The Institute of AI and Neural Theory is an independent research initiative developing formal frameworks for understanding how distributed systems coordinate belief and action under uncertainty. Our work begins from a structural observation. Biological organisms, artificial systems, and human institutions all face the same architectural problem. Each must convert distributed local information into coordinated systemic action while preserving the local distinctions that make the system worth coordinating. The constraints any working solution must satisfy are substrate-independent. The implementations that satisfy them are not.

​

Current research develops this framework across several domains. We study consequence-sensitive compression in belief arbitration, asking which inferential structure must be preserved when action stakes vary. We examine the developmental conditions under which arbitration capacity matures, identifying the regimes that produce expertise rather than over-resolution. We build substrate-agnostic frameworks for evaluating arbitration across biological, artificial, and institutional systems. Our approach combines formal modeling with empirically testable predictions, drawing on cognitive science, computational neuroscience, machine learning theory, and philosophy of mind.