A small, growing catalog. Each title earns its place. Notebooks, manuals, field guides — the kind of writing that takes a position and defends it.
A working journal of how an architecture decision becomes a product, becomes a company, becomes a brand. Drafted live in the Cognielo chat module — the same surface our users will see, used by the founder to keep an honest record of what was decided when, by whom, and why.
Adaptix is not a how-to guide. It is closer to a captain's log: the actual sequence of trade-offs, the dead ends, the calls that turned out to be right for reasons different from the reasons we made them. The patent provisional, the threshold-recovery cluster, the lane-guard architecture, the brand voice, the pricing model — every load-bearing decision in Elo AI's first year is recorded here.
The current draft lives in our internal repo at ~/Desktop/Book/entries/. Publication target late 2026. Pre-orders are not open yet — we will not take money for a book that hasn't shipped.
The shelf is short by design. The next titles will live here as they take shape. We would rather publish three good books than thirty filler ones. If you have a topic you wish someone would write a book on — privacy-preserving AI, threshold cryptography for non-cryptographers, the legal-tech-software seam — we are listening.
Software companies are not normally book publishers. We are publishing because the kind of software we make requires a kind of explanation that does not fit on a marketing page. The substrate-locality covenant, the methodology-as-claim approach, the operator-blind P2P transport — these are ideas that need slow, careful pages to make sense of.
A book also forces a different discipline. A landing page can be wrong and you fix it next week. A book has to defend itself five years from now in a coffee shop. We want the things we are saying about our own software to survive that test.