Next pilot — International School of Beijing · Aug 5–9, 2026 · applications open
Curriculum & Philosophy

What we believe.
What we teach.
Why.

This is our curriculum and our philosophy — eight principles for what school should look like when every student has a tutor smarter than their teacher in their pocket. The things AI can't teach a kid — agency, taste, craft, courage, judgment — are what's left, and what we spend the days building. The water, not the container. Less about transmission, more about the things that decide a life. Less résumé, more eulogy. Less waiting room, more arena. The credit belongs to the one who is actually in it.

The eight principles

The whole curriculum, on one screen.

I

Agency over compliance

Students choose what to build on day one. Mentors refuse to hand them a brief.

 II

Make things — in front of people

Build. Present. Share. Reflect. Repeat. The week is the loop, compressed.

 III

Practitioners, not credentials

Taught by people who've actually shipped — founders, engineers, designers.

 IV

AI is leverage, not a crutch

Tools handle the floor. Students spend the saved time on the part only they can do.

 V

The examined life, early

Thirty minutes a day, no screen, one notebook, one honest question.

 VI

Risk early, while it's cheap

Public demos are non-optional. Courage is a muscle built through micro-reps.

 VII

Small choices, real systems

Small promises kept in public, every day. The system is the artifact.

 VIII

Be water — first principles

Every project stress-tested with five questions. Graded on whether they asked.
How the principles cascade into a day

Eight principles, seven moving parts.

The principles aren't a poster on the wall. They're built into the architecture of every cohort — when, with whom, and on what students actually spend their hours.

07:30 – 09:30 · every weekday

AI mornings — clear the floor.

Two hours with an AI tutor that knows the student. Foundational skill acceleration — reading, math, prompting, writing, code basics. The same floor school spends twelve years on, cleared before lunch.

IV Leverage I Agency VII Systems
10:00 – 16:00 · every weekday

Build afternoons — the workshop.

Teams pick a real problem they care about and ship a real AI product. The whole day's craft work — design, code, user research, iteration — happens here with mentors in the room.

II Make IV Leverage VIII Be water
throughout the build block

Practitioner 1:1s — standing next to you.

Founders, engineers, and designers who've actually shipped sit alongside students while they're stuck. Not office hours — real pair-work, real reviews, the question you couldn't ask yourself.

III Practitioners V Examined life
30 min · every afternoon

The quiet room — one honest question.

Phones away. Notebooks out. A single examined-life question on the wall. Mentors answer it out loud first, badly and honestly, so students see what a real reflection actually looks like.

V Examined life VII Systems
end of every day

Daily public commits — small promises, kept.

Every kid posts a daily log to the cohort: what they shipped today, what they learned, what they'd do differently. The system is the artifact. The product is the proof the system worked.

VII Systems I Agency
at the close of every cohort

Demo day — the public reckoning.

Live pitch to a real panel — practitioners, investors, parents. Strongest teams win prizes. Every kid ships something they're not sure about, in front of a room, and survives it. That's the callus we're after.

II Make VI Risk early I Agency
where the work is

Cohorts that travel — the school meets the work.

A cohort doesn't need a campus. It needs a room, a teaching team, and a real problem worth shipping. The next cohort runs at ISB Beijing in August. Future cohorts will run where the work is.

VIII Be water VI Risk early III Practitioners
Principle I

Agency over compliance.

What school does

Rewards obedience, grades the rubric — then wonders why graduates can't decide anything without permission.

What Arena does

Hands students a blank slate on day one. Mentors refuse to write the brief. The first hour is uncomfortable on purpose.

The quiet, steady power to decide what story you'll tell about your own life.
In the curriculum

Students choose what to build on day one. The moment a kid moves from tell me what to do to I think I'll try this, the rest of the week unlocks. Every cohort ends with a "promise to self" — one small thing they commit to keeping for ninety days. The shipped product is the artifact. The deeper artifact is the kid who proved to themselves they could cast the small vote anyway.

What students actually do
  • On the first morning — pitch your idea to the cohort. Three minutes. No deck.
  • Early on — interview three potential users yourself, in person or by cold email.
  • At the close — write your "promise to self" for the next 90 days. Print it. Sign it.
Principle II

Make things — in front of people.

What school does

Grades a five-paragraph essay no one will ever read and calls that learning.

What Arena does

The week is one loop, compressed: build, present, share, reflect, repeat. Every day ends with a demo. Every demo is a small public reckoning.

Taste, judgment, courage — can't be taught from a textbook. They have to be lived.
In the curriculum

First days: pick a problem you actually care about. Last day: present a working product to a room of real adults — practitioners, parents, peers who will not pity-clap. The artifact is a thing on the internet with your name on it, that someone other than your mom can use.

What students actually do
  • By midway — ship a working prototype to a public URL. Anyone can use it.
  • Toward the end — record a 90-second demo film. Post it.
  • Demo day — 5-minute live pitch + Q&A to an 8-person panel.
Principle III

Practitioners, not credentials.

