What a kid can ship in a year
Ten concrete things an 8-to-16-year-old can actually build, finish, and put in front of a real user inside one year at Hi, Bot.
May 25, 2026 · Hi, Bot
People ask us, often, what a kid actually leaves a year of Hi, Bot with. The other posts on this blog answer the philosophical version of that question — the rubric is the artifact, the output is the rubric, etc. This post is the concrete version. Ten things, plausible for an eight-to-sixteen-year-old, that we have seen kids ship inside a year.
None of these is a stretch goal. None of these requires the kid to be a prodigy. Each of these has been done by a kid in this age range, in a room not unlike ours, with a mentor sitting next to them and the verify habit in their pocket.
1. A Pokédex for a sibling
A camera input, a model call, an answer back. The kid points her phone at a bug in the yard, the app says what it is and tells her one weird thing about it. Her sibling uses it for the rest of the summer. Total build time, with a mentor: about three weekends. Total joy delivered: hundreds of bug identifications by a six-year-old.
2. A tuned agent that wins a small tournament
A kid picks a game — a board game, a card game, a simple strategy game — and tunes an agent to play it well. She runs it in a small tournament against other kids' agents. She loses the first round and learns more from the loss than from any class she will take in fifth grade. By round three, she's winning. The trophy is the agent file, with her name on it.
3. A daily explainer for a younger sibling
A short, one-topic post written for a specific person who actually asked the question. Volcanoes one week. The Treaty of Westphalia another. The kid writes it herself, uses AI as a research assistant, edits for the ear, sometimes records it as a one-minute podcast. Twenty issues in a year. The sibling is the subscriber. The grandparents are the subscribers nobody asked.
4. A card game her cousin will actually play
Designed, prototyped, illustrated, printed at the local print shop. The cards live in a small box in her cousin's house. The cousin plays it with her friends. The kid does a v2 a few months later, after watching the rules break in real games. This is the entire product cycle, at twelve.
5. A trivia game for the group chat
Ten questions about her favorite show, deployed to a link her friends can open on their phones. She runs a leaderboard for her own group, which is the correct scope for a leaderboard. The friends play it during lunch on Monday. She updates the questions for season two when it drops.
6. A study tool that quizzes her on the stuff she keeps getting wrong
This is the one the parents like. The kid builds a small tool that takes the questions she missed on her actual quizzes, turns them into flashcards, and quizzes her again on a schedule. By the end of the cohort it's helping with two of her classes. By next year, she is the kid in her grade who knows how to build her own study tools. This will compound for the rest of her academic life.
7. A real children's book for a five-year-old in her life
Eight pages, an actual ending, illustrations she made or generated and then edited. Read aloud at bedtime in somebody's house. The five-year-old hands it back: can you read it again. The kid has shipped a real artifact to a real audience, which is something most adults have never done.
8. A neighborhood dog tracker
A small website that lets her log every dog she sees on her walks. Name, breed (guessed, then identified by the model when she's not sure), location, photo if the owner is okay with it. A year in, she has 140 dogs. She has also, as a side effect, learned what a database is, what a form is, how to host a thing, and how to ask a stranger politely if she can take a picture of their dog.
9. A science-fair project that didn't come from a kit
Materials inventoried, hypothesis written by the kid, AI used as a research librarian, results graphed, presentation built. The judges can tell. Kits are obvious. A project that started as a question the kid actually asked is obvious in the other direction. We have watched this distinction win science fairs and we have watched kids beam about it for months.
10. Three paying users for a small product
This is the stretch one. Not every kid does this. The ones who do, ship something so specific it has three customers — three classmates who pay a dollar a month for a thing she made; three parents in her co-op who pay five dollars for a tool that solves a problem her own parents had. Three paying users teaches more about product, support, and the responsibility of building for people than ten finished projects with no users.
If you read this list and your reaction is that's a lot for a year, fair. It is. The thing to know is that no single kid does all ten. Most do two or three, deeply. A few do six. Almost every kid who stays the year does one thing somebody else uses, which is the milestone we actually care about.
The other thing to know is that the list is not aspirational. None of these are demos. None of these are screenshots. Each one ends with another human using the thing. That's the bar.
We will not promise your kid will ship the Pokédex. We will promise that the room is set up so the Pokédex is reachable for her, if she wants it. That's the deal we offer. Everything else is hers to make.
The waitlist is open. The room is small on purpose.
› Keep reading
Get the next dispatch.
Occasional, never spammy.