Mar 6
Loading video player...
Most people use Gemini like a chat tool. But when you combine **Gemini + Gems + NotebookLM** intentionally, it stops being a chatbot and becomes a system. In this video, I show how to design a 3-layer AI workflow where: - NotebookLM holds stable knowledge - Gem holds thinking rules - Gemini handles execution Using a fictional product example, I walk through how to build a source-of-truth notebook, design a reusable Gem, and run multiple tasks without rebuilding context each time. This is about reliability, not prompts. Relevant Video: 3 Ways to Design Gemini Gems: https://youtu.be/cIH-iS0J0b4 10 Practical Gem Ideas for Everyday Work: https://youtu.be/irSKKXp3zx8 Master NotebookLM in 16 Minutes: https://youtu.be/5Tnc6hQDbvE Gemini Deep Dive (Workspace + Visual features): https://youtu.be/aFG2es9M9VI 00:00 Hook — From Chatting to Systems 00:39 The Three Thinking Layers 02:07 Two Design Principles 02:44 Set up Projects: NotebookLM & Gem 06: 58 Scenario 1 → Same notebook & gem + simple task 08:07 Scenario 2 → Same notebook & gem + enriched session 11:05 Scenario 3 →Two notebooks, same gem, expanded system 14:58 Final Thoughts