Give Me 9 Minutes, I'll Make You AI-Native

About this video

Jeff Su lays out three levels of working with AI — AI curious (free tier, uses chatbots only when reminded or stuck), AI literate (pays for AI, keeps a prompts database, knows which feature and model to use), and AI native (has redesigned workflows assuming an AI collaborator exists) — and argues most professionals are stuck at level two. Habit one, leave AI breadcrumbs: instead of disposable chat threads, copy the conversation’s URL (Command or Control+L) and hyperlink it (Command or Control+K) into the document where the output is actually used, with a short context note beside each link explaining why it matters; his presentation-prep doc keeps links to the chats that optimized his prompt, brainstormed the outline, applied storytelling principles, and refined talking points, and his Notion project pages link out to the corresponding ChatGPT and Claude projects. The organizing principle is to file information by where you will use it, not where you found it, and his rule of thumb is to anchor any conversation that took more than 10 minutes or produced something you will reference again. Habit two, build an AI swipe file system: maintain a curated library of excellent examples so prompts start from a model of good work instead of a bare instruction — share saved business proposals and ask the AI to analyze the key patterns in structure and tone, then apply those patterns to your new content. He says he used this at Google for important presentations, uploading slide decks from McKinsey, Bain, and BCG so senior leaders from those firms assumed he had worked there. Save strong work to the swipe file whenever you encounter it, start narrow with two to three repeated use cases such as presentations, emails, or reports, and organize folders by use case rather than source or date. Habit three, AI-first task planning: before a big piece of work, break it into steps and microtasks, decide for each whether it is manual or AI-assisted, and if AI, specify the exact tool.

Key takeaways

Anchor reusable AI chats where the work lives: paste the chat hyperlink (with a note on why it matters) into the document or project page that uses its output.
Keep a swipe file of excellent work organized by use case, and prompt AI to first extract the patterns from your examples before applying them to new content.
Plan AI use before starting any project over an hour: list the microtasks, tag each AI or manual, and name the exact tool for each AI task.
Match tools to task types in the plan — source-grounded tools for fact-checking, creative models for drafting and structuring.
Maintain a central prompts database by use case and build toward 10 to 15 battle-tested daily prompts instead of collecting random ones.

Episode details: published December 4, 2025 · 8:55 long. Summary written by Self Growth Videos from the episode itself. Business results and any income figures mentioned are Jeff Su’s own claims, shared for education — they are not typical or guaranteed.

← Previous: Master 80% of Notion with this ONE Feature · Next: Master Gemini 3.1 for Work in 12 Minutes (2026) →

All Jeff Su videos · Jeff Su profile · Jeff Su (official site)