About Cupel and its author

Cupel began as a technical answer to a narrow question: once AI can pass the exams and produce the portfolios that credentials were built to certify, how do we verify that a human can still catch its mistakes? That question turned out to sit inside a much bigger one — what happens to people, careers, and the meaning of expertise as knowledge work becomes cheap to automate. Answering the narrow question honestly requires engaging with the bigger one, in public, with evidence.

This site is where I do that. It publishes three things: the Cupel framework itself (an open specification, not a product), Field Notes — commentary that reacts to new data and developments as they land, and a running State of the Conversation page that links the primary sources worth reading on AI provenance, credentialing, and the human impact of automated knowledge work.

I am not trying to build a company out of this. Cupel is open-source (Apache 2.0) and trademark-protected (UK IPO No. UK00004352899) so that the name keeps meaning something as the project grows, but the framework itself belongs to whoever wants to build on it. What I want is for this to be a credible, evidence-first place to follow and contribute to this conversation — and for that work to be identifiably mine.

I write and review everything published here. Disagreement, corrections, and counter-evidence are welcome — the fastest way to reach me is by opening a discussion or issue on GitHub or by emailing contact@cupel.foundation.