01
What can a room do that the internet can't?
ObservedMost of what a community does already happens online. Announcements, arguments and links all live in group chats and feeds.
InferringA gathering that only repeats the internet is a group chat with snacks. A room earns its place by doing what a feed cannot.
InferringOur guess is three things. Trust: people commit to people they have sat across a table from. Mixing: an ecologist and a developer never meet in the same chat, but they can share a room. Local problems: a small trader's or a clinic's problems are never posted anywhere.
02
Who is it for?
ObservedTwo crowds usually show up. Students arrive in groups, often for food and swag. People with a few years of work behind them mostly stay home.
ObservedThe people with the most interesting problems rarely come at all: the ecologist with years of survey sheets, the accountant who knows where small-business credit breaks, the lawyer reading property filings.
InferringSo the room is for people who hold problems and can work on them, inside tech and outside it. Students matter as future builders, not as the measure of success.
03
Why would anyone give their time?
PremisePeople give their time for some mix of money, learning, good company, recognition, purpose and opportunity.
ObservedSmall payments can do worse than none. They replace people's own reasons with a price.
ObservedPrograms built on ranks, missions and points lose the people with other options first.
InferringSo no pay, no points. What a room can offer instead is scarce and real: a seat in a smaller room, an introduction, a wider audience for your work. Those go to people who contribute, by presenting, shipping, mentoring or bringing a problem worth working on.
04
What does it take for something to get built?
InferringFour things: a real problem, people with different skills, enough time together, and tools they can use.
ObservedA meetup gives you people but not time. The internet gives you time but not trust. Hackathons give you time, usually around an invented problem.
ObservedAI has changed the last ingredient. Someone who has never written code can turn a spreadsheet into a dashboard over a weekend, if someone sits beside them.
InferringCode is no longer the hard part. The hard part is getting a real problem, different hands and enough time into the same place. That is the work.
05
What we don't know yet
- Will people come for ideas and company alone, with no money involved?
- Can people who don't code really build now, or does every table still need a developer?
- Will a local business trust a room of strangers with a real problem?
- Is a seat, an introduction and an audience enough to keep someone around for months?
- Does what a room makes outlive the room?
06
How we hold a room
Frugal before flashy.
Spend on the people, not the banners.
Small rooms, no audience.
If you came to watch, the room is too big.
Different hands.
Biology, rivers, hardware, law, games. One table, one problem.
The place teaches.
Local culture is not the backdrop. It's half the faculty.
The work is public. The people are not.
Names go on what ships, never on a leaderboard.
The human is the infrastructure.
Sleep, food and rest are part of the design, not a break from it.
07
Reading by the fire
Books and papers that shaped these questions. They sharpen the questions; they don't prove our answers.
Goliath's Curse
On how much of human history was lived in small, mobile, loosely held groups, and what changed when we settled.
Situated Learning: Legitimate Peripheral Participation
Learning as a way into shared work, not a transfer of facts. A newcomer needs a real task, not just a seat.
Cognition in the Wild
A ship's navigation team, where the thinking lives in people, tools and the room together.
Pay Enough or Don't Pay at All
Why a small payment can do worse than none.
Beyond Markets and States
Shared things can be governed by the people who use them, without a single centre of authority.