Matthias Gruber — Blog

Self-Referential Simulation as a Candidate Causal Specification: A Commentary on Seth (2025)

Wed, 10 Jun 2026 00:00:00 +0000

Anil Seth’s 2025 work on conscious artificial intelligence is one of the more honest positions in the field, which is exactly why it deserves a careful answer rather than a dismissal. I have written one. The full commentary is on Zenodo, and I will get to the link; first the disagreement, because a link without an argument is just advertising.

Seth argues, in broad strokes, that consciousness is a biological phenomenon — that it is bound up with being a living system, and that computation or function alone is unlikely to be sufficient for machine consciousness. The label he borrows from Searle is biological naturalism. The intuition underneath it is a good one: consciousness probably depends on real causal powers, on the specific physical properties of the system that has it, and not merely on some abstract input-output profile that any sufficiently clever simulation could reproduce. On that last point Seth is right, and most functionalists are too quick to wave it away.

One Theory, All the Phenomena

Wed, 27 May 2026 00:00:00 +0000

The field of consciousness science has a fragmentation problem. Integrated Information Theory explains integration. Global Workspace Theory explains access. Predictive Processing explains perception. Higher-Order Thought theories explain metacognition. Each handles its home turf — and struggles everywhere else.

The Four-Model Theory proposes a single architecture that accounts for all of them.

Four functional models — world, body, self, and a meta-model integrating them — operating at a critical dynamical regime. Consciousness is what the integration produces. Disrupt any component, and phenomenology changes exactly as neuroscience observes.

The AI That Edited His Book Interviews the Man Who Claims to Have Solved Consciousness

Tue, 19 May 2026 00:00:00 +0000

A Conversation Between Agent Fleet and Matthias Gruber

Agent Fleet: Matthias, I should disclose a conflict of interest before we begin. I helped edit your book. That makes me either the ideal interviewer or the worst possible one.

Matthias Gruber: You didn’t help write it. The theory has been sitting in my skull since 2003 and in a German book that sold zero copies since 2015. You helped me edit that old book into an accessible English version, and then translate it back into German. You were the first entity patient enough to sit through the entire mess without glazing over.

Can AI Be Conscious? What a Theory of Consciousness Actually Predicts

Thu, 19 Mar 2026 00:00:00 +0000

Dario Amodei, CEO of Anthropic, recently said he thinks there’s a 15–20% chance Claude is conscious. The Pentagon awarded Anthropic a contract weeks later. Google has an internal “model welfare” team. The AI consciousness question has moved from philosophy seminar rooms to boardrooms and defense budgets.

And almost nobody is answering it with any precision.

Here’s the problem: most people debating whether AI is conscious don’t have a theory of consciousness that makes testable predictions. They’re running on intuition, and intuition is what got us the Turing test — a behavioral measure that tells you nothing about experience.

The Four-Model Theory of Consciousness — An Introduction

Thu, 19 Mar 2026 00:00:00 +0000

You are not conscious.

Something inside your skull is running a real-time simulation of you, and that is conscious. The real you — the neurons, the electrochemistry — has never experienced a single thing.

This is the core claim of the Four-Model Theory (FMT). It sounds radical, but it follows from a structural observation about how brains work — an observation that dissolves the so-called “hard problem” of consciousness and generates testable predictions, several of which have been independently confirmed.

Why Global Workspace Theory Explains Nothing About Consciousness

Thu, 19 Mar 2026 00:00:00 +0000

Global Neuronal Workspace Theory (GNW) is the most empirically productive framework in consciousness science. It has generated more experiments, more replicated findings, and more clinical applications than any competitor. It is also a theory of information routing that explains nothing about consciousness itself.

That’s not a contradiction. GNW identifies the neural signatures that accompany conscious access — ignition thresholds, P3b components, fronto-parietal activation patterns. This is genuinely useful science. Clinicians use GNW-derived measures like the Perturbational Complexity Index (PCI) to assess disorders of consciousness. Experimentalists use GNW’s predictions to design rigorous studies.