MY FIRST AI TEAM — THE WEALTH EXPERIMENT 1. CHOOSE ONE RECURRING JOB Task: Who the finished work is for: One small finished output: Approved inputs I have: Acceptance checks: What I want this to make room for: Time limit for the whole first-week test: 2. BRIEF TWO ROLES Maker: prepare one small draft from approved inputs. Return the draft, source references, assumptions, and open questions. Reviewer: compare the draft AND original inputs against the acceptance checks. Return evidence for pass/needs-work and the smallest necessary corrections. My decision: verify important findings and approve the output. Sending, publishing, spending, deleting, and changes to live systems wait for my decision. 3. TRY A FIRST WEEK [ ] Choose a small sample and quality bar. Time a comparable manual run if it fits; otherwise leave the baseline unknown. [ ] Adapt the two role briefs and agree on the handoff. [ ] Try one complete maker-to-reviewer handoff. Record all active time. [ ] Check the result myself. Include failures and corrections in the effort. [ ] Choose keep, change, or pause. Leave one next action for the next session. These are planning steps, not a promise of task completion or time savings. Stop at my limit and shrink the scope if needed. 4. RECORD EACH ATTEMPT Date / sample: Was the output accepted? Which checks passed or failed? Manual baseline minutes (unknown is okay): One-time setup minutes: Briefing and handoff minutes: Review minutes: Correction / failure minutes: Additional tool costs: Actual customer response, if relevant: What remains unknown: Decision and reason: Next small action: WHEN I STALL Open the last version. Write one thing I learned and one five-minute action. Shrink the test or intentionally pause when it does not fit my available time or energy. Build an editable plan: https://wealthexperiment.live/tools/ai-team-planner/ Compare equivalent finished work: https://wealthexperiment.live/tools/productivity-calculator/ Routines and perspective: https://wealthexperiment.live/ai-team/#waymaker-practice