You’re Not Behind (Yet): How to Learn AI in 19 Minutes

TL;DR

Follow this 5-phase, 3-month plan to go from AI beginner to automating entire workflows without getting overwhelmed.

Key Points

• 1.Phase 1 (Week 1) — Foundations: Pin Claude/ChatGPT as a tab, use voice dictation instead of typing, download the mobile app, and auto-record meetings with tools like Grain or Fathom.
• 2.Phase 2 (Week 2) — AI as Coach: Don't ask AI to do your work yet. Use it as a thought partner by giving it your role and goals, then letting it interview you to find high-leverage priorities.
• 3.Phase 3 (Weeks 3–4) — AI as Worker: Apply the 10-80-10 rule: you do the first 10%, AI does the middle 80%, you review the final 10%. Always provide rich context (transcripts, competitor examples, strategy docs) to avoid generic output.
• 4.Phase 4 (Months 1–2) — AI as System: Build a prompt library by iterating prompts like a recipe — test output, fix what's weak (e.g. "hooks under 20 words," "no rhetorical questions"), save as a keyboard shortcut using tools like Text Expander.
• 5.Phase 5 (Month 4+) — AI as Infrastructure: Automate repetitive tasks using Zapier or Make.com (simple connectors), then N8N for advanced workflows — e.g. auto-generating weekly student progress reports from call transcripts and Slack messages.
• 6.The key skill separating good from bad AI users is taste — if the AI output feels like cringe, that's your quality bar working; give feedback like you would to a junior intern rather than accepting the slop.