Skip to main content

Community of Practice

A living, people-centered organism that compounds. A room full of people actively applying a craft, sharing what is working, and making each other better in real time. Not a meetup. Not a Discord. A thing that grows.

What It Is​

A community of practice is a group of people who share a craft and are actively pursuing mastery of it together. The word "practice" is load-bearing. These are people doing the thing, not just talking about the thing.

Three signals that a room is a real community of practice:

  • People are building something when they walk in. Their laptops are open. They have a project, a client, a workflow they are trying to improve. They came here because being around other practitioners is how they get better faster.
  • Field notes are the currency. Nobody is selling a course from the front of the room. Nobody is expert at something that is a month old. People trade what they tried this week, what worked, what broke, what they read at 2am that changed their thinking.
  • Money flows into the room, not out. Members are earning from the craft, and they pass work and clients to each other. The community is not a lead-magnet funnel. It is the place where the economy of the craft actually happens.

A group chat that only talks, a conference that only lectures, and a meetup where everyone is trying to sell each other something are three common things that are not communities of practice, even when they call themselves one.

Why It Matters For Applied AI​

Nobody is an expert in applied AI right now. The ground moves every month. Tools that mattered in February are obsolete by April. Practices that were best-in-class six months ago are the reason people feel stuck today.

In a landscape like this, comparing field notes with other practitioners is not a nice-to-have. It is the only way to stay current. A single person alone will eat months of lost time that a community member would have saved in a five-minute conversation.

This is why co-teaching is the new self-teaching and why compounding docs matter more than polished ones. The community is the update mechanism.

The Anti-Patterns​

A community of practice degrades when any of these creep in:

  • Grift. Someone using the trust of the room to sell something that does not actually work. One bad actor poisons the signal for months.
  • Talking-about-talking. When conversation drifts from "here is what I tried this week" to "here is what I think about the industry," energy leaves the room.
  • Gatekeeping as status. Charging for access, rank-based permissions, or "you cannot speak unless you have built X." The point is to make everyone better, not to preserve a hierarchy.
  • Consumer mode. When the majority of the room is there to receive rather than contribute. Every community of practice has a contributor-to-consumer ratio it can sustain. Past a certain point, contributors burn out and leave.

The health check: does a new member walk in, offer one real thing within their first few visits, and get something real back? If yes, the community is alive. If new members observe silently and never contribute, the community is becoming an audience.

How To Build One​

You do not build a community of practice by declaring it. You build it by running the room it requires.

  1. Pick a craft tight enough to have shared problems. "AI" is too wide. "Applied AI for founders and operators" is tight enough that people recognize each other's bottlenecks.
  2. Open with a real thing, not a pitch. Every event starts with someone showing their actual work, in progress, with the rough edges visible. This sets the floor.
  3. Make field-note sharing the default ritual. Round-the-room updates. Live problem-solving. Small demos over big talks. The Applied AI Live format was designed around this.
  4. Protect the signal. Be willing to say no to sponsorships, speakers, or formats that turn the room into an audience. The soul of the community is in what you refuse.
  5. Close the economic loop. Pass work to members. Introduce clients to practitioners. A community of practice where nobody is earning from the craft will not survive the first hard month. A community of practice where members routinely make each other money is nearly impossible to kill.

What Makes a Community Magnetic​

When a community of practice is working, people can feel it the moment they walk in. A few ingredients reliably produce that magnetism:

  • Truth-seeking as the default posture. The room is allergic to performative confidence and comfortable with "I do not know yet." Real learning happens when ego is cheap.
  • On the actual frontier edge. What gets shared tonight is not last year's playbook. It is what people tried this week. The room keeps you current because everyone else is.
  • Floor raised, ceiling uncapped. New members get the benefit of every good pattern the community has discovered. The floor rises fast. There is also no cap on how far a member can take their craft or their career. The strongest practitioners keep leveling up inside the community and pulling others with them.
  • Rising average economic viability. Learning is the starting condition. As the community compounds, more members are earning real money from the craft, and the community network is how opportunity flows. Pass clients. Share introductions. Close the economic loop.
  • Inclusive and welcoming. High standards and open doors coexist. The door is open to anyone who shows up in good faith, willing to contribute. The standard is on what you do in the room, not where you came from.

Austin is the proof point. Bordeaux is the international one. Everything else is the playbook.

The Role of Applied AI Society​

Applied AI Society wants to help you have an easier time starting communities of practice around applied AI. The chapter playbooks are an open-source specification for how to run one. The philosophy canon is the shared ethos. The public case studies and field notes are the compounding knowledge layer.

If you are thinking about starting one in your city, talk to us. That is what we are here for.

Further Reading​