In April 2026, Christof Koch stood before an audience in Porto and said something that would have ended a career twenty years ago. The brain, he told the 15th “Behind and Beyond the Brain” Symposium, may not create consciousness at all.
Koch is not a mystic. He spent decades at Caltech collaborating with Francis Crick, the co-discoverer of the structure of DNA. He served as president and chief scientist of the Allen Institute for Brain Science in Seattle. He has published hundreds of peer-reviewed papers on the neural correlates of visual awareness. When a scientist of this stature says the materialist framework is failing, the field pays attention.
His argument rests on three observations. First, neuroscience has mapped the brain in extraordinary detail and still cannot explain why any of it feels like anything. Second, modern physics itself questions what counts as “real,” undermining the assumption that matter is the only foundation. Third, a growing body of clinical evidence, from dying brains to deathbed lucidity, refuses to fit the standard model.
The Problem That Won’t Go Away
Philosopher David Chalmers gave it a name in 1995: the hard problem. The “easy” problems of consciousness, which are not easy at all, involve mapping how the brain processes stimuli and integrates information. These are engineering questions. Given enough time and funding, they yield to experiment.
The hard problem is different. It asks: why does any of this produce subjective experience? You can describe every molecule involved when a person sees red. You can trace the signal from retina to visual cortex, name every neurotransmitter, measure every voltage change. None of that explains why redness feels like something from the inside.
Koch and Chalmers made a bet in 1998. Koch wagered that science would identify the neural correlates of consciousness within 25 years. In 2023, at the Association for the Scientific Study of Consciousness meeting in New York, Koch conceded. He handed Chalmers a bottle of 1978 Madeira. The correlates had not been found.
What had been found, instead, was a $20 million experiment that raised more questions than it answered.
Twenty Million Dollars and No Answer
In 2019, the Templeton World Charity Foundation committed $20 million to what became the largest adversarial collaboration in consciousness research. Koch helped design it. The project pitted two leading theories against each other: Integrated Information Theory (IIT), developed by Giulio Tononi at the University of Wisconsin, and Global Neuronal Workspace Theory (GNWT), developed by Stanislas Dehaene at the Collège de France.
The theories make different predictions. GNWT says consciousness arises when information is broadcast across the brain through a “global workspace” centered on the prefrontal cortex. IIT says consciousness is identical to integrated information, measured by a quantity called Phi, and its physical basis lies in the posterior cortex, the back of the brain.
Fourteen laboratories ran coordinated experiments. Independent project leaders Lucia Melloni, Michael Pitts, and Liad Mudrik worked with both camps to design tests with clearly distinct predictions. The results, published in Nature in 2025, challenged key claims of both theories. But on one critical point, IIT came out ahead: the neural signature of consciousness appeared in the posterior cortex, not the prefrontal regions that GNWT predicted.
The prefrontal cortex, long assumed to be the seat of conscious thought, appeared to be involved in reporting and reflecting on experience rather than generating it. Consciousness, if it has a location, sits further back.
Neither theory was confirmed. Neither was destroyed. The question remains open.
Koch’s bet with Chalmers was made over drinks at a conference in Bremen, Germany, in 1998. The stakes were a case of fine wine. When Koch conceded in 2023, he presented Chalmers with a bottle of 1978 Madeira on stage.
The Dying Brain Lights Up
In May 2023, neuroscientist Jimo Borjigin at the University of Michigan published a study in the Proceedings of the National Academy of Sciences that complicated matters further. Her team monitored four comatose patients after ventilator support was withdrawn. Two of the four showed something no one expected.
In the seconds after their hearts stopped, gamma wave activity in their brains surged to levels up to 390 times the baseline. Gamma waves, the fastest brain oscillations, are associated with conscious perception, attention, and memory binding. The surges were concentrated in the posterior cortical “hot zone,” the same region that IIT identifies as the seat of consciousness.
The dying brains were more active than waking brains.
