How Kids Create Art, Music and Games With AI in PlayWorlds
A nine-year-old we know recently made her first video game. It took her about ninety minutes from idea to playable build. The game stars a fox in a snowy forest who has to collect glowing crystals while avoiding owls. There's original concept art for the fox, bright orange, blue scarf, slightly mischievous expression. There's an original music score that plays during levels, gentle strings during exploration, a faster percussion track when the owls show up. She invited two friends from school to play it and they spent the rest of the evening adding new levels together.
This is what creating with AI looks like for kids in 2026. Not a single tool. Not a single skill. A whole creative practice that spans visual art, music, design and storytelling, all made possible by AI handling the technical layer while kids handle the direction.
This post is about the three creative mediums kids work in inside PlayWorlds and what each one teaches them.
Concept art: kids design characters with AI image generation
A patient creative coach who helps kids turn ideas into playable worlds. Read more →
The first thing most kids want to do when starting a new game is design the main character. Before AI, this was a wall. Drawing the character you imagined, actually drawing it, recognizably, with the personality intact, takes years of technique that most kids don't have at seven or eight.
In PlayWorlds, kids design characters by describing them and iterating with AI image generation. The flow looks like this:
The kid says (out loud or in text): "I want a fox character. He should be orange and small. He has goggles and a leather jacket."
Leonardo Da Vinci, our AI guide, generates four variations. Each one interprets the description slightly differently, different shades of orange, different jacket styles, different goggle designs.
The kid picks the closest match and gives feedback: "I like this one, but make the jacket blue, give him a longer tail and make him look more brave."
Leonardo Da Vinci regenerates with the changes. Four new variations.
The kid keeps iterating. Maybe five rounds. Maybe fifteen. Eventually the character looks like the one in their head.
This loop teaches kids something important: their first description is rarely complete. The character in their imagination has details they haven't articulated yet. By looking at AI-generated variations and reacting to them, kids discover what they actually wanted. It's a kind of creative archaeology, uncovering the vision they had all along, one iteration at a time.
By the time the character makes it into the game, the kid has had dozens of small creative decisions stacked together. They've developed taste. They've practiced specificity. They've learned the value of feedback.
AI-generated music: kids compose game soundtracks
Music is the medium where AI's impact on kids' creativity might be most dramatic. Composing music traditionally requires either years of formal training or a lot of patience with software that's hard to use. Most kids who imagine soundtracks for their games have never had a way to actually make them.
Inside PlayWorlds, kids compose original music for their games using AI music tools. The flow is similar to character design:
The kid describes the mood: "Spooky cave music with weird heartbeat drums and high creepy strings."
The AI composes a short piece. The kid listens.
The kid gives feedback: "Make it slower. Add more drums in the middle. The strings should be quieter."
The AI iterates. The kid listens to the new version.
This continues until the music feels right.
Kids who go through this loop a few times start hearing music differently. They notice instruments they hadn't paid attention to before. They develop opinions about tempo and mood. They learn the language of describing music, fast or slow, high or low, dense or sparse, cheerful or ominous. None of this requires music theory. It just requires lots of reps in a fast feedback loop.
The result is that kids build emotional fluency with music as a creative medium. They know what kind of music their game needs because they've tested options against their own ears and refined toward something that feels right. That's a real creative skill and most adults don't have it.
Full game worlds: kids build multiplayer games through conversation
The biggest thing kids create with AI in PlayWorlds is the games themselves, full multiplayer worlds with characters, music, mechanics and rules. This is where everything comes together.
A kid starts with an idea, often vague. "I want to make a game where you're a fox and you collect things in the forest." Leonardo Da Vinci asks questions: What are you collecting? Why? What gets in your way? How do you win?
Through conversation, the idea sharpens. The fox is collecting glowing crystals. Owls are trying to steal them back. The win condition is collecting fifty crystals before sunrise. There are platforms to jump on and caves to hide in.
Leonardo Da Vinci generates the world. The kid walks around in it. They notice things they want to change, the forest is too dark, the owls move too fast, the crystals are too far apart. They tell Leonardo Da Vinci. Leonardo Da Vinci iterates.
Eventually the kid invites friends. The friends play through the level. They give feedback, "the boss is too easy" or "I love the music when the owls show up." The kid takes the feedback and iterates again.
This is real game design. Real iteration. Real collaboration. It's the same loop professional game designers run, just compressed into a few hours and made accessible to a child without scripting skills.
What kids learn that they don't even know they're learning
The skills kids develop through this kind of creation are the skills that matter most in an AI-native world:
Multidisciplinary thinking. Making a game means working in visuals, music and systems all at once. Kids who do this regularly stop seeing creative fields as separate. They develop the polymath mindset that's becoming the most valuable kind of mind to have.
Iterative confidence. First outputs are starting points, not final answers. Kids who internalize this become better creators in every medium for the rest of their lives. They stop being attached to first drafts and start being committed to keeping going until something is great.
Taste. The single most important creative skill in the AI era. Kids develop it by making thousands of small choices about what's good and what's not, across visual art, music and game design, over a long time.
Communication. Describing what you want clearly enough for an AI (or anyone else) to deliver it is the most general-purpose skill on this list. Kids who get reps describing creative outputs to AI become much better communicators in every other domain.
Collaboration. Multiplayer creation means kids build with friends. They learn to combine their visions into something neither could have made alone. They learn to give and receive creative feedback. They learn that creation is often a team sport.
Why this matters now
Kids are growing up in a world where everyone will eventually create with AI. Some adults will use it well; many won't. The ones who use it well will be the ones who developed creative judgment, taste, iteration, prompting, multidisciplinary fluency, early.
The kids using PlayWorlds today are building exactly those skills, through the activity they love most: making games to play with friends.
We don't think of PlayWorlds as just a gaming platform. We think of it as the place where the next generation of creators figures out how to work with AI, by working with AI on something they care about, before they ever step into a classroom or a workplace.
The fox game took ninety minutes. The skills it taught will compound for the next sixty years.
Want your kid to start creating with AI through play? Join the PlayWorlds waitlist.
Related reading: