The New Knowledge Worker: Thriving in the Age of AI Agents

Knowledge work is changing fast. I gave a talk today about what it actually means to work effectively in 2026, when AI agents can write code, run research, manage memory, and coordinate with tools on your behalf.

The talk covers the Agentic AI Stack — the mental model I use to think about how humans and AI systems work together — and practical strategies for becoming a more effective knowledge worker in the age of agents.

The diagrams and materials from this talk are here.

Timestamps

• 00:00:00 Introduction
• 00:02:42 The Progression of AI: 2022 to 2026
• 00:11:06 Introducing the Agentic AI Stack
• 00:11:57 Layer 1 — The Model (Frontier vs. Open LLMs)
• 00:12:22 Layer 2 — The Harness (Claude Code, Codex & more)
• 00:13:52 Layer 3 — Memory (Markdown, CLAUDE.md, Obsidian)
• 00:16:33 Layer 3 — Skills
• 00:18:51 Layer 3 — Connectors & Other Features
• 00:19:54 Layer 4 — Agentic Knowledge Work Use Cases
• 00:24:26 Agentic Brains
• 00:26:57 Demo: Codex Chief of Staff
• 00:32:15 Mindset
• 00:38:33 Read the Docs
• 00:40:37 Downsides & Pitfalls of AI
• 00:43:08 Q&A / Wrap-up

Notes

https://github.com/lawwu/ai-tutorials/blob/main/04-new-knowledge-worker/NOTES.md

Resources

• Reflections on AI: 2022 to 2026
• Top 10 Human Skills in the Age of AI
• Karpathy’s LLM Wiki
• gbrain
• Claude Code Docs