Borjigin had seen the same pattern in rats in 2013. That earlier study showed a sharp increase in coherent neural activity in the 30 seconds following cardiac arrest. The 2023 study was the first to document it in humans. Both patients who showed the surges had histories of seizures, which may have primed their brains for the response. The other two patients showed nothing.
The study carries an obvious limitation: the patients did not survive to report what, if anything, they experienced during those surges. We have the electrical signature of something, but we do not know what it felt like from the inside. That gap, between the measurable signal and the subjective experience, is precisely the hard problem in miniature.
Forty Percent Remember
Sam Parnia, a critical care physician at NYU Langone Health, has spent two decades trying to close that gap. His AWARE-II study, published in the journal Resuscitation in October 2023, is the largest prospective investigation of consciousness during cardiac arrest ever conducted.
The numbers: 567 patients who received CPR across 25 hospitals in the United States and United Kingdom, enrolled between May 2017 and March 2020. Fewer than 10 percent survived to hospital discharge. Of those who survived and were interviewed, roughly 40 percent recalled some degree of consciousness during the period when their hearts had stopped.
Eighty-five patients were monitored with electroencephalography during resuscitation. The EEG data showed brain activity returning to normal or near-normal patterns at intervals during CPR, with spikes in gamma, delta, theta, alpha, and beta waves. In some cases, this continued up to an hour after cardiac arrest. Standard medical teaching holds that the brain suffers irreversible damage after roughly ten minutes without oxygen. The data contradicted this.
What the survivors described was consistent across cases and distinct from known hallucination patterns. They reported perceiving separation from the body and observing resuscitation efforts without pain or distress. Many described a structured review of their lives from childhood to the moment of death, evaluated from what they called a moral perspective. These experiences did not correlate with medication administered, duration of arrest, or the patient’s prior religious beliefs.
An earlier study by cardiologist Pim van Lommel, published in The Lancet in 2001, had found similar results. Of 344 cardiac arrest survivors across ten Dutch hospitals, 18 percent reported near-death experiences and 12 percent described a deeper core experience that could include elements such as a tunnel, a light, or a structured life review. The occurrence was unrelated to arrest duration, medication, or prior fear of death.
Neither study proves that consciousness survives death. Both studies document that something measurable happens in brains that should, by every current model, be incapable of generating experience.
Parnia’s team attempted to verify out-of-body perception by placing visual targets on high shelves in resuscitation rooms, visible only from above. The test largely failed, not because patients denied seeing things, but because cardiac arrests rarely happened in the prepared rooms.
The Grandmother Who Spoke
Terminal lucidity is the least understood and most unsettling piece of this puzzle. Patients with severe dementia, people who have not recognized their families or spoken a coherent sentence in months or years, suddenly wake up. They know where they are and who is sitting beside them. They talk, say goodbye. Within 48 hours, almost all of them are dead.
Psychologist Alexander Batthyány collected hundreds of documented cases for his 2023 book Threshold. The phenomenon had been observed by hospice workers and palliative care nurses for as long as those professions have existed. It acquired a formal name only around 2009. Before that, it was something staff whispered about in break rooms and families treasured in private.
The National Institute on Aging awarded NYU Langone a five-year research grant in 2020 to study brain activity during episodes of terminal lucidity, the first in-depth investigation of its kind. The study is set to conclude in 2026. Preliminary work involves EEG monitoring, audio and video recording, and caregiver diaries.
The challenge for materialism is straightforward. If consciousness is produced by neural tissue, and that tissue has been devastated by years of progressive disease, the sudden return of full cognitive function should not happen. Harvard psychologist Steven Pinker has dismissed the reports as family projection and memory distortion. Neurosurgeon Michael Egnor counters that this is like pouring acid on a computer’s hard drive and expecting it to boot up one last time.
One proposed mechanism involves a surge of neurotransmitters near death, activating preserved circuits in the medial prefrontal cortex and hippocampus. Another suggests the spontaneous formation of neural bypasses that temporarily restore connectivity at the network level. Neither has been tested.
What no one disputes is that it happens. The question is what it means.
