AIs are already telling your story to those who don't know you yet. In these notes, I analyze how to govern that narrative: essays, analyses, and technical guides on semantic architecture, Generative Engine Optimization (GEO), and the transition from the web of content to the web of agents. For those who want to be read by machines as fact, not hypothesis.
Many professionals want their own AI assistant, a bot that speaks on their behalf, and imagine they need to "train" it on their data. That is the wrong part of the problem. A bot's intelligence is borrowed from an already capable model, while what makes it reliable is another layer: the verifiable knowledge of the professional.
From marketing to human customers to Business to Agent (B2A): why the next recipient is an AI agent that reads, evaluates and acts - and why brands win by declaring their identity in structured data. With the Patagonia case (owner: "The Earth").
The operational guide to B2A: from the Generative Web to the Agentic Web, how to become agent-ready, the B2C/B2B/B2A comparison, the agentic commerce protocol stack and the impact on CAC. With a disambiguation of B2A.
The underlying principle behind all other notes. What entity linking is, why it is not entity resolution, and how it defines the Semantic Entity Architect role.
How to build a machine-readable semantic identity for AI engines without a Wikidata item, using JSON-LD, Schema.org, llms.txt and official profile mappings.
Why customers look for suppliers and solutions by asking AI, and what a business needs to be found and cited correctly in AI-generated responses.
If you share a name with other people, AI systems can merge your histories. The fix isn't a better bio — it's building a verifiable entity.
Nonprofits are among the least readable entities for AI systems. Being found and cited accurately doesn't require more content — it requires structure.