I've been working on a new approach to AI alignment diagnostics—one that models intelligence as a system navigating a structured conceptual space, where reasoning processes correspond to discrete transitions through a graph distributed over a continuous 3D space. This perspective allows us to visualize alignment failures as topological breakdowns—failures of...
My recent paper, Why AI Alignment Is Not Reliably Achievable Without a Functional Model of Intelligence: A Model-Theoretic Proof argues that alignment under conceptual novelty isn’t something you can train into a system or enforce externally. It must be structurally guaranteed from within—and that guarantee depends on the presence of...
We’re inviting submissions for a workshop at AGI-2025 that challenges foundational assumptions about AGI and alignment using a formal functional model. Purpose. This workshop invites submissions of 2-page briefs about any model of intelligence of your choice, to explore whether a functional model of intelligence can be used to very...