What school does

Mistakes a teaching certificate for the right to teach a thing. Hires well-intentioned generalists who've never built what they're teaching.

What Arena does

Staffs every week with people who've actually shipped — a founder who's raised, an engineer who's deployed, a designer whose work you've used.

You wouldn't learn investing from someone who's only read about it.
In the curriculum

Mentorship is real — not office hours, but a person standing next to you while you're stuck. Students walk out with a handful of adults who now know their name, and a clearer sense of what good actually looks like in the wild.

What students actually do
  • Early on — pair-code with a YC-backed founder for two hours.
  • Mid-cohort — design review with a design lead from a company you've heard of.
  • At the close — get a written, no-holds-barred code review from a senior engineer.
Principle IV

AI is leverage, not a crutch.

What school does

Bans the tool that will define these kids' working lives — and then tests them on what the tool already does better.

What Arena does

AI in every seat from minute one. Not as a crutch — as a co-pilot. Students learn to prompt, critique, catch the hallucination.

A twelve-year-old can now ship what was a college-senior project a few years ago. That's the new floor.
In the curriculum

Students spend the time the tool gave them back on the part only they can do — the taste call, the design call, the what should we actually build call. The artifact is a kid who can sit down at a blank screen with an idea and have a working version by dinner.

What students actually do
  • AI mornings — adaptive tutor pushes them through the basics 2.6× faster than a classroom.
  • Build afternoons — Cursor + Claude as pair-programmer; ship code you couldn't have written a year ago.
  • Every day — critique the AI output. Catch the hallucination. Learn where it fails.
Principle V

The examined life, started early.

What school does

Optimizes the résumé column and skips the eulogy column entirely.

What Arena does

Thirty minutes a day, no screen, a notebook, one honest question. Mentors model how they answer — badly, honestly.

Whose life is it?
In the curriculum

The artifact is a small private document each student leaves with — not a plan, but a list of questions they now know to keep asking themselves on the Tuesdays no one will ever notice.

What students actually do
  • Daily quiet room — 30 min, no screen. One question on the wall.
  • The three questions — What did I enjoy today? What did I do because someone was watching? What would I work on if no one ever found out?
  • At the close — a letter to your future self. Read it again in a year.
Principle VI

Risk early, while it's cheap.

What school does

Penalizes the wrong answer and rewards the safe one — which trains exactly the kid the world will punish hardest.

What Arena does

Public demos are non-optional. Cold-emailing the practitioner you want to learn from is a Tuesday assignment.

Courage isn't a trait you're born with. It's a muscle built through micro-reps.
In the curriculum

Every kid ships something they're not sure about, in front of a room, and survives it. The artifact is a callus. A cohort won't make them fearless — they'll just be slightly less afraid of the thing they were most afraid of when they walked in. That's the rep that compounds.

What students actually do
  • Early on — cold-email a stranger who's an expert in your problem. Ask for 15 minutes.
  • Mid-cohort — show your half-broken prototype to real users. Take the feedback live.
  • Demo day — pitch to investors with real prizes on the line.
Principle VII

Small choices, real systems.

What school does

Optimizes for the test in May and teaches nothing about the Tuesday in November.

What Arena does

Every morning is a small promise kept in public. Every kid posts a daily commit: what they built, what they learned, what they'd do differently.

The person you become at forty is mostly the person who showed up on a few thousand Tuesdays in their twenties.
In the curriculum

Kids leave with a one-page personal operating system they wrote themselves — morning ritual, weekly review, the friend who has permission to call them out. The system is the artifact. The project is the proof the system worked once.

What students actually do
  • End of every day — post a daily commit to the cohort: built, learned, would-do-differently.
  • At the close — run a public post-mortem on what shipped. No flinching.
  • Take-home — write your one-page personal operating system. Pin it above your desk.
Principle VIII

Be water — first principles over inherited form.

What school does

Trains masters of execution and strangers to understanding.

What Arena does

Every project stress-tested with five questions: how, why, what principle, when does it fail, how do I adapt.

The most advanced thing in any room will still, always, be a human being who knows who they are.
In the curriculum

Students aren't graded on the answer — they're graded on whether they asked. The artifact is a thinking move they now own. The kid who can ask why are we doing it this way in any room, for the rest of their life — that's who Arena is trying to send back into the world.

What students actually do
  • Early on — run the five questions on your project. Defend each level.
  • Mid-cohort — find a competing approach. Steelman it. Then decide.
  • Toward the end — pivot something material based on user feedback. Show your work.
The whole curriculum, in one paragraph

Arena is a school for students ages twelve to twenty-two — built around shipping real AI products alongside practitioners who've actually built things, learning to ask why until they hit principle, and starting — early, before the script of inherited ambition hardens around them — the only project that compounds for the rest of a life: figuring out whose life it is, and casting the small vote anyway. They walk out with a working product, a small system they wrote themselves, a handful of adults who now know their name, and a callus where their fear of public demos used to be.

Build it with us → See the ISB pilot · Aug 5–9