Beyond Prediction -- Structuring Epistemic Integrity in Artificial Reasoning Systems
Year: 2025
Type: preprint
Abstract: This paper develops a comprehensive framework for artificial intelligence systems that operate under strict epistemic constraints, moving beyond stochastic language prediction to support structured reasoning, propositional commitment, and contradiction detection. It formalises belief representation, metacognitive processes, and normative verification, integrating symbolic inference, knowledge graphs, and blockchain-based justification to ensure truth-preserving, auditably rational epistemic agents.
Source: arXiv (Cornell University)
Author: Wright, Craig Steven
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6948229050