Teaching practical AI.
A publication for people learning to build with AI. For professionals tired of the noise, and beginners who don't know where to start.
Recent notes
- № 06
What is NLWeb?
How Microsoft's open protocol turns your website into an endpoint AI agents can talk to natively, and why your Schema.org data is the input it needs.
NLWeb9 min →9 min → - № 05
What is AEO?
How Google's answer boxes, voice assistants, and AI Overviews pick which page to quote directly, and what Lovable handles for you so you can focus on writing in an extractable shape.
AEO8 min →8 min → - № 04
What is GEO?
How AI search engines pick which pages to quote, and what Lovable's SEO & AI search review handles for you so you can focus on writing in a way they can lift.
GEO7 min →7 min → - № 03
What is SEO?
How search engines actually find and rank pages, and what Lovable's SEO & AI search review handles for you so you can focus on the writing.
SEO7 min →7 min → - № 02
Building Hermóðr, my own agentic OS on Hermes
Everyone's talking about Hermes. I'm building on top of it: a Norse-themed agent OS where each agent is a god with one job.
HERMES3 min →3 min → - № 01
Four tools, four layers.
Softr, Lovable, Claude desktop, Claude Code CLI. What each one actually is, and the layer of the stack it owns.
MEMO6 min →6 min →
I'm Ken Ove Ferbu. I write about practical AI, drawing on 15+ years in IT and the past year building with Claude, Lovable and Softr.
Notes and walkthroughs, with the prompts and files behind each one. No hustle vocabulary. No AI-generated humans. Just the work, taught honestly.