Doug Finke's Experiment: Giving AI a Persistent Memory

Episode 1 · Doug Finke's Experiment: Giving AI a Compounding Personal Knowledge Base (2026-09-08, 3 posts)

Microsoft MVP Doug Finke is running a public experiment using a minimal harness of just five tools, Markdown, and an LLM to let agents accumulate knowledge instead of forgetting after each session, with a live stream scheduled for Thursday.

• Live experiment: a tiny harness, five tools and Markdown to give your AI a personal knowledge library — dfinke · 2026-09-08
• Doug Finke to Livestream 'Own Your Intelligence: Personal AGI' Experiment — dfinke · 2026-09-10
• Five tools, Markdown, one LLM: making agents compound knowledge you own — dfinke · 2026-09-10

Episode 2 · Doug Finke's Personal AGI Live Experiment: Five Tools Plus Markdown (2026-09-10, 9 posts)

On September 10–11, Doug Finke, a 16-time Microsoft MVP and author of ImportExcel, PSAI and "PowerShell for Developers", ran a public live experiment "Own Your Intelligence: Personal AGI" at a New York Agentic AI community Meetup, tackling the problem that AI sessions are disposable and agents forget everything when a session ends.

Confirmed

• The experiment was deliberately minimal: no frameworks, just a small harness, five basic tools, Markdown files and one LLM.
• In his warm-up posts, Finke argued that most AI conversations are one-shot—agents finish the task and forget what they learned the moment the session ends—and the experiment asked "what if it didn't have to be that way."
• The core idea: your work becomes a library, and the AI only needs to find the right book at the right moment—turning a user's work output into a knowledge base that can be continuously accumulated, organized and retrieved, giving the agent long-term memory instead of resetting to zero.
• Finke described the experiment as having "no framework, no magic", only just-enough mechanisms, to see whether the AI starts behaving as if it has a brain the user truly owns, running on the user's own infrastructure and available anytime.

Why it matters

Mainstream AI assistants generally lack cross-session personal knowledge accumulation. The experiment validates a minimal viable path—simple tools plus Markdown—for giving agents persistent memory, which is a useful reference for both personal knowledge management and the agent developer community.

• 16x Microsoft MVP livestreams Personal AGI: five tools and Markdown to make agents compound knowledge — dfinke · 2026-09-10
• Personal AGI experiment: a tiny harness, five tools, Markdown, one LLM to give agents memory — dfinke · 2026-09-10
• No frameworks, no magic: turning your work into a library the AI can actually search — dfinke · 2026-09-10
• Live Experiment: Building a Personal AGI with a Markdown Knowledge Library — dfinke · 2026-09-11
• Microsoft MVP Demos Personal AGI: Five Tools, Markdown and One LLM — dfinke · 2026-09-11
• A minimal agent memory harness: five tools, Markdown, one LLM — no frameworks — dfinke · 2026-09-11
• 16x Microsoft MVP Demos a Minimal Harness That Lets AI Agents Own Your Knowledge — dfinke · 2026-09-11
• Agent Forgets Everything After Each Session? A Minimal Fix, Demoed Live — dfinke · 2026-09-11
• 'No Frameworks, No Magic': Turn Your Work Into a Library AI Can Search — dfinke · 2026-09-11