Epistemic Keys — Before AI Decides, Who Defines What Counts?

An invitation into a distinction that AI governance repeatedly overlooks: an individual can remain the final approver of a decision while the criteria determining what may count as an admissible possibility have been fixed elsewhere.

Epistemic Keys are six small public, interpretive instruments arising from my work in second-order cybernetics, Infosomatics, Sapiopoiesis and Trajectory AI. They do not substitute a verdict for judgment. They make selected assumptions about criteria, sourcehood, delegation and consequence visible enough to become questionable.

Admissibility Gate asks whether a consequential option has been excluded by criteria the responsible person could not revise. Sourcehood Map traces where a governing criterion originates. Trajectory Lens distinguishes increased performance inside a fixed frame from the transformation of that frame. Consequence Return asks where responsibility can materially return. Foresight Stress Test examines how an apparently successful solution behaves when its premises shift. Delegation Boundary asks what cannot legitimately be handed to an automated executor.

Each instrument permits examination across research, corporate governance, education and other institutional settings. The resulting prompts are exploratory philosophical aids, not psychometric tests, legal advice or validated measures of autonomy.

The questions are deliberately available as a public threshold. Their deeper conceptual grounds remain within my independent scholarly corpus, accessible through Sapiognosis and The Epistemic Core.

Individual research membership (Free / Paid / Founding) is handled on Substack; choose the existing purchase route inside the instrument above.

For partner-governed institutional exploration, the separate Epistemic Core pathway is linked above. Institutional access is subject to independent assessment and agreement; all pre-existing scholarly work remains independently authored.

Dr Leon Tsvasman — Cybernetician & AI Epistemologist