IWhat sticks
How people actually learn
The real theory of how people learn: memory, feedback, why learning styles are a myth. AI then accelerates every rep. The method travels; the prompts don't.
Inaugural cohort · February 14–20, 2027 · San Diego · Invitation-only
AI amplifies experts.
Learn how to learn anything.
You spend a week in residence, learning a field you've never worked in — one next door to yours, or a world away — with AI as your accelerator. By Friday, you stand behind what you built and teach it to the room for five minutes. That's the most direct way to de-risk your career.
Request an invitation
One email when applications open. No spam.
Why this, why now
When a prompt can conjure your decade of work, what's left that's yours?
The decade you spent becoming good at something is now a prompt away for anyone with a chat window. You watch a model draft what used to take you a week, and you wonder what you're still for.
The fear is that your edge suddenly looks optional, and your job description could change faster than you can retrain. Most people feel that and either ignore it or doomscroll.
White-collar work is getting sorted into two curves. One belongs to the people getting on the AI curve, compounding small wins month after month. The other belongs to everyone watching from the window, waiting to see how it shakes out. Five or ten years out, those two groups land in very different places.
Getting on the curve is how you de-risk your career, and it's trainable. You practice in one week with coaches to guide you, walking into a field you don't know and making it yours. The ability you build is the asset, and it stays yours no matter what the models do next.
Who this week is for
An expert in one domain,
a stranger in the next.
You're at least a few years into your craft, and often a couple of decades: a scientist, an engineer, a financier, a marketer, a lawyer, an educator. You can feel AI changing the ground under the field you've mastered. This week is your chance to be a beginner again, in a field you've never worked in, where your seniority doesn't follow you.
You choose the field, you set the questions, and you work hard on purpose. The people who love the sound of that are exactly who we look for.
All week you'll be reading, thinking, writing, and arguing your way through the new field. That's how knowledge fields are learned, and it's where AI accelerates the loop, pressing on what you actually understand.
The one rule
Your new field is yours to choose, with one rule: it can't be your day job under a new name. Both directions from there are good ones.
Next door. A marketing executive takes on behavioral science; a quant takes on epidemiology; an operations manager takes on the math of scheduling.
A world away. A surgeon takes on game theory; a physicist takes on historiography.
You'll aim at a field when you apply, and it can keep sharpening right up to arrival. Pull it off once, next door or a world away, and you can learn anything.
The room
We cap the week at twenty participants and curate the room for a diversity of knowledge fields. You spend the week alongside people making the same kind of jump, and you leave with peers who understand the jump from the inside.
What you leave with
One week in residence, three things you carry forward
Five days on how people learn, then AI accelerates it. Two coaches in the room, and up to twenty peers spanning many different fields.
IIWhat you can show
An artifact you defend
A journal club, a written explainer, a small model or analysis, a teaching deck. You pick the form; we push on whether it proves you learned.
IIIWhat you can't fake
A five-minute teach
Day 5: you teach your new domain to the room and show your artifact as evidence. If they understand, you own it. Where you can't, that's your honest gap.
Where this goes
This is year one.
The February retreat is the inaugural one. Here's what it grows into.
- 01
The retreat
February 14–20, 2027, twenty people in residence. One week with coaches to guide you as you take on a field you've never worked in and come out owning it.
- 02
The community of learners
Everyone this work touches, from waitlist to alumni, seeds a standing community of learners. We think agents can do the looking-after: introductions across fields, learning sprints, the upkeep that usually burns out volunteer organizers. That part is still taking shape.
- 03
Again, next year
The plan is to run this annually. People who come this year come back with a new field to take on, and they pull the next cohort in with them. The longer game: pounding away at the problem of learning in an age of AI.
The week, day by day
On Monday, the field is new to you. By Friday, you're teaching it.
You arrive Sunday afternoon to settle in. We take over one Airbnb house for the week, and the program runs Monday 9 am through Friday early afternoon; checkout is Saturday. The syllabus is tailored to the room and keeps evolving until the retreat.
- Mon–Wed
The teaching days
Three days with both of us at the front of the room: short blocks on how learning works, each paired with hands-on practice. Never lecture-heavy.
Why learning is the career skill that survives AI · levels of questioning · analogies, and where they break · writing as recall · teaching to expose what you don't yet own.
