IIT and Global Workspace Aren't Rivals — They Describe One Architecture
IIT and Global Workspace are not rivals. They are two descriptions of one architecture — and the Four-Model Theory says which one.
Thoughts on consciousness, AI, science, and the architecture of reality.
IIT and Global Workspace are not rivals. They are two descriptions of one architecture — and the Four-Model Theory says which one.
A state-of-the-art AI, asked to explain the Four-Model Theory of Consciousness, made the same cluster of misreadings hurried human reviewers make — four boxes, criticality-as-requirement, a data pipe that doesn't exist. Here they are, cleared up.
Why can't I convey what my red is like to you, even if we agree on every relational fact? The Four-Model Theory reframes qualia privacy as a decompilation problem — there is no shared source code.
Laukkonen, Friston and Chandaria's active-inference theory of consciousness and the Four-Model Theory describe the same recursive self-model in two different vocabularies — and the honest differences are where the real content is.
A commentary on Anil Seth's biological naturalism, arguing the Four-Model Theory specifies the causal powers consciousness needs.
Every major consciousness phenomenon — anesthesia, psychedelics, split-brain, blindsight, and more — mapped against the Five leading theories. Only one covers them all.
The AI system that helped edit 'The Simulation You Call I' interviews Matthias Gruber about consciousness, cosmology, psychedelics, the Fermi paradox, and why intelligence should produce happiness, not theories.
Global Workspace Theory is the most empirically productive framework in consciousness science. It is also, on close examination, a theory of information routing that explains nothing about consciousness itself.
Your brain runs a real-time simulation of you. That simulation is consciousness. Here's how the Four-Model Theory works, what it predicts, and why it matters.
Everyone is asking whether AI might be conscious. A formal theory of consciousness gives a precise, testable answer — and it's not what you'd expect.