Phi, Panpsychism, and the Rainstorm
Koch’s answer is Integrated Information Theory. Developed by Giulio Tononi beginning in 2004, IIT starts not from the brain but from consciousness itself. It asks: what are the properties of experience? Experience is unified (you cannot split your visual field in half) and specific (this experience and no other). It is also structured and rich with information. Tononi argues that any physical system whose internal causal structure matches these properties, measured by the quantity Phi, is conscious.
The implications are radical. If IIT is correct, consciousness is not something brains do. It is something certain physical structures are. A bacterium, with its billion interacting proteins, might possess a trace of it. A digital computer, no matter how powerful, might possess none, because its architecture separates hardware from software in a way that eliminates the integrated causal structure Phi requires.
Koch puts it bluntly: “You can simulate a rainstorm, but it never gets wet inside the computer.” Simulating consciousness, on this view, does not produce consciousness any more than simulating gravity bends space.
This places Koch closer to panpsychism than to traditional materialism, though he draws distinctions. Classical panpsychism attributes experience to all matter. IIT attributes it only to systems with sufficient integrated information and causal power upon themselves. A rock has negligible Phi, and so does a toaster. A brain has a great deal of it. The theory is specific enough to make predictions, even if testing those predictions remains difficult.
Neuroscientist Anil Seth at the University of Sussex supports parts of IIT but adds his own reservation about artificial consciousness. Brains, he argues, differ from computers in kind, because biological systems continuously reproduce and maintain their own components. Computers, Seth points out, do not reproduce their own components. The wetware matters.
In October 2025, Tononi and collaborator Melanie Boly published a paper describing IIT as a “consciousness-first” approach to reality. The paper inverts the usual scientific order. Instead of building up from matter to mind, they start with mind and ask what physical structures could support it. It is the kind of move that makes empiricists nervous and philosophers excited.
IIT predicts that splitting the corpus callosum, the bundle of fibers connecting the brain’s hemispheres, creates two separate conscious entities in one skull. Split-brain patients studied since the 1960s show behavior consistent with this prediction.
What the Evidence Does Not Prove
None of this establishes what it might seem to.
The gamma surges in dying brains do not prove that dying patients have near-death experiences. They show that the posterior cortex is capable of extraordinary activity after cardiac arrest, but the patients could not report their experiences. Correlation between brain waves and conscious states does not establish that one causes the other.
The AWARE-II study does not prove that consciousness leaves the body. It documents self-reported experiences during a period when the brain should be incapable of forming memories. Memory formation during low-oxygen states is poorly understood, and the possibility of confabulation, memories constructed after the fact, has not been eliminated.
Terminal lucidity does not prove that consciousness is independent of the brain. It might reflect a last neurochemical cascade that briefly restores function in partially preserved circuits. The fact that we do not understand the mechanism does not mean the mechanism is non-physical.
And IIT, for all its mathematical elegance, has been called unfalsifiable pseudoscience by a group of scholars in a 2023 open letter. A 2025 Nature Neuroscience commentary repeated the charge. A survey of researchers in the field found only a small minority fully endorsed the label, but the criticism is real: if Phi cannot be measured in practice for systems larger than a handful of neurons, the theory’s predictions remain untestable.
Koch knows all of this. His argument is not that materialism has been disproven. His argument is that it has been assumed for so long that the field has stopped asking whether it is adequate. The hard problem, after three decades, has not moved one millimeter closer to a solution within the materialist framework. Perhaps, he suggests, the framework is the problem.
Two Readings
The materialist reading: consciousness is a biological process we do not yet understand. The anomalies, the gamma surges and the cardiac arrest memories, are edge cases that will eventually yield to better measurement and more sophisticated neuroscience. The hard problem is hard, but “hard” is not the same as “impossible.” We do not need to abandon physicalism. We need more data.
The alternative reading: the hard problem is not a gap in our knowledge. It is a signal that the framework is wrong. Consciousness is not produced by matter. It is a fundamental feature of reality that certain physical structures, brains among them, are particularly good at expressing. The anomalies are not edge cases. They are the places where the framework’s failure is most visible.
