Notes
Short pieces, one question at a time. Each has the same shape. It gives the answer in brief, what is happening, why it matters, what to watch and every source I used. I draft them with AI assistance and review and edit each one myself, as set out in how I work.
What do these notes on AI cover?
They cover AI for people who make business decisions. The subjects are regulation, agents and how to read vendor claims. Each note takes one question and has the same shape. It gives the answer in brief, then what is happening, why it matters, what to watch and every source used. I draft them with AI assistance and review and edit each one myself.
-
The EU AI Act timeline as it stands in October 2026
What already applies and what moved to 2027 and 2028. The next date is 2 December 2026.
Regulation
-
The same question, put to ChatGPT, Claude, Gemini and Perplexity
On the same 749 prompts, four assistants split on whether to name a brand 27.2% of the time. Figures from LLM Scout.
Research
-
AI agents and the Model Context Protocol, in plain terms
What agentic means, what MCP is, who governs it now and where agents still fail.
Agents
-
How to read an AI benchmark claim
What a score measures, how scores mislead and seven questions to ask a vendor.
Evaluation
-
Your analytics is hiding your AI search revenue
GA4 gave Claude 0.1% of conversions. 10.6% of customers named it. Why analytics misses AI search and the question to add.
Measurement
-
Fable 5.2 for business: what changed, what it costs and when to use it
What changed, what it costs on each plan, the 16 November deadline and when to pick it over Opus 5.5.
Models
Also here
If you are looking for the underlying documents, the reading list collects the primary sources these notes rely on. New notes are also published to the RSS feed.
AI strategy / Implementation / Training
Work with me
For businesses that want AI built into their day-to-day work. It starts with a free 30-minute discovery call.