Insights
Writing on the agentic web
Longer pieces on specific questions - comparisons, status reports and a few arguments. Each one takes a position rather than surveying opinions. No newsletter, no gated downloads.
Web3 vs Web4: What Actually Changes
Web3 changed who owns the data. Web4 changes who acts on it. A concrete comparison across ownership, execution, identity and payments.
How AI Agents Will Change the Internet
Most of the web assumes a human reader with a mouse. Agents break that assumption - and with it the business model that funded it.
What Is Agentic AI?
Agentic AI is not a bigger model. It is a model placed in a loop with tools, memory and a goal - and that changes what can go wrong.
Do AI Agents Need Blockchains? A Sober Look
Four problems agents actually have, and an honest assessment of which of them a blockchain solves - and which it does not.
Autonomous Commerce in 2026: What Already Works
Separating the parts of agent-driven commerce that ship today from the parts that are still slide decks. A status report, not a forecast.
When Machines Answer the Phone
Voice is the oldest customer interface and the last one to be automated convincingly. What changes for callers, staff and the business.
Digital Identity for AI Agents
An agent acting for you needs to prove three things: who it is, who it acts for, and what it was allowed to do. None is solved by an API key.
Can AI Agents Make Payments?
Technically yes, and several rails already work. The open questions are authorisation, dispute handling and who is liable when an agent is wrong.
Provenance of AI-Generated Content
Detection is losing. Provenance - signed, verifiable claims about where a file came from - is the approach that scales. How it works today.
The Protocols the Agentic Web Runs On
MCP, A2A and x402 solve three different problems: tool access, agent-to-agent messaging and machine payments. What each one actually specifies.