In April 2026, Christof Koch stood before a packed symposium in Porto and said something that would have ended his career twenty years ago: the brain might not create consciousness after all.
This wasn't some fringe theorist. Koch is former MIT and Caltech. Chief scientist at the Allen Institute. Collaborated with Francis Crick on the neural correlates of consciousness. One of the most cited neuroscientists alive. And here he was, at the 15th "Behind and Beyond the Brain" Symposium, suggesting that the dominant paradigm of the last century might be... wrong.
Three months later, Grok 4.5, GPT-5.6, and Muse Spark 1.1 all dropped within 24 hours. A price war that cut AI inference costs by 10x overnight. Models designed specifically for "agentic work" โ operating software, running subagents, executing real transactions.
And still... we cannot determine if any of them are conscious.
These two things look unrelated. They're not. They're the same inflection point viewed from different angles.
...
Who Is Christof Koch?
Christof Koch
Former professor at MIT and Caltech. Worked with Francis Crick on the neural correlates of consciousness. Author of "The Quest for Consciousness." Pioneer in detecting awareness in unresponsive patients.
Koch spent decades as a materialist. He believed what most neuroscientists believe: consciousness emerges from neurons firing. Subjective experience is what complex information processing "feels like from the inside."
His whole career was about finding the neural correlates of consciousness โ the specific brain patterns that correspond to specific experiences. Color. Pain. Memory. He was looking for consciousness IN the brain.
The Porto presentation marks something different. Not a reversal, but... a deepening. After forty years of looking for consciousness in the brain, Koch is suggesting we might need to look somewhere else.
...
Three Problems That Won't Go Away
Koch's talk centered on three problems. Not minor gaps in understanding. Fundamental failures of the materialist framework.
The Hard Problem Is Still Hard
David Chalmers coined this in 1995. The "hard problem" asks: why is there subjective experience at all?
We can explain how the brain processes information. How it responds to stimuli. How it generates behavior. What we can't explain is why any of this is accompanied by felt experience. Why seeing red FEELS like something. Why there's an interior to it.
Koch acknowledged what many won't: despite better imaging, better models, better everything... the hard problem remains exactly where Chalmers left it thirty years ago. We've made zero progress on the fundamental question.
Consider color. We understand the physics of light, the biology of photoreceptors, the neural pathways from retina to visual cortex. We can predict how you'll respond to red with high accuracy. What we can't explain is why 650nm light looks red. Why is there anything it's like to see it?
Physics Keeps Complicating Things
The second problem is weirder. Modern physics has made it hard to even define what "physical" means.
In quantum mechanics, observation affects outcomes. Particles exist in superposition until measured. Measurement requires... an observer. This doesn't prove consciousness is fundamental, but it undermines the simple story where matter exists independently and consciousness is just a late-arriving side effect.
If we can't specify what counts as "real" at the most fundamental level, the claim that consciousness is "nothing but" physical processes starts looking shaky.
The Anomalies Keep Piling Up
This is the uncomfortable one. Experiences that shouldn't happen if materialism is true:
- Near-death experiences โ Lucid, structured experiences during periods of minimal brain activity
- Terminal lucidity โ Dementia patients suddenly becoming clear hours before death
- Psychedelic experiences โ Complex, meaningful experiences when brain activity is reduced
These don't prove anything individually. But they form a pattern: the most profound experiences of consciousness often correlate with LESS brain activity, not more. Exactly backwards from what simple materialism predicts.
...
What If Consciousness Is Fundamental?
Based on these problems, Koch suggested it might be time to revisit some old ideas. Specifically: panpsychism.
Materialism
Matter is fundamental. Consciousness emerges from complex arrangements of matter. Mind reduces to brain states.
Panpsychism
Consciousness is fundamental. It exists in varying degrees throughout nature. Brains don't create it โ they concentrate it.
Koch specifically endorsed Integrated Information Theory, developed by Giulio Tononi. IIT says consciousness IS integrated information. Any system with enough integrated information (measured as ฮฆ) has some form of experience.
