Research Series
Governed Epistemic Transformation and the Reinforcement Learning Contour
A framework for shifting AI-assisted inquiry from document production to governed recursive learning — epistemic contours, a differentiation principle, and human acceptance authority as structure rather than procedure.
Research questions
- When AI has made scientific text cheap to produce, what is the right unit of scientific analysis — the document, or the mechanism that produced it?
- How can a research cycle be judged by the beneficial, traceable differentiation it produces between epistemic contours rather than by the volume of output?
- What role-based architecture makes provenance, temporality and human acceptance authority structural, so that recursion is governed rather than merely automated?
Constituent works
-
PreprintEmpirical StudyGoverned Local Experiment
Bounded Recursive Self-Improvement Under Human Governance: Evidence from a Seven-Cycle Reinforcement Learning Contour Experiment
Empirical follow-up reporting a bounded seven-cycle scaffolding-level experiment with four promoted states, one retained failed transition, and two NO_CHANGE outcomes under separate human promotion authority. The record does not establish AGI, unbounded improvement, autonomous promotion, permanent saturation, or external production validation.
-
PreprintConceptual Framework
The Reinforcement Learning Contour: From Document Production to Governed Epistemic Transformation in AI-Assisted Scientific Inquiry
Generative artificial intelligence has sharply reduced the technical cost of producing scientific text. Increased production capacity does not, however, entail cumulative knowledge, conceptual novelty, or epistemic progress. This paper introduces the Reinforcement Learning Contour (RLC) as a conceptual and architectural framework for shifting AI-assisted research from repetitive content generation toward governed recursive learning… Version note: current registry metadata is Version 2; the archived PDF and this summary preserve Version 1.
Evidence status & validation boundary
The series now contains two separately labelled, non-peer-reviewed works: a Version 2 conceptual-framework preprint and a Version 2 empirical preprint reporting a bounded local scaffolding-level experiment. The empirical record supports a narrow claim about the tested governance architecture; it does not establish AGI, unbounded or exponential recursive growth, autonomous promotion, permanent saturation, or external production validation.