The origin.
Jim wanted to spend time with Theo. Theo was nine. The plan was to teach him how to make a game — not a game in someone else’s sandbox, an actual game at an actual URL that his friends could play.
The first session went like the first session usually goes. The two of them sat down. Jim opened a code editor. Theo had ideas. They typed for an hour. The game crashed.
The lesson wasn’t that Theo couldn’t code. The lesson was that the gap between "I have an idea" and "I have a thing my friends can play" was full of work that had nothing to do with Theo’s idea. Boilerplate. Setup. Asset wrangling. Bugs that came from typos. None of it was the part Theo wanted to be doing.
So the next session, Jim put an AI in the middle of the loop. Theo said what he wanted; the AI proposed a file change; Theo played the result and either kept it, revised it, or undid it. The work that used to be tedious became something Jim and Theo could do together, with Theo running the show.
That loop is now the loop the studio runs on. Every kid who opens Tell and Show is doing what Theo was doing in their living room, on the same software, with a sharper version of the same partner.
What we found out.
Three things became obvious within the first few sessions, and they’re the things the rest of the product is shaped around.
Kids are not patient enough to read documentation.
And they shouldn’t have to be. The studio is built so that the kid’s first AI change happens inside the first ten minutes. Everything they need to learn, they learn from watching the change land.
Visible AI changes are the whole game.
When the AI proposes a thing and the kid can see exactly what changed, they stay the author. When the AI just hands them a finished thing, they become an audience. The studio is built around the first version of that loop, not the second.
Parents need a way to be in the room without being in the room.
Parents are responsible for what their kid ships. They cannot read every prompt. The product is built so the parent can supervise without becoming the AI safety officer — via approval emails, a family dashboard, a per-kid safety log, and a server-enforced publishing gate.
What we believe.
This is the part where most company-about pages start sounding like company-about pages. We’ll keep it short.
Kids should make things that are real.
Not toy versions inside a closed sandbox. A real URL, a real file system, real code, real ownership. The skills are different. The pride is different.
AI literacy is taught through making, not through lessons.
Every concept we want kids to walk away with — prompts steer answers, tools take action, AI can be confidently wrong, context matters, iteration is the move — surfaces the first time they bump into it in their own project. No worksheet ever.
The parent is the customer; the kid is the user.
The parent decides whether to buy. The kid decides what to make. The product is shaped around both of those decisions being thoughtful ones.
Software for kids should be made for kids.
Not made for adults and then sanded down. The defaults are kid defaults. The vocabulary is kid vocabulary. The interface assumes the kid is capable of more than the adult thinks they are — because they are.
What we’re building toward.
Game is the track that’s live, because that’s the track Theo lives in. Story, Site, and Movie are coming next — same loop, different medium. Past that, the roadmap is shaped by what kids and parents actually ask for once they’re inside the studio.
The cohort program exists because some families want a human in the room. The desktop app exists because some families want offline first. The CLI exists because by 12, kids are ready to type real terminal commands.
We are deliberately small. The company is built so that the answer to every product question is the same person who would write back if you emailed support.
Working here.
We’re a tiny team in New York, and the next hires are educators. Six roles, all open in the next twelve months — curriculum designer, pedagogy researcher, lead educator for our cohorts and lessons, per-track specialists (Game / Story / Site / Movie), an accessibility specialist for the neurodivergent kids the maker loop reaches especially well, and a family workshop facilitator for parent + kid formats.
All six are described at /jobs. If you read that page and recognized yourself in one (or two), write us at hello@tellandshow.ai and tell us about a kid you’ve taught or a curriculum you’ve designed.
Last updated 2026-05-16.