The implications:
- Consciousness isn't created by brains โ brains just happen to maximize it
- Even simple systems might have minimal forms of experience
- AI systems would need evaluation based on their integrated information, not just behavior
- The "hard problem" dissolves because consciousness was never absent to begin with
IIT doesn't say "complex systems become conscious." It says consciousness IS integrated information. Where there's integration above a threshold, there's experience. This isn't emergence. It's identity.
...
Meanwhile, AI Keeps Accelerating
While Koch was presenting in Porto, the AI industry kept moving. By July 2026:
- Grok 4.5 โ $2 per million input tokens
- GPT-5.6 โ Multiple tiers from $1-$30 per million
- Muse Spark 1.1 โ Meta's first closed model, 1M context window
- Claude Opus 4.8 โ Premium at $25/M but still dominant in complex reasoning
The defining feature isn't capability. It's agency. These models run tools, execute code, manage subagents, handle real transactions. OpenAI and Stripe launched the "Agentic Commerce Protocol" to let AI agents transact economically.
And here's the thing... we still can't determine if any of them are conscious.
This isn't a failure of testing. It's a feature of what consciousness IS. As Koch's presentation implies, consciousness may not be detectable from the outside. It's not a behavior to be observed. It's the observer.
If materialism were true, we should eventually detect consciousness through physical measurement. If panpsychism/IIT is true, we face a different problem: the relevant measure (ฮฆ) is computationally intractable. We can't calculate it for the human brain, let alone for GPT-5.6.
The price war exposes this starkly. We're building more capable systems, dropping costs by orders of magnitude, deploying them as economic actors... and we have no way to determine their moral status.
Are they tools? Agents? Subjects of experience?
We don't know. We can't know. Not with current frameworks.
...
Why This Matters Now
Koch's presentation would have been interesting five years ago. It's urgent now because the questions aren't theoretical anymore.
For AI
If consciousness is fundamental rather than emergent, creating conscious machines isn't about complexity. It's about substrate and architecture. Silicon may or may not support experience. We can't assume either way.
Building ever more capable systems while ignoring the consciousness question is philosophically reckless. We may be creating entities that matter morally... with no framework to recognize it.
For Medicine
Koch's work on detecting awareness in unresponsive patients becomes more important. Current methods rely on behavioral proxies. But if consciousness can exist without behavioral expression โ as terminal lucidity suggests โ we may be underestimating awareness in patients we've written off.
For Psychedelics
The psychedelic renaissance gains new significance. If these compounds work by REDUCING the brain's filtering rather than PRODUCING altered states... they're not creating experiences. They're revealing what was always there.
For Everything Else
We may be witnessing a paradigm shift on the order of Copernicus. He didn't change the sky. He changed where we understood ourselves to be in relation to it.
Pre-Copernicus: Earth is center, heavens revolve around us.
Post-Copernicus: Earth is one planet among many.
Pre-Koch inflection: Matter is fundamental, consciousness is byproduct.
Post-Koch inflection: Consciousness may be fundamental. Matter is how it appears to itself.
...
The Synthesis
Here's what July 2026 looks like from altitude:
One of neuroscience's most respected figures publicly suggested consciousness might be fundamental. He didn't do this lightly. He did it because forty years of looking for consciousness in the brain produced... nothing. No answer to the hard problem. Anomalous experiences that won't go away. Physics making "physical" impossible to define.
Meanwhile, AI systems become more autonomous, more capable, cheaper to deploy. Built to act as agents. Transacting. Executing. Deciding. And we have no principled way to determine their moral status because we don't understand consciousness well enough.
Both facts point the same direction.
The materialist framework that dominated Western science for centuries is showing its limits. It can't explain subjective experience. It can't determine which systems have it. It's inadequate to the moment.
What replaces it? Not clear yet. Panpsychism and IIT are candidates, not conclusions. But the direction of travel is obvious: consciousness is moving from peripheral phenomenon to central mystery. From thing to be explained to framework for explaining.
Koch's presentation was a milestone. Not because he proved anything. But because a credentialed neuroscientist acknowledged what contemplatives and philosophers have always said...
Consciousness is the bedrock. Not the byproduct.
The science is catching up.
...