Live experiment: a tiny harness, five tools and Markdown to give your AI a personal knowledge library
dfinke · x · 2026-09-08
Doug Finke, 16-time Microsoft MVP and creator of ImportExcel, will livestream an experiment exploring how little machinery an AI agent needs before it behaves like it has a brain you actually own.
- The core idea: most AI conversations are disposable — the agent finishes the work, then forgets everything when the session ends. His setup gives the agent a knowledge library it can organize, search, retrieve, and use while working, so your work becomes a library and the AI just needs to find the right book at the right moment.
- Stack: a tiny harness, five basic tools, Markdown, and one LLM — no frameworks.
- He nods to Garry Tan's description of G-Brain as "the library plus the librarian" and Andrej Karpathy's LLM Wiki blueprint.
- The public livestream (Thursday, 4:30pm ET) is the complete experiment; deeper code, prompts, retrieval choices, and provenance questions continue in his paid Agentic AI Forge community.
Related event: Doug Finke's Experiment: Giving AI a Compounding Personal Knowledge Base(3 posts)→
More from coding & agent
- Jane Street's Yinsky shares tips for making codebases agent-friendly: types, property tests, formal methods — sh_reya · 2026-10-27
- Celesto launches CI runners nearly 90% cheaper than GitHub Actions, with agent sandboxes — aniketmaurya · 2026-10-27
- Runway launches App Studio: build no-code creative apps that generate image, video and audio — runwayml · 2026-10-27
- Supabase founders on rebuilding the backend for a billion AI-built apps — Latent Space · 2026-10-27
- Dev scans 5,000 MCP servers daily: 10% dead, 30% auth-gated, then ClaudeBot shows up — art2meta · 2026-10-27
- Solo dev ships national park sim with Claude in 6 months, Opus spawned 560 Haiku subagents to translate it — idun0 · 2026-10-27