For nonprofits and associations

Most organizations use AI to go faster. Practice using it to think better.

You inherit the archive of a project at a fictional nonprofit: meeting transcripts, board decks, budgets, emails, field notes. Something went wrong, or went right for reasons nobody wrote down. Your job is to figure out what really happened, using AI as a thinking partner rather than a shortcut.

Why a simulated archive?

Every leader inherits messy context: a strategic plan built on assumptions nobody tested, a summary that says the opposite of the research underneath it, a spreadsheet whose first tab is careful and whose later tabs are wishful. Nobody is taught to read that inheritance. Courses about AI mostly teach you to write a prompt and save an hour. That is useful, and it is not where AI changes how an organization decides.

A simulated archive is a flight simulator for judgment. It is realistic enough to argue with and safe enough to be wrong in. No real member, donor or client data is involved, so you can practice freely and bring the habits back to work. Each vault is built on purpose: some documents hold up, some are noise, and some are true but shaded by someone protecting a budget or a reputation. The real lesson is never in the headline. People who cross-reference find it; people who skim do not.

What you practice

The point is not the tool. It is a set of thinking moves people have used for decades, which AI makes cheap enough to run on every important decision. In the Thinking Studio you pick a move and an AI coach works the archive with you. It reads documents, asks the next question and points you to what to read side by side. It will not hand you the answer.

  • Assumption → hypothesis

    Find what a plan quietly takes for granted, then rewrite it as something you could test and be wrong about.

  • Inversion and premortem

    Ask how you would guarantee failure, then check the record for which of those things actually happened.

  • Six Thinking Hats

    Walk one decision through facts, feelings, risks, benefits, alternatives and process, and notice the hat nobody wore.

  • Systems thinking

    Trace the loop behind the outcome and tell a symptom fix from a leverage point.

  • Red team / steelman

    Build the best case for the option that was rejected, and attack the one that was chosen.

  • Cynefin sense-making

    Decide whether a problem was clear, complicated, complex or chaotic, and whether the response fit.

How it works

  1. Open a vault and read the premise. Browse the folders like you would on your first week in a new job.
  2. Pick a thinking move and work the problem with the coach. Every claim should trace back to a document.
  3. Submit a mission, such as “list the five load-bearing assumptions and turn each into a testable hypothesis.” You get a score, the strongest move you made, and one specific thing to fix.
  4. When you get it, the debrief shows what the vault was built to teach.
  5. Download any vault as a folder for Obsidian or Claude Cowork and run the same moves with your own tools.

Build a vault that mirrors your organization

The vault generator lets a team describe its own kind of scenario, such as a membership decline, a stalled system migration or a credential losing value, and tune the simulation: the balance of evidence, noise and motive, the number of planted errors, how deeply the lesson is buried, and which thinking moves it should exercise. Sealed mode hides the answer even from the person who configured it, so a leadership team can build one and investigate it together.

Who it is for

Executive directors, program leads, association staff and boards who want AI to strengthen judgment, not only throughput. The scenarios are written for the nonprofit and association world, including the Washington, D.C., Maryland, Virginia and Chicago sectors.

Start with the free sample vault.

No account needed. One mission and two thinking moves are free.

Open the sample