In July 2026, Anthropic showed that a language model holds thoughts it never says out loud - and that you can read them. That is a genuine breakthrough. But reading the thinking is not the same as knowing how and where the thoughts were formed.
The architecture of identity - what a system loves, ignores, fears and blindly assumes. Identity drives interpretation, and so intent.
Ethos DNA. The quasi-genetic constituents whose content, architecture and dynamics terraform an ethos from within i-Space.
The study of the ethoic substrate - what it determines, how it is structured, and how it may be read.
The deliberate re-formation of that substrate. Already practised everywhere, named almost nowhere, and licensed by no one.
i-Space Field Lexicon: A Framework and Priority Statement. Barry E James. Working paper, deposited open access 22 July 2026, CC BY-NC-ND 4.0. Indexed in OpenAIRE.
10.5281/zenodo.21498878 Read the deposit
The framework was not built for AI. It was developed over years to understand ethos in people and in the institutions they build - what it is, how it forms, and why it is so rarely visible to those inside it. The convergence with machine interpretability was not sought. It arrived.
The author originated the ethos innovation behind the regulatory sandbox concept, proposed to the FSA in 2012 and since adopted across more than fifty jurisdictions. The artefact was copied everywhere, the method that produced it now has a name.
We are now planning the inaugural i-Space conference 2026. If you're interested in contributing or attending let me know and we'll keep you updated.
Occasional, substantive, updates. No list sharing, ever.