An independent lab for practical AI
Open-source tools, research notes, and practical experiments for people building with AI.
Agents / Models / Economics / Governance
01 / The open lab
See how AI agents work.
And why they fail.
Watch a failure. Inspect the cause. Apply one repair and replay. Three browser-based simulations make the invisible parts of an agent system visible.
02 / Observations & explanations
A missing response does not mean an action failed. Follow one request all the way to two tickets—and the repair that prevents it.
Token prices are only the beginning. Count the tools, retries, and failed attempts behind a useful result.
Start with a real task and a clear definition of success. A useful comparison ends with a decision you can explain.
Start with jurisdiction, your role, and the intended use. Learn to separate binding rules, proposals, and voluntary guidance.
03 / The reading room
One paper in depth. Four more in perspective. What the evidence says, where it stops, and why it matters for builders.
Read the collectionConnecting a plan to observations from the world.
2022 · In depthLearning when an external tool is useful.
2023 · Research noteHaving context and using it are different abilities.
2023 · Research noteCarrying feedback into the next attempt.
2023 · Research noteChecking whether an agent reliably did the right thing.
2024 · Research noteKeep a thread of curiosity open
Get new writing in your feed reader, or follow the projects on GitHub.