You leave day three with your artifact committed: the thing you'll defend Thursday and show Friday.
- Thursday
Build day
Heads-down work on your artifact, AI throughout, with a 1:1 defense of your work against one of us. The evening is the celebration dinner.
- Friday
Festival of Teaching
Five minutes at the front: teach the room your new domain, artifact as proof. Then a closing debrief, and the method goes home with you.
Each day ends with a short reflection, written by hand, no AI — Thursday night belongs to the celebration dinner. Writing it yourself is the one thing the model can't do for you: noticing your own thinking, and asking the question that turns exposure into ownership.
Your coaches
We've lived the domain jumps you're here to learn.
Eric Ma
Senior Principal Data Scientist, Moderna
Eric started wet lab in biochemistry and microbiology, then jumped into computation and kept going: Bayesian stats, graph theory, deep learning, diffusion models. Each jump was its own foreign field. He's spent the last five years running this loop inside drug discovery at Moderna, where the work now moves two to three times the load and reaches a first prototype roughly ten times faster. These days he's running the same loop on quantum computing, in public. He's building the retreat to hand that curve to professionals in every field.
Read Eric's learning journey
Daniel Chen
Data Science Lecturer, University of British Columbia
Daniel studies how people actually learn to work with data. He teaches data science at the University of British Columbia, authored the book Pandas for Everyone, and has been a longtime Carpentries instructor since 2014. He brings practical pedagogy experience to the table.
On working with Eric and Dan
On working with Eric and Dan
Eric Ma is one of the most curious, inspired and inspiring people I've had the chance to meet and work with. He is always exploring new concepts with a wonderful combination of seriousness and humor, and his enthusiasm is contagious. I couldn't dream of a better partner in discovery.
Brice ChallamelHead of AI Strategy and Adoption, OpenAI · formerly VP, AI Products & Innovation, ModernaThe details
A week to focus.
In person, San Diego, February 14–20, 2027. Lodging and meals are included; travel to and from the venue is your own.
- Cohort
- Inaugural cohort · up to 20 participants
- Dates
- Sun Feb 14 – Sat Feb 20, 2027 (program Mon–Fri)
- Location
- San Diego, California (venue to be announced)
- Format
- One week, in person
- Lodging
- One Airbnb house for the whole cohort; single rooms are subject to availability. You're welcome to arrange your own housing, at your own expense.
- What you bring
- Your own AI subscription and favorite chat harness. The methods are tool-agnostic.
- Admission
- Curated, by interview.
- Investment
- $4,997 USDCovers lodging, meals from Monday breakfast through Friday lunch, and materials to keep forever. Travel to and from the venue is your own. How meals are served — a private chef or delivered meals — is still being decided; either way, they're covered.
Request an invitation
Applications open to the waitlist first · February 14–20, 2027
From the calls
Asked before you ask.
The things people ask on the interview calls, answered straight.
What's the demographic of the group?
A small, curated room: capped at twenty, put together for a mix of knowledge fields, by interview. The current cohort skews science and tech, and everyone in it is a working professional, from a few years out of school to around twenty years in. What everyone shares is a field they've mastered and a field they're ready to be a beginner in, and no two people bring the same home turf. If your field isn't science or tech, the mix needs you.
Is there a corporate versus non-corporate rate?
Not right now. Tuition is one rate, $4,997, whether you pay it yourself or your employer does.
How should I get prepared?
The real preparation is choosing your field: the thing you'll spend the week learning, outside your expertise. A shortlist is enough, it can change right up to arrival, and we'll help you gather real source material for it. Bring the AI subscription you already pay for and the chat app you like; the methods are tool-agnostic, and there's no pre-course and no skills test.
The only named reading is the introduction of How Learning Works, your course textbook. You get it on arrival, and reading it before or during Day 1 are both fine. Over-preparing is welcome; knowing very little about your topic is fine too. By Friday you'll stand behind a small artifact and teach it to the room for about five minutes.
Request an invitation
Up to twenty seats, admitted by application.
Applications open to the waitlist first. One email when they open, one reminder before they close.
When applications open, we'll invite you to a short interview. That's how we curate the room. And the waitlist itself is where the community of learners starts.