Two Consciousness Theories Failed the Same Brain Scan, and Panpsychism Still Does Not Own the Universe

The headline that will not die says scientists have decided the universe is conscious. They have not. What they have done, in the last two years, is more interesting and less mystical. They forced two famous theories of consciousness to write their predictions down before the brain scans came back. Both theories missed. A third project is now arguing, on paper, about what should be tested next. A conference that has been running since 1994 opens in San Diego on October 11, 2026, with a plenary that asks whether life or consciousness came first. None of that is a verdict that electrons feel pain. It is a field that finally agreed to lose in public.

Panpsychism is the philosophical bet sitting beside that lab work, not the result of it. The claim is that consciousness is not a late trick invented by brains. It is a basic feature of the physical world, present in some thin form wherever there is matter, the way mass and charge are present. Durham philosopher Philip Goff has been the bet’s most patient translator. He does not pretend physics already says this. He says physics never described the stuff it measures. In a 2025 UnHerd interview he put the point on Bertrand Russell: physics, in that tradition, is mathematical structure. Stephen Hawking had already written the line Goff keeps borrowing, on the last page of A Brief History of Time: even a finished physics would be a set of equations, and would not say what breathes fire into them. Goff’s proposal, which he also makes in a TEDx talk, is that the fire might be experience itself. Very simple conscious entities, on this view, are what the equations are about. Brains do not create consciousness from zero. They organize something that was already there.

Critics call that a category error with a dignified name. If an electron is «a little bit» conscious, they argue, you have not explained the kind of consciousness that hurts, plans, grieves, or writes a methods section. You have renamed the mystery and distributed it across the periodic table. The hard problem — why any physical process should feel like something from the inside — does not shrink because you spread it. It gets a larger address. Anil Seth, at the University of Sussex, has argued the opposite direction for years: experience is what a biological prediction machine does, a controlled hallucination the brain builds to stay alive. On that view, asking whether the cosmos is aware is like asking whether the weather is sad. The question sounds deep because the nouns are large.

The scanner did not pick a winner

For a long time the fight lived in journals. Then the Templeton World Charity Foundation paid for a different kind of argument. The Cogitate Consortium, a theory-neutral collaboration across laboratories, asked the defenders of Integrated Information Theory and Global Neuronal Workspace Theory to register predictions before the data arrived. Integrated Information Theory, associated with Giulio Tononi and Christof Koch, treats consciousness as the amount of irreducible information a system generates, and it predicted a posterior «hot zone» where sustained synchronization would track conscious content. Global Neuronal Workspace Theory, associated with Stanislas Dehaene, treats consciousness as a broadcast. Information becomes conscious, on that account, when it ignites across prefrontal and parietal networks and is made available to the rest of the brain.

The paper landed in Naturaleza on April 30, 2025. The experiment used functional MRI, magnetoencephalography, and, in patients who already had electrodes implanted for clinical reasons, intracranial EEG. People looked at images that stayed on screen for different lengths of time. The consortium’s own account of the seven-year project is blunt about what the glossy PDF does not show: nobody won. Information about what the person was consciously seeing showed up in visual, ventrotemporal, and inferior frontal cortex. Sustained responses in occipital and lateral temporal cortex tracked how long the stimulus lasted. Some content-specific synchronization appeared between frontal areas and early visual cortex.

What did not appear is the part that matters. The sustained synchronization inside the posterior hot zone, at the scale Integrated Information Theory needed, was not there. The ignition at stimulus offset that Global Neuronal Workspace Theory needed, and a robust prefrontal code for certain conscious features, was not there either. Koch has not dressed the result up as a victory. Dehaene’s camp has argued that this particular task was a limited round, and that other tests still to be run could restore a role for prefrontal cortex. A letter signed by a large group of scholars had already called Integrated Information Theory pseudoscience, in part because its math leans toward a panpsychist reading in which any sufficiently integrated system, not just a brain, has some experience. That fight was aired in Nature Neuroscience in April 2025. Seth and Alex Gómez-Marín answered that a bold theory is allowed to be wrong, and that consciousness research has a duty to risk being wrong in the open.

The follow-on is called INTREPID. It is another adversarial collaboration, this time setting Integrated Information Theory against predictive-processing accounts, including active inference and a view called neurorepresentationalism. A review by Andrew Corcoran, Andrew Haun, Reinder Dorman, Tononi, Karl Friston, Cyriel Pennartz, and the consortium is in press at Neuroscience & Biobehavioral Reviews (doi: 10.1016/j.neubiorev.2026.106742). It is not a second scanner verdict. It is the argument, written down in advance, about which predictions would actually distinguish the theories, including whether conscious content changes when specific neurons are inactivated rather than merely quiet. That is a harder test than asking where a signal happens to light up. Turning a neuron off and asking whether the experience changes is closer to a cause than a photograph.

A conference, not a coronation

The Science of Consciousness, the long interdisciplinary meeting that began in 1994, reconvenes at Paradise Point in San Diego from October 11 through October 16, 2026. The program is not a panpsychist rally. Plenary sessions run from quantum biophotonics and psychedelics to a panel that asks whether artificial intelligence is conscious, and another that puts the question in the oldest form it has: life first, or consciousness first. Edward Boyden, Susan Schneider, and Stuart Hameroff are among the names on the speaker list. The point of the week is the same point Cogitate forced in the scanner. Say what would count against you before the data arrives.

None of this licenses the sentence «scientists now believe the universe has a conscience.» Conscience is a moral word. Consciousness, in this literature, is the bare fact that there is something it is like to be a system. Goff’s camp thinks that fact may sit at the bottom of physics. Seth’s camp thinks it is a biological computation we have not finished describing. Carlo Rovelli, the physicist, has argued in public debates that the hard problem is not the kind of obstacle the panpsychists think it is. Patricia Churchland wants the evidence, not the metaphysics. Bernardo Kastrup argues from the other shore, that mind is the thing we actually have and matter is the theory. A September 2026 discussion among those positions, hosted by the Institute of Art and Ideas, is a better picture of the field than a meme about a conscious cosmos.

El Naturaleza paper did not referee that argument. It did something the philosophy had avoided. It made two leading empirical theories pay a forfeit on predictions their own authors had signed. Integrated information is not dead, and the workspace theory is not dead. Both are now theories with a published miss, which is a more scientific condition than a theory with only friends. Whether the cosmos itself is aware remains a bet about what the equations leave out. Whether a consciousness theory can survive an adversarial protocol is no longer a bet. It is a method, and the method has already embarrassed its first two contestants.

The AEGIS Alliance has followed the stranger edges of that cosmology before, including the argument that our universe may ride an expanding bubble whose dark energy will not sit still. The consciousness argument belongs in the same drawer: large claims, small respect for official consensus, and a duty to say which measurement would hurt. More of that reporting lives on the Ciencia desk.