Method
How a rough observation becomes a published piece, and how to read the labels.
Evidence labels
Each piece carries exactly one label. If a piece mixes kinds of evidence, I split it.
- Lived experience
- I was there. Details are anonymized.
- Verified
- Checked against named sources, listed with the date checked.
- Assumption
- A working hypothesis, not yet tested.
- Illustrative example
- An invented scenario. Numbers are not real.
Four kinds of piece
- Note. A short observation or concept. May stay private.
- Research. A topic investigated from sources, with the sources listed and dated.
- Experiment. A hands-on test of an AI tool, agent or workflow on a real buying task. Protocol first, results second.
- Case study. A real or illustrative problem: Problem, Background, Constraints, Alternatives, Analysis, Decision, Result, Lessons. Not every case needs every part.
How a piece matures
Pieces carry a stage, borrowed from digital gardens. Seedling means early and likely to change. Growing means taking shape. Evergreen means I would stand behind it today. Behind the scenes, an item moves from Inbox to Researching, Analyzed, Ready to publish and Published.
Confidentiality
Real projects are described in general terms, such as "a leading energy company". Names, prices, quantities and anything else that could identify a client are removed or changed, and the piece says when they have been. Anything private is never published.
Two languages
Every piece is published in English and in Türkçe. A piece appears in a language only when its translation is ready.
How AI is used here
AI helps with search, summarizing, drafting and translating. It does not replace checking a source, and it does not decide what a piece claims. Where an experiment involves AI output, the piece says what was checked by hand.
Corrections
If something is wrong, tell me and I will fix it and note the change on the piece.