February 10, 2026Traits
A trait model for effective AI
A four-layer framework for evaluating what makes an AI system genuinely effective — from governance and architecture to behavior and raw capability.
Research
Notes and frameworks from the team, published as we develop them.
January 28, 2026World view
Why data ownership has to come first
A short note on why we treat data sovereignty as a precondition, not a feature.
January 12, 2026World view
Notes on building AI that defers to you
Some early thinking on what it actually means for a system to operate under a user’s declared worldview.
December 20, 2025Morality
A note on moral frames, not moral defaults
Why we ask the system to reason from a stated moral frame instead of shipping one baked in.
December 5, 2025Generalization and portability of learning
Portability without degradation
Early notes on what it takes to move a model and its data between providers without losing fidelity.