Full report
Personal
RAG vs Markdown wiki
Which method suits a knowledge base better: a RAG system, or the Markdown LLM Wiki idea used through Obsidian?
Compared4
Open3
Obsidian / Markdown / Vector searchResearch note
Catalyst · Bloom on arrival
Lab / Explorations in progress
Small experiments. Useful outcomes.
Every experiment I have run, with what it was testing, what came out, and what it still has not answered. Tools and products live on the portfolio.
- Experiments
- 3
- Aspects compared
- 12
- Open limits
- 10
- Published as
- Research notes
The exploration index
Notes from the lab.
3 experiments
- Full report
Personal
AI Behavioral Runtime
Can a defined set of principles aim a model's reasoning in a better direction, without touching the reasoning itself?
Compared4Open4Reasoning runtime / Gemini 3.1 ProResearch note
- Full report
Personal
Workflow Runtime
Should the model design a workflow for each problem, or run a workflow that was designed and tested in advance?
Compared4Open3Agent orchestration / ClaudeResearch note
How the lab works
Think. Make. Refine.
Catalyst / Identity
Independent parts. One point of discovery.
Six outer nodes joined into three pairs around one coral nucleus. Bloom on arrival, Ripple on interaction, still the rest of the time.
- 01
Think.
Start with a useful question. Name the recurring problem before choosing a form.
- 02
Make.
Give the idea a tangible form. Keep the composition quiet enough for the work to stay readable.
- 03
Refine.
Learn from what happens next. Keep what is useful. Drop what is only decorative.
