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