An AI game maker for kids: what it is good at, and what to know first

The pattern is familiar. A ten-year-old has a clear, detailed, ambitious idea for a game. They start, they hit the point where something has to be programmed, and the idea dies there. Not because they lack the ability to think about games — they think about games constantly — but because the gap between the idea and a working script is months of syntax.

Block-based tools bridge some of that and they are genuinely good at what they do. The limit is that the block set defines the game, so the ambitious idea gets flattened into whatever the blocks can express.

Flockbay takes a different route. The code is written by Director 1, in GDScript, inside a real Godot project. The child describes what should happen and it happens.

First they pick the kind of game. The app asks 2D or 3D before it makes anything, shown as two pictures, and there is no default. A lot of the games children describe are flat ones — a platformer, something seen from above — and picking 2D means there is a game running immediately: sky, a band of scenery, ground with a ledge in it, a platform off to one side, and a person who walks, falls and jumps under the keys they are already pressing.

Then they build the world by hand. In a flat game that is dragging ground where they want it and putting sprites down; in 3D it is terrain, water, weather, buildings, props and characters, placed and moved and rearranged. This is the part that holds attention, because it is spatial and immediate and looks like something within minutes. Ground they drag is walkable as soon as they let go, so nothing has to be set up before they can try it.

Then they describe the rules. "When you touch the water you go back to the start." "There are three keys and the door needs all of them." Director 1 writes the code and they press play.

And then the important part, which is that they play it and decide it is wrong, and go back in. That judging loop is the actual education. It is game design, and it does not require them to write a line of anything.

It is a desktop application for Mac or Windows, downloaded and installed. It is not a website, not an iPad app, and not something that runs on a school Chromebook. It needs a real computer because the game runs locally on the copy of Godot bundled inside it.

It needs an internet connection to build, because Director 1, asset generation and world data all come over the network. Playing a finished game does not need one.

It uses an AI that writes code and generates assets from what the user types. That is worth understanding before handing it over, in the same way you would want to understand any tool with a text box connected to a model.

It is not a curriculum. It is not designed for classroom management, and there is no teacher dashboard, no class accounts and no seat pricing, because it is a per-person desktop app.

It is also not a toy. The engine underneath is full size and the app assumes someone with an attention span and an opinion. A determined twelve-year-old will get a long way. A six-year-old will probably want something simpler and there are good tools for that.

Age suitability is a parent's call and we are not going to make it for you. What we can be exact about is what the tool does.

The download is free and building the world is free, which is enough to find out whether it holds their interest. Indie at $10 a month unlocks the AI assistance — Director 1 writing the gameplay logic, and generating textures, sounds, props and rigs — and the free plan includes a limited trial of it.

Not programming syntax, and it would be dishonest to imply otherwise. If the goal is for a child to learn to write code, a tool that writes the code is the wrong choice and there are good ones aimed squarely at that.

What it does teach is game design, which is a real discipline and a harder one. Why a level is confusing. Why an enemy is unfair rather than difficult. How to change one thing and see what it did.

It also teaches finishing, which is the rarest skill in this whole hobby. A small game that someone else played is worth more than three ambitious folders.