I start with learning objectives. Then I teach in public.
I test real second-brain and knowledge-system approaches (including current Astra-style and Obsidian + Claude methods), publish the comparisons, and extract the frameworks that actually survive use.
The public work becomes the proof. The extracted frameworks become the offers. The deeper implementations become SecondBrain XL.
Everything here is built from that loop: test → compare → synthesize → productize → implement.
See the public work