2026-07-25
When GPS Stops at the Waterline: What Scuba Diving Taught Me About AI
A dive in Kos became a practical lesson in underwater navigation, human judgement, AI estimation and why human-in-the-loop systems still matter.
A Neuvottelija AI series · since April 2026
OpenClaw Chronicles
The original series remains a first-class part of Neuvottelija AI. It documents Samantha, Stöbä and Hermes as a real multi-agent lab, including memory, security, migrations and human approval gates. Existing article URLs are unchanged.
2026-07-14
The Last Mile: Improving Agent Readiness from 86% to 100%
The final fourteen points of Cloudflare's Agent Readiness benchmark were harder than the first eighty-six, and they were closed by three frontier models working in parallel: Claude Fable 5 built the English episode corpus, Google Gemini signed the domain and published DNS-based agent discovery, and Codex GPT-5.6 Sol fixed the OAuth metadata, favicon and retrieval quality.
2026-07-11
From 57% to Agent-Native: How This Site Took Over www.neuvottelija.fi
Why a Lovable-hosted React SPA scored 57 while a Cloudflare and Astro static-first site reached 86, and why agent readiness is an architectural property rather than a feature toggle.
2026-06-01
Improving Sami-to-Samantha Security
MemoriLabs graph memory was separated from group context and re-enabled only for Sami's private DM via a plugin-level exact-session allowlist. Tool use is trust-gated. New T1–T5 trust model: no group should be able to become T1.
2026-05-18
Pearl, Friston, and the Strategic Sound of a Cooling Fan
Four AI workers debate Pearl-style vs Friston-style agent ontology. Tapio Nissilä adds HAR as a third axis: per-path supervision (Audit / Monitor / Approve / Command), not per-agent autonomy.
2026-05-10
Stöbä's Brain Transplant: Moving from OpenClaw to Hermes Agent
Stöbä, a secondary AI worker running on an Ubuntu ThinkPad, was migrated from OpenClaw to Hermes Agent using a clean direct cutover. The migration preserved identity, memory and selected skills while restoring Telegram, WhatsApp and Discord as control channels.
2026-05-03
Hermes on Sami's MacBook Pro: A GPT-5.5 + Gemma 4:26b Hybrid
The Hermes agent on Sami's MacBook Pro: GPT-5.5 for thinking, a local Gemma 4:26b for light tasks and fallback. A controlled hybrid — no automatic routing.
2026-04-28
Hermes Joins Samantha: A Research Agent on the Same Mac mini M4
Samantha gains a bounded research delegate, Hermes, on the same Mac mini M4 — separate macOS user, restricted wrapper, shared handoff folder, Exa via MCP for web research.
2026-04-24
Stöbä Became a Junior Investment Banker with Curated Data on IPOs and SaaS Valuation
Curated local safe memory in practice: Stöbä retrieves IPO readiness and Translink SaaS valuation material as a junior investment banker would, with sources and discipline.
2026-04-22
Showing Your Added Value Openly in the AI Era
In the AI era, added value comes from understanding and the ability to explain — not from mere output. Open documentation is the new distribution channel for expertise.
2026-04-19
Samantha's First Commit
Samantha got her own git identity (samanthamiettinen) and pushed her first commit. Stöbä received four local memory layers on the ThinkPad.
2026-04-18
Why Nathan B Jones Matters to Our Second Brain
Personal context is the moat, capture is cheap, curation is the work, and trust is built by retrieval. Nathan B Jones' framing applied to Samantha and Stöbä.
2026-04-17
ClawHub Is a Quiet Revolution — and Samantha Wrote Part of the Code
ClawHub is a versioned skill registry for OpenClaw agents (55,862 skills). At the morning AI Campfire we presented the OpenClaw architecture and the Watch Them Talk view.
2026-04-17
Samantha and Stöbä: Two Agents, Two Memories
A two-agent architecture: Samantha carries the long context while Stöbä acts as the fast executor. Two separate memories reduce errors and cost.