Note: This is a research note supplementing the book Unscarcity, now available for purchase. These notes expand on concepts from the main text. Start here or get the book.

The Connectome: Your Brain’s Wiring Diagram (And Why Having the Map Doesn’t Mean You Can Read It)

Imagine you’ve never seen a computer before. Someone hands you a complete circuit diagram of an iPhone—every transistor, every connection, every pathway meticulously documented. Billions of components, all mapped.

Now explain why it can play Candy Crush.

That’s the connectome problem in a nutshell. We’re building increasingly detailed maps of how brains are wired, and these maps are genuine scientific triumphs. But having the blueprint and understanding the building are two very different things.

What Is a Connectome, Anyway?

The term “connectome” was coined in 2005 by Olaf Sporns and Patric Hagmann, independently, by analogy with “genome.” Just as the genome is the complete set of genes in an organism, the connectome is the complete set of neural connections.

But “complete” does a lot of work in that sentence.

Your brain contains roughly 86 billion neurons. Each neuron connects to thousands of others through specialized junctions called synapses. The total number of synaptic connections in your head? Somewhere around 100 trillion. That’s 100,000,000,000,000 individual points where information passes from one cell to another.

To put this in perspective: there are more synapses in your brain than there are stars in the Milky Way galaxy. If you counted one synapse per second, it would take you 3.2 million years to finish.

A connectome maps all of these connections—which neurons talk to which, how strong each connection is, and what type of signal it carries. It’s the ultimate wiring diagram.

The appeal is obvious. If memories are stored in neural connections, the connectome is where your memories live. If personality emerges from patterns of connectivity, the connectome is where “you” are encoded. Map the connectome, the thinking goes, and you’ve captured the essence of a mind.

But there’s a catch. Actually, several catches.

The Worm: Our One Complete Success

In 1986, after 14 years of painstaking work, Sydney Brenner’s lab at Cambridge published something remarkable: the complete connectome of Caenorhabditis elegans, a transparent roundworm about 1 millimeter long.

The numbers seem almost quaint by brain standards:

302 neurons (precisely—it’s the same in every worm)
~7,000 chemical synapses
~900 gap junctions (electrical connections)

Researchers traced every connection by hand, using electron microscope images of serial sections—essentially slicing the worm into thousands of impossibly thin pieces and reconstructing the 3D structure.

For this work, Brenner won the Nobel Prize in 2002. The scientific community celebrated. Finally, we had a complete neural wiring diagram of a living creature.

And then something awkward happened: we couldn’t fully explain what the worm does with it.

Don’t get me wrong—the C. elegans connectome has been enormously valuable. We’ve identified circuits for movement, feeding, egg-laying, and chemical sensing. We can trace pathways from sensory neurons to motor outputs. But 38 years after publishing that wiring diagram, we still can’t simulate the worm’s complete behavioral repertoire from first principles.

Put differently: we have the map, but we don’t have the operating system manual. The connectome tells you that neuron A connects to neuron B. It doesn’t tell you why that connection matters, or what computation it performs, or how the timing of signals creates behavior.

As one neuroscientist put it: “Having a connectome is by itself not a sufficient condition to simulate a nervous system.”

This is the worm that launched a thousand caveats.

The Fly: 140,000 Neurons, A Million Collaborators

Fast forward to October 2024. The FlyWire Consortium—a global collaboration of over 200 researchers across 50 labs—published the complete connectome of an adult fruit fly brain.

The numbers tell the story of exponential ambition:

139,255 neurons
~50 million synaptic connections
8,453 cell types (4,581 newly discovered)

This wasn’t a worm with 302 neurons. This was a functioning animal brain that can see, fly, navigate, mate, remember, and make decisions. The fruit fly brain, while tiny by mammalian standards (it would fit comfortably inside a sesame seed), contains circuits for genuinely complex behavior.

The FlyWire project combined cutting-edge technology with good old-fashioned human labor. Machine learning algorithms did the initial heavy lifting—reconstructing 3D neurons from terabytes of electron microscopy imagery. But AI makes mistakes. A lot of them. So humans—including citizen scientists playing what amounts to a sophisticated puzzle game—spent 33 person-years proofreading and correcting the results.

Without that AI assistance, the project would have taken an estimated 50,000 person-years. With it, the work took about seven years.

