A Grand Unified Neuroalgorithmic Theory of Mind

The Blueprint for Artificial General Intelligence (AGI)

An end-to-end cognitive-affective architecture which integrates:


  • a predictive coding engine driving a global workspace for conscious attention to provide the cognitive bottleneck necessary for a serial train of thought;

  • holographic associative long-term storage incorporating unconscious experience and somatic markers for unconscious re-experience of memories;

  • subconscious short-term working memory via anticipatory predictive caching;

  • phased dimensionality bridges for symbol grounding, reversibly converting between meaning and experience;

  • system-wide Hebbian plasticity providing robust temporal credit assignment for radical one-shot learning;

  • and a dynamic hyperparameter control loop for system self-tuning;

  • supporting true tabula rasa self-initiating and self-sustaining operation, intrinsic motivation, and goal-directed behavior;

  • all built upon a data format which is a unit of hyperdimensional subjective experience.

  • feeling pathways: hyperdimensional quale (~300 GB/sec)

  • thinking pathways: meaningful embeddings (~3 GB/sec)

  • tuning pathways: adjust gains, enable learning (<100 MB/sec)


Technical Briefings


The Computation of Thought: A Practical Architecture for Artificial General Intelligence

Introduction: The Reverse Engineering of the Brain’s Core Algorithm

AGI Engineering Overview: Data Formats, Data Flow, Speeds and Feeds

Documentation


A Grand Unified Neuroalgorithmic Theory of Mind and Blueprint for Artificial General Intelligence (AGI)

The Brain’s Algorithm of Cognition and Affect, Its Logical Derivation from Neuroscientific First Principles, and Resynthesis into a Neuroalgorithmic Computational Architecture and Formalism for Artificial General Intelligence.

January, 2026—‍‍ ‍Jeremy S. Meredith‍ —‍‍ ‍version 1.0.3

Strategic Inquiries