best a tool can do.