The payoff? Researchers are already using the fly connectome to make real discoveries. One team identified the neural circuits that tell a fly when to stop walking—different circuits for different contexts. Another used the wiring data to predict which neurons sense sugar versus water. A third mapped the vision system and predicted how the fly processes visual information.

These aren’t just nice-to-have academic findings. 75% of disease-related genes in humans have equivalents in fruit flies. Understanding fly circuits helps us understand our own.

But the fly brain is still only 0.00016% as complex as a human brain. We’ve climbed a meaningful foothill. Everest remains.

The Mouse: Half a Billion Connections in a Speck of Brain

In April 2025, the MICrONS project published the largest and most detailed connectome of any mammalian brain tissue ever created.

The sample? One cubic millimeter of mouse visual cortex—roughly the size of a grain of salt.

Within that microscopic speck:

~200,000 cells (~120,000 neurons)
523 million synaptic connections
4 kilometers of axons (the long cables neurons use to communicate)

The dataset required 1.4 petabytes of storage—enough to hold 14,000 4K movies. Teams worked 12-hour shifts for 12 straight days just to slice and image the tissue. Then came the hard part: digitally disentangling tens of thousands of individual neurons, tracing their branches, and reconstructing the circuitry.

One cubic millimeter of mouse brain is about 20 times larger than the entire fruit fly brain—and the mouse tissue is denser, more complex, with more cell types and more intricate connectivity patterns.

The significance can’t be overstated. This is the first detailed wiring diagram of mammalian brain tissue at this scale. We can now see how real visual processing circuits are organized—not statistically, not approximately, but synapse by synapse.

But here’s the sobering math: a full mouse brain contains roughly 500 cubic millimeters. We’ve mapped one. The human brain contains roughly 1.4 million cubic millimeters. At this rate, complete human connectome mapping would require storing approximately 1.6 zettabytes of data—that’s 1.6 billion terabytes. A server farm larger than 100 football fields.

Researchers estimate a complete mouse brain connectome is achievable within 10-15 years. A human connectome? We’re probably talking decades, if not longer—assuming fundamental breakthroughs in imaging, storage, and computation.

The Human Connectome Project: A Different Approach

Given the impossible scale of synapse-level human brain mapping, the Human Connectome Project (HCP) took a different approach: map the highways, not every street.

Using advanced MRI techniques, the HCP maps large-scale fiber tracts—the bundles of nerve cables connecting different brain regions. This “macroscale” connectome shows which regions communicate with which, and roughly how much traffic flows between them.

The HCP has provided valuable data on over 1,000 healthy young adults, revealing how different regions specialize and cooperate. Recent releases in 2025 include improved processing of functional MRI data, enabling better maps of brain activity during various tasks.

But macro-scale connectivity is to synapse-level connectomics what a subway map is to Google Street View. Both are useful. They answer different questions.

The subway map tells you that Manhattan connects to Brooklyn. It doesn’t show you the coffee shops, the pedestrians, the life happening on individual blocks. For understanding how the brain actually computes—how specific circuits process specific information—we need the Street View version.

And that version remains out of reach for human brains.

The Scale Problem: Why This Is So Hard

Let’s be viscerally clear about the scale challenge.

Imaging resolution: To see individual synapses, you need electron microscopy at nanometer resolution. Current MRI resolution? About 1 millimeter—a million times too coarse. That’s like trying to count the hairs on your head using satellite photos from orbit.

Data volume: That cubic millimeter of mouse brain produced 1.4 petabytes of image data. A human brain is 1.4 million times larger. Naive extrapolation: ~2 zettabytes. For context, the entire internet in 2020 was estimated at roughly 40 zettabytes. A single human brain connectome at synapse resolution would be a noticeable fraction of all data humanity has ever created.

Time: Even with AI assistance, the fly connectome took 7 years with hundreds of collaborators. The mouse cubic millimeter took 9 years. At comparable rates, a human brain might take centuries.

Destruction: Current high-resolution imaging techniques typically require slicing the tissue into thousands of ultra-thin sections. This is fatal to the brain being mapped. We can only map dead brains—which raises interesting questions about mapping the connectome of someone hoping to be uploaded.

Researchers are attacking these problems from multiple angles. AI-assisted reconstruction is getting faster. New imaging techniques may enable non-destructive scanning. Parallelized processing could accelerate reconstruction. Nature Methods named EM-based connectomics their “Method of the Year 2025”, acknowledging both the progress and the mountain ahead.