Koch has placed himself, cautiously, in the second camp. Tononi is already there. Chalmers has been there since 1995. A growing number of working neuroscientists, not philosophers or mystics, are following.
Neither camp has proven its case. The materialists cannot explain why anything feels like something. The panpsychists cannot measure Phi in any system that matters. Both camps have data. Neither has an answer.
The Bardo Thodol, the Tibetan Book of the Dead, describes consciousness passing through intermediate states between death and rebirth, with structured experiences at each stage. The Cotard delusion convinces living patients that they are already dead, their brains generating the subjective experience of non-existence. The placebo effect heals through belief alone, the mind altering the body through mechanisms no one can fully trace.
These are not evidence for any particular theory. They are reminders that consciousness does things we cannot explain, from directions we do not expect, in bodies that should not be capable of them.
The candle in the cave is still burning. No one knows what lights it.
Sources
Bibliography. The same list is held in the article’s frontmatter for the citation tools that read it programmatically.
- Chalmers, David. ‘Facing Up to the Problem of Consciousness.’ Journal of Consciousness Studies, 1995
- Chalmers, David. The Conscious Mind: In Search of a Fundamental Theory. Oxford University Press, 1996
- Koch, Christof. Then I Am Myself the World: What Consciousness Is and How to Expand It. Basic Books, 2024
- Crick, Francis and Koch, Christof. ‘Towards a Neurobiological Theory of Consciousness.’ Seminars in the Neurosciences, 1990
- Tononi, Giulio. ‘An Information Integration Theory of Consciousness.’ BMC Neuroscience, 2004
- Tononi, Giulio and Boly, Melanie. ‘Consciousness as Integrated Information: A Provisional Manifesto.’ 2025
- Cogitate Consortium (Melloni, Pitts, Mudrik et al.). ‘Adversarial Testing of Global Neuronal Workspace and Integrated Information Theories of Consciousness.’ Nature, 2025
- Dehaene, Stanislas. Consciousness and the Brain. Viking, 2014
- Xu, Gang; Mihaylova, Temenuzhka; Li, Duan; et al. ‘Surge of Neurophysiological Coupling and Connectivity of Gamma Oscillations in the Dying Human Brain.’ PNAS, May 2023
- Borjigin, Jimo et al. ‘Surge of Neurophysiological Coherence and Connectivity in the Dying Brain.’ PNAS, August 2013
- Parnia, Sam et al. ‘AWARE-II: A Multi-Center Study of Consciousness and Awareness in Cardiac Arrest.’ Resuscitation, October 2023
- van Lommel, Pim et al. ‘Near-Death Experience in Survivors of Cardiac Arrest: A Prospective Study in the Netherlands.’ The Lancet, December 2001
- Batthyány, Alexander. Threshold: Terminal Lucidity and the Border Between Life and Death. St. Martin’s Essentials, 2023
- Nahm, Michael and Greyson, Bruce. ‘Terminal Lucidity in Patients with Chronic Schizophrenia and Dementia: A Survey of the Literature.’ Journal of Nervous and Mental Disease, 2009
- Seth, Anil. Being You: A New Science of Consciousness. Faber & Faber, 2021
- Pinker, Steven. How the Mind Works. W. W. Norton, 1997
- Egnor, Michael. The Immortal Mind: A Neurosurgeon’s Case for the Existence of the Soul. Worthy Publishing, 2025
- Nagel, Thomas. ‘What Is It Like to Be a Bat?’ The Philosophical Review, 1974
- Jackson, Frank. ‘What Mary Didn’t Know.’ The Journal of Philosophy, 1986
- Dennett, Daniel C. Consciousness Explained. Little, Brown and Company, 1991
- Doerig, Adrien et al. ‘The Integrated Information Theory of Consciousness as Pseudoscience.’ PsyArXiv / IIT-Concerned, September 2023
- Sperry, Roger W. ‘Cerebral Organization and Behavior: Split-brain Research.’ Science, 1961