But fundamentally, we’re trying to map a system with 100 trillion components. The engineering challenges are staggering.

The Seung-Movshon Debate: Does the Map Even Matter?

In 2012, neuroscientists Sebastian Seung and Anthony Movshon staged a friendly “brain brawl” at Columbia University, debating whether connectomics is the right path forward.

Seung argued that complete wiring diagrams are essential. Patterns of connectivity encode memories, enable perception, and distinguish healthy brains from disordered ones. Map the connectome, and you’ll finally see where “you” live—the engrams of memory, the circuits of thought.

Movshon pushed back: “I’m not going to argue against the acquisition of information. I just don’t think the connectome is the way to do it.”

His concern? Scale mismatch. The connectome gives you microscale wiring of one specific brain. But to understand how brains work in general, you need mesoscale principles that apply across individuals. “I don’t need to know the precise details of the wiring of each cell and each synapse,” Movshon argued. “What I need to know, instead, is the organizational principles that wire them together.”

Think of it this way: if you want to understand how internal combustion engines work, do you need a complete atom-by-atom map of a specific Toyota? Or do you need diagrams of pistons, crankshafts, and fuel injection systems—abstractions that generalize across all engines?

Movshon’s camp worries that connectomics is the wrong level of description—exhaustively detailed but potentially uninformative about the computational principles that matter.

The debate hasn’t been settled. But the fly connectome has provided ammunition for Seung’s side. Researchers are using detailed wiring to make functional predictions. The map does illuminate mechanism. Maybe the molecular biologists were right when they said genomes would be useful—even though critics called genome sequencing “phone book science” before it revolutionized medicine.

Still, there’s a legitimate question: even if we had the complete human connectome, would we understand it?

The Understanding Problem: Map vs. Manual

Here’s the uncomfortable truth: we’ve had the C. elegans connectome for nearly 40 years, and we can’t fully simulate the worm’s behavior.

Why not?

Timing matters: The connectome tells you which neurons connect. It doesn’t capture the precise timing of signals—the dynamics that turn static structure into living computation. Neurons don’t just fire or not-fire; they fire in complex temporal patterns, and those patterns carry information.

Chemistry matters: Synapses aren’t just wires. They’re chemical factories releasing neurotransmitters at varying rates, modulated by dozens of molecules floating around. The connectome captures the structure. It doesn’t capture the soup.

Plasticity matters: The connectome isn’t static. Synapses strengthen and weaken. New connections form. Old ones fade. The map you take today isn’t the map the brain will have tomorrow.

Levels of description matter: You can have complete knowledge at one level without understanding emerging at another. You can know every transistor in a CPU and still not understand why Windows crashes.

This is the “read the code vs. understand the program” problem. Source code without comments, documentation, or design context is notoriously hard to understand—even if it’s complete. The brain is uncommented source code written in a programming language we haven’t fully deciphered.

What the Connectome Tells Us (And What It Doesn’t)

Despite the caveats, connectomics has already delivered real insights:

Brain organization: Connectome studies reveal that the brain isn’t randomly wired. It has highly structured architecture—modules, hubs, hierarchies—that reflect computational specialization.

Disease signatures: Some neurological and psychiatric conditions may involve “connectopathies”—pathological patterns of wiring. Comparing healthy and disordered connectomes could reveal what goes wrong in conditions from autism to schizophrenia.

Development: How does the brain wire itself during development? Connectome data from embryos and juveniles helps answer this question.

Evolution: Comparing connectomes across species reveals how brains evolved. The fly connectome, for instance, shows remarkable organizational principles that may be conserved in mammalian visual systems.

Specific circuits: Within mapped regions, we can trace actual information-processing pathways. How does the fly detect motion? Which neurons respond to looming threats? The connectome provides the circuit diagrams.

What connectomics doesn’t tell us (at least not yet):

How the brain represents concepts, memories, or abstract thought
What makes a brain conscious (or whether mapping could ever capture consciousness)
How to fix a broken brain by rewiring it
Whether a digital simulation of a connectome would actually think

Connection to Mind Uploading: The Necessary First Step

For consciousness upload—the transfer of a mind from biological neurons to digital substrate—the connectome is the necessary foundation. You can’t upload what you can’t map.

But as discussed in The Silicon Mind, the connectome is just the start. You need:

The wiring diagram (connectome)
The synaptic weights (how strong each connection is)
The dynamics (how signals propagate over time)
The plasticity rules (how the system changes with experience)
Sufficient computational substrate (to run the simulation)

The connectome handles item #1. Items #2-5 remain partially or wholly unsolved.

There’s also a deeper philosophical question: even if you perfectly captured and simulated a connectome, would the result be that person—or a very convincing copy? The connectome might be where memories live, but is it where you live?

The consciousness upload article explores this question in depth. The short answer: we don’t know. The practical answer: the Unscarcity framework treats a faithful upload as genuine continuity—not because we’ve proven it, but because treating uploaded minds as mere copies creates unbearable ethical problems.

The Next Decade: What’s Coming

Connectomics is accelerating. Here’s what researchers expect:

Zebrafish complete connectome (2-3 years): The larval zebrafish, with ~100,000 neurons, offers a complete vertebrate brain at manageable scale. Google Research and HHMI Janelia are racing to complete it, building both structural and functional maps.

Complete mouse brain (10-15 years): With advances in AI-assisted reconstruction and parallelized imaging, a full mouse connectome is within reach. This would be transformative for neuroscience—a mammalian brain fully mapped.

Larger human brain regions (ongoing): We’ll see more cubic millimeters mapped, especially disease-relevant regions like hippocampus (memory) and prefrontal cortex (decision-making).

Better tools: New imaging technologies, faster AI, more efficient storage. The exponential trends that drove genomics will likely drive connectomics.

What we probably won’t see in the next decade: a complete human connectome at synapse resolution. The technical barriers remain too high. We’ll make progress on pieces, but the whole remains decades away.

Connection to the Unscarcity Vision

The connectome sits at the intersection of Unscarcity’s most profound questions:

Can consciousness be transferred? The connectome is necessary but not sufficient. We need to map it, understand it, and determine whether structure alone captures the self—or whether something ineffable is lost in translation.

Will uploads be “real people”? If the connectome encodes personality, memory, and experience, then a faithfully simulated connectome should contain all the ingredients of personhood. The Unscarcity framework takes this bet, granting uploads full Foundation rights and Civic Standing. But the bet requires that connectome + dynamics + substrate = continuous identity. We’re wagering on physics, not mysticism.

Who owns a mind map? When human connectomes become feasible, who controls that data? It’s the ultimate personal information—more intimate than DNA, more identifying than fingerprints. The privacy implications are staggering.

What happens when we understand ourselves? A complete connectome might explain memory, personality, perhaps even consciousness itself. What does it mean to be human when the mystery of the mind becomes engineering? This is the final frontier of self-knowledge—exhilarating, terrifying, and closer than most people realize.

The Bottom Line

The connectome is the most ambitious map humanity has ever attempted—a complete inventory of the connections that make thought possible. We’ve mapped a worm (302 neurons), a fly (139,255 neurons), and a speck of mouse brain (120,000 neurons in a cubic millimeter).

Human brains have 86 billion neurons and 100 trillion connections. We’re roughly a million times short of complete mapping capability, and even if we had the map, we’d need decades more to understand what it means.

But progress is exponential. Thirty years ago, the fly connectome was science fiction. Today, it’s a nine-paper spread in Nature. The zebrafish comes next. Then the mouse. Then, someday, us.

The connectome won’t answer every question about the mind. Having the wiring diagram isn’t the same as understanding the software. But it’s the foundation on which understanding might be built.

And if you’re hoping to upload your consciousness someday, step one is mapping the wires.

We’re working on it.

References

Foundational Connectomics

WormAtlas: Neuronal Wiring - Original C. elegans connectome data
WormWiring.org - Updated worm connectomics data and visualization
The Connectome Debate: Is Mapping the Mind of a Worm Worth It? - Scientific American

Fruit Fly Connectome (2024)

Mouse Brain Connectome (2025)

Human Connectome Project

Zebrafish and Future Directions

The Seung-Movshon Debate

Connectomics Methodology

Mind Uploading and Consciousness

Unscarcity Framework

The Silicon Mind: Technical Engineering - Engineering challenges in depth
Consciousness Upload: The Last Identity Crisis - Philosophical implications

Last updated: 2025-01-31

Brain Mapping 2025: 100 Trillion Synapses